Guidelines for Management of Acute Cervical Spinal Injuries
-
Upload
muhammad-irfan -
Category
Documents
-
view
177 -
download
4
Transcript of Guidelines for Management of Acute Cervical Spinal Injuries
INTRODUCTION
Guidelines for Management of Acute Cervical Spinal Injuries
Spinal cord injuries occur approximately 14,000 times peryear in North America, and most involve the cervicalspine region. Most, although not all, of these injuries will
include cervical spine fracture-dislocations. Patients who sus-tain cervical spinal cord injuries usually have lasting, oftendevastating, neurological deficits and disability. In addition,tens of thousands of patients each year will sustain traumaticcervical spine injuries without spinal cord injury. The man-agement of these patients and their injuries, spinal cord, andvertebral column has not been standardized nor is it consis-tent within a single institution, from one center to another, oramong centers within a geographic region. Treatment strate-gies are usually dependent on the experiences of institutionsor individual health care providers, on physician training, andon the resources available at the treatment facility. Manage-ment can affect outcome in these patients; therefore, cliniciansworldwide strive to provide the “best and most timely care.”Often, we may not be fully aware of what the “best care” may be,or whether “timeliness” matters. In many circumstances, “bestcare” likely encompasses a variety of treatment strategies, allwith acceptable success rates and reasonable inherent risks.
The Section on Disorders of the Spine and PeripheralNerves of the American Association of Neurological Surgeonsand the Congress of Neurological Surgeons has long beeninterested in seeking answers to the key management issuesassociated with acute spine and spinal cord injuries. Identifi-cation of “best care” strategies is desired for all aspects of thecare of patients with acute cervical injury. Such strategies caninclude care and transport of the patient before admission tothe facility, neurological and radiographic assessment, medi-
cal management of spinal cord injury, closed reduction ofcervical fracture-dislocations, and specific treatment options,both operative and nonoperative, for each specific cervicalinjury type known to occur from the occiput through the firstthoracic level. The leadership of the Spine Section charged uswith the task of generating guidelines on the management ofpatients with acute cervical spine and cervical spinal cordinjuries. We began, in May 2000, by identifying 22 topic areasand generating questions around which recommendationswould be formed. We followed a meticulous process foundedin evidence-based medicine. Published scientific evidencewas searched for and relied on rather than expert opinion ortraditional practices. In the course of developing these recom-mendations, new methodology had to be created for classify-ing evidence on clinical assessment, resulting in a substantialcontribution to the guidelines literature as a whole. We firstconvened in September 2000; 1 year later, the task wascompleted.
The authors and the Joint Section on Disorders of the Spineand Peripheral Nerves hope that these guidelines will definethe variety of assessment and treatment options available to aclinician in the management of an individual patient, providedirection within the broad scope of clinical practice derivedfrom medical evidence, highlight what is known regardingspecific issues, and, importantly, define what is not knownand stimulate additional research.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
Drawing by Leonardo da Vinci of the human cranium and spinal canal. Courtesy, Dr. Edwin Todd, Pasadena, California. S1
FOREWORD
On behalf of the American Association of NeurologicalSurgeons/Congress of Neurological Surgeons JointSection on Disorders of the Spine and Peripheral
Nerves, it is my great privilege to introduce these Guidelinesfor the Management of Acute Cervical Spine and Spinal CordInjuries. These guidelines represent the initial installment of amore comprehensive guidelines initiative from the Joint Sec-tion on behalf of all practicing neurosurgeons and their pa-tients. The Section is grateful to the small working group whodevoted considerable time and effort to the generation of thisoutstanding document. We would like to formally recognizethe Joint Section on Trauma for their important collaborationon this project. The Section would also like to acknowledgeand thank the parent organizations, the American Associationof Neurological Surgeons and the Congress of NeurologicalSurgeons, for their guidance of and support for this project,most notably through the efforts of the American Associationof Neurological Surgeons/ Congress of Neurological Sur-geons Guidelines Committee.
The Section is also deeply indebted to Michael Apuzzo andthe staff of Neurosurgery for their advice and editorial assis-tance in preparing this document for publication. The appli-cation of Neurosurgery’s rigorous peer-reviewed editorial pro-cess has clearly enhanced the quality, balance, and stature ofthis document. Perhaps most importantly, Neurosurgery hasprovided an extraordinary vehicle for the widespread dissem-ination and ultimate incorporation of these guidelines to im-prove the care and enhance the outcomes of patients withtraumatic cervical spine and spinal cord injuries.
One of the truly important functions of organized neuro-surgery is the generation of evidenced-based clinical practiceguidelines. Properly developed, such guidelines can answerimportant questions, resolve uncertainty, identify areas ofdeficient knowledge and opportunities for future scientificinvestigation, standardize treatment, and improve the qualityof care and the outcomes for patients. The now widely dis-seminated head trauma guidelines, for example, have clearlymade a difference in the outcomes of patients with severehead injury.
Guidelines development is a highly structured process withrigorous methodological criteria and exacting standards. It isa time-, labor-, and resource-intensive process that has servedas a significant obstacle to more widespread guidelines de-velopment throughout neurosurgery. In the past, clinicalpractice guidelines have been developed by publicly sup-ported epidemiologists and methodologists who understoodstudy design, data analysis, and the guidelines process, butnot the disease. This absence of context and clinical perspec-tive significantly limited the value and relevance of theirresults. Alternatively, clinician-generated guidelines oftentook the form of methodologically flawed consensus panelsand expert opinion, also of limited value.
The Joint Spine Section recognized the importance ofevidence-based clinical practice guidelines and the challenges
of their development. The appropriate clinical expertise, strictadherence to established methodological standards for guide-lines development, and considerable resource investment forthe development, dissemination, and maintenance of theguidelines documents were deemed crucial to our guidelinesinitiative. Cervical spine and spinal cord injury was chosen asthe initial guidelines topic because of the personal, social, andeconomic devastation of these injuries, their complex nature,and the high level of uncertainty, as reflected in wide practicevariations, as to the value and indications for many of theaspects of evaluation and treatment.
The clinical practice guidelines contained in this supple-ment to Neurosurgery represent a remarkable effort. Theyaddress the key issues related to the evaluation and manage-ment of these complex conditions that are relevant to thetreating physician. In every chapter, the pertinent issues aresuccinctly stated, the published data are comprehensivelypresented in the evidentiary tables, and the evidence is thor-oughly discussed and critically evaluated throughout the text.The linkage between the quality of the evidence and thestrength of the recommendations was not a “black box” pro-cess but an open, deliberative exercise by skilled expertsguided by a rigorous set of standards.
Despite the strength and potential value of this document,it is important to acknowledge the inherent limitations ofclinical practice guidelines. This, or any other, evidence-basedclinical practice guidelines document does not represent thedefinitive source of knowledge on the stated topic. Rather, itrepresents recommendations of varying strength and cer-tainty based on an analysis of the best available publisheddata. These data, however, are often conflicting, flawed, orincomplete, and there are unavoidable elements of potentialbias from subjectivity, perspective, and experience of the in-dividuals and group involved in the analysis and interpreta-tion of these data. In essence, proof is a relative term based onthe interpretation of evidence. Furthermore, it is subject todifferent standards. A relevant example comes from the fieldof jurisprudence, where the standard of proof (i.e., guilt) forcriminal trials is “beyond a reasonable doubt,” whereas thestandard for civil courts must simply reflect “a preponderanceof evidence” or “more likely than not.” These different stan-dards evolved because of the perceived different conse-quences of a wrongful verdict. Moreover, as Stephen Haineslikes to note, the verdict “not guilty” does not mean innocent;it merely says not proved. Such are the vagaries associatedwith the interpretation of even scientific evidence. Principledpeople can look at the same evidence and come to differentconclusions subject to their own personal perspective, expe-rience, and stake in the result.
Nevertheless, the Joint Section and Guidelines Develop-ment Group went to great lengths to identify and avoid—or atleast minimize—these potential problems. The working groupadopted the most widely recognized and rigorous standardsfor guidelines development. A diverse panel of experts with
S-ii Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
expertise in spine, trauma, and epidemiology brought rele-vant clinical, scientific, and methodological competence toenhance both the analytical and the deliberative aspects of thisprocess. Periodic outside reviews were routinely obtained fortopics or areas of contention or uncertainty to add additionalperspective and balance. Above all, the process was account-able and transparent at every stage.
Ultimately, we offer these guidelines as a living documentto those professionals who treat patients with traumatic spinalinjury. We hope each practitioner will critically evaluate theseguidelines and come to his or her own conclusion on how,whether, and when to implement its recommendations. Itmay be used either as a reference or as a basis for standard-ized protocols of evaluation and management of the patients
with traumatic spinal injury. For clinical and basic scienceresearchers, we hope that it will identify and catalyze scien-tific investigation in areas of deficient knowledge. As a Sec-tion, we stand firmly behind this important document andwill continuously update the recommendations as newknowledge and understanding is developed. We sincerelybelieve that these guidelines can improve the care and en-hance the outcomes of patients with traumatic injuries to thecervical spine and spinal cord.
Paul C. McCormickChair, Joint Section on Disorders of the
Spine and Peripheral NervesNew York, New York
Foreword
METHODOLOGY
Methodology of Guideline Development
The evolution of medical evidence has occurred rapidlyduring the past 50 years. From initial reports, which areanecdotal in nature, to large-scale randomized con-
trolled trials, medical evidence is variable. From the evidence,and influenced by personal experience, clinicians choose waysto manage disease. The medical specialties have pioneered theuse of evidence from experimental trials to support clinicalpractice decisions. The surgical specialties have lagged behindin the development of large-scale studies of surgical proce-dures and perioperative management. However, the high costof medical care, together with variations in practice fromregion to region, has given rise to an interest in developingstrategies for linking practice to underlying evidence. Duringthis process, it has become clear that the variability of theevidence must somehow be reflected in any recommenda-tions derived from it.
In the 1980s, criteria were developed for use in selectingevidence for developing treatment recommendations. In aformal document, Clinical Practice Guidelines: Directions for aNew Program (4), the Institute of Medicine addressed suchissues as definition of terms, specification of key attributes ofgood guidelines, and certain aspects of planning for imple-mentation and evaluation (4). The key intent was to promotestandardization and consistency in guideline development.Several key concepts were espoused:
1. A thorough review of the scientific literature should pre-cede the development of guidelines.
2. The available scientific literature should be searched byusing appropriate and comprehensive search terminology.
3. The evidence should be evaluated and weighted to reflectthe scientific validity of the methodology used to generatethe evidence.
4. There should be a link between the available evidence andthe recommendations, with the strength of the recommen-dations reflecting the strength of the evidence, in turnreflecting the scientific certainty (or lack thereof) of theevidence.
5. Empirical evidence should take precedence over expertjudgment in the development of guidelines.
6. Expert judgment should be used to evaluate the quality ofthe literature and to formulate guidelines when the evi-dence is weak or nonexistent.
7. Guideline development should be a multidisciplinary pro-cess, involving key groups affected by the recommendations.
To develop the guidelines for managing acute cervicalspine and spinal cord injuries, we used the evidence-basedapproach reflected in those concepts, rather than a consensus-based approach with input from experts based on the litera-ture and their personal experience. We used a strict process ofliterature review, ranking the published articles by strength of
study design. Every effort was made to maintain objectivityand avoid the influence of personal or professional biasthrough a methodology defined in advance. The methodologywe chose follows the recommendations of the Institute ofMedicine Committee to Advise the Public Health Service onClinical Practice Guidelines (4), as outlined in detail below.
METHODOLOGY FOR DEVELOPING GUIDELINES
Literature search
The first step was to undertake an extensive literaturesearch for each clinical question addressed. The computerizeddatabase of the National Library of Medicine was searched foravailable English-language literature on human studies pub-lished during the past 25 years. The search terms reflected theclinical question in as much detail as possible, as described ineach chapter. Abstracts were reviewed, and clearly relevant ar-ticles were selected for evaluation. Each article was evaluated bystudy type (e.g., therapy, diagnosis, or clinical assessment).
Evaluating the strength of the therapy literature
Evidence can be generated by several different study de-signs. The strongest study protocol, when well designed andexecuted, is the randomized controlled trial. The prospectiv-ity, presence of contemporaneous comparison groups, andadherence to strict protocols reduce sources of systematicerror (or bias). The randomization process reduces the influ-ence of unknown aspects of the patient population that mightaffect the outcome (random error).
The next strongest study designs are the nonrandomizedcohort study and the case-control study. These designs alsocompare groups that receive specific treatments, but in anonrandomized fashion. In the nonrandomized cohort study,an established protocol for patient treatment is followed andgroups are compared in a prospective manner, provided thattheir allocation to the treatment group is not determined bycharacteristics that would not allow them to receive either ofthe treatments being studied. The patients in each group havethe disorder of interest (e.g., spinal cord injury) and receivedifferent interventions, and the differences in outcome arethen studied. In the case-control study, the patients aregrouped by outcome (e.g., functional ability), and their treat-ment (e.g., surgery versus no surgery) is evaluated for arelationship. These studies are more subject to systematic andrandom error and are therefore less compelling than the ran-domized controlled trial. However, a randomized controlledtrial with significant design flaws that threaten its validitymay be classified as a weaker study. Least strong evidence isgenerated by published series of patients with the same orsimilar disorder followed for outcome, but not compared as totreatment. Also in this category are the case report, expert
S2 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
opinion, and the randomized controlled trial so flawed thatthe conclusions are uncertain. These statements regardingstudy strength refer to studies on treatment, but patient man-agement also includes diagnosis and clinical assessment.These aspects of patient care require clinical studies that aredifferent in design, generating evidence about choices of di-agnostic tests and clinical measurement.
Evaluating the strength of the diagnostic test literature
To be useful, diagnostic tests must be reliable and valid.Reliability refers to the test’s stability in repeated use, underthe same circumstances. Validity describes the extent to whichthe test reflects the “true” state of affairs, as measured bysome “gold standard” reference test. Accuracy reflects thetest’s ability to determine which patient does and which doesnot have the suspected disorder. Overall, the test must beaccurate in picking out the true-positives and true-negatives,with the lowest possible false-positive and false-negativerates. The attributes of diagnostic tests are represented bysensitivity, specificity, positive predictive value, and negativepredictive value. These values may be calculated by using aBayesian 2 � 2 table (Table 1).
From Table 1, the components of accuracy can be expressedand calculated as shown in Table 2. A characteristic of diag-nostic tests is that these attributes do not always rise together,but in general, the accuracy values should be higher than 70%for the test to be considered useful. The issue of reliability ofthe test will be discussed below.
Evaluating the strength of the patientassessment literature
There are two points to consider when patient assessmentis key in the patient management paradigm: the initial assess-ment (e.g., the patient’s condition in the trauma room) and theultimate assessment, or outcome. All assessment tools,whether they are radiographic, laboratory, or clinical, requirethat the measurement be reliable. In assessments carried outby mechanical or electronic equipment, reliability is ensuredby calibrating the devices regularly. In assessments carriedout by observers, reliability is ensured by verifying agreement
TABLE 1. Bayesian Table for Calculating the Attributes of Diagnostic Test Literature
Test Resulta“Gold Standard”
Patient Has Injury Patient Has No Injury
Positive:appears to have injury
True-positive(a)
False-positive(b)
(a � b)
Negative:appears to have no injury
False-negative(c)
True-negative(d)
(c � d)
(a � c) (b � d) (a � b � c � d)a Per cervical spine x-ray.
TABLE 2. Calculation for Level of Accuracy of aDiagnostic Test
Test Attribute FormulaProbabilityStatement
Sensitivity aa � c
If a patient has apositive x-ray,how likely ishe/she to havethe injury ofinterest?
Specificity db � d
If a patient has anegative x-ray,how likely ishe/she to nothave the injuryof interest?
Positive predictive value aa � b
If a patient hasan injury, howlikely is he/sheto have apositive test?
Negative predictive value dc � d
If a patient doesnot have aninjury, howlikely is he/sheto have anegative test?
Accuracy a � da � b � c � d
Methodology of Guideline Development
S3Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
between different observers of the same assessment and bythe same observer at different times. Because a certain amountof agreement between observers or observations is expectedto occur by chance alone, a statistic has been developed tomeasure agreement between observations or observers be-yond chance: the index of concordance, or the kappa (�)statistic (3). Once again, the Bayesian 2 � 2 table can be usedto understand and to calculate � (Table 3).
From these numbers, � is calculated by:
� �N�a � d� � �n1 f1 � n2 f2�
N2 � �n1 f1 � n2 f2�
or
� �2�ad � bc�
�n1 f2 � n2 f1�
Values of � indicate the strength of agreement betweenobservers or observations, as shown in Table 4 (5). We ratedeach article on clinical assessment for its adherence to therules of reliability and noted the exact � value. These valueswere linked to the strength of recommendations, as describedbelow.
Linking evidence to guidelines
The concept of linking evidence to recommendations hasbeen further formalized by the American Medical Associationand many specialty societies, including the Congress of Neu-
rological Surgeons, the American Association of NeurologicalSurgeons, and the American Academy of Neurology (1, 2, 6,7). This formalization involves designating a specific relation-ship between the strength of evidence and the strength of theresulting recommendations to avoid ambiguity. In the para-digm for therapeutic maneuvers, evidence derived from thestrongest clinical studies (well-designed randomized con-trolled trials) generates Class I evidence. Class I evidence isused to support recommendations of the strongest type,called Practice Standards, indicating a high degree of clinicalcertainty. Nonrandomized cohort studies, randomized con-trolled trials with design flaws, and case-control studies (com-parative studies with less strength) are designated Class IIevidence. Class II evidence is used to support recommenda-tions called Practice Guidelines, reflecting a moderate degreeof clinical certainty. Other sources of information, includingobservational studies (e.g., case series and expert opinion), aswell as randomized controlled trials with flaws so serious thatthe conclusions of the study are truly in doubt, generate ClassIII evidence. Class III evidence supports Practice Options, re-flecting unclear clinical certainty. These categories of evidenceare summarized in Table 5. The general term for all of therecommendations is Practice Parameters. Because so few Prac-tice Standards exist, the term more commonly used to de-scribe the whole body of recommendations is Practice Guide-lines. Thus, we have named this document Guidelines for theManagement of Acute Cervical Spine and Spinal Cord Injuries.
One of the practical difficulties encountered in implement-ing this methodology is that a poorly designed randomizedcontrolled trial might take precedence over a well-designedcase-control or nonrandomized cohort study. We attemptedto avoid this pitfall by carefully evaluating the quality of thestudy as well as its type.
These criteria apply to practice guidelines (parameters) fortreatment. To assess literature pertaining to prognosis, diag-nosis, and clinical assessment, completely different criteria
TABLE 3. Bayesian Table for Calculating the Reliability ofPatient Assessment Literature
Observer#2
Observer #1
Yes No
Yes Agree(a)
Disagree(b)
(a � b) � f1
No Disagree(c)
Agree(d)
(c � d) � f2
(a � c) � n1 (b � d) � n2 (a � b � c � d) � N
TABLE 4. Relationship between the Index of Concordance(�) and the Strength of Agreement between Observersor Measurements
Value of � Strength of Agreement
�0 Poor
0–0.20 Slight
0.21–0.40 Fair
0.41–0.60 Moderate
0.61–0.80 Substantial
0.81–1.00 Almost perfect
TABLE 5. Classification of Evidence onTherapeutic Effectiveness
Evidence Class Description
Class I Evidence from one or more well-designed, randomized controlled clinicaltrials, including overviews of such trials.
Class II Evidence from one or more well-designedcomparative clinical studies, such asnonrandomized cohort studies, case-control studies, and other comparablestudies, including less well-designedrandomized controlled trials.
Class III Evidence from case series, comparativestudies with historical controls, casereports, and expert opinion, as well assignificantly flawed randomizedcontrolled trials.
S4 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
must be used. Criteria for prognosis have been developed andwere widely used in the publications of prognostic indicatorsin severe traumatic brain injury (1). We have not addressedissues of prognosis in this document. For diagnosis, the issueis the ability of the diagnostic test to successfully distinguishbetween patients who have and those who do not have thedisease or finding of interest. Classification of evidence ondiagnostic tests is illustrated in Table 6. Measures of clinicalassessment must be both reliable and valid. The assessmentmust be performed reliably by different observers and by thesame observer at different times. To be valid, the clinicalassessment, like diagnostic tests described above, must ade-quately represent the true condition of the patient. This latteraspect is difficult to measure, so most clinical assessments aregraded according to their reliability (Table 7).
For each question addressed in these guidelines, we as-sessed the study type in each identified article and assigned aclassification according to the scheme outlined above. Thesedesignations are listed in the evidentiary tables in eachchapter.
GUIDELINES DEVELOPMENT PROCESS
A group of individuals with interest and expertise in thetreatment of patients with cervical spine injuries and/or thedevelopment of guideline practice parameters was assembledunder the auspices of and with the support of the Joint Sectionon Disorders of the Spine and Peripheral Nerves of the Amer-ican Association of Neurological Surgeons/Congress of Neu-
rological Surgeons. The group reflected expertise in spinalneurosurgery, neurotrauma, and clinical epidemiology. Theissues chosen for inclusion in the document were those con-sidered pertinent to the management of patients with acutecervical spine and/or spinal cord injury (e.g., transport, med-ical management, treatment of specific fracture/dislocationpatterns, vascular injury, and prophylaxis for thromboem-bolic events).
A MEDLINE search of literature published from January1966 to January 2001 was carried out by using the searchterms described in each chapter. The search was limited tohuman subjects and included English language literature forall but one of the chapters. Additional articles were foundthrough the reference lists in the articles found, as well asfrom other sources known to the authors. Articles were re-jected on the basis of irrelevance to the clinical question athand. Case reports were included if there was insufficientmaterial from case series. All articles were evaluated accord-ing to the medical evidence-based protocol outlined above.Recommendations were derived for therapy, diagnosis, andclinical assessment. Chapters written by the primary authorswere rewritten by a different set of authors, and the finalproduct was agreed on by consensus. On occasion, the as-sessed quality of the study design was so contentious and theconclusions so uncertain that we assigned a lower classifica-tion than might have been expected without such a detailedreview. In every way, we sought to adhere to the Institute ofMedicine criteria for searching, assembling, evaluating, andweighting the available medical evidence and linking it to thestrength of the recommendations presented in this document.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.Email: [email protected]
TABLE 6. Classification of Evidence on Diagnosis
Evidence Class Description
Class I Evidence provided by one or more well-designed clinical studies of a diversepopulation using a “gold standard”reference test in a blinded evaluationappropriate for the diagnosticapplications and enabling theassessment of sensitivity, specificity,positive and negative predictive values,and, where applicable, likelihood ratios.
Class II Evidence provided by one or more well-designed clinical studies of a restrictedpopulation using a “gold standard”reference test in a blinded evaluationappropriate for the diagnosticapplications and enabling theassessment of sensitivity, specificity,positive and negative predictive values,and, where applicable, likelihood ratios.
Class III Evidence provided by expert opinion,studies that do not meet the criteria forthe delineation of sensitivity, specificity,positive and negative predictive values,and, where applicable, likelihood ratios.
TABLE 7. Classification of Evidence on Clinical Assessment
Evidence Class Description
Class I Evidence provided by one or more well-designed clinical studies in whichinterobserver and intraobserverreliability is represented by a � statisticof 0.80 or greater.
Class II Evidence provided by one or more well-designed clinical studies in whichinterobserver and intraobserverreliability is represented by a � statisticof 0.60 or greater.
Class III Evidence provided by one or more well-designed clinical studies in whichinterobserver and intraobserverreliability is represented by a � statisticof less than 0.60.
Methodology of Guideline Development S5
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
REFERENCES
1. Bullock R, Chesnut RM, Clifton GL, Ghajar J, Marion DW, Narayan RK,Newell DW, Pitts LH, Rosner MJ, Walters BC, Wilberger JE: Guidelinesfor the management of severe traumatic brain injury: Intracranial pres-sure treatment threshold. J Neurotrauma 17:493–495, 2000.
2. Bullock R, Chesnut RM, Clifton G, Ghajar J, Marion DW, NarayanRK, Newell DW, Pitts LH, Rosner MJ, Wilberger JW: Guidelines forthe management of severe head injury: Brain Trauma Foundation.Eur J Emerg Med 3:109–127, 1996.
3. Cohen J: A coefficient of agreement for nominal scales. EducPsychol Meas 20:37–46, 1960.
4. Field MJ, Lohr KN (eds): Clinical Practice Guidelines: Directions for aNew Program—Committee to Advise the Public Health Service on Clin-ical Practice Guidelines: Institute of Medicine. Washington DC, Na-tional Academy Press, 1990.
5. Landis JR, Koch GG: The measurement of observer agreement forcategorical data. Biometrics 33:159–174, 1977.
6. Rosenberg J, Greenberg MK: Practice parameters: Strategies forsurvival into the nineties. Neurology 42:1110–1115, 1992.
7. Walters BC: Clinical practice parameter development in neurosur-gery, in Bean JR (ed): Neurosurgery in Transition: The SocioeconomicTransformation of Neurological Surgery. Baltimore, Williams &Wilkins, 1998, pp 99–111.
Semischematic drawing of the cervical vertebrae shows that the cervical musculature stabilizes the cervical spinal column.Reproduced from, Keele KD, Pedretti C: Leonardo da Vinci: Corpus of the Anatomical Studies in the Collection of Her Maj-esty the Queen at Windsor Castle. London, Harcourt Brace Jovanovich, 1979.
S6 Guidelines for Management of Acute Cervical Spinal Injuries
CHAPTER 1
Cervical Spine Immobilization before Admission to the Hospital
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• All trauma patients with a cervical spinal column injury or with a mechanism of injury having the potential
to cause cervical spine injury should be immobilized at the scene and during transport by using one ofseveral available methods.
• A combination of a rigid cervical collar and supportive blocks on a backboard with straps is effective inlimiting motion of the cervical spine and is recommended. The long-standing practice of attempted cervicalspine immobilization using sandbags and tape alone is not recommended.
RATIONALE
Early management of the patient with a potential cervicalspinal cord injury begins at the scene of the accident. Thechief concern during the initial management of patients
with potential cervical spine injuries is that neurological func-tion may be impaired by pathological motion of the injuredvertebrae. It is estimated that 3 to 25% of spinal cord injuriesoccur after the initial traumatic insult, either during transit orearly in the course of management (14, 15, 42, 48, 81, 97).Many cases have been reported that had a poor outcomebecause of mishandling of cervical spine injuries (12, 51, 81,97). As many as 20% of spinal column injuries involve mul-tiple noncontinuous vertebral levels; therefore, the entire spi-nal column is potentially at risk (38, 39, 66, 73). Consequently,complete spine immobilization has been used in spinal care,before admission to the hospital, to limit motion until injuryhas been ruled out (2, 5, 27, 40, 66, 73, 76, 100, 104). During thelast 30 years, the neurological status of spinal cord-injuredpatients arriving in emergency departments has dramaticallyimproved. During the 1970s, most patients (55%) referred toregional spinal cord injury centers arrived with completeneurological lesions. In the 1980s, however, most spinal cord-injured patients (61%) arrived with incomplete lesions (46).This improvement in the neurological status of patients hasbeen attributed to the development of emergency medicalservices (EMS) initiated in 1971, and the care (including spineimmobilization) rendered by EMS personnel before the pa-tient reaches the hospital (2, 45, 46, 103). Spine immobilizationis now an integral part of preadmission management and isadvocated, for all patients with potential spine injury aftertrauma, by EMS programs nationwide and by the AmericanCollege of Surgeons (1, 2, 5, 6, 16, 32, 70, 93).
Recently, the use of spine immobilization for all traumapatients, particularly those with a low likelihood of traumatic
cervical spinal injury, has been questioned. It is unlikely thatall patients rescued from the scene of an accident or site oftraumatic injury require spine immobilization (34, 50, 69, 76).Some authors have developed and advocate a triage systembased on clinical criteria to select patients for preadmissionspine immobilization (13, 32, 74).
Several devices are available for immobilizing the patientwith a potential spine injury during transportation to thehospital. However, the optimal device has not yet been iden-tified by careful comparative analysis (17, 21, 27, 53, 61, 64, 94,99). The recommendations of the American College of Sur-geons consist of a hard backboard, a rigid cervical collar,lateral support devices, and tape or straps to secure the pa-tient, the collar, and the lateral support devices to the back-board (3, 5). A more uniform, universally accepted method forspine immobilization for trauma patients with potential spineinjury may reduce the cost and improve the efficiency ofpreadmission spinal injury management (13, 32, 74). Al-though spine immobilization is typically effective in limitingmotion, it has been associated with morbidity in a smallpercentage of cases (4, 9, 18, 19, 26, 55, 90, 100). These issuesare the subject of this review on the use and effectiveness ofpreadmission spine immobilization.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of literature published from 1966 to 2001 was per-formed. The search was limited to the English language andhuman studies. The medical subject heading “spinal immobi-lization” produced 39 articles. A second search, combining theexploded terms “spinal injuries” and “immobilization,”yielded 122 articles. A third search, combining the explodedterms “spinal injuries” and “transportation of patients,”yielded 47 articles. A fourth search, combining the exploded
S7Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
terms “spinal injuries” and “emergency medical services,”produced 119 articles. Additional references were culled fromthe reference lists of the articles. Finally, members of theauthor group were asked to contribute articles known to themon the subject matter that were not found by other searchmethods. Duplicate references were discarded. The abstractswere reviewed, and articles unrelated to the specific topicwere eliminated. This process yielded 101 articles for thisreview, which are listed in the reference list. Articles used toformulate this guideline are summarized in Table 1.1.
SCIENTIFIC FOUNDATION
Pathological motion of the injured cervical spine may createor exacerbate cervical spinal cord or cervical nerve root injury(38–40, 66, 73, 96). This potential has led to the use of spineimmobilization for trauma patients who have sustained acervical vertebral column injury or experienced a mechanismof injury that could result in cervical spinal column injury (5,6, 27, 33, 34, 40, 66, 73, 74, 76, 104).
Kossuth (56, 57) is credited with pioneering the currentlyaccepted methods of protecting and immobilizing the cervicalspine during extrication of patients with acute injury. Far-rington (36, 37) championed the concept of preadmissionimmobilization. Dick and Land (30) noted in their review ofspine immobilization devices that techniques of preadmissionspine immobilization appeared as early as 1971 in standardEMS texts and in the American Academy of Orthopedic Sur-geons Committee on Injuries Emergency text (2). Initially, thepreferred method for immobilizing the cervical spine was touse a combination of a soft collar and a rolled-up blanket (21).Later, in 1974, Hare introduced a more rigid extrication collar.Hare’s contribution launched an era of innovation for spineimmobilization devices (27).
Currently, in North America, spine immobilization is one ofthe most frequently performed procedures in the preadmis-sion care of patients with acute trauma (2, 6, 7, 27, 38, 40, 66,73, 76, 98, 104). Although clinical and biomechanical evidencedemonstrates that spine immobilization limits pathologicalmotion of the injured spinal column, there is no Class I orClass II medical evidence to support spinal column immobi-lization in all patients after trauma. Although immobilizationof an unstable cervical spinal injury makes good sense, andClass III evidence reports exist of neurological worseningwith failure of adequate spine immobilization, no case-controlstudies or randomized trials address the effect of spine im-mobilization on clinical outcomes after cervical spinal columninjury (6, 27, 32, 40, 42, 48, 50, 66, 69, 73, 96). The issue isimportant; tens of thousands of patients with trauma aretreated with spine immobilization each year, but few of themwill have spinal column injuries or instability (39, 74, 83).
Other considerations in the use of preadmission spine im-mobilization include the cost of equipment, the time andtraining of EMS personnel to apply the devices, and theunnecessary potential morbidity for patients who do not needspine immobilization (4, 9, 18, 19, 26, 27, 55, 58, 84, 90, 100). Aswith many interventions in the practice of medicine, spineimmobilization has been instituted in preadmission manage-
ment of trauma patients with potential spinal injuries on thebasis of principles of neural injury prevention and years ofclinical experience, but without supportive scientific evidencefrom rigorous clinical trials. For a variety of both practical andethical reasons, it may be impossible to obtain this informa-tion in clinical trials.
In 1989, Garfin et al. stated that no patient should be extri-cated from a crashed vehicle or transported from an accidentscene without spinal stabilization (40). The authors creditedstabilization of the cervical spine as a key factor in decliningpercentages of complete spinal cord injury lesions, from 55%in the 1970s to 39% in the 1980s, and in the significant reduc-tion of mortality in patients with multiple injuries that includecervical spine injuries. Unfortunately, no Class I or Class IImedical evidence supports their claims.
Few articles have directly evaluated the effect of preadmis-sion spine immobilization on neurological outcome after in-jury. Several Class III evidence reports cite lack of immobili-zation as a cause of neurological deterioration among acutelyinjured trauma patients transported to medical facilities fordefinitive care (12, 40, 51, 62, 81). The most pertinent study isToscano’s (96) retrospective case series report. Toscano, in1988, reported that 32 (26%) of 123 trauma patients sustainedmajor neurological deterioration in the period between injuryand admission at the hospital. The author attributed neuro-logical deterioration to patient mishandling and cited the lackof spine immobilization after traumatic injury as the primarycause. The report supports the need for spine immobilizationof trauma patients with potential spinal column injuries be-fore admission to the hospital.
In contrast, a collaborative, 5-year retrospective chart re-view reported by the University of New Mexico and theUniversity of Malaya challenges this position. Hauswald et al.(50) analyzed only patients with acute blunt spine or spinalcord injuries. At the University of Malaya, none of the 120patients they managed were immobilized with spinal ortho-ses during transport. All 334 patients managed at the Univer-sity of New Mexico were initially treated with spine immo-bilization. The hospitals were reportedly comparable inphysician training and clinical resources. Two independentphysicians, blinded to the participating hospital, character-ized the neurological injuries into two groups: disabling andnondisabling. Data were analyzed by logistic regression tech-niques, with hospital, patient age, sex, anatomic level of in-jury, and injury mechanism as variables. Neurological deteri-oration after injury was less frequent in patients with spinalinjuries in Malaya, who were not treated with formal spineimmobilization during transport (odds ratio, 2.03; 95% confi-dence interval, 1.03–3.99; P � 0.04), than in patients in NewMexico, who were managed with spinal column immobiliza-tion techniques. Even with the analysis limited to cervicalspine injuries, no significant protective effect from spine im-mobilization was identified. The authors theorized that be-cause the initial injury is of tremendous force, additionalmovement of the spine by the patient or rescuers is insuffi-cient to cause further injury. However, they noted that be-cause of the small sample size, the benefit of spine immobili-zation might not be statistically measurable in their study.
S8 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
This report has been challenged, and several flaws have beenidentified. Patients who died at the scene or during transportwere excluded from analysis. Injuries were not matched byseverity of neurological injury or by type of spinal columninjury. The mechanisms of injury differed dramatically in thetwo populations. Malayan patients were immobilized or heldimmobile during transport, but spinal orthoses as immobili-zation devices were not used. For these reasons and others,the conclusions drawn by the authors are questionable (27,76).
Evidence in the literature evaluating the effectiveness ofpreadmission spine immobilization is sparse. The article byHauswald et al. (50) was published in 1998 after a period duringwhich universal spine immobilization after trauma had beenapplied in the United States and North America. Ethical andpractical issues preclude a contemporary clinical trial designedto study the effectiveness of preadmission spine immobilizationcompared with no immobilization, primarily because spine im-mobilization for trauma patients is perceived as essential withminimal risk and is already widely used. Intuitively, the use ofpreadmission spine immobilization is a rational means of limit-ing spinal motion in spine-injured patients in an effort to reducethe likelihood of neurological deterioration caused by patholog-ical motion at the site(s) of injury.
The consensus from all articles reviewed (Class III evidence),from an anatomic and biomechanical perspective and from time-tested clinical experience with traumatic spinal injuries, is that allpatients with cervical spinal column injuries, or those with thepotential for a cervical spinal injury after trauma, should betreated with spinal column immobilization until injury has beenexcluded or definitive management has been initiated. Althoughthere is insufficient medical evidence to support a treatmentstandard or a treatment guideline, practitioners are stronglyencouraged to provide spine immobilization to spine-injuredpatients (or those with a likelihood of spinal injury) until defin-itive assessment can be accomplished.
Orledge and Pepe (76) in their commentary on theHauswald findings (50) point out some limitations of thearticle, but they also suggest that it raises the issue of a moreselective evidence-based protocol for spine immobilization.Should all trauma patients be managed with spine immobili-zation until spinal injury has been excluded, or should immo-bilization be selectively used for patients with potential spinalinjury on the basis of well-defined clinical criteria? Whichclinical criteria should be used? After the Hauswald report,several prospective studies supported the use of clinical find-ings as indicators of the need for preadmission spine immo-bilization after trauma (33–35). Several EMS systems now useclinical protocols to help decide which patients should bemanaged with spine immobilization after trauma (43, 102).
Domeier et al. (32–34), in a multicenter prospective study of6500 trauma patients, found that the application of clinicalcriteria (altered mental status, focal neurological deficit, evi-dence of intoxication, spinal pain or tenderness, or suspectedextremity fracture) was predictive of most patients with cer-vical spinal injuries that required immobilization. The predic-tive value of their criteria held true for patients with high- orlow-risk mechanisms of injury. They suggested that clinical
criteria, rather than the mechanism of injury, be evaluated asthe standard for the use of spine immobilization.
Brown et al. (13) examined whether EMS providers couldaccurately apply clinical criteria to evaluate the cervical spinesof trauma patients before transport to a definitive care facility.The criteria included the presence of pain or tenderness of thecervical spine, the presence of a neurological deficit, an al-tered level of consciousness, evidence of drug use or intoxi-cation (particularly alcohol, analgesics, sedatives, or stimu-lants), and/or the presence of other significant trauma thatmight act as a distracting injury. Immobilization of the cervi-cal spine was initiated if any one of six criteria was present.The clinical assessment of trauma patients by EMS providerswas compared with the clinical assessment provided by emer-gency physicians. The providers (EMS technicians and emer-gency physicians) were blinded to each other’s assessments.Agreement between EMS and physician providers was ana-lyzed by � statistic. Five hundred seventy-three patients wereincluded in the study. The assessments matched in 79% of thecases (n � 451). For 78 patients (13.6%), the EMS clinicalassessment indicated spine immobilization but the physi-cian assessment did not. For 44 patients (7.7%), the physi-cian’s clinical assessment indicated spine immobilization butthe EMS assessment did not. For the individual components,� ranged from 0.35 to 0.81. For the decision to immobilize, �was 0.48. The EMS clinical assessments were generally morein favor of immobilization than the physician’s clinical assess-ments. Brown et al. concluded that EMS and physician clinicalassessments to rule out cervical spinal injury after traumahave moderate to substantial agreement. The authors recom-mended, however, that systems that allow EMS personnel todecide whether to immobilize patients after trauma shouldprovide attentive follow-up of those patients to ensure appro-priate care and to provide immediate feedback to the EMSproviders. Meldon et al. (71), in an earlier study, found sig-nificant disagreement between the clinical assessments andsubsequent spine immobilization of patients by EMS techni-cians and physicians. They recommended further researchand education before widespread implementation of thispractice.
Clinical criteria to select appropriate patients for spine im-mobilization are being studied in Michigan (102) and havebeen implemented in Maine (43) and San Mateo County, CA(88). Recommendations regarding the adoption of EMS pro-tocols for preadmission spine immobilization await definitivestudies of safety and efficacy (23). EMS personnel who makethese assessments require intensive education and careful,quality-assurance scrutiny to ensure that trauma patients withpotential spinal injuries are appropriately triaged and man-aged. Until further studies can be undertaken, the availableClass III studies support the use of spine immobilization forall patients with potential cervical spinal injury after trauma.
Devices and techniques for preadmissionspine immobilization
Preadmission spine immobilization is effective in limitingspinal motion during transportation of the patient (7, 27, 40,
Preadmission Cervical Spine Immobilization S9
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE1.
1.Su
mm
ary
ofR
epor
tson
Prea
dmis
sion
Cer
vica
lSp
ine
Imm
obili
zati
ona
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Mar
kens
onet
al.,
1999
(61)
An
eval
uatio
nof
the
Ken
dric
kex
tric
atio
nde
vice
bfo
rpe
diat
ric
spin
alim
mob
iliza
tion.
IIIK
endr
ick
extr
icat
ion
devi
cepr
ovid
esex
celle
ntst
atic
and
dyna
mic
imm
obili
zatio
n.
Perr
yet
al.,
1999
(77)
An
expe
rim
enta
lev
alua
tion
of3
imm
obili
zatio
nde
vice
sco
mpa
red
duri
ngsi
mul
ated
vehi
cle
mot
ion.
Nec
km
otio
nw
asju
dged
by3
phys
icia
ns.
IIISu
bsta
ntia
lam
ount
sof
head
mot
ion
can
occu
rdu
ring
sim
ulat
edve
hicl
em
otio
n,
rega
rdle
ssof
the
met
hod
ofim
mob
iliza
tion.
Mov
emen
tof
the
trun
kca
nha
veth
esa
me
effe
ctas
head
mot
ion
onm
otio
nac
ross
the
neck
.
Bau
eran
dK
owal
ski,
1998
(9)
Ast
udy
ofth
eef
fect
ofsp
inal
imm
obili
zatio
nde
vice
son
pulm
onar
yfu
nctio
nin
15m
en.
IIISi
gnifi
cant
rest
rict
ion
ofpu
lmon
ary
func
tion
may
resu
ltfr
omsp
inal
imm
obili
zatio
n.
Maw
son
etal
.,19
98(6
3)A
pros
pect
ive
stud
yto
dete
rmin
eth
eas
soci
atio
nbe
twee
nim
mob
iliza
tion
and
pres
sure
ulce
rsin
39SC
Ipa
tient
s.
IIITi
me
spen
ton
back
boar
dis
sign
ifica
ntly
asso
ciat
edw
ithpr
essu
reul
cers
deve
lopi
ng
with
in8
d.
Hau
swal
det
al.,
1998
(50)
5-yr
retr
ospe
ctiv
ech
art
revi
ewof
patie
nts
with
acut
etr
aum
atic
SCI
from
2ce
nter
s.N
one
ofth
e12
0pa
tient
sat
the
Uni
vers
ityof
Mal
aya
had
spin
alim
mob
iliza
tion
with
orth
otic
devi
ces
duri
ngtr
ansp
ort.
All
334
patie
nts
atth
eU
nive
rsity
ofN
ewM
exic
odi
d.Th
e
hosp
itals
wer
eco
mpa
rabl
e.N
euro
logi
cal
inju
ries
wer
eas
sign
edto
2ca
tego
ries
,di
sabl
ing
orno
tdi
sabl
ing,
by2
blin
ded
phys
icia
ns.
Dat
aw
ere
anal
yzed
usin
gm
ultiv
aria
telo
gist
ic
regr
essi
on.
Ther
ew
asle
ssne
urol
ogic
aldi
sabi
lity
inth
eM
alay
sian
patie
nts
(OR
,2.
03;
95%
CI,
1.03
–3.9
9;P
�0.
04).
Res
ults
wer
esi
mila
rw
hen
the
anal
ysis
was
limite
dto
patie
nts
with
cerv
ical
inju
ries
(OR
,1.
52;
95%
CI,
0.64
–3.6
2;P
�0.
34).
IIIO
ut-o
f-ho
spita
lim
mob
iliza
tion
has
little
effe
cton
neur
olog
ical
outc
ome
inpa
tient
sw
ith
blun
tsp
inal
inju
ries
.
The
asso
ciat
ion
betw
een
spin
alco
lum
nm
ovem
ent
and
the
pote
ntia
lfo
rSC
Ire
mai
ns
uncl
ear.
Bla
yloc
k,19
96(1
1)A
pros
pect
ive
stud
yto
dete
rmin
eth
eas
soci
atio
nbe
twee
nim
mob
iliza
tion
and
pres
sure
ulce
rsin
32SC
Ipa
tient
s.
IIIPr
essu
reso
res
deve
lope
d,m
ostly
inpa
tient
sw
how
ere
turn
edaf
ter
3h.
Mos
tof
thos
e
with
out
sore
sw
ere
turn
ed�
2h
afte
rim
mob
iliza
tion.
John
son
etal
.,19
96(5
2)M
easu
red
imm
obili
zatio
nan
dco
mfo
rton
10-p
oint
scal
e.Th
eva
cuum
splin
tw
as
com
pare
dw
ithba
ckbo
ard.
IIIV
acuu
msp
lints
are
mor
eco
mfo
rtab
lean
dfa
ster
toap
ply
than
back
boar
dsan
dpr
ovid
e
asi
mila
rde
gree
ofim
mob
iliza
tion.
Vac
uum
splin
tsar
eno
tri
gid
enou
ghfo
rex
tric
atio
nan
dar
em
ore
expe
nsiv
e.
Rod
gers
and
Rod
gers
,19
95
(84)
Cas
ere
port
ofm
argi
nal
man
dibu
lar
nerv
epa
lsy
due
toco
mpr
essi
onby
ace
rvic
alha
rd
colla
r.
IIITh
eco
llar
was
rem
oved
;th
epa
lsy
reso
lved
unev
entfu
llydu
ring
the
next
2d.
Cha
net
al.,
1994
(19)
Apr
ospe
ctiv
est
udy
ofth
eef
fect
sof
spin
alim
mob
iliza
tion
onpa
inan
ddi
scom
fort
in21
volu
ntee
rsaf
ter
30m
in.
All
subj
ects
deve
lope
dpa
in.
IIISt
anda
rdsp
inal
imm
obili
zatio
nm
aybe
aca
use
ofpa
inin
anot
herw
ise
heal
thy
subj
ect.
Liew
and
Hill
,19
94(5
9)2
case
repo
rts
ofsi
gnifi
cant
occi
pita
lpr
essu
reul
cera
tion
asso
ciat
edw
ithth
eus
eof
hard
cerv
ical
colla
r.
IIIPr
essu
reul
cers
may
occu
rw
ithth
eus
eof
hard
cerv
ical
colla
rs.
Maz
olew
ski,
1994
(64)
Ast
udy
tote
stth
eef
fect
iven
ess
ofst
rapp
ing
tech
niqu
esin
redu
cing
late
ral
mot
ion
ona
back
boar
din
labo
rato
ryin
19ad
ults
.
IIISt
rapp
ing
shou
ldbe
adde
dto
the
tors
oto
redu
cela
tera
lm
otio
non
aba
ckbo
ard.
Plai
sier
etal
.,19
94(7
8)A
pros
pect
ive
eval
uatio
nof
cran
iofa
cial
pres
sure
offo
urdi
ffere
ntce
rvic
alor
thos
esin
20
adul
ts.
Pres
sure
was
mea
sure
dat
the
occi
put,
man
dibl
e,an
dch
in.
Opi
nion
son
com
fort
wer
eal
soco
llect
ed.
IIITh
eN
ewpo
rtor
Mia
mi-
Jco
llars
have
favo
rabl
esk
inpr
essu
repa
ttern
san
dsu
peri
or
patie
ntco
mfo
rt.
Rap
hael
and
Cho
tai,
1994
(82)
Ara
ndom
ized
,si
ngle
-blin
d,cr
osso
ver
stud
yof
9pa
tient
ssc
hedu
led
for
elec
tive
spin
al
anes
thes
ia.
The
cere
bros
pina
lflu
idpr
essu
rein
the
lum
bar
suba
rach
noid
spac
ew
as
mea
sure
dw
ithan
dw
ithou
ta
Stifn
eck
cerv
ical
colla
rap
plie
d.
IIITh
ere
was
asi
gnifi
cant
elev
atio
nof
cere
bros
pina
lflu
idpr
essu
rein
7of
the
patie
nts
stud
ied
whe
nth
ece
rvic
alco
llar
was
appl
ied
(P�
0.01
).
Cha
ndle
ret
al.,
1992
(20)
Aco
mpa
riso
nof
the
rigi
dce
rvic
alex
tric
atio
nco
llar
with
Am
mer
man
halo
orth
osis
in20
men
.
IIIA
mm
erm
anha
loor
thos
isan
dsp
ine
boar
dpr
ovid
edsi
gnifi
cant
lybe
tter
imm
obili
zatio
n,
equi
vale
ntto
halo
vest
.
Ros
en,
1992
(87)
Aco
mpa
riso
nof
4ce
rvic
alco
llars
in15
adul
tvo
lunt
eers
,by
goni
omet
ry.
IIIV
acuu
msp
lint
cerv
ical
colla
rre
stri
cted
rang
eof
mot
ion
ofth
ece
rvic
alsp
ine
mos
t
effe
ctiv
ely.
S10 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE1.
1.C
onti
nued
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Scha
ferm
eyer
etal
.,19
91
(89)
Ast
udy
toas
sess
the
rest
rict
ive
effe
cts
of2
spin
alim
mob
iliza
tion
stra
ppin
g
tech
niqu
eson
the
resp
irat
ory
capa
city
and
forc
edvi
tal
capa
city
of51
child
ren.
IIISp
inal
imm
obili
zatio
nsi
gnifi
cant
lyre
duce
dre
spir
ator
yca
paci
tyas
mea
sure
dby
FVC
in
heal
thy
patie
nts
6–15
yrol
d.Th
ere
isno
sign
ifica
ntbe
nefit
ofon
est
rapp
ing
tech
niqu
e
over
the
othe
r.
Schr
iger
etal
.,19
91(9
1)A
stud
yco
mpa
ring
the
flat
back
boar
dw
ithoc
cipi
tal
padd
ing
inac
hiev
ing
neut
ral
posi
tion
in10
0he
alth
yvo
lunt
eers
.
IIIO
ccip
ital
padd
ing
plac
esth
ece
rvic
alsp
ine
inm
ore
neut
ral
alig
nmen
t.
Coh
enet
al.,
1990
(22)
Ast
udy
anal
yzin
gth
eR
EDin
64pa
tient
s.III
RED
isan
effe
ctiv
esp
inal
imm
obili
zatio
nde
vice
with
adva
ntag
esov
ercu
rren
tly
avai
labl
ede
vice
s.
Bar
ney
and
Cor
dell,
1989
(8)
Eval
uate
dpa
inan
ddi
scom
fort
duri
ngim
mob
iliza
tion
onri
gid
spin
ebo
ards
in90
patie
nts.
IIISp
ine
boar
dsm
ayca
use
disc
omfo
rt.
Tosc
ano,
1988
(96)
Prev
entio
nof
neur
olog
ical
dete
rior
atio
nbe
fore
adm
issi
onto
hosp
ital.
Ret
rosp
ectiv
e
revi
ewof
123
patie
nts;
32of
123
sust
aine
dm
ajor
neur
olog
ical
dete
rior
atio
nfr
om
inju
ryto
adm
issi
on.
IIIA
ppro
pria
teha
ndlin
gof
patie
nts
with
spin
alin
jury
afte
rtr
aum
aca
nre
duce
maj
or
neur
olog
ical
dete
rior
atio
ndu
eto
path
olog
ical
mot
ion
ofve
rteb
ral
colu
mn.
Gra
zian
oet
al.,
1987
(44)
Ara
diog
raph
icco
mpa
riso
nof
prea
dmis
sion
cerv
ical
imm
obili
zatio
nm
etho
dsw
ith
the
shor
tbo
ard
tech
niqu
ein
45vo
lunt
eers
.
IIITh
esh
ort-
boar
dte
chni
que
prov
edto
besi
gnifi
cant
lybe
tter
(P�
0.05
)th
anK
endr
ick
extr
icat
ion
devi
cean
dco
llars
.
Lina
res
etal
.,19
87(6
0)A
stud
yof
32SC
Ipa
tient
sto
dete
rmin
ew
heth
erpr
essu
reso
res
are
asso
ciat
edw
ith
prol
onge
dim
mob
iliza
tion.
IIITh
ere
isa
stro
ngas
soci
atio
nbe
twee
n1–
2h
ofim
mob
iliza
tion
and
the
deve
lopm
ent
of
pres
sure
sore
s.
McG
uire
etal
.,19
87(6
8)A
radi
ogra
phic
eval
uatio
nof
mot
ion
ofth
eth
orac
olum
bar
spin
ein
aca
dave
rw
ithan
unst
able
thor
acol
umba
rsp
ine,
and
apa
tient
with
aT1
2–L1
frac
ture
disl
ocat
ion.
IIIEx
trem
em
otio
nat
anun
stab
leth
orac
olum
bar
spin
ese
gmen
tca
noc
cur
duri
ngth
e
logr
oll
man
euve
r.Th
eba
ckbo
ard
and
the
Scoo
pst
retc
her
offe
red
adeq
uate
stab
iliza
tion
for
thor
acol
umba
rsp
ine
inst
abili
ty.
McC
abe
and
Nol
an,
1986
(65)
Ara
diog
raph
icco
mpa
riso
nof
the
4ce
rvic
alco
llars
in7
adul
ts.
IIIPo
lyet
hyle
ne-1
prov
ided
mos
tre
stri
ctio
nin
flexi
on.
Clin
eet
al.,
1985
(21)
Ara
diog
raph
icco
mpa
riso
nof
7m
etho
dsof
cerv
ical
imm
obili
zatio
nin
97ad
ults
.III
The
shor
t-bo
ard
tech
niqu
eap
pear
edto
besu
peri
orto
the
3co
llars
stud
ied.
The
colla
rs
prov
ided
noau
gmen
tatio
nof
imm
obili
zatio
nov
erth
atpr
ovid
edby
the
shor
tbo
ard
alon
e.
Podo
lsky
etal
.,19
83(7
9)St
atic
tria
lus
ing
goni
omet
ryco
mpa
ring
soft
colla
r,ha
rdco
llar,
extr
icat
ion
colla
r,
Phila
delp
hia
colla
r,bi
late
ral
sand
bags
and
tape
,an
dth
eco
mbi
natio
nof
sand
bags
,
tape
,an
dth
ePh
ilade
lphi
aco
llar
in25
norm
alad
ult
volu
ntee
rs.
IIIH
ard
foam
and
plas
ticco
llars
wer
esu
peri
orto
soft
colla
rs.
Sand
bags
and
tape
in
com
bina
tion
with
ari
gid
cerv
ical
colla
rw
asth
ebe
stm
eans
ofth
ose
eval
uate
dto
limit
cerv
ical
spin
em
otio
n.Th
ead
ditio
nof
aPh
ilade
lphi
aco
llar
was
sign
ifica
ntly
mor
e
effe
ctiv
ein
redu
cing
neck
exte
nsio
n(P
�0.
01),
from
15to
7.4
degr
ees,
ach
ange
of
49.3
%.
The
com
bina
tion
ofsa
ndba
gsan
dta
peal
one
does
not
allo
wsu
ffici
ent
rest
rict
ion
ofm
otio
n,pa
rtic
ular
lyin
exte
nsio
n.
aSC
I,sp
inal
cord
inju
ry;
FVC
,fo
rced
vita
lca
paci
ty;
RED
,R
usse
llex
tric
atio
nde
vice
;O
R,
odds
rati
o;C
I,co
nfid
ence
inte
rval
.b
Ken
dric
kex
tric
atio
nde
vice
,Fer
no-W
ashi
ngto
n,In
c.,W
ilmin
gton
,OH
;New
port
colla
r(n
owkn
own
asA
spen
colla
r),F
ijiM
anuf
actu
ring
,Inc
.,Lo
ngB
each
,CA
;Sco
opst
retc
her,
Fern
o-W
ashi
ngto
n,W
ilmin
gton
,OH
;Mia
mi-
J,Je
rom
eM
edic
al,M
oore
stow
n,N
J;St
ifnec
kri
gid
colla
r,A
llian
ceM
edic
al,R
usse
lvill
e,M
O;P
olye
thyl
ene-
1,A
llian
ceM
edic
al,R
usse
lvill
e,M
O;A
mm
erm
anha
loor
thos
is,A
mm
erm
anTr
aum
aSy
stem
s,Pa
cific
Palis
ades
,CA
;Rus
sell
extr
icat
ion
devi
ce,M
illa
Mit
chel
l&C
o.,N
ewSo
uth
Wal
es,A
ustr
alia
;Phi
lade
lphi
aco
llar,
Phila
delp
hia
Col
lar
Co.
,W
estv
ille,
NJ.
Preadmission Cervical Spine Immobilization S11
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
66, 73, 104). Various devices and techniques are available toprovide immobilization of the cervical spine. Attempts todefine the best method have been hampered by physical andethical constraints (17, 27, 53, 61, 64, 94, 99).
Ways of measuring the efficacy of spine immobilizationdevices vary among investigators. Comparative studies of thevarious devices have been performed on healthy volunteers,but none have been tested in a large number of patients withspinal injury. It is difficult to extrapolate normative data toinjured patients with spinal instability (17, 20, 24, 27, 29, 49,52, 53, 58, 65, 67, 77, 94, 98, 99).
Several methods have been used to measure movement ofthe cervical spine. They include clinical assessment, plumblines, photography, radiography, cinematography, computedtomography, and magnetic resonance imaging. Roozmon etal. (85) summarized the problems inherent in each methodand concluded that there was no satisfactory noninvasivemeans of studying neck motion, particularly if one is to quan-tify movement between individual vertebral segments.
The position in which the injured spine should be placedand held immobile, the “neutral position,” is poorly defined(25, 28, 75, 88, 92). Schriger defined the neutral position as thenormal anatomic position of the head and torso that oneassumes when standing and looking ahead (90). This positioncorrelates to 12 degrees of cervical spine extension on a lateralradiograph. Schriger comments that the extant radiographicdefinition of neutral position was based on radiographicstudy of patients who were visually observed to be in theneutral position. Schriger et al. (91) used this position in theirevaluation of occipital padding on spine immobilization back-boards. De Lorenzo et al. (28), in their magnetic resonanceimaging study of 19 adults, found that a slight degree offlexion equivalent to 2 cm of occiput elevation produces afavorable increase in spinal canal/spinal cord ratio at levelsC5 and C6, a region of frequent unstable cervical spine inju-ries. Backboards have been used for years in extricating andimmobilizing spine-injured patients. Schriger et al. (91) ques-tioned the ability of a flat board to allow neutral positioning ofthe cervical spine. They compared spine immobilization byusing the flat backboard with and without occipital paddingin 100 adults. Clinical observation and assessment were usedto determine the neutral position of the cervical spine. Theauthors found that occipital padding combined with a rigidbackboard places the cervical spine in a better neutral positionthan a flat backboard alone (91, 93). McSwain (70) determinedthat more than 80% of adults require 1.3 to 5.1 cm of paddingto achieve neutral positioning of the head and neck relative tothe torso and noted that physical characteristics and musculardevelopment alter the cervical-thoracic angle, thus affectingpositioning. This makes it impossible to dictate specific rec-ommendations for padding.
In general, spine immobilization consists of a cervical collar,supports on either side of the head, and the long and shortbackboards with associated straps to attach and immobilizethe entire body to the board (27). Garth (41) proposed perfor-mance standards for cervical extrication collars, but thesestandards have not been uniformly implemented. A variety ofdifferent cervical collars is available. Several studies compare
collars alone or combined with other immobilization devicesby a wide range of assessment criteria (17, 19, 20, 24, 94, 99).
Podolsky et al. (79), in 1983, evaluated the efficacy of cer-vical spine immobilization techniques by using goniometricmeasurements. Twenty-five healthy volunteers lying supineon a rigid emergency department resuscitation table wereasked to actively move their necks as far as possible in sixways: flexion, extension, rotation to the right and left, andlateral bending to the right and left. Control measurementswere made with no device, and then measurements wererepeated after immobilization in a soft collar, hard collar,extrication collar, Philadelphia collar (Philadelphia Collar Co.,Westville, NJ), bilateral sandbags joined with 3-inch-widecloth tape across the forehead attached to either side of theresuscitation table, and the combination of sandbags, tape,and a Philadelphia collar. Hard foam and hard plastic collarswere better at limiting cervical spine motion than soft foamcollars. Neither collars alone nor sandbags and tape in com-bination provided satisfactory restriction of cervical spinemotion. For all six cervical spine movements, sandbags andtape immobilization were significantly better than any of theother methods of attempted cervical spine immobilizationused alone. The authors found that sandbags and tape com-bined with a rigid cervical collar were the best means of thoseevaluated to limit cervical spine motion. Adding a Philadel-phia collar to the sandbag and tape construct significantlyreduced neck extension (P � 0.01), from 15 degrees to 7.4degrees, a change of 49.3%. Collar use had no significantadditive effect for any other motion of the cervical spine.Sandbags as adjuncts to cervical spine immobilization requiremore rather than less attention from care providers (54). Sand-bags are heavy, and, if the extrication board is tipped side toside during evacuation and transport, the sandbags can slide,resulting in lateral displacement of the patient’s head andneck with respect to the torso. Sandbags can be taped to theextrication board, but because they are small compared withthe patient, this can be difficult and/or ineffective. Finally,sandbags must be removed before initial lateral cervical spinex-ray assessment because they can obscure the radiographicbony anatomy of the cervical spine. For these reasons (54) andthe findings by Podolsky et al. (79), use of sandbags and tapealone to attempt to immobilize the cervical spine is notrecommended.
In 1985, Cline et al. (21) compared methods of cervical spineimmobilization used in preadmission transport. The authorsfound that strapping the patient to a standard short board wasmore effective than cervical collar use alone. They noted nosignificant differences among the rigid collars they tested.McCabe and Nolan (65) used radiographic assessment tocompare four different collars for their ability to restrict mo-tion in flexion-extension and lateral bending. They found thatthe Polyethylene-1 collar (Alliance Medical, Russelville, MO)provided the most restriction of motion of the cervical spine,particularly for flexion. Rosen (87), in 1992, used goniometricmeasurements to compare limitation of cervical spine move-ment of four rigid cervical collars on 15 adults. Of the fourdevices they tested, the vacuum splint cervical collar pro-
S12 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
vided the most effective restriction of motion of the cervicalspine.
Graziano et al. (44) compared preadmission cervical spineimmobilization methods by measuring cervical motion radio-graphically in the coronal and sagittal planes in 45 immobi-lized adults. In this study, the Kendrick extrication device(Ferno-Washington, Inc., Wilmington, OH) and the Extrica-tion Plus-One device (Medical Specialties, Inc., Charlotte, NC)were nearly as effective in limiting cervical motion as theshort immobilization board. Both devices were more effectivethan a rigid cervical collar alone.
Cohen et al. (22), in 1990, described the Russell extricationdevice (RED) (Milla Mitchell & Co., New South Wales, Aus-tralia) for immobilization of patients with potential spineinjuries. The RED was comparable to the short immobilizationboard for preadmission spine immobilization. Chandler et al.(20) compared a rigid cervical extrication collar with the Am-merman halo orthosis (Ammerman Trauma Systems, PacificPalisades, CA) in 20 men. The Ammerman halo orthosis com-bined with a rigid spine board provided significantly bettercervical spine immobilization than a cervical collar and spineboard. The Ammerman halo orthosis and spine board wasequivalent to the standard halo vest immobilization device.
Perry et al. (77) evaluated three cervical spine immobiliza-tion devices during simulated vehicle motion in six adults.Neck motion was assessed by three neurologists and neuro-surgeons as to whether motion was “clinically significant.”The authors found that substantial head motion occurredduring simulated vehicle motion regardless of the method ofimmobilization. The authors observed that the efficacy ofcervical spine immobilization was limited unless the motionof the head and the trunk was also effectively controlled.Mazolewski (64) tested the effectiveness of strapping tech-niques to reduce lateral motion of the spine of adults re-strained on a backboard. Subjects were restrained on awooden backboard with four different strapping techniques.The backboard was rolled to the side, and lateral motion ofthe torso was measured. The author found that additionalstrapping securing the torso to backboard reduced lateralmotion of the torso.
Finally, the traditional method of moving a patient onto along backboard has typically involved the logroll maneuver.The effectiveness of this transfer technique has been ques-tioned (31, 87). Significant lateral motion of the lumbar spinehas been reported (68, 95). Alternatives to the logroll maneu-ver include the HAINES method and the multihand or fire-man lift method (4, 5, 47). In the HAINES method (acronymfor High Arm IN Endangered Spine), the patient is placedsupine, the upper arm away from the kneeling rescuer isabducted to 180 degrees, the near arm of the patient is placedacross the patient’s chest, and both lower limbs are flexed. Therescuer’s hands stabilize the head and neck and the patient isrolled away onto an extrication board or device (47). Themultihand or fireman lift method involves several rescuers oneither side of the patient; the rescuers slide their arms under-neath the patient and lift the patient from one position toanother onto an extrication board or device.
This review depicts the evolution of techniques availablefor providing preadmission spine immobilization of spine-injured patients during transport and underscores their diver-sity. These studies are limited by the fact that none of thestudies evaluates the full range of available devices usingsimilar criteria. Overall, it seems that a combination of rigidcervical collar immobilization with supportive blocks on arigid backboard with straps to secure the entire body of thepatient is most effective in limiting motion of the cervicalspine after traumatic injury (5). The long-standing practice ofattempted spine immobilization using sandbags and tapealone is insufficient.
Safety of preadmission spine immobilization devices
Despite obvious benefits, cervical spine immobilization hasa few potential drawbacks. Immobilization can be uncomfort-able, it takes time to apply, application may delay transport,and it is associated with modest morbidity (4, 9, 18, 19, 26, 90,100).
Chan et al. (19) studied the effects of spine immobilizationon pain and discomfort in 21 healthy adults. Subjects wereplaced in backboard immobilization for 30 minutes, andsymptoms were chronicled. All subjects developed pain,which was described as moderate to severe in 55% of volun-teers. Occipital headache and sacral, lumbar, and mandibularpain were the most frequent complaints. In a later study,Chan et al. (18) compared spine immobilization on a back-board to immobilization with a vacuum mattress-splint de-vice in 37 healthy adults. The authors found that the fre-quency and severity of occipital and lumbosacral pain wassignificantly higher during backboard immobilization than onthe vacuum mattress-splint device. Johnson et al. (52) per-formed a prospective, comparative study of the vacuum splintdevice versus the rigid backboard. The vacuum splint devicewas significantly more comfortable than the rigid backboardand could be applied more quickly. The vacuum splint deviceprovided better immobilization of the torso. The rigid back-board with head blocks was slightly better at immobilizingthe head. Vacuum splint devices, however, are not recom-mended for extrication because they are reportedly not rigidenough, and they are more expensive. At a cost of approxi-mately $400, the vacuum splint device is roughly three timesmore expensive than a rigid backboard (18).
Hamilton and Pons (49) studied the comfort level of 26adults on a full-body vacuum splint device compared with arigid backboard, with and without cervical collars. Subjectsgraded their immobilization and discomfort. No statisticallysignificant difference was found between the vacuum splintdevice and collar combination and the backboard and collarcombination for flexion and rotation. The vacuum splint-collar combination provided significantly better immobiliza-tion in extension and lateral bending than the backboard-collar combination. The vacuum splint alone provided bettercervical spine immobilization in all neck positions exceptextension than the rigid backboard alone. A statistically sig-nificant difference in subjective perception of immobilizationwas noted; the backboard alone was less effective than the
Preadmission Cervical Spine Immobilization S13
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
three alternatives. In conclusion, the vacuum splint device,particularly when used with a cervical collar, is an effectiveand comfortable alternative to a rigid backboard (with orwithout the collar) for cervical spine immobilization.
Barney and Cordell (8) evaluated pain and discomfort dur-ing immobilization on rigid spine boards in 90 trauma pa-tients and found that rigid spine boards cause discomfort.Padding the rigid board improves patient comfort withoutcompromising cervical spine immobilization (101). Minimiz-ing the pain of immobilization may decrease voluntary move-ment and therefore decrease the likelihood of secondary in-jury (19).
Cervical collars have been associated with elevated intra-cranial pressure (ICP). Davies et al. (26) prospectively ana-lyzed ICP in a series of injured patients managed with theStifneck rigid collar (Alliance Medical). ICP rose significantly(P � 0.001; mean, 4.5 mm Hg) when the collar was firmly inplace. The authors cautioned that because head-injured pa-tients may also require cervical spine immobilization, it isessential that secondary insults producing raised ICP be min-imized. Kolb et al. (55) also examined changes in ICP after theapplication of a rigid Philadelphia collar in 20 adult patients.ICP averaged 176.8 mm H2O initially and increased to anaverage of 201.5 mm H2O after collar placement. Althoughthe difference in ICP of 24.7 mm H2O was statistically signif-icant (P � 0.001), it remains uncertain that it has clinicalrelevance. Nonetheless, this modest increase in pressure maybe important in patients who already have elevated ICP.Plaisier et al. (78), in 1994, prospectively evaluated craniofa-cial pressure with the use of four different cervical orthoses.The authors found small changes in craniofacial pressure(increases) but no significant differences among the four collartypes.
Spine immobilization increases the risk of pressure sores.Linares et al. (60) found that pressure sores were associatedwith immobilization (patients who were not turned duringthe first 2 hours after injury). The development of pressuresores was not related to mode of transportation to hospital orto the use of a spinal board and sandbags during transporta-tion. Mawson et al. (63) prospectively assessed the develop-ment of pressure ulcers in 39 spinal cord-injured patients whowere immobilized immediately after injury. The length oftime on a rigid spine board was significantly associated withthe development of decubitus ulcers within 8 days of injury (P� 0.01). Rodgers and Rodgers (84) reported a marginal man-dibular nerve palsy caused by compression by a hard collar.The palsy resolved uneventfully during the next 2 days. Blay-lock (11) found that prolonged cervical spine immobilizationmay result in pressure ulcers. Improved skin care (keepingthe skin dry), proper fitting (avoid excessive tissue pressure),and the appropriate choice of collars (those that do not trapmoisture and do not exert significant tissue pressure) canreduce this risk (10, 11).
Cervical spine immobilization may also increase the risk ofaspiration and may limit respiratory function. Bauer andKowalski (9) examined the effect of the Zee Extrication Device(Zee Medical Products, Irvine, CA) and the long spinal boardon pulmonary function. They tested pulmonary function in 15
healthy, nonsmoking men by using forced vital capacity,forced expiratory volume in 1 second, the ratio of forcedexpiratory volume in 1 second to the forced vital capacity, andforced midexpiratory flow (25–75%). They found a significantdifference (P � 0.05) between before-strapping and after-strapping values for three of the four functions tested whenon the long spinal board. Similarly, significant differenceswere found for three of the four parameters when using theZee Extrication Device. These differences reflect a markedpulmonary restrictive effect of appropriately applied entire-body spine immobilization devices.
Totten and Sugarman (97) evaluated the effect of whole-body spine immobilization on respiration in 39 adults. Respi-ratory function was measured at baseline, once immobilizedwith a Philadelphia collar on a rigid backboard, and whenimmobilized on a Scandinavian vacuum mattress with a vac-uum collar. The comfort levels of each of the two methodswere assessed on a visual analog scale. Both immobilizationmethods restricted respiration by an average of 15%. Theeffects were similar under the two methods, although theforced expiratory volume in 1 second was lower on the vac-uum mattress. The vacuum mattress was significantly morecomfortable than the wooden backboard (4).
In conclusion, cervical spine immobilization devices aregenerally effective in limiting motion of the cervical spine butmay be associated with important but usually modest mor-bidity. Cervical spine immobilization devices should be usedto achieve the goals of safe extrication and transport butshould be removed as soon as it is safe to do so.
SUMMARY
Spine immobilization can reduce untoward movement ofthe cervical spine and can reduce the likelihood of neurolog-ical deterioration in patients with unstable cervical spine in-juries after trauma. Immobilization of the entire spinal col-umn is necessary in these patients until a spinal column injury(or multiple injuries) or a spinal cord injury has been ex-cluded, or until appropriate treatment has been initiated.Although not supported by Class I or Class II medical evi-dence, this effective, time-tested practice is based on anatomicand mechanical considerations in an attempt to prevent spinalcord injury and is supported by years of cumulative traumaand triage clinical experience.
It is unclear whether the spines of all patients with traumamust be immobilized during preadmission transport. Manypatients do not have spinal injuries and therefore do notrequire such intervention. The development of specific selec-tion criteria for those patients for whom immobilization isindicated remains an area of investigation.
The variety of techniques used and the lack of definitiveevidence to advocate a uniform device for spine immobiliza-tion make it difficult to formulate recommendations for im-mobilization techniques and devices. It seems that a combi-nation of rigid cervical collar with supportive blocks on arigid backboard with straps is effective at achieving safe,effective spine immobilization for transport. The long-
S14 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
standing practice of attempting to immobilize the cervicalspine with sandbags and tape alone is not recommended.
Cervical spine immobilization devices are effective but canresult in patient morbidity. Spine immobilization devicesshould be used to achieve the goals of spinal stability for safeextrication and transport. They should be removed as soon asdefinitive evaluation is accomplished and/or definitive man-agement is initiated.
KEY ISSUES FOR FUTURE INVESTIGATION
The optimal device for immobilization of the cervical spineafter traumatic vertebral injury should be studied in a pro-spective fashion. A reliable in-field triage protocol to be ap-plied by EMS personnel for patients with potential cervicalspine injuries after trauma needs to be developed.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Alexander RH, Proctor HJ: Advanced Trauma Life Support Coursefor Physicians: ATLS. Chicago, American College of Surgeons,1993, ed 5, pp 21–22.
2. American Academy of Orthopedic Surgeons, Committee on In-juries: Emergency Care and Transportation of the Sick and Injured.Chicago, American Academy of Orthopedic Surgeons, 1971, pp111–115.
3. American College of Surgeons, Committee on Trauma: AdvancedTrauma Life Support. Chicago, American College of Surgeons,1993, p 201.
4. American College of Surgeons, Committee on Trauma: AdvancedTrauma Life Support. Chicago, American College of Surgeons,1993, pp 214–218.
5. American College of Surgeons, Committee on Trauma: Spineand spinal cord trauma, in Advanced Trauma Life Support Programfor Doctors: ATLS. Chicago, American College of Surgeons, 1997,ed 6, pp 215–242.
6. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support: Advanced Pre-hospital Care. Englewood Cliffs,Prentice-Hall, 1998, ed 2, p 120.
7. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support for Paramedics and Advanced EMS providers. UpperSaddle River, Brady, 1998, ed 3, p 153.
8. Barney R, Cordell W: Pain associated with immobilization onrigid spine boards. Ann Emerg Med 18:918, 1989 (abstr).
9. Bauer D, Kowalski R: Effect of spinal immobilization devices onpulmonary function in the healthy, non-smoking man. AnnEmerg Med 17:915–918, 1988.
10. Black CA, Buderer NM, Blaylock, B, Hogan BJ: Comparativestudy of risk factors for skin breakdown with cervical orthoticdevices. J Trauma Nurs 5:62–66, 1998.
11. Blaylock B: Solving the problem of pressure ulcers resultingfrom cervical collars. Ostomy Wound Manage 42:26–33, 1996.
12. Bohlman HH: Acute fractures and dislocations of the cervicalspine: An analysis of three hundred hospitalized patients andreview of the literature. J Bone Joint Surg Am 61A:1119–1142,1979.
13. Brown LH, Gough JE, Simonds WB: Can EMS providers ade-quately assess trauma patients for cervical spinal injury?Prehosp Emerg Care 2:33–36, 1998.
14. Brunette DD, Rockswold GL: Neurologic recovery followingrapid spinal realignment for complete cervical spinal cord in-jury. J Trauma 27:445–447, 1987.
15. Burney RE, Waggoner R, Maynard FM: Stabilization of spinalinjury for early transfer. J Trauma 29:1497–1499, 1989.
16. Butman A, Vomacka R: Part 1: Spine immobilization. Emergency23:48–51, 1991.
17. Carter VM, Fasen JA, Roman JM Jr, Hayes KW, Petersen CM:The effect of a soft collar, used as normally recommended orreversed, on three planes of cervical range of motion. J OrthopSports Phys Ther 23:209–215, 1996.
18. Chan D, Goldberg RM, Mason J, Chan L: Backboard versusmattress splint immobilization: A comparison of symptoms gen-erated. J Emerg Med 14:293–298, 1996.
19. Chan D, Goldberg R, Tascone A, Harmon S, Chan L: The effectof spinal immobilization on healthy volunteers. Ann EmergMed 23:48–51, 1994.
20. Chandler DR, Nemejc C, Adkins RH, Waters RL: Emergencycervical-spine immobilization. Ann Emerg Med 21:1185–1188,1992.
21. Cline JR, Scheidel E, Bigsby EF: A comparison of methods ofcervical immobilization used in patient extrication and trans-port. J Trauma 25:649–653, 1985.
22. Cohen A, Bosshard R, Yeo JD: A new device for the care of acutespinal injuries: The Russell Extrication Device (RED). Paraplegia28:151–157, 1990.
23. Cone DC, Wydro GC, Mininger CM: Current practice in clinicalcervical spinal clearance: Implication for EMS. Prehosp EmergCare 3:42–46, 1999.
24. Cooke M: Spinal boards. J Accid Emerg Med 13:433, 1996 (let-ter).
25. Curran C, Dietrich AM, Bowman MJ, Ginn-Pease ME, King DR,Kosnik E: Pediatric cervical-spine immobilization: Achievingneutral position? J Trauma 39:729–732, 1995.
26. Davies G, Deakin C, Wilson A: The effect of a rigid collar onintracranial pressure. Injury 27:647–649, 1996.
27. De Lorenzo RA: A review of spinal immobilization techniques.J Emerg Med 14:603–613, 1996.
28. De Lorenzo RA, Olson JE, Boska M, Johnston R, Hamilton GC,Augustine J, Barton R: Optimal positioning for cervical immo-bilization. Ann Emerg Med 28:301–308, 1996.
29. Dick T: Comparing the short-board technique. Ann Emerg Med18:115–116, 1989 (letter).
30. Dick T, Land R: Spinal immobilization devices: Part 1—Cervicalextrication collars. J Emerg Med Serv JEMS 7:26–32, 1982.
31. Dick T, Land R: Full spinal immobilizers. J Emerg Med ServJEMS 8:34–36, 1983.
32. Domeier RM: Indications for pre-hospital spinal immobilization:National Association of EMS Physicians Standards and ClinicalPractice Committee. Prehosp Emerg Care 3:251–253, 1999.
33. Domeier RM, Evans RW, Swor RA, Hancock JB, Fales W,Krohmer J, Fredericksen SM, Shork MA: The reliability of pre-hospital clinical evaluation for potential spinal injury is notaffected by the mechanism of injury. Prehosp Emerg Care 3:332–337, 1999.
34. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Fredericksen SM: Prehospital clinical findings associated withspinal injury. Prehosp Emerg Care 1:11–15, 1997.
Preadmission Cervical Spine Immobilization S15
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
35. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Fredericksen SM: Prospective validation of out-of-hospital spi-nal clearance criteria: A preliminary report. Acad Emerg Med4:643–646, 1997 (letter).
36. Farrington JD: Death in a ditch. Bull Am Coll Surg 52:121–130,1967.
37. Farrington JD: Extrication of victims: Surgical principles.J Trauma 8:493–512, 1968.
38. Fenstermaker RA: Acute neurologic management of the patientwith spinal cord injury. Urol Clin North Am 20:413–421, 1993.
39. Frohna WJ: Emergency department evaluation and treatment ofthe neck and cervical spine injuries. Emerg Med Clin North Am17:739–791, 1999.
40. Garfin SR, Shackford SR, Marshall LF, Drummond JC: Care ofthe multiply injured patient with cervical spine injury. ClinOrthop 239:19–29, 1989.
41. Garth G: Proposal for the establishment of minimum perfor-mance specifications for cervical extrication collars. Presented atthe 14th Annual Meeting of the American Society for Testing andMaterials, Skeletal Support Committee, West Conshohocken,PA, 1988.
42. Geisler WO, Wynne-Jones M, Jousse AT: Early management ofthe patient with trauma to the spinal cord. Med Serv J Can22:512–523, 1966.
43. Goth P: Spinal Injury: Clinical Criteria for Assessment and Manage-ment. Augusta, Medical Care Development Publishing, 1994.
44. Graziano AF, Scheidel EA, Cline JR, Baer LJ: A radiographiccomparison of pre-hospital cervical immobilization methods.Ann Emerg Med 16:1127–1131, 1987.
45. Green BA, Eismont FJ, O’Heir JT: Spinal cord injury: A systemsapproach—Prevention, emergency medical services and emer-gency room management. Crit Care Clin 3:471–493, 1987.
46. Gunby P: New focus on spinal cord injury. JAMA 245:1201–1206, 1981.
47. Gunn DB, Eizenberg N, Silberstein M, McMeeken JM, Tully EA,Stillman BC, Brown DJ, Gutteridge GA: How should an uncon-scious person with a suspected neck injury be positioned?Prehospital Disaster Med 10:239–244, 1995.
48. Hachen HJ: Emergency transportation in the event of acutespinal cord lesion. Paraplegia 12:33–37, 1974.
49. Hamilton RS, Pons PT: The efficacy and comfort of full-bodyvacuum splints for cervical-spine immobilization. J Emerg Med14:553–559, 1996.
50. Hauswald M, Ong G, Tandberg D, Omar Z: Out-of-hospitalspinal immobilization: Its effect on neurologic injury. AcadEmerg Med 5:214–219, 1998.
51. Jeanneret B, Magerl F, Ward JC: Over distraction: A hazard ofskull traction in the management of acute injuries of the cervicalspine. Arch Orthop Trauma Surg 110:242–245, 1991.
52. Johnson DR, Hauswald M, Stockhoff C: Comparison of a vac-uum splint device to a rigid backboard for spinal immobiliza-tion. Am J Emerg Med 14:369–372, 1996.
53. Jones SL: Spine trauma board. Phys Ther 57:921–922, 1977.54. Kilburn MP, Smith DP, Hadley MN: The initial evaluation and
treatment of the patient with spinal trauma, in Batjer HH, LoftusCM (eds): Textbook of Neurological Surgery: Principles and Practice.Philadelphia, Lippincott Williams & Wilkins (in press).
55. Kolb JC, Summers RL, Galli RL: Cervical collar-induced changesin intracranial pressure. Am J Emerg Med 17:135–137, 1999.
56. Kossuth LC: Removal of injured personnel from wrecked vehi-cles. J Trauma 5:704–705, 1965.
57. Kossuth LC: The initial movement of the injured. Mil Med132:18–21, 1967.
58. Lerner EB, Billittier AJ IV, Moscati RM: The effects of neutralpositioning with and without padding on spinal immobilizationof healthy subjects. Prehosp Emerg Care 2:112–116, 1998.
59. Liew SC, Hill DA: Complication of hard cervical collars in multi-trauma patients. Aust N Z J Surg 64:139–140, 1994.
60. Linares HA, Mawson AR, Suarez E, Biundo JJ: Association be-tween pressure sores and immobilization in the immediate post-injury period. Orthopedics 10:571–573, 1987.
61. Markenson D, Foltin G, Tunik M, Cooper A, Giordano L, FittonA, Lanotte T: The Kendrick extrication device used for pediatricspinal immobilization. Prehosp Emerg Care 3:66–69, 1999.
62. Marshall LF, Knowlton S, Garfin SR, Klauber MR, EisenbergHM, Kopaniky D, Miner ME, Tabbador K, Clifton GL: Deterio-ration following spinal cord injury: A multi-center study.J Neurosurg 66:400–404, 1987.
63. Mawson AR, Biundo JJ Jr, Neville P, Linares HA, Winchester Y,Lopez A: Risk factors for early occurring pressure ulcers follow-ing spinal cord injury. Am J Phys Med Rehabil 67:123–127, 1988.
64. Mazolewski P, Manix TH: The effectiveness of strapping tech-niques in spinal immobilization. Ann Emerg Med 23:1290–1295,1994.
65. McCabe JB, Nolan DJ: Comparison of the effectiveness of differ-ent cervical immobilization collars. Ann Emerg Med 15:50–53,1986.
66. McGuire RA Jr: Protection of the unstable spine during transportand early hospitalization. J Miss State Med Assoc 32:305–308,1991.
67. McGuire RA Jr, Degnan G, Amundson GM: Evaluation of cur-rent extrication orthoses in immobilization of the unstable cer-vical spine. Spine 15:1064–1067, 1990.
68. McGuire RA Jr, Neville S, Green BA, Watts C: Spinal instabilityand the log-rolling maneuver. J Trauma 27:525–531, 1987.
69. McHugh TP, Taylor JP: Unnecessary out-of-hospital use of fullspinal immobilization. Acad Emerg Med 5:278–280, 1998 (letter).
70. McSwain NE Jr: Spine management skills, in Pre-Hospital TraumaLife Support. Akron, Educational Direction, 1990, ed 2, pp 225–256.
71. Meldon SW, Brant TA, Cydulka RK, Collins TE, Shade BR:Out-of-hospital cervical spine clearance: Agreement betweenemergency medical technicians and emergency physicians.J Trauma 45:1058–1061, 1998.
72. Deleted in proof.73. Muhr MD, Seabrook DL, Wittwer LK: Paramedic use of a spinal
injury clearance algorithm reduces spinal immobilization in theout-of-hospital setting. Prehosp Emerg Care 3:1–6, 1999.
74. Nypaver M, Treloar D: Neutral cervical spine positioning inchildren. Ann Emerg Med 23:208–211, 1994.
75. Olson CM, Jastremski MS, Vilogi JP, Madden CM, Beney KM:Stabilization of patients prior to interhospital transfer. Am JEmerg Med 5:33–39, 1987.
76. Orledge JD, Pepe PE: Out-of-hospital spinal immobilization: Is itreally necessary? Acad Emerg Med 5:203–204, 1998.
77. Perry SD, McLellan B, McIlroy WE, Maki BE, Schwartz M, FernieGR: The efficacy of head immobilization techniques during sim-ulated vehicle motion. Spine 24:1839–1844, 1999.
78. Plaisier B, Gabram SG, Schwartz RJ, Jacobs LM: Prospectiveevaluation of craniofacial pressure in four different cervical or-thoses. J Trauma 37:714–720, 1994.
79. Podolsky S, Baraff LJ, Simon RR, Hoffman JR, Larmon B, AblonW: Efficacy of cervical spine immobilization methods. J Trauma23:461–465, 1983.
80. Deleted in proof.
S16 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
81. Prasad VS, Schwartz A, Bhutani R, Sharkey PW, Schwartz ML:Characteristics of injuries to the cervical spine and spinal cord inpolytrauma patient population: Experience from a regionaltrauma unit. Spinal Cord 37:560–568, 1999.
82. Raphael JH, Chotai R: Effects of the cervical collar on cerebro-spinal fluid pressure. Anaesthesia 49:437–439, 1994.
83. Rimel RW, Jane JA, Edlich RF: An educational training programfor the care at the site of injury of trauma to the central nervoussystem. Resuscitation 9:23–28, 1981.
84. Rodgers JA, Rodgers WB: Marginal mandibular nerve palsy dueto compression by a cervical hard collar. J Orthop Trauma9:177–179, 1995.
85. Roozmon P, Gracovetsky SA, Gouw GJ, Newman N: Examiningmotion in the cervical spine: Part I—Imaging systems and mea-surement techniques. J Biomed Eng 15:5–12, 1993.
86. Deleted in proof.87. Rosen PB, McSwain NE Jr, Arata M, Stahl S, Mercer D: Compar-
ison of two new immobilization collars. Ann Emerg Med 21:1189–1195, 1992.
88. San Mateo County, California: EMS System Policy Memorandum#F-3A. 1991.
89. Schafermeyer RW, Ribbeck BM, Gaskins J, Thomason S, HarlanM, Attkisson A: Respiratory effects of spinal immobilization inchildren. Ann Emerg Med 20:1017–1019, 1991.
90. Schriger DL: Immobilizing the cervical spine in trauma: Shouldwe seek an optimal position or an adequate one? Ann EmergMed 28:351–353, 1996.
91. Schriger DL, Larmon B, LeGassick T, Blinman T: Spinal immo-bilization on a flat backboard: Does it result in neutral position ofthe cervical spine? Ann Emerg Med 20:878–881, 1991.
92. Smith MG, Bourn S, et al.: Ties that bind: Immobilizing theinjured spine. J Emerg Med Serv JEMS 14:28–35, 1989.
93. Stauffer ES: Orthotics for spinal cord injuries. Clin Orthop 102:92–99, 1974.
94. Suter R, Tighe T, et al.: Thoracolumbar spinal instability duringvariations of the log-roll maneuver. Prehospital Disaster Med7:133–138, 1992.
95. Swain A, Dove J, Baker H: ABCs of major trauma: PartI—Trauma of the spine and spinal cord. BMJ 301:34–38, 1990.
96. Toscano J: Prevention of neurological deterioration before ad-mission to a spinal cord injury unit. Paraplegia 26:143–150, 1988.
97. Totten VY, Sugarman DB: Respiratory effects of spinal immobi-lization. Prehosp Emerg Care 3:347–352, 1999.
98. Tuite GF, Veres R, Crockard HA, Peterson D, Hayward RD: Useof an adjustable, transportable, radiolucent spinal immobiliza-tion device in the comprehensive management of cervical spineinstability: Technical note. J Neurosurg 85:1177–1180, 1996.
99. Wagner FC Jr, Johnson RM: Cervical bracing after trauma. MedInstrum 16:287–288, 1982.
100. Walsh M, Grant T, Mickey S: Lung function compromised byspinal immobilization. Ann Emerg Med 19:615–616, 1990 (let-ter).
101. Walton R, DeSalvo JF, Ernst AA, Shahane A: Padded vs unpad-ded spine board for cervical spine immobilization. Acad EmergMed 2:725–728, 1995.
102. Washtenaw/Livingston County Medical Control Authority: Spi-nal Injury Assessment and Immobilization: EMS Protocols. Ann Ar-bor, Washtenaw/Livingston County Medical Control Authority,1997.
103. Waters RL, Meyer PR Jr, Adkins RH, Felton D: Emergency,acute, and surgical management of spine trauma. Arch PhysMed Rehabil 80:1383–1390, 1999.
104. Worsing RA Jr: Principles of pre-hospital care of musculoskeletalinjuries. Emerg Med Clin North Am 2:205–217, 1984.
Sketch of a dissection showing the head falling forward, as hap-pens in some cases of destruction of the ligaments, associatedwith disease of the joints between the atlas and axis and occipi-tal bones. From, Hilton J: On Rest and Pain: A Course of Lec-tures on the Influence of Mechanical and Physiological Rest inthe Treatment of Accidents and Surgical Diseases, and the Diag-nostic Value of Pain. New York, Wood, 1879, 2nd ed.
Preadmission Cervical Spine Immobilization S17
CHAPTER 2
Transportation of Patients with Acute Traumatic CervicalSpine Injuries
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Expeditious and careful transport of patients with acute cervical spine or spinal cord injuries is
recommended, from the site of injury by the most appropriate mode of transportation available to thenearest capable definitive care medical facility.
RATIONALE
Definitive assessment, resuscitation, and care of thepatient with an acute traumatic cervical spine injurycannot be rendered at the accident scene. Optimal
care for patients with spinal injury includes initial resuscita-tion, immobilization, extrication, and early transport of thepatient to a medical center with the capability for diagnosisand treatment (3–5, 9, 11). Delay in transportation to a defin-itive treatment center is associated with less favorable out-comes, longer hospitalizations, and increased costs (7, 8, 11).
Several modes of transportation are available to transfer thespinal injury patient, primarily land (ambulance) and air (he-licopter or fixed-wing aircraft). Selection of the ideal mode oftransportation for an individual patient depends on the pa-tient’s clinical circumstances, distance, geography, and avail-ability. The goal is to expedite efficient, safe, and effectivetransportation, without unfavorable effects on patient out-come. These issues provide the rationale for this guideline.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of literature published from 1966 to 2001 was per-formed. The search was limited to the English language.Medical subject headings were used in combination with“spinal injury” and “transport.” The first search term (key-word and exploded subject heading) yielded 8,493 articles.The second search term (as keyword) yielded 12,437 articles.A search combining both terms provided 44 articles. The 44abstracts were reviewed, and additional references wereculled from the reference lists of the remaining articles. Fi-nally, members of the author group were asked to contributearticles on the subject matter that were known to them but notfound by other search means. This process yielded 13 articlesthat were directly relevant to the subject of transportationof patients with spinal injuries. All articles provide ClassIII medical evidence. Pertinent articles are summarized inTable 2.1.
SCIENTIFIC FOUNDATION
Safe, rapid transport of the spine-injured patient to a med-ical facility for definitive care has long been a fundamentaltenet of emergency medical service (EMS) care delivery. Norandomized clinical trials to establish the necessity or effec-tiveness of this strategy have been performed. A search of theliterature does not provide Class I or Class II medical evi-dence in support of this practice.
One of the basic principles of preadmission spinal care isthe early transfer of the injured person to a center with theresources and expertise to manage the patient with an acutecervical spine or spinal cord injury (3–5, 9, 11). Early compli-cations can be prevented and improved neurological out-comes have been reported when early transfer to a specializedspinal cord injury (SCI) center is accomplished (5, 11). Duringtransportation, every effort must be made to limit untowardspinal motion and to preserve neurological function (12).
Several options exist for the transportation of patients to adefinitive care facility. The selection of the mode of transpor-tation is based on the patient’s clinical status and what isreasonable and available to achieve the goals of rapid transfer,while maintaining effective medical support for the patientand proper spinal immobilization for patients at risk.
In 1974, Hachen (5) described the creation of a nationwideemergency transportation protocol for patients with spinalinjury. The protocol had been implemented in Switzerland in1968. All SCI patients in Switzerland were immediately trans-ported to the National Spinal Injuries Center in Geneva by theSwiss Air Rescue Organization. In the 10-year follow-up ofthis protocol published in 1977, Hachen (6) reported that earlytransport from the site of the accident to the SCI center underclose medical supervision was associated with no patientdeath during transport. Before 1968, many deaths occurredduring transport, secondary to acute respiratory failure, be-fore definitive care could be provided. After 1968, patientswere transported rapidly with an onboard anesthetist whoprovided respiratory, cardiac, and hemodynamic monitoring,
S18 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
resuscitation, and nasotracheal intubation as necessary. Theaverage time for the rescue operation was reduced from 4.5hours to 50 minutes. There was a significant reduction incardiovascular and respiratory morbidity and mortality. Themortality rate for complete quadriplegic patients droppedfrom 32.5% in 1966 to 6.8% in 1976, and for incomplete cer-vical cord-injured patients from 9.9 to 1.4% during the sameperiod. Hachen (5, 6) concluded that survival and outcome forpatients with acute spinal cord injuries was enhanced by awell-organized medical system, rapid medically supervisedtransfer by helicopter to a specialized center, and then defin-itive care in an SCI facility for aggressive management in theintensive care unit (ICU) setting.
Zach et al. (13), in 1976, described their experience with 117acute SCI patients managed according to prospective protocol inthe Swiss Paraplegic Center in Basel, Switzerland. All patientswere treated in the ICU setting with aggressive medical man-agement and cardiac and blood pressure support. Outcomeswere stratified by initial injury and time of admission afterinjury. Of cervical spinal cord injuries managed in this fashion,62% were improved at last follow-up, no patient with a cervicallevel injury worsened, and 38% were unchanged. Of patientswho arrived within 12 hours of injury, 67% improved, comparedwith their initial neurological condition. Of patients admittedbetween 12 and 48 hours of injury, 59% showed neurologicalimprovement. When admission occurred more than 48 hours
after injury, improvement was seen in only 50% of patients. Theauthors concluded that early transport and “immediate medicalspecific treatment of the spinal injury” seemed to facilitate neu-rological recovery.
In 1984, Tator et al. (11) reported their experience with 144patients with acute spinal cord injuries treated between 1974 and1979 at the Acute Spinal Cord Injury Unit (ASCIU) at Sunny-brook Medical Centre in Toronto, Ontario, Canada. They founda marked reduction in both morbidity and mortality after acutespinal cord injury for the group of patients managed during theperiod 1974 to 1979 compared with a similar group managedduring the period 1947 to 1973, which was before the creation ofa dedicated regional spinal cord injury unit. Reasons cited forthese improvements included earlier transport to the ASCIUafter trauma and better definitive management at arrival.
In a subsequent 1993 publication comparing ASCIU pa-tients managed from 1974 to 1981 with the patients managedfrom 1947 to 1973, Tator et al. (10) noted a statistically signif-icant difference in time from injury to arrival, 5 hours forASCIU patients compared with 13 hours for the pre-ASCIUgroup. They found a significant decrease in the severity ofspinal cord injury (65% complete cervical lesions in pre-ASCIU patients compared with 46% for ASCIU patients) andnoted fewer complications, shorter hospital stays, and lowerexpenses for patients managed under the new ASCIU para-digm. Their findings support the advantages of early trans-
TABLE 2.1. Summary of Reports on Transportation of Patients with Spinal Injuriesa
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Tator et al.,1993 (10)
A study of 201 ASCI patients, ICU care, hemodynamic supportcompared with 351 previous patients.
III Less severe cord injuries due to immobilization, resuscitation, and earlytransfer to ICU setting.
Armitage et al.,1990 (1)
Case reports of 4 patients who developed respiratory problemsduring airplane transport.
III Airplane air is less humid, and measures to optimize humidity andpulmonary function in high cervical injury patients may be required duringtravel.
Boyd et al.,1989 (2)
A prospective cohort study to determine the effectiveness of airtransport for major trauma patients when transferred to a traumacenter from a rural emergency room.
III Patients with severe multiple injury from rural areas fare better withhelicopter EMS than ground EMS.
Burney et al.,1989 (3)
Retrospective review of the means of transport and type ofstabilization used for all patients with ASCI.
III ASCI patients can be safely transported by air or ground using standardprecautions. Distance and extent of associated injury are the bestdeterminants of mode of transport.
Tator et al.,1984 (11)
A retrospective review of results of innovations between 1974 and1979 at Sunnybrook Medical Centre in Toronto.
III Patients transferred to the specialized unit earlier, with consequent markedreduction in complications and cost of care.
Hachen, 1977(6)
A study of 188 ASCI patients managed in ICU, aggressive treatmentof hypotension, respiratory insufficiency.
III Reduced morbidity and mortality with early transfer, attentive ICU care andmonitoring, and aggressive treatment of hypotension and respiratory failure.
Zach et al.,1976 (13)
A study of 117 ASCI patients at Swiss center, ICU setting, aggressivemedical therapy (Rheomacrodexb � 5 d; dexamethasone � 10 d).
III Improved neurological outcome with aggressive medical treatment. Betteroutcome for early referrals.
Hachen, 1974(5)
Retrospective review of effectiveness of emergency transportation ofspinal injury patients in Switzerland. From 1965 to 1974, all ASCIpatients were immediately transported by air to ASCI center.Mortality reduced to zero during transport. Average time for therescue operation reduced from 4.5 h to 50 min. Significantreduction in cardiovascular and respiratory morbidity/mortality.
III Mortality and morbidity of patients with acute spinal injury is reduced by awell-organized medical response with smooth and rapid transfer byhelicopter to a specialized SCI center.
a ASCI, acute spinal cord injury; ICU, intensive care unit; EMS, emergency medical services; SCI, spinal cord injury.b Rheomacrodex, dextran 40 (Medisan, Parsippany, NJ).
Patient Transport in Acute Traumatic Cervical Spine Injuries
S19Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
port to a regional, specialized SCI center for definitive com-prehensive care of patients with SCI.
Burney et al. (3) reviewed the means of transport and type ofstabilization used for all patients with acute spinal cord injuriestransferred to the University of Michigan Medical Center from1985 to 1988 to determine the effect of these variables on impair-ment and neurological improvement. Sixty-one patients werereviewed. Twenty-five patients were transported by ground am-bulance (41%), 33 by helicopter (54%), and 3 by fixed-wingaircraft (5%). Forty-three patients (70.5%) had cervical spineinjuries, 11 (18%) had thoracic spine injuries, and 7 (11.5%) hadlumbar spine injuries. Fifty-one patients (84%) were transferredwithin 24 hours of injury. A variety of standard methods ofstabilization were used during transport. No patient sustainedan ascending neurological injury as a result of early transport.The level of function improved before discharge in 26 (43%) of 61patients. Patients transported to the medical center within 24hours of injury were more likely to show improvement (25 of 51patients) than those transported after 24 hours (1 of 10 patients).There was no significant difference in the probability of im-provement between ground transport (8 of 25 patients) or airtransport (18 of 36 patients). The authors concluded that acuteSCI patients could be safely transported, with standard precau-tions, by air or ground. They found that distance and the extentof the patient’s associated injuries were the best determinants ofthe mode of transport.
Rural areas reportedly account for 70% of fatal accidents, andrural mortality rates for people involved in motor vehicle acci-dents are four to five times higher than in urban areas. A pro-spective cohort study by Boyd et al. (2) examined the effective-ness of air transport of major trauma patients when transferredto a trauma center from a rural emergency room. The studyconsisted of 872 consecutive trauma patients admitted afterlong-distance transfer. The authors found that air transport wasassociated with a 25.4% reduction in predicted mortality (Z �3.95; P � 0.001). The benefit of helicopter EMS transport wasrealized only in major trauma patients with a probability ofsurvival of less than 90%. Thus, the benefits identified with earlyhelicopter EMS transport were directly related to injury severity.It is unclear whether these findings can be extrapolated to pa-tients with spine and/or spinal cord injuries, because the authorsdid not stratify injuries by body systems in their report.
Neither land nor air transport has been reported in theliterature to adversely affect the outcome for spine-injuredpatients when properly executed. One note of caution wasoffered by Armitage et al. (1). They described four spine-injured patients who developed respiratory distress/failureduring airplane transport. They noted that because patientswith cervical spinal cord injuries may have severely reducedpulmonary performance, measures to optimize oxygenation,humidification, and pulmonary function should be under-taken for these patients, particularly during air transport.
SUMMARY
The patient with an acute cervical spine or spinal cordinjury should be expeditiously and carefully transported from
the site of injury to the nearest capable definitive care medicalfacility. The mode of transportation chosen should be basedon the patient’s clinical circumstances, distance from the tar-get facility, and geography to be traveled, and should be themost rapid means available. Patients with cervical spinal cordinjuries have a high incidence of airway compromise andpulmonary dysfunction; therefore, respiratory support mea-sures should be available during transport. Several studiessuggest that rates of morbidity and mortality of SCI patientsdecreased after the advent of sophisticated transport systemsto dedicated SCI centers. These studies all provide Class IIImedical evidence on this issue.
KEY ISSUES FOR FUTURE INVESTIGATION
Development and refinement of transportation protocolsfor patients with cervical spine and spinal cord injury shouldbe undertaken and could be accomplished by using a large,prospectively collected data set. From these data, case-controlor comparative cohort studies could be structured to generateClass II evidence.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Armitage JM, Pyne A, Williams SJ, Frankel H: Respiratory prob-lems of air travel in patients with spinal cord injuries. BMJ300:1498–1499, 1990.
2. Boyd CR, Corse KM, Campbell RC: Emergency interhospitaltransport of the major trauma patient: Air versus ground.J Trauma 29:789–794, 1989.
3. Burney RE, Waggoner R, Maynard FM: Stabilization of spinalinjury for early transfer. J Trauma 29:1497–1499, 1989.
4. Guttman L: Initial treatment of traumatic paraplegia and tetra-plegia, in Spinal Injuries: Proceedings of the Symposium held in theRoyal College of Surgeons of Edinburgh, June 7–8, 1963. Edinburgh,The Royal College of Surgeons, 1967.
5. Hachen HJ: Emergency transportation in the event of acute spinalcord lesion. Paraplegia 12:33–37, 1974.
6. Hachen HJ: Idealized care of the acute injured spinal cord inSwitzerland. J Trauma 17:931–936, 1977.
7. Neville S, Watts C: Management of the unstable cervical spine intransport: A re-evaluation. Aeromed J Sept/Oct: 32, 1987.
8. Rutledge G, Sumchai A: A safe method for transportation ofpatients with cervical spine injuries. Aeromed J Sept/Oct: 33,1987.
9. Stover S, Fine PR: Spinal Cord Injury: The Facts and Figures. Bir-mingham, University of Alabama at Birmingham, 1986, p 45.
10. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Changes in epidemiology of acute spinal cord injury from 1947 to1981. Surg Neurol 40:207–215, 1993.
11. Tator CH, Rowed DW, Schwartz ML, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–294, 1984.
12. Toscano J: Prevention of neurological deterioration before admis-sion to a spinal cord injury unit. Paraplegia 26:143–150, 1988.
13. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cord injuryin the Swiss Paraplegic Centre. Paraplegia 14:58–65, 1976.
S20 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
CHAPTER 3
Clinical Assessment after Acute Cervical Spinal Cord Injury
RECOMMENDATIONSNEUROLOGICAL EXAMINATION:Standards: There is insufficient evidence to support neurological examination standards.Guidelines: There is insufficient evidence to support neurological examination guidelines.Options: The American Spinal Injury Association international standards for neurological and functional
classification of spinal cord injury are recommended as the preferred neurological examination tool forclinicians involved in the assessment and care of patients with acute spinal cord injuries.
FUNCTIONAL OUTCOME ASSESSMENT:Standards: There is insufficient evidence to support functional outcome assessment standards.Guidelines: The Functional Independence Measure is recommended as the functional outcome assessment tool
for clinicians involved in the assessment and care of patients with acute spinal cord injuries.Options: The modified Barthel index is recommended as a functional outcome assessment tool for clinicians
involved in the assessment and care of patients with acute spinal cord injuries.
RATIONALE
Acute traumatic spinal cord injury (SCI) affects 12,000 to14,000 people in North America each year. The func-tional consequences of an acute SCI (ASCI) are vari-
able; therefore, the initial clinical presentation of patients withASCI is a key factor in determining triage and therapy and inpredicting prognosis. Consistent and reproducible neurolog-ical assessment scales are necessary to define the acutelyinjured patient’s neurological deficits and to facilitate com-munication with caregivers regarding patient status. Prognos-tic information provided by comparing people with injurieswith the historical outcomes of patients with similar injuries isof value to patients and families. The evaluation of newtherapies proposed for the treatment of ASCI requires the useof accurate, reproducible neurological assessment scales andreliable functional outcome measurement tools, not only tomeasure potential improvement after therapy, but to deter-mine its functional significance. For these reasons, the clinicalneurological assessment and the determination of functionalabilities are important aspects of the care of patients withASCI. The purpose of this review of the medical literature isto determine which neurological assessment scales and whichfunctional impairment tools are the most useful in the care ofpatients with ASCI.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasperformed. The search was limited to the English languageand human studies. The terms “spinal cord injury” or “spinal
injury” combined with the term “classification or assessment”yielded 17,923 references. A second search with the terms“scale” or “weights and measures” or “index” or “abstractingand indexing” combined with the terms “spinal cord injury”or “spinal injury” yielded 337 references. These 337 referencesand the 17,923 references from the broader search were im-ported into a database, and duplicates were eliminated. Arti-cles germane to this topic were selected by reviewing theirtitles and abstracts. Additional references were culled fromthe reference lists of the remaining articles. Finally, membersof the author group were asked to contribute articles knownto them on the subject matter that were not found by othersearch means.
A total of 53 articles were accessed, reviewed, graded, andincluded in this review. There is no Class I medical evidencein the literature on this topic. There are two Class II compar-ative analyses of functional outcome scales. Twenty-sevenpertinent articles are summarized in Tables 3.1 and 3.2.
SCIENTIFIC FOUNDATION
A variety of assessment systems are available for docu-menting neurological status of patients after ASCI. They in-clude the Frankel scale, the modified Frankel scale, Lucas andDucker’s neurotrauma motor index, the Sunnybrook scale, theBotsford scale, the Yale scale, the National Acute Spinal CordInjury Study (NASCIS) scale, the American Spinal Injury As-sociation (ASIA) scale, and the ASIA/International MedicalSociety of Paraplegia (IMSOP) international standards forneurological and functional classification of SCI scale (1–4,6–8, 11–13, 15, 16, 28, 37, 42, 47).
S21Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Several of these assessment scales have been refinedthrough serial iterations (1–4, 28, 41, 42, 47). A few are widelyused, but others have not attained general acceptance and
recognition. Ideally, the clinical neurological assessment ofASCI patients should be uniform, reproducible, and thor-ough, but easy to use. The assessment tool must be detailed
TABLE 3.1. Summary of Reports on Neurological Examination Scalesa
Series (Ref. No.) Description of Study Evidence Class Conclusions
Jonsson et al., 2000(33)
A study of the interrater reliability of the ASIA ISCSCI-92.Physicians and physiotherapists classified 23 patientsaccording to the ISCSCI-92 and calculated � values.
III Indicates a weak interrater reliability for scoring incomplete ASCIlesions using the 1992 ASIA standards.
Cohen et al., 1998(18)
A test of the ASIA ISCSCI-92. Participants completed apretest and posttest in which they classified 2 patients whohad an ASCI.
III Further revision of the ASIA 1992 standards and more training wasneeded to ensure accurate classification of SCI.
El Masry et al.,1996 (26)
A study to assess the reliability of the ASIA and NASCISmotor scores. The motor scores of 62 consecutive ASCIpatients were retrospectively reviewed.
III The differences in correlation coefficients between the ASIA motorscore and the NASCIS motor score were not statisticallysignificant. The ASIA and NASCIS motor scores can both be usedfor the neurological quantification of motor deficit and motorrecovery.
Wells and Nicosia,1995 (51)
A comparison of the Frankel scale, Yale scale, MIS, MBI,and FIM in 35 consecutive ASCI patients.
III The best assessment tool is a combination of two scales, onebased on neurological impairment and one on functionaldisability.
Waters et al., 1994(50)
An assessment of strength using motor scores derived fromASIA compared with motor scores based onbiomechanical aspects of walking in predicting ambulatoryperformance in 36 ASCI patients.
III The ASIA scoring system compared favorably with thebiomechanical scoring system. ASIA motor score stronglycorrelates with walking ability.
Davis et al., 1993(19)
A prospective study of 665 ASCI patients to determine thereliability of the Frankel and Sunnybrook scales.
III Demonstrated high interrater reliability of Frankel and Sunnybrookscales. Both scales correspond to total sensory and motor functionbut are insensitive to walking and bladder function.
Bednarczyk andSanderson, 1993(5)
A study comparing ASIA scale, NACIS scale, and BB(wheelchair basketball) Sports Test in 30 ASCI patientsclassified by the same examiner.
III ASIA scale showed the greatest discrimination in grouping subjectswith ASCI. NASCIS scale had negative correlation with ASIA scaleand BB sports test.
Botsford and Esses,1992 (7)
Description of a new functionally oriented scale withassessment of motor and sensory function, rectal tone, andbladder function.
III Botsford scale was sensitive for detecting improvement in functionover time after ASCI.
Priebe and Waring,1991 (44)
A study of the interobserver reliability of the 1989 revisedASIA standards assessed by quiz given to 15 physicians.
III The interobserver reliability for the revised ASIA (1989) standardswere improved compared with previous versions, but less thanoptimal. Changes were recommended.
Bracken et al.,1990 (10)
Multicenter North American trial examining effects ofmethylprednisolone or naloxone in ASCI (NASCIS II).
III forneurologicalassessment
Motor scores of 14 muscles on 5-point scale, right side of bodyonly. Sensory scores of pinprick and light touch, 3-point scale,bilateral. No interrater reliability comparison.
Lazar et al., 1989(36)
A prospective study of the relationship between earlymotor status and functional outcome after ASCI in 78patients. Motor status was measured by the ASIA MIS, andfunctional status was evaluated with the MBI.
III The MIS correlated well with functional status for quadriplegicpatients, poorly for paraplegic patients. Individual differences inambulation limit its predictive usefulness.
Bracken et al.,1985 (11)
Multicenter North American trial examining effects ofmethylprednisolone in ASCI (NASCIS I).
III forneurologicalassessment
Motor scores of 14 muscles on 6-point scale. Right side of bodyonly. Sensory scores of pinprick and light touch, 3-point scale,bilateral. No interrater reliability comparison.
Tator et al., 1982(47)
Initial description of the Sunnybrook scale, a 10-gradenumerical neurological assessment scale.
III Improvement from the Frankel scale. Motor grading subdividedbut not very sensitive.
Chehrazi et al.,1981 (15)
Initial description of the Yale scale and its use in a groupof 37 patients with ASCI.
III Provides assessment of the severity of ASCI.
Lucas and Ducker,1979 (37)
Initial description of a motor classification of patients withSCI and its use in 800 patients.
III Allows the clinical researcher to evaluate current treatments andassess the potential of new treatment regimes.
Bracken et al.,1978 (12)
Description of 133 ASCI patients classified using motorand sensory scales developed by Yale Spinal Cord InjuryStudy Group.
III Considerable discrepancy between motor and sensory impairmentscales among patients with greater motor than sensory loss.
Frankel et al., 1969(28)
The first clinical study of the Frankel scale to assessneurological recovery in 682 patients treated with posturalreduction of spinal fractures.
III First neurological examination scale for ASCI.
a ASIA, American Spinal Injury Association; ISCSCI, international standards for neurological and functional classification of incomplete spinalcord injury; SCI, spinal cord injury; ASCI, acute SCI; NASCIS, National Acute Spinal Cord Injury Study; MBI, modified Barthel index; MIS, motorindex score; FIM, functional independence measurement.
S22 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
and precise to specifically document a given patient’s injuryand must provide descriptive measurement scales that allowdetermination of loss or gain of function with time and ther-apy. Finally, there must be measurement of the patient’sfunctional abilities relative to the neurological examination todocument whether losses or gains have meaningful signifi-cance to the patient and to accurately determine outcome.Whatever assessment system is used, it must have interob-server reliability. Difficulties exist when clinicians use poorlydefined measurement tools or different methods of neurolog-ical assessment to describe the same patient, hindering thedefinition (and potentially the management) of that patient bydifferent clinicians and the comparison of that patient withother patients with similar injuries. The accurate assessmentof both the neurological status and the functional skills ofASCI patients is essential for patient management, the con-duct of research studies, and comparisons of clinical thera-peutic trials.
Numerous assessment scales have been used to evaluatepatients with SCIs. Scales may be divided into two generaltypes. The first type is examination-specific and focuses on theneurological deficits that are a result of SCI. These scales usethe motor and sensory examination primarily (or exclusively)to assign a numerical value or letter grade (1, 2, 6–12, 15, 28,37, 42, 47). The second type of scale focuses on functionalskills, including a patient’s ability to care for himself or her-self, participate in personal hygiene, transfer, or ambulate (3,4, 14, 22–24, 26, 29, 34, 35, 38, 41, 45). In general, the first typeof scale is used for the acute assessment of patients with SCI,
whereas both assessment scales are used to define the chron-ically injured patient. More contemporary assessment scalesincorporate both neurological examination scores and func-tional outcome scores in their overall definition of individualpatients (3, 4, 51).
Neurological examination scales
Frankel et al. (28) provided the first report of a stratifiedneurological scale used to characterize patients with ASCI in1969. The authors used a five-grade scale, A to E, to defineSCIs in 682 patients managed at the Stoke Mandeville Hospi-tal between 1951 and 1968. Grade A patients had completemotor and sensory lesions. Grade B patients had sensory-onlyfunction below the level of injury. Grade C patients had motorand sensory function below the level of injury, but their motorfunction was useless. Grade D patients had motor functionuseful, but not normal function below the level of SCI. GradeE patients had recovery with no motor, sensory, or sphincterdisturbance. The Frankel scale, as it became known, waswidely adopted for use in the description of SCI patients andin assessment of their therapy (outcome) in the 1970s and1980s (1, 2, 28, 29, 42). It was easy to use, was based solely onmotor and sensory function, and required very little patientassessment before classification into one of five grades. How-ever, differentiation between patients classified into Grades Cand D was imprecise. These were broad injury groups withconsiderable range within each injury grade. The sensitivity ofthe Frankel scale to change in serial measurements, particu-
TABLE 3.2. Summary of Reports on Functional Outcome Scalesa
Series (Ref. No.) Description of Study Data Class Conclusions
Field-Fote et al., 2001 (27) SCI-FAI offered as functional assessment scale forgait assessment.
III Reliable and relatively sensitive measure of walking ability inpatients with ASCI. Interrater reliability good. No � values offered.
Kucukdeveci et al., 2000(35)
To determine the reliability and validity of the MBIin Turkey.
III Adaptation of the MBI successful in Turkey as long as its limitationsare recognized. � values �0.5.
Ditunno et al., 2000 (23) WISCI offered as index for ambulation skills after SCIin pilot study.
III Good reliability, excellent interrater reliability but needs assessmentin clinical settings.
Yavuz et al., 1998 (53) Assessment of the relationship of two functionaltests, FIM and QIF, to ASIA scores.
III Strong correlation between FIM and QIF to ASIA scores.
Catz et al., 1997 (14) SCIM offered as new disability scale for spinal cordlesions. Thirty patients assessed with SCIM and FIM.
III SCIM more sensitive than FIM.
Hamilton et al., 1994 (31) Assessment of interrater agreement of FIM in 1018patients in 89 UDS hospitals.
II � values for 7-level FIM ranged from 0.53 to 0.66. � values higherin subset of UDS hospitals with experienced rehabilitationclinicians, 0.69–0.84.
Dodds et al., 1993 (25) Assessment of reliability of FIM in characterizing11,102 UDS rehabilitation patients.
III FIM has high internal consistency and adequate discriminativecapabilities and was good indicator of burden of care.
Hamilton et al., 1991 (32) Interrater agreement assessment of FIM in 263patients in 21 UDS hospitals.
II � values for 7-level FIM ranged from 0.61 to 0.76; mean, 0.71.
Shah et al., 1989 (45) Description of MBI. III The MBI has greater sensitivity and improved reliability than theoriginal version, without additional difficulty or implementationtime.
Gresham et al., 1986 (29) Assessment of QIF as functional scale, comparedwith Barthel index.
III The QIF was more sensitive and reliable than the Barthel index.
a SCI, spinal cord injury; ASCI, acute SCI; FAI, functional ambulatory inventory; MBI, modified Barthel index; WISCI, walking index for spinalcord injury; FIM, functional independence measurement; QIF, quadriplegic index of function; ASIA, American Spinal Injury Association; SCIM,spinal cord independence measure; USD, Uniform Data System.
Clinical Assessment after Acute Cervical Spinal Cord Injury S23
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
larly among patients in Grades C and D was poor. Significantimprovement in patient function could occur over time with-out the patient advancing a Frankel grade (1, 2, 7). Modifica-tions of the Frankel scale were offered in an attempt to solvethis shortcoming, but the use of the Frankel scale as either anacute assessment tool or an outcome measure has been largelyabandoned because of its lack of sensitivity (3, 4, 6, 7, 13, 21,41, 47–49, 51).
Institutions and investigators have proposed a variety ofneurological assessment scales for SCI patients over the years(1–4, 6–12, 15, 28, 37, 42, 47). In 1978, Bracken et al. (12)described the SCI severity scale developed at the Yale Uni-versity School of Medicine. This scale combines motor andsensory function in selected muscle groups and dermatomes.Its primary focus is to distinguish between complete andincomplete SCIs. The sensory severity scale ranged from onepoint to seven points, and the motor severity scale rangedfrom one point to five points. The authors reported a strongcorrelation between the two scales and noted that “changescores” in motor and sensory function correlated with out-come at discharge compared with at admission. Their pro-posed assessment scales did not assess bowel or bladderfunction and were complicated by grouping all patients to oneof five possible motor scales and into one of seven possiblesensory scales. This scale was difficult to memorize and hardto apply at the bedside.
In 1979, Lucas and Ducker (37) at the Maryland Institute forEmergency Medical Services developed a scoring system forpatients with ASCI. Their scale was based on motor functionat and below the level of injury (Lucas and Ducker’s neuro-trauma motor index) and was used to evaluate more than 800patients collected by the Nationwide Spinal Cord Injury Reg-istry. The scale was later modified for a prospective study ofSCI treatment regimes used at the Maryland Institute forEmergency Medical Services. The authors chose 14 musclesfor examination and used mathematical analysis to predict amotor outcome score based on the initial motor examinationand an empirically derived understanding of the recoveryrate of individual injury subtypes. The scoring system waslimited in that many patients were excluded from the analysis(only 436 of more than 800 patients were analyzed), the stan-dard error of the predicted recovery score was large, and thecalculations were cumbersome. Their scoring system waslater modified by ASIA into a motor index score (24, 41).
In the early 1980s, three different SCI neurological assess-ment scales were introduced, none of which gained popularacceptance (15, 34, 47). In 1980, Klose et al. (34) described theUniversity of Miami Neuro-spinal Index, which consisted oftwo subscales, one motor and one sensory. The motor scalewas scored on a scale of 0 to 5 points for 44 muscle groups,resulting in a possible range of scores from 0 to 220 points.Sensory scoring was a three-point scale for pinprick and vi-bratory sensation in 30 dermatomes on each side of the body.Initial interobserver reliability was high among three physicaltherapists who examined 10 neurologically stable patients inthe rehabilitation setting. Further studies were planned todetermine the efficacy of the University of Miami Neuro-spinal Index in the acute setting and as an outcome tool. The
Yale Scale was reported in 1981 by Chehrazi et al. (15) at theYale New Haven Medical Center. The scale used the BritishMedical Research Council’s gradation (0 to 5 points) of musclestrength by using 10 selected muscle groups from each side ofthe body. Sensory function was scored on a two-point scalefor superficial pain, position sense, and deep pain. Bladderand bowel functions were not scored. In 1982, the SunnybrookCord Injury Scales for assessing neurological injury and re-covery from SCI were proposed by Tator et al. (47). A 10-pointnumerical neurological assessment scale was offered, whichrepresented an improvement on the Frankel scale in howsensory losses were classified. However, like the Frankelscale, motor grading was not very sensitive. The differentia-tion between Grades 3 to 5 and Grades 6 to 8, correspondingto Frankel Grades C and D, remained relatively imprecise andresulted in large, heterogeneous groups of patients. Bladderand bowel functions were not assessed.
In 1984, ASIA (1) generated standards for the neurologicalclassification of spinal injury patients. The neurological as-sessment used a 10-muscle group motor index score (scale of0 to 5 points) and incorporated the Frankel classification as thefunctional abilities assessment tool. The sensory examinationwas not scored, but the most cephalad level of normal sensa-tion was noted. These standards were revised in 1989 toprovide better, more specific sensory level determinations (2).In 1991, Priebe and Waring (44) examined the interobserverreliability of the revised ASIA standards (1989 version). Pa-tient examples, in quiz format, were given to house staff andfaculty of a department of physical medicine and rehabilita-tion. Respondents were asked to classify each patient withrespect to sensory level, motor level, zone of injury, andFrankel classification according to the 1984 ASIA standards.Two months later, the respondents were asked to completeanother quiz by using the 1989 ASIA standards. Although thepercentage of correct answers improved by using the 1989ASIA standards, the authors conclude that interobserver reli-ability was “less than optimal” with a � coefficient of 0.67,indicating agreement between observers but only within therange of fair agreement.
Botsford and Esses (7) introduced a new functionally ori-ented neurological grading system that incorporated motorand sensory function, rectal tone, and bladder control. Themotor assessment score, on a scale of 0 to 5 points, assessedflexor and extensor groups at major joints (hence a “morefunctional” motor assessment). Sensory function was gradedon a 10-point scale (0 to 10 points) divided into five categories.Voluntary rectal contraction was scored on a 10-point scale (0,5, or 10 points). Bladder function was divided into “normal”and “not normal” and assigned 5 points. The authors appliedtheir proposed grading system to a historical group of pa-tients who had initially been assessed and classified accordingto the Frankel scale. They concluded that the new gradingsystem was more sensitive for detecting improvement in theneurological examination and in functional performance overtime.
Two national ASCI studies (NASCIS I and II) were accom-plished in the late 1980s and early 1990s in examination ofmethylprednisolone as a treatment for patients with ASCI
S24 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
(8–11). Investigators at multiple centers in North Americaused a motor assessment scale (NASCIS scale) that evaluatedmotor function in 14 muscle groups graded from 1 to 6 points(NASCIS I) (8, 11), or 0 to 5 points (NASCIS II) (9, 10). Scoresfor the right and left sides of the body were obtained inde-pendently. Sensory function was divided into pinprick andlight touch in dermatomes C2 through S5 and scored on ascale of 0 to 3 points. Functional abilities were not assessed inthe NASCIS I and II studies. Like most other neurologicalassessment scoring systems before the NASCIS scales, therewas no documentation of interobserver reliability, despite thelarge numbers of patients examined and entered into theNASCIS I and II trials.
In 1992, ASIA (3) generated new standards for neurologicaland functional classification of SCI in conjunction withIMSOP. These standards replaced the revised 1989 version.The new assessment recommendations included motor indexscores, sensory examination scores (scale of 0 to 2 points), andthe ASIA impairment scale (modified Frankel classification),and incorporated the Functional Independence Measure(FIM). FIM is a functional assessment tool and is used toassess the effect of SCI on the patient’s functional abilities. Itquantifies the extent of individual disability and complementsthe neurological assessment by providing scoring for activi-ties of eating, grooming, bathing, dressing upper body, dress-ing lower body, and toileting (20–22, 24). Improvements inneurological function over time or with treatment (as docu-mented by neurological examination scales) can be measuredin terms of functional or meaningful improvement to thepatients with the addition of FIM in the assessment battery.
Davis et al. (19) measured the interobserver reliability of theFrankel classification and the Sunnybrook scale by experi-enced personnel who were provided with concise definitions.The authors demonstrated high interobserver reliability of theFrankel classification and Sunnybrook scales (Pearson corre-lation coefficients, 0.71–0.91), with 94 to 100% intraobserveragreement. Values of the � statistic were not provided. Theauthors concluded that both assessment systems corre-sponded well to total sensory and motor function in SCIpatients but were insensitive to ambulation skills and bladderfunction.
In 1993, Bednarczyk and Sanderson (5) reported on theability of three different classification systems to describe SCIpatients and to compare the correlation between the threescales when applied by a single trained observer. They com-pared the NASCIS scale with the ASIA scale and the BB(wheelchair basketball) Sports Test. The authors found thatthe ASIA scale had the greatest discrimination in groupingsubjects with SCI into mixed-injury categories and intoincomplete-injury categories. The BB Sports Test had a posi-tive correlation with the ASIA scale (Spearman rank correla-tion coefficient, 0.81). The NASCIS scale had a negative cor-relation with both the ASIA scale (Spearman coefficient,�0.66) and the BB Sports Test (Spearman coefficient, �0.48).In contrast, El Masry et al. (26) retrospectively assessed 62consecutive ASCI patients and compared ASIA and NASCISmotor scores with conventional motor examinations. Theseauthors found that both motor assessment scales were repre-
sentative of the conventional motor scores reported for thesepatients and could be used to quantify motor deficits andrecovery after ASCI.
Lazar et al. (36), in 1989, evaluated the relationship betweenearly motor status and functional outcome after SCI prospec-tively in 52 quadriplegic and 26 paraplegic patients. Motorstatus was measured within 72 hours of injury and quantifiedwith the ASIA motor index score. Functional status was eval-uated with the modified Barthel index (MBI). A senior phys-ical therapist completed ASIA motor index score and MBIassessments on each patient at admission to the spinal cordintensive care unit and every 30 days during rehabilitation.The authors found that early motor function correlated wellwith average daily improvement in functional status, includ-ing self-care and mobility (P � 0.001). The initial motor indexscore strongly correlated with the functional status of quad-riplegic patients at admission (P � 0.001), at 60 days (P �0.001), and at rehabilitation discharge (P � 0.001), but hadpoor correlation in paraplegic patients. The ASIA motor indexscore correlated significantly with the MBI self-care subscoreat 60 days and at discharge (P � 0.01), but not with the MBImobility subscore. Lazar et al. concluded that the ASIA motorindex score is a useful tool in predicting function duringrehabilitation, although individual differences in ambulation,particularly for patients with paraplegia, limit the predictiveusefulness of this index.
Waters et al. (50), in 1994, compared the strength of 36 acuteSCI patients as determined by the ASIA motor score withmotor scores based on biomechanical aspects of walking topredict ambulatory performance. The authors found that theASIA scoring system compared favorably with the biome-chanical scoring system and was a relatively simple clinicalmeasure that correlated strongly with walking ability. In 1995,Marino et al. (40) compared the ASIA motor level and theupper extremity motor score (UEMS) with the neurologicallevel of injury in 50 quadriplegic patients. At 12 months afterinjury, quadriplegia index of function (QIF) assessments wereobtained. Spearman rank correlations were calculated. Theauthors found that the motor level was more highly correlatedwith the UEMS and the QIF than with the neurological levelof injury. The UEMS had the highest correlation to the QIFfeeding score, 0.78. Marino et al. concluded that the ASIAmotor level and UEMS better reflect the severity of impair-ment and disability. Similarly, Ota et al. (43) compared theASIA motor scores and neurological level of injury with FIMin 100 Frankel Grade A and B patients. The authors found thatthe motor score reflected the patients’ disability as deter-mined by FIM better than the ASIA level of injury.
Wells and Nicosia (51) compared the usefulness and limi-tations of five different SCI scoring systems applied by asingle skilled observer in the assessment of 35 consecutive SCIpatients: the Frankel classification, the Yale scale, the ASIAmotor index score, the MBI, and the FIM score. The authorsfound that the Frankel classification correlated strongly withthe Yale scale and the ASIA motor index scores, but weaklywith MBI and FIM. These three assessment scales shared afocus on impairment measurement. The MBI and the FIMscore correlated strongly with each other but weakly with the
Clinical Assessment after Acute Cervical Spinal Cord Injury S25
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
other scales, and they shared a focus on disability. Wells andNicosia (51) concluded that one classification system or scalealone does not adequately describe SCI patients in both theacute and follow-up settings. They favored a combination oftwo scales to characterize ASCI patients, one based on neu-rological impairment and the other on functional disability.
Jonsson et al. (33) evaluated interobserver reliability of the1992 ASIA standards for neurological and functional classifi-cation of SCI. Two physicians and two physiotherapists clas-sified 23 SCI patients according to the 1992 recommendations.Values for the � statistic for pinprick scores varied from 0 to0.83 (poor to very good), from 0 to 1.0 for light touch scores,and from 0 to 0.89 for motor function. They found weakinterobserver reliability for scoring patients with incompleteSCIs. Cohen et al. (17) performed further tests of reliability ofthe 1992 ASIA standards. The 106 professionals in the fieldof SCI completed tests before and after formal review of the1992 ASIA standards, in which they classified two SCI pa-tients by sensory and motor levels, zone of partial preserva-tion, ASIA impairment scale, and completeness of injury.Percentage “correct” was calculated for each tested criterion.The authors reported that participants had very little diffi-culty in correctly classifying the patient with a complete SCIbut had variable success characterizing the patient with anincomplete SCI. They concluded that further refinement of the1992 ASIA standards and more training in their applicationwas required.
In 1996, ASIA/IMSOP (4) provided a revised version of theinternational standards for neurological and functional clas-sification of spinal injury (an update of their 1992 recommen-dations). Further refined by input from numerous interna-tional organizations, the combination of the ASIA impairmentscale, the ASIA motor index score, the ASIA sensory scale,and FIM is considered to be the most representative assess-ment and classification tool for patients with ASCI. It wasconsidered to be an improvement on the 1992 standards,which were subject to criticism (4, 13, 17, 18, 25, 33, 41, 43, 44).
Functional outcome scales
Functional outcome scales are nonspecific measures of hu-man performance ability relevant to medical rehabilitation,that is, how a person functions with activities of everyday life.Several scales have been developed in an effort to accuratelycharacterize an injured person’s functional skills and disabil-ities to quantify his or her functional independence (3, 4, 14,21–23, 25, 27, 29, 30, 34, 35, 38, 39, 41, 43, 45, 46). These scalesattempt to determine a patient’s ability or inability to liveindependently. Scales for functional rating include the Barthelindex, the MBI, the FIM, the QIF, the spinal cord indepen-dence measure (SCIM), the walking index for SCI (WISCI),and the SCI functional ambulation inventory (SCI-FAI) (3, 4,14, 21–23, 25, 27, 29, 30, 34, 35, 38, 39, 41, 43, 45, 46). Thesescales are applicable to a wide range of nervous system dis-orders, but the QIF, the SCI-FAI, and the SCIM are morespecific for patients with SCI (14, 27, 29). All of these scaleshave been successfully used to characterize SCI patients (3, 4,14, 21–23, 25, 27, 29, 30, 34, 35, 38, 39, 41, 43, 45, 46).
Among many available functional assessment scales, theBarthel index has been one of the most popular (35, 38, 52). Ithas been used for both characterizing individual patients andin evaluating the efficacy of various rehabilitation programs.The Barthel index consists of 10 ratable patient skill items.Values are assigned to each item (0, 5, or 10 points) based onthe amount of physical assistance required to perform eachtask. A Barthel index total score ranges from 0 to 100 points (0,fully dependent; 100, fully independent). In the original ver-sion, each item is scored in three steps (38). The MBI, with afive-step scoring system, seems to have greater sensitivity andimproved reliability than the original version, without exam-ination difficulty or an increase in implementation time. Shahet al. (45) found the internal consistency reliability coefficientfor the MBI to be 0.90, compared with 0.87 for the originalindex. In another study, Kucukdeveci et al. (35) evaluated thereliability and validity of the MBI in 50 inpatient rehabilitationSCI patients in Turkey. Patients were assessed by the MBI atadmission and at discharge. Reliability was tested by usinginternal consistency, interobserver reliability, and the intra-class correlation coefficient. Construct validity was assessedby association with impairments (ASIA) and by Rasch analy-sis. The internal consistency coefficient was 0.88. The level ofagreement between two observers was sufficient (� statistic,0.5). The intraclass correlation coefficient was 0.77. However,Rasch analysis revealed that bladder and bowel items of theMBI misfit the construct. The authors concluded that adapta-tion of the MBI is useful in assessing SCI patients in Turkey,as long as its limitations are recognized.
The FIM was developed to provide uniform assessment ofthe severity of patient disability and medical rehabilitationoutcome (20–22, 24). It is an 18-item, seven-level scale de-signed to assess the severity of patient disability, estimateburden of care, and determine medical rehabilitation func-tional outcome. The FIM has emerged as a standard assess-ment instrument for use in rehabilitation programs for dis-abled persons (4, 20–22, 24, 25, 30–32, 41, 43, 51, 53). In 1993,Dodds et al. (25) assessed FIM with respect to validity andreliability in characterizing 11,102 general rehabilitation pa-tients in the Uniform Data System from the Pacific Northwest.They compared admission and discharge FIM scores andassessed for validation by using several hypotheses. The au-thors found high overall internal consistency and that FIMidentified significant functional gains in patients over time.FIM discriminated patients on the basis of age, comorbidity,and discharge destination. The authors concluded that FIMhad high internal consistency and adequate discriminativecapabilities, and it was a good indicator of burden of care.
Hamilton et al. (31, 32) assessed interobserver agreement ofthe seven-level FIM in two separate reports. In the 1991 report(32), two or more pairs of clinicians assessed each of 263patients undergoing inpatient medical rehabilitation at 21hospitals in the United States subscribing to the Uniform DataSystem for medical rehabilitation. Criteria were intraclass cor-relation coefficient (ICC) (analysis of variance) for total FIM,and FIM subscores 0.90 or higher (5 of 6 subscores must be�0.90; no ICC could be �0.75). Values of the � statistic (un-weighted) for individual FIM items must be 0.45 or higher for
S26 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
at least 15 of the 18 items. The total FIM ICC was 0.97.Subscore ICCs were: self-care, 0.96; sphincter control, 0.94;mobility (transfers), 0.96; locomotion, 0.93; communications,0.95; and social cognition, 0.94. The FIM item � mean was 0.71(range, 0.61–0.76). On the basis of their methods of analysis,the authors concluded that the seven-level FIM has goodclinical interobserver agreement (32). In 1994, Hamilton et al.(31) reported interobserver reliability among clinicians from89 rehabilitation facilities in the United States within theUniform Data System; 1018 patients were characterized. Byanalysis methods similar to those reported above, total FIMICC was 0.96. Subscore ICCs ranged from 0.89 to 0.96. FIMitem � coefficients ranged from 0.53 to 0.66. For a subset ofinstitutions that met Uniform Data System reliability criteria,values of the � statistic ranged from 0.69 to 0.84. The authorsconcluded that FIM is reliable when used by trained andtested inpatient medical rehabilitation clinicians. Stineman etal. (46) used FIM and a function-based strategy to generatefunctional outcome benchmarks among 3604 SCI patients.They found that most patients whose motor-FIM scores atadmission were above 30 were able to groom, dress the upperextremity, manage bladder function, use a wheelchair, andtransfer from bed to chair by the time of discharge fromrehabilitation. Most patients with scores above 52 attainedindependence in all but the most difficult FIM tasks, such asbathing, tub transfers, and stair climbing. The authors con-cluded that these “FIM item attainment benchmarks” may beuseful in counseling patients, predicting outcome, and antic-ipating patient care needs after discharge.
The QIF was developed in 1980 because the Barthel indexwas deemed to be too insensitive to document the small butsignificant functional gains made by quadriplegics (tetraple-gics) during medical rehabilitation (29, 53). The QIF is com-prised of variables that are each weighted and scored (trans-fers, grooming, bathing, feeding, dressing, wheelchairmobility, bed activities, bladder and bowel program, andunderstanding of personal care). A final score ranging from 0to 100 points is derived that characterizes each patient’s func-tional abilities and serves as a reference for future assessment.Gresham et al. (29) tested the QIF on a group of 30 completequadriplegic patients at admission to and discharge frominpatient medical rehabilitation. Resultant scores were com-pared with those simultaneously obtained by the Barthel in-dex and the Kenny self-care evaluation. The QIF was found tobe more sensitive for patient functional improvement (46%)than that defined by the Barthel index (20%) or the Kennyself-care evaluation (30%). The QIF was also tested for reli-ability. Ratings by three different nurses, working indepen-dently, were found to be significantly positively correlated forall subscores (P � 0.001). Gresham et al. (29) concluded thatthe QIF provides a useful option in choosing a functionalassessment instrument for quadriplegic patients.
Yavuz et al. (53) compared ASIA scores, the QIF, and theFIM in 29 subjects with cervical SCI. The same examiner usedall three scales at admission to and discharge from the reha-bilitation center. The authors identified strong correlation ofASIA scores to both the FIM and the QIF. Feeding and dress-ing categories of QIF showed an even stronger correlation to
ASIA motor scores; however, statistical significance was thesame for corresponding categories of FIM and QIF. The per-centage of recovery on ASIA motor scores was significantlycorrelated only to gain in QIF scores, not in FIM scores. Theauthors recommended that additions to the FIM may be use-ful to improve sensitivity, particularly in the feeding, dress-ing, and bed activity categories.
Catz et al. (14) developed a new disability scale specific forpatients with spinal cord pathology, SCIM, and compared itwith FIM in the assessment and characterization of 30 pa-tients. Two pairs of trained staff members recorded scores 1week after admission and every month thereafter during hos-pitalization. The authors found remarkable consistency be-tween each pair of observers for all tasks assessed (� coeffi-cient, 0.66–0.98). The authors found the SCIM to be moresensitive than FIM to changes in function of spinal cord lesionpatients: SCIM detected all functional changes detected byFIM, but FIM missed 26% of changes detected by SCIM scor-ing. The authors concluded that SCIM may be a useful instru-ment for assessing functional changes in patients with lesionsof the spinal cord.
The WISCI was proposed as a scale to measure functionallimitations in walking of patients after SCI (23). This scaleincorporates gradations of physical assistance and devicesrequired for walking after paralysis of the lower extremitiessecondary to SCI. The purpose of the WISCI is to documentchanges in functional capacity with respect to ambulation in arehabilitation setting. A pilot study of the WISCI was com-pleted using video clips of patients walking. Raters at eightinternational centers completed the assessment. The concor-dance for the pilot data was significant. Interobserver reliabil-ity revealed 100% agreement. The authors conclude that theWISCI scale showed good validity and reliability but neededfurther evaluation before it could serve as a useful tool forclinical studies (23).
Finally, the SCI-FAI is a functional observational gait-assessment instrument developed at the University of Miamithat addresses three key domains of walking function inindividuals with SCI: gait parameters/symmetry, assistivedevice use, and temporal-distance measures (27). The authorsassessed its validity and reliability in a study of 22 patientswith incomplete SCIs examined by four trained raters. Inter-observer reliability was good for all four raters (ICC range,0.850–0.960). A moderate correlation (Pearson correlation co-efficient, 0.58) was found between change in gait score andlower extremity strength. The authors concluded that theSCI-FAI is a reliable, valid, and relatively sensitive measure ofwalking ability in individuals with SCI.
SUMMARY
A variety of injury classification schemes have been used todescribe patients who have sustained SCIs. There are twogeneral types of assessment scales: neurological examinationscales and functional outcome scales. The most accurate andmeaningful description of SCIs, in the acute setting and infollow-up, seems to be accomplished by using a neurologicalscale in conjunction with a functional outcome scale. At
Clinical Assessment after Acute Cervical Spinal Cord Injury S27
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
present, the most used and studied neurological assessmentscales are the ASIA scores, including the motor index scores,sensory scores, and the ASIA impairment scale. After manyrevisions and several refinements, these scales are easy toapply and are reliable.
The 1996 ASIA recommendations for international stan-dards of neurological and functional classification of SCI in-clude the ASIA scales, as noted, and the FIM, which has beenstudied extensively as a functional outcome tool. It seems tobe the best functional outcome scale used to describe disabil-ity among SCI patients, both early and late after injury. It iseasy to administer and is valid and reliable. Interobserveragreement with the FIM has been high in several studies, withreported values of the � statistic of 0.53 to 0.76.
KEY ISSUES FOR FUTURE INVESTIGATION
Any future investigation of or clinical trial involving SCIpatients must include both a neurological examination scaleand a functional outcome assessment. Therapeutic trials ofSCI patients should include reliable neurological and func-tional scoring systems and should verify the validity andinterobserver reliability of those scoring scales as part of theinvestigational paradigm.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. American Spinal Injury Association. Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1984.
2. American Spinal Injury Association. Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1989.
3. American Spinal Injury Association/International Medical Soci-ety of Paraplegia: Standards for Neurological and Functional Classi-fication of Spinal Cord Injury, Revised 1992. Chicago, AmericanSpinal Injury Association, 1992.
4. American Spinal Injury Association/International Medical Soci-ety of Paraplegia: International Standards for Neurological and Func-tional Classification of Spinal Cord Injury, Revised 1996. Chicago,American Spinal Injury Association, 1996.
5. Bednarczyk JH, Sanderson DJ: Comparison of functional andmedical assessment in the classification of persons with spinalcord injury. J Rehabil Res Dev 30:405–411, 1993.
6. Benzel EC, Larson SJ: Functional recovery after decompressivespine operation for cervical spine fractures. Neurosurgery 20:742–746, 1987.
7. Botsford DJ, Esses SI: A new scale for the clinical assessment ofspinal cord function. Orthopedics 15:1309–1313, 1992.
8. Bracken MB, Collins WF, Freeman DF, Shepard MJ, Wagner FW,Silten RM, Hellenbrand KG, Ransohoff J, Hunt WE, Perot PL Jr,Grossman RG, Green BA, Eisenberg HM, Rifkinson N, GoodmanJH, Meagher JN, Fischer B, Clifton GL, Flamm ES, Rawe SE:Efficacy of methylprednisolone in acute spinal cord injury. JAMA251:45–52, 1984.
9. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC, MarshallLF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr,Wilberger JL, Winn HR, Young W: Methylprednisolone or nalox-one treatment after acute spinal cord injury: 1-year follow-updata—Results of the Second National Acute Spinal Cord InjuryStudy. J Neurosurg 76:23–31, 1992.
10. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized, controlled trial of meth-ylprednisolone or naloxone in the treatment of acute spinal-cordinjury: Results of the Second National Acute Spinal Cord InjuryStudy. N Engl J Med 322:1405–1411, 1990.
11. Bracken MB, Shepard MJ, Hellenbrand KG, Collins WF, Leo LS,Freeman DF, Wagner FC, Flamm ES, Eisenberg HM, GoodmanJH, Perot PL Jr, Green BA, Grossman RG, Meagher JN, Young W,Fischer B, Clifton GL, Hunt WE, Rifkinson N: Methylpred-nisolone and neurological function 1 year after spinal cord injury:Results of the National Acute Spinal Cord Injury Study.J Neurosurg 63:704–713, 1985.
12. Bracken MB, Webb SB Jr, Wagner FC: Classification of the sever-ity of acute spinal cord injury: Implications for management.Paraplegia 15:319–326, 1978.
13. Capaul M, Zollinger H, Satz N, Dietz V, Lehmann D, Schurch B:Analyses of 94 consecutive spinal cord injury patients using ASIAdefinition and modified Frankel score classification. Paraplegia32:583–587, 1994.
14. Catz A, Itzkovich M, Agranov E, Ring H, Tamir A: SCIM: SpinalCord Independence Measure—A new disability scale for patientswith spinal cord lesions. Spinal Cord 35:850–856, 1997.
15. Chehrazi B, Wagner FC Jr, Collins WF Jr, Freeman DH Jr: A scalefor evaluation of spinal cord injury. J Neurosurg 54:310–315, 1981.
16. Cheshire DJ: A classification of the functional end-results of injuryto the cervical spinal cord. Paraplegia 8:70–73, 1970.
17. Cohen ME, Ditunno JF Jr, Donovan WH, Maynard FW Jr: A test ofthe 1992 International Standards for Neurological and FunctionalClassification of Spinal Cord Injury. Spinal Cord 36:554–560, 1998.
18. Cohen ME, Sheehan TP, Herbison GJ: Content validity and reli-ability of the International Standards for Neurological Classifica-tion of Spinal Cord Injury. Top Spinal Cord Inj Rehabil 4:15–31,1996.
19. Davis LA, Warren SA, Reid DC, Oberle K, Saboe LA, Grace MG:Incomplete neural deficits in thoracolumbar and lumbar spinefractures: Reliability of Frankel and Sunnybrook scales. Spine18:257–263, 1993.
20. Ditunno JF Jr: Functional assessment measures in CNS trauma.J Neurotrauma 9[Suppl 1]:S301–S305, 1992.
21. Ditunno JF Jr: New spinal cord injury standards, 1992. Paraplegia30:90–91, 1992.
22. Ditunno JF Jr: American spinal injury standards for neurologicaland functional classification of spinal cord injury: Past, presentand future—1992 Heiner Sell Lecture of the American SpinalInjury Association. J Am Paraplegia Soc 17:7–11, 1994.
23. Ditunno JF Jr, Ditunno PL, Graziani V, Scivoletto G, Bernardi M,Castellano V, Marchetti M, Barbeau H, Frankel HL, D’AndreaGreve JM, Ko H-Y, Marshall R, Nance P: Walking Index for SpinalCord Injury (WISCI): An international multicenter validity andreliability study. Spinal Cord 38:234–243, 2000.
24. Ditunno JF Jr, Young W, Donovan WH, Creasey G: The Interna-tional Standards Booklet for Neurological and Functional Classi-fication of Spinal Cord Injury: American Spinal Injury Associa-tion. Paraplegia 32:70–80, 1994.
S28 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
25. Dodds TA, Martin DP, Stolov WC, Deyo RA: A validation of thefunctional independence measurement and its performanceamong rehabilitation inpatients. Arch Phys Med Rehabil 74:531–536, 1993.
26. El Masry WS, Tsubo M, Katoh S, El Miligui YH, Khan A: Valida-tion of the American Spinal Injury Association (ASIA) motorscore and the National Acute Spinal Cord Injury Study (NASCIS)motor score. Spine 21:614–619, 1996.
27. Field-Fote EC, Fluet GG, Schafer SD, Schneider EM, Smith R,Downey, Ruhl CD: The Spinal Cord Injury Functional Ambula-tion Inventory (SCI-FAI). J Rehabil Med 33:177–181, 2001.
28. Frankel HL, Hancock DO, Hyslop G, Melzak J, Michaelis LS,Ungar GH, Vernon JD, Walsh JJ: The value of postural reductionin the initial management of closed injuries of the spine withparaplegia and tetraplegia. Paraplegia 7:179–192, 1969.
29. Gresham GE, Labi ML, Dittmar SS, Hicks JT, Joyce SZ, StehlikMA: The Quadriplegia Index of Function (QIF): Sensitivity andreliability demonstrated in a study of thirty quadriplegic patients.Paraplegia 24:38–44, 1986.
30. Hamilton BB, Granger CV, Sherwin FS, Zielezny M, Tashman JS:A Uniform National Data System for Medical Rehabilitation, inFuhrer MJ (ed): Rehabilitation Outcomes: Analysis and Measurement.Baltimore, Paul H. Brookes Publishing Co., 1987, pp 137–147.
31. Hamilton BB, Laughlin JA, Fiedler RC, Granger CV: Interraterreliability of the 7-level Functional Independence Measure (FIM).Scand J Rehabil Med 26:115–119, 1994.
32. Hamilton BB, Laughlin JA, Granger CV, Kayton RM: Interrateragreement of the seven-level Functional Independence Measure(FIM). Arch Phys Med Rehabil 72:790, 1991.
33. Jonsson M, Tollback A, Gonzalez H, Borg J: Inter-rater reliabilityof the 1992 international standards for neurological and func-tional classification of incomplete spinal cord injury. Spinal Cord38:675–679, 2000.
34. Klose KJ, Green BA, Smith RS, Adkins RH, MacDonald AM: Universityof Miami Neuro-Spinal Index (UMNI): A quantitative method for de-termining spinal cord function. Paraplegia 18:331–336, 1980.
35. Kucukdeveci AA, Yavuzer G, Tennant A, Suldur N, Sonel B,Arasil T: Adaptation of the modified Barthel Index for use inphysical medicine and rehabilitation in Turkey. Scand J RehabilMed 32:87–92, 2000.
36. Lazar RB, Yarkony GM, Ortolano D, Heinemann AW, Perlow E,Lovell L, Meyer PR: Prediction of functional outcome by motorcapability after spinal cord injury. Arch Phys Med Rehabil 70:819–822, 1989.
37. Lucas JT, Ducker TB: Motor classification of spinal cord injurieswith mobility, morbidity and recovery indices. Am Surg 45:151–158, 1979.
38. Mahoney FI, Barthel DW: Functional evaluation: The BarthelIndex. Md State Med J 14:61–65, 1965.
39. Marino RJ, Huang M, Knight P, Herbison GJ, Ditunno JF Jr, SegalM: Assessing selfcare status in quadriplegia: Comparison of theQuadriplegia Index of Function (QIF) and the Functional Inde-pendence Measure (FIM). Paraplegia 31:225–233, 1993.
40. Marino RJ, Rider-Foster D, Maissel G, Ditunno JF: Superiority ofmotor level over single neurological level in categorizing tetra-plegia. Paraplegia 33:510–513, 1995.
41. Maynard FM Jr, Bracken MB, Creasey G, Ditunno JF Jr, DonovanWH, Ducker TB, Garber SL, Marino RJ, Stover SL, Tator CH, WatersRL, Wilberger JE, Young W: International Standards for Neurologi-cal and Functional Classification of Spinal Cord Injury: AmericanSpinal Injury Association. Spinal Cord 35:266–274, 1997.
42. Maynard FM Jr, Reynolds GG, Fountain S, Wilmot C, Hamilton R:Neurological prognosis after traumatic quadriplegia: Three-yearexperience of California Regional Spinal Cord Injury Care Sys-tem. J Neurosurg 50:611–616, 1979.
43. Ota T, Akaboshi K, Nagata M, Sonoda S, Domen K, Seki M, ChinoN: Functional assessment of patients with spinal cord injury:Measured by the motor score and the Functional IndependenceMeasure. Spinal Cord 34:531–535, 1995.
44. Priebe MM, Waring WP: The interobserver reliability of the re-vised American Spinal Injury Association standards for neurolog-ical classification of spinal injury patients. Am J Phys MedRehabil 70:268–270, 1991.
45. Shah S, Vanclay F, Cooper B: Improving the sensitivity of the BarthelIndex for stroke rehabilitation. J Clin Epidemiol 42:703–709, 1989.
46. Stineman MG, Marino RJ, Deutsch A, Granger CV, Maislin G: Afunctional strategy for classifying patients after traumatic spinalcord injury. Spinal Cord 37:717–725, 1999.
47. Tator CH, Rowed DW, Schwartz ML: Sunnybrook Cord InjuryScales for assessing neurological injury and neurological recov-ery, in Tator CH (ed): Early Management of Acute Spinal CordInjury. New York, Raven Press, 1982, pp 17–24.
48. Toh E, Arima T, Mochida J, Omata M, Matsui S: Functionalevaluation using motor scores after cervical spinal cord injuries.Spinal Cord 36:491–496, 1998.
49. Waters RL, Adkins RH, Yakura JS: Definition of complete spinalcord injury. Paraplegia 9:573–581, 1991.
50. Waters RL, Adkins RH, Yakura JS, Vigil D: Prediction of ambu-latory performance based on motor scores derived from stan-dards of the American Spinal Injury Association. Arch Phys MedRehabil 75:756–760, 1994.
51. Wells JD, Nicosia S: Scoring acute spinal cord injury: A study ofthe utility and limitations of five different grading systems. J Spi-nal Cord Med 18:33–41, 1995.
52. Yarkony GM, Roth E, Lovell L, Heinemann A, Katz RT, Wu Y:Rehabilitation outcomes in complete C5 quadriplegia. Am J PhysMed Rehabil 69:73–76, 1988.
53. Yavuz N, Tezyurek M, Akyuz M: A comparison of two functional testsin quadriplegia: The Quadriplegia Index of Function and the FunctionalIndependence Measure. Spinal Cord 36:832–837, 1998.
Clinical Assessment after Acute Cervical Spinal Cord Injury S29
CHAPTER 4
Radiographic Assessment of the Cervical Spine inAsymptomatic Trauma Patients
RECOMMENDATIONSSTANDARDS: Radiographic assessment of the cervical spine is not recommended in trauma patients who are
awake, alert, and not intoxicated, who are without neck pain or tenderness, and who do not havesignificant associated injuries that detract from their general evaluation.
RATIONALE
Spinal cord injury (SCI) is a potentially devastating con-sequence of acute trauma and can occur with improperimmobilization of an unstable cervical spine fracture.
Immobilization of an injured patient’s cervical spine aftertrauma is now standard care in most emergency medicalservices (EMS) systems. Immobilization of the cervical spineis maintained until spinal cord or spinal column injury isruled out by clinical assessment and/or radiographic survey.Radiographic study of the cervical spine of every traumapatient is costly and results in significant radiation exposureto a large number of patients, few of whom will have a spinalcolumn injury. The purpose of this review is to define whichradiographic studies are necessary in the assessment of thecervical spine in asymptomatic patients after trauma.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasperformed. MEDLINE subject headings and keywords “spi-nal cord injury,” “spinal fractures,” or “spinal injuries” re-sulted in 7994 matches. Combination with the keyword “cer-vical” resulted in 1844 matches. The search was limited tohuman studies and the English language, resulting in 1268articles. Combination with the keywords “clearance,” “diag-nosis,” or “radiography” yielded 184 matches. The titles andabstracts of these 184 articles were reviewed. All articles fo-cusing on clinical decision-making in the diagnosis of cervicalspine injury in adult patients with trauma were included.Additional references were culled from the reference lists ofthe remaining articles. Finally, members of the author groupwere asked to contribute articles known to them on the subjectmatter that were not found by other search means. The bib-liography developed by the EAST (Eastern Association for theSurgery of Trauma) practice parameter workgroup for cervicalspine clearance was reviewed (15), as was the reference listdeveloped by the NEXUS (National Emergency X-radiographyUtilization Study) group (5, 7).
Nine large, prospective cohort studies were identified;these studies provide Class I evidence. No randomized con-trolled trials in the literature addressed this issue. Manysmaller studies, case series, and retrospective cohort studieswere identified, which provide corroborating Class II andClass III evidence. This guideline was generated from thesearticles. The 13 articles most germane to this issue are sum-marized in Table 4.1.
SCIENTIFIC FOUNDATION
A missed cervical spine injury can result in devastatingneurological injury. For this reason, radiographic assessmentof the cervical spine is liberally used in patients after acutetrauma. Cervical spine radiographs are relatively inexpensiveand are easy to obtain. Computed tomography (CT) andmagnetic resonance imaging may also be used to evaluate theselected spine in certain circumstances. These studies aremore expensive but remain widely available. Because theoverall incidence of cervical spinal column injury in the gen-eral population of trauma patients is low, many patients areexposed to unnecessary ionizing radiation and may be immo-bilized unnecessarily, sometimes for long periods. For theseconcerns and others (e.g., financial, resource allocation, anduse), the issue of radiographic assessment of asymptomaticpatients after trauma has been raised. A number of investiga-tors have proposed that asymptomatic patients do not requireradiographic assessment of the cervical spine after trauma (2,5–7, 9, 14, 16). Asymptomatic patients after trauma are de-fined as those patients who meet all of the following criteria:
1. Are neurologically normal. These patients must have aGlasgow Coma Scale score of 15 and must not have any ofthe following: a) disorientation to person, place, or time; b)inability to remember three objects at 5 minutes; c) delayedor inappropriate response to external stimuli; or d) anyfocal motor or sensory deficit.
2. Are not intoxicated. Patients should be considered intoxi-cated if they have: a) a recent history of intoxication orintoxicating ingestion; b) evidence of intoxication on clini-cal examination; or c) laboratory evidence for the presence
S30 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
of drugs that alter the level of alertness, including bloodalcohol levels higher than 0.08 mg/dl.
3. Do not have neck pain or midline tenderness. Midlinetenderness is present if the patient complains of pain onpalpation of the posterior midline neck from the nuchalridge to the first thoracic vertebra.
4. Do not have an associated injury that is distracting to thepatient. Significant distracting injuries have been definedas: a) long bone fractures; b) visceral injuries requiringsurgical consultation; c) large lacerations, degloving, orcrush injuries; d) large burns; and e) any other injury thatmight impair the patient’s ability to participate in a generalphysical, mental, and neurological examination (5).
On the basis of these criteria, approximately one-third of traumapatients evaluated in emergency rooms or trauma centers areasymptomatic (range, 14–58%) (2, 5, 7, 9, 14, 16–18). Avoidance ofradiographic assessment in this patient population will result in adecrease in unnecessary radiation exposure, less patient time im-mobilized and confined in an uncomfortable position, and a signif-icant savings in both cost and resources (13).
Establishing a treatment standard for a therapeutic inter-vention requires the existence of at least one randomizedcontrolled study. However, a treatment standard for the use-fulness of a diagnostic test can be established with evidencederived from well-designed clinical studies that include a“diverse population using a ‘gold standard’ reference test in ablinded evaluation appropriate for the diagnostic applicationsand enabling the assessment of sensitivity, specificity, positiveand negative predictive values, and where applicable, likeli-hood ratios” (19). In evaluating the role of the radiographicassessment of asymptomatic trauma patients, we may con-sider the clinical examination to be a diagnostic test. X-rayimaging studies of the cervical spine may be considered thegold standard in this circumstance, because we are attemptingto ascertain whether the clinical examination can accuratelypredict the results of the radiographic assessment in a givenpopulation of patients. The population in question should berepresentative of the trauma population evaluated at anygiven emergency room or trauma center.
The literature reviewed included nine large cohort studiesthat included a representative trauma population, definedsymptomatic and asymptomatic patients by the criteria listedabove, and reported the incidence of spinal injury in thesegroups of patients as detected by subsequent radiographicassessment alone or by imaging of the cervical spine supple-mented by clinical follow-up (2, 5, 6, 9, 14, 16–18). All ninestudies were judged to provide Class I evidence, allowing theestablishment of a treatment standard. Many case series andretrospective cohort studies exist and provide corroboratingClass II and Class III evidence. These investigations are sum-marized in Table 4.1 and will be briefly discussed below.
The largest study addressing this issue encompassed 34,069patients evaluated at 21 emergency rooms across the United
States (5). All patients were studied with standard three-viewcervical radiography supplemented by CT, magnetic reso-nance imaging, or other studies as needed. Of 1818 patientsfound to have spinal injuries, 576 were considered to beclinically significant. Two patients of the 576 were prospec-tively assigned to the “asymptomatic” group. One patient hada probable injury at C2 that was not treated because thepatient refused treatment. Clinical follow-up of this patientrevealed no sequelae. The second patient had a laminar frac-ture of C6. He subsequently developed paresthesias in thearm and underwent surgery. Taking the worst-case scenario,and assuming that both of these patients were truly asymp-tomatic (the second patient later developed paresthesias), andthe injuries were truly significant (the first patient’s injurywas probably not significant given his subsequent clinicalcourse), the negative predictive value of an asymptomaticexamination was 99.9%. In contrast, the positive predictivevalue of a “symptomatic” examination was 1.9% (5).
Bayless and Ray (2), in 1989, studied a consecutive series of228 patients who had received “significant blunt head injury.”Patients were classified as symptomatic or asymptomatic atadmission to the hospital. All patients were observed for atleast 24 hours in the hospital and were assessed with at leasta three-view cervical spine x-ray series. A chart review wasperformed 2 years after admission, and any subsequent hos-pital visits were noted. Of the 228 patients, 211 were judged tohave adequate three-view cervical spine series. Of these, 122were judged asymptomatic and none had a significant spineinjury (2). Hoffman et al. (6) performed a prospective study of974 consecutive patients with blunt trauma evaluated at auniversity emergency room. All patients underwent at least athree-view cervical spine x-ray series supplemented with CT,oblique views, or flexion/extension views on the basis ofphysician judgment. Quality assurance logs, risk managementrecords, and hospital charts from subsequent admissionswere also reviewed. Of the 974 patients included in the study,353 were judged asymptomatic and none were identified tohave had a cervical spine injury (6).
Kreipke et al. (9) performed a prospective study involving 860consecutive acute trauma patients who arrived at a Level Itrauma center. All patients underwent five-view cervical radiog-raphy supplemented with CT and/or flexion/extension viewswhen required. Of these, 324 patients were judged asymptom-atic and none had a cervical spine injury detected on radio-graphic assessment (9). Neifeld et al. (14) prospectively studied886 trauma patients admitted at an urban emergency room. Allpatients were studied with a five-view cervical spine series. Of241 patients who were asymptomatic, none had a spine injurydetected radiographically.
Roberge et al. (16), in 1988, reported the results of a prospec-tive study involving all patients who received a five-view cervi-cal spine series while in an urban emergency room. Of 467patients studied, 155 were judged to be asymptomatic, and nonewere found to have a spinal injury. Ross et al. (17) prospectively
Radiographic Spinal Assessment in Asymptomatic Trauma Patients
S31Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE4.
1.Su
mm
ary
ofR
epor
tson
Rad
iogr
aphi
cSp
inal
Ass
essm
ent
ofA
sym
ptom
atic
Trau
ma
Pati
ents
a
Seri
es
(Ref
.N
o.)
Des
crip
tion
ofSt
udy
Evid
ence
Cla
ssC
oncl
usio
ns
Hof
fman
etal
.,20
00(5
)Pr
ospe
ctiv
est
udy
of34
,069
patie
nts.
4309
asym
ptom
atic
.
2ha
d“c
linic
ally
sign
ifica
ntin
juri
es”.
All
patie
nts
radi
ogra
phed
.
Neg
ativ
epr
edic
tive
valu
e:99
.9%
.
Posi
tive
pred
ictiv
eva
lue:
1.9%
.
Not
e:1
of2
“mis
sed
inju
ries
”di
dno
tre
ally
have
a“s
igni
fican
tin
jury
,”as
hew
as
untr
eate
dan
dha
dno
sequ
elae
with
clin
ical
follo
w-u
p.Th
eot
her
patie
ntde
velo
ped
pare
sthe
sias
inhi
sar
man
dw
asfo
und
toha
vea
lam
inar
frac
ture
ofC
6.
IX
-ray
sno
tne
cess
ary
inas
ympt
omat
icpa
tient
s.
Gon
zale
set
al.,
1999
(4)
2176
patie
nts
pros
pect
ivel
yst
udie
dw
ithsc
reen
ing
exam
inat
ion
and
x-ra
ys.
1in
jury
was
dete
cted
bypl
ain
x-ra
ysin
anot
herw
ise
asym
ptom
atic
patie
nt;
how
ever
,pl
ain
x-ra
ys
mis
sed
13in
juri
esov
eral
l.
IPl
ain
film
radi
ogra
phy
does
not
impr
ove
sens
itivi
ty
(com
pare
dw
ithth
eph
ysic
alex
amin
atio
n)fo
rth
ede
tect
ion
ofce
rvic
alsp
ine
inju
ryin
asym
ptom
atic
patie
nts.
Rot
het
al.,
1994
(18)
Pros
pect
ive
stud
yof
682
patie
nts
adm
itted
toem
erge
ncy
depa
rtm
ent
with
trau
ma.
96w
ere
asym
ptom
atic
,no
neha
din
jury
.
Ove
rall
inci
denc
eof
inju
ryw
as2%
.
All
patie
nts
radi
ogra
phed
.
Follo
w-u
pcl
inic
alvi
sit
betw
een
30an
d15
0d
post
inju
ry,
achi
eved
in43
%.
Neg
ativ
epr
edic
tive
valu
eof
asym
ptom
atic
exam
inat
ion:
100%
.
Posi
tive
pred
ictiv
eva
lue
ofsy
mpt
omat
icex
amin
atio
n:2.
7%.
IX
-ray
slik
ely
not
nece
ssar
yin
asym
ptom
atic
patie
nts.
Lind
sey
etal
.,19
93(1
0)16
86pa
tient
sst
udie
dre
tros
pect
ivel
y,59
7pa
tient
sst
udie
dpr
ospe
ctiv
ely.
Ato
tal
of49
patie
nts
with
cerv
ical
spin
ein
juri
esw
ere
iden
tifie
d(o
vera
llin
cide
nce
2.1%
).
No
patie
ntw
ithan
inju
ryw
asas
ympt
omat
ic.
III
The
tota
lnu
mbe
rsof
sym
ptom
atic
and
asym
ptom
atic
patie
nts
are
not
repo
rted
,pr
eclu
ding
the
calc
ulat
ion
ofne
gativ
eor
posi
tive
pred
ictiv
eva
lues
.
Asy
mpt
omat
icpa
tient
sdo
not
requ
ire
radi
ogra
phic
imag
es.
Hof
fman
etal
.,19
92(6
)97
4bl
unt
trau
ma
patie
nts
pros
pect
ivel
yst
udie
d.
Ove
rall
inci
denc
eof
cerv
ical
spin
ein
jury
was
2.8%
.
Of
353
aler
t,as
ympt
omat
icpa
tient
s,no
neha
da
sign
ifica
ntsp
ine
inju
ry.
Follo
w-u
p:ra
diog
raph
sne
gativ
ein
all
353.
Cha
rts,
qual
ityas
sura
nce
logs
,an
dri
skm
anag
emen
tre
cord
sre
view
edw
ith3-
mo
follo
w-u
p.
Neg
ativ
epr
edic
tive
valu
eof
asym
ptom
atic
exam
inat
ion:
100%
.
Posi
tive
pred
ictiv
eva
lue
ofsy
mpt
omat
icex
amin
atio
n:4.
5%.
IA
sym
ptom
atic
patie
nts
dono
tre
quir
ece
rvic
alsp
ine
x-ra
ys.
Ros
set
al.,
1992
(17)
Pros
pect
ive
stud
yof
410
patie
nts
seen
attr
aum
ace
nter
.
196
patie
nts
had
asym
ptom
atic
exam
inat
ion,
none
had
inju
ry.
All
patie
nts
stud
ied
with
plai
nx-
rays
,C
Tus
edas
nece
ssar
y.
Neg
ativ
epr
edic
tive
valu
e:10
0%.
Posi
tive
pred
ictiv
eva
lue:
6.1%
.
IR
adio
grap
hyno
tm
anda
tory
for
asym
ptom
atic
patie
nts.
Mai
npo
int
ofpa
per
was
that
mec
hani
smof
inju
ryis
not
a
valu
able
pred
icto
rof
inju
ry.
McN
amar
aet
al.,
1990
(12)
Ret
rosp
ectiv
ere
view
of28
6pa
tient
sju
dged
tobe
“hig
hri
sk”
bym
echa
nism
ofin
jury
.
178
wer
eas
ympt
omat
ic,
none
had
cerv
ical
spin
ein
jury
.
108
wer
esy
mpt
omat
ic,
5ha
dce
rvic
alsp
ine
inju
ry.
Cha
rtfo
llow
-up
perf
orm
edto
dete
rmin
ein
cide
nce
ofin
jury
.
Neg
ativ
epr
edic
tive
valu
efo
ras
ympt
omat
icex
amin
atio
n:10
0%.
Posi
tive
pred
ictiv
eva
lue
for
sym
ptom
atic
exam
inat
ion:
4.9%
.
III
Man
ypa
tient
sex
clud
edgi
ving
topo
or
docu
men
tatio
n,se
lect
popu
latio
nfo
llow
-up
inad
equa
te(fi
lms
not
done
onev
eryo
ne,
node
laye
d
char
tre
view
).
Cer
vica
lsp
ine
x-ra
ysno
tne
cess
ary
inas
ympt
omat
ic
patie
nts.
S32 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE4.
1.C
onti
nued
Seri
es
(Ref
.N
o.)
Des
crip
tion
ofSt
udy
Evid
ence
Cla
ssC
oncl
usio
ns
Bay
less
and
Ray
,19
89(2
)Se
ries
of22
8pa
tient
s,21
1w
ithco
mpl
ete
stud
ies.
Ove
rall
inci
denc
eof
sign
ifica
ntsp
inal
inju
ryw
as1.
7%.
Of
122
aler
t,as
ympt
omat
icpa
tient
s,no
neha
da
sign
ifica
ntin
jury
.
Follo
w-u
p:x-
rays
nega
tive
inal
l12
2.
Cha
rts
revi
ewed
for
any
subs
eque
ntre
fera
ble
visi
tsw
ithin
2yr
.
Neg
ativ
epr
edic
tive
valu
eof
asym
ptom
atic
exam
inat
ion:
100%
.
Posi
tive
pred
ictiv
eva
lue
ofsy
mpt
omat
icex
amin
atio
n:3%
.
IA
sym
ptom
atic
patie
nts
dono
tre
quir
ece
rvic
alsp
ine
x-ra
ys.
Kre
ipke
etal
.,19
89(9
)Pr
ospe
ctiv
est
udy
of86
0pa
tient
spr
esen
ting
totr
aum
ace
nter
.
324
asym
ptom
atic
,no
neha
din
jury
.
All
patie
nts
radi
ogra
phed
.
Neg
ativ
epr
edic
tive
valu
eof
asym
ptom
atic
exam
inat
ion:
100%
.
Posi
tive
pred
ictiv
eva
lue
ofsy
mpt
omat
icex
amin
atio
n:4%
.
IX
-ray
sno
tne
cess
ary
inas
ympt
omat
icpa
tient
s.
Mir
vis
etal
.,19
89(1
3)40
8pa
tient
sst
udie
dw
ithst
anda
rdx-
rays
and
CT.
Tota
lpo
pula
tion
seen
was
4135
patie
nts.
241
patie
nts
unde
rwen
tC
Tbe
caus
eof
“sus
pici
ous”
x-ra
ys,
failu
reto
visu
aliz
eex
trem
esof
C-s
pine
,or
for
conf
irm
atio
nof
know
nfr
actu
re.
Of
thes
e24
1,13
8pa
tient
sw
ere
clin
ical
ly
asym
ptom
atic
.
CT
serv
edas
“gol
dst
anda
rd.”
Non
eof
138
patie
nts
had
acl
inic
ally
rele
vant
inju
ry(a
lthou
gh1
had
ano
ndis
plac
edC
7
tran
sver
sepr
oces
sfr
actu
rew
hich
was
trea
ted
with
aco
llar)
.
Neg
ativ
epr
edic
tive
valu
eof
asym
ptom
atic
exam
inat
ion:
99.3
–100
%.
Posi
tive
pred
ictiv
eva
lue
ofsy
mpt
omat
icex
amin
atio
n:12
.6%
.
II
Sele
ctpo
pula
tion
gold
stan
dard
may
befa
lse
end
poin
t
Clin
ical
lyre
leva
ntce
rvic
alsp
ine
inju
ryis
extr
emel
y
unco
mm
onin
asym
ptom
atic
patie
nts.
X-r
ays
may
be
unne
cess
ary.
Nei
feld
etal
.,19
88(1
4)Pr
ospe
ctiv
est
udy
of88
6pa
tient
s24
4as
ympt
omat
icpa
tient
s,no
neha
din
jury
.
All
patie
nts
radi
ogra
phed
.
Neg
ativ
epr
edic
tive
valu
e:10
0%.
Posi
tive
pred
ictiv
eva
lue:
6.2%
.
IA
sym
ptom
atic
patie
nts
dono
tre
quir
ex-
rays
.
Rob
erge
etal
.,19
88(1
6)Pr
ospe
ctiv
est
udy
invo
lvin
g46
7tr
aum
apa
tient
s.
155
asym
ptom
atic
patie
nts
wer
eas
ympt
omat
ic,
none
had
asp
ine
inju
ry.
312
wer
esy
mpt
omat
ic,
8ha
dsp
ine
inju
ries
.
All
patie
nts
“sch
edul
edto
follo
w-u
p”in
surg
ery
clin
ic,
auth
ors
stat
eth
atno
mis
sed
inju
ries
have
been
iden
tifie
d.
Neg
ativ
epr
edic
tive
valu
eof
asym
ptom
atic
exam
inat
ion:
100%
.
Posi
tive
pred
ictiv
eva
lue
ofsy
mpt
omat
icex
amin
atio
n:2.
5%.
IA
sym
ptom
atic
patie
nts
dono
tre
quir
ex-
rays
.
Bac
hulis
etal
.,19
87(1
)18
23of
4941
trau
ma
patie
nts
stud
ied
with
plai
nx-
rays
.
94pa
tient
sfo
und
toha
vein
juri
es.
All
wer
esy
mpt
omat
ic.
No
asym
ptom
atic
patie
ntha
da
radi
ogra
phic
ally
dete
ctab
lein
jury
.
IIIA
sym
ptom
atic
patie
nts
dono
tre
quir
ex-
rays
.
aC
T,co
mpu
ted
tom
ogra
phy;
C-s
pine
,ce
rvic
alsp
ine.
Radiographic Spinal Assessment in Asymptomatic Trauma Patients S33
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
studied 410 trauma patients admitted to a trauma center in 1992.All patients underwent a three-view cervical spine series sup-plemented as needed with CT, flexion/extension views, fluoros-copy, and radionucleotide bone scans. Of 196 patients judged tobe asymptomatic, none had a cervical spine injury diagnosedwith these imaging modalities. Roth et al. (18), in 1994, prospec-tively studied 682 patients evaluated at a military hospital afterblunt trauma. All patients underwent a three-view cervical spinex-ray assessment. A subsequent chart review revealed no missedinjuries (the hospital was the only military hospital within aradius of 2500 miles), and 45% of patients were successfullycontacted 30 to 150 days after the initial evaluation for additionalclinical follow-up. Of 96 asymptomatic patients, none had acervical spine injury. Recently, Gonzales et al. (4) studied a seriesof 2176 patients evaluated in an emergency room after trauma;1768 were judged to be asymptomatic, but the criteria used wereslightly different from those described above. Three injurieswere later detected in this group of 1768 patients, but two ofthese patients were not truly asymptomatic (one had a sternalfracture, multiple rib fractures, and a splenic hematoma; theother had significant facial fractures), and the third patient’sinjuries were radiographically occult. The third patient’s injurieswere detected by CT and were treated with a collar.
In addition to these studies that provide Class I evidence,other studies have been reported that provide corroboratingClass II and Class III evidence germane to this issue (1, 10, 12,13). For example, Mirvis et al. (13) studied 241 patients with ahistory of blunt trauma who were assessed with cervical spinex-rays supplemented by CT. Aside from a single nondis-placed transverse process fracture of C7 (which was not seenon conventional x-rays), none of the 138 patients deemed to beasymptomatic had a significant spinal injury. McNamara et al.(12) performed a retrospective review of 286 trauma patientsevaluated in an urban emergency room. Of 178 patients char-acterized as asymptomatic, none had a spinal injury detectedwith subsequent radiographic assessment. Bachulis et al. (1)surveyed a prospectively acquired database of 4941 consecu-tive patients evaluated after blunt traumatic injury. Of 1823patients who underwent radiographic assessment of the cer-vical spine, 94 were found to have a spinal injury. All patientswith spinal injuries were symptomatic. No asymptomatic pa-tient had a spinal injury. Lindsey et al. (10) reviewed 2283consecutive trauma patients and determined that no patientwith a spinal injury was asymptomatic.
Case reports exist describing asymptomatic patients whohave harbored potentially unstable spinal injuries (11, 20). Forexample, Woodring and Lee (20) reviewed 216 patients whohad cervical spine injuries diagnosed with CT. They reportedthat 11 of these 216 patients were not reported to be symp-tomatic in the medical record. It is unclear why these 11patients were subjected to computed tomographic evaluationof the cervical spine if they were asymptomatic. These authorsalso reported a 61% false-negative rate with the use of plainx-rays in this population. Woodring and Lee encouraged theliberal use of CT based on the mechanism of injury. Noobjective definitive conclusion can be drawn from this retro-spective case series of a very select patient population. One
must question the usefulness of radiographic assessment ofany patient given a 61% false-negative rate.
Other authors have refuted the contention that “mechanismof injury” is a reliable predictor of cervical spine injury (8, 17).Mace (11) reported the case of a 51-year-old man who wasawake and alert, was neurologically intact, and had no com-plaints of neck pain or other associated distracting injury, butwas found to have an unstable fracture of C2. It is importantto note, however, that the patient had no history of trauma butwas being evaluated for a sore throat. Cervical spine x-rayswere obtained to rule out a peritonsillar abscess. From thesereports, it is clear that potentially unstable spinal injuries mayexist in asymptomatic patients (even those presenting withsore throats). However, these injuries are so rare that they donot appear in even the largest population-based studies. Theexperience of Davis et al. (3) is illustrative. They described theetiology of 34 missed cervical spine injuries in a series of32,117 trauma patients evaluated at a group of emergencyrooms servicing San Diego County. No missed injury oc-curred in an asymptomatic patient in their study.
SUMMARY
Clinical investigations that provide Class I evidence involv-ing nearly 40,000 patients, plus Class II and III evidencestudies involving more than 5000 patients, convincingly dem-onstrate that asymptomatic patients do not require radio-graphic assessment of the cervical spine after trauma. Thecombined negative predictive value of cervical spine x-rayassessment of asymptomatic patients for a significant cervicalspine injury is virtually 100% (2, 4–6, 9, 14, 16–18).
In contrast, the reported incidence of cervical spine injuriesin the symptomatic patient ranged from 1.9 to 6.2% in theseClass I evidence studies. Symptomatic patients require radio-graphic study to rule out the presence of a traumatic cervicalspine injury before the cervical spine immobilization is dis-continued (2, 4–6, 9, 14, 16–18). The type and extent of radio-graphic assessment of symptomatic patients after trauma isreviewed in Chapter 5.
KEY ISSUES FOR FUTURE INVESTIGATION
None.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Bachulis BL, Long WB, Hynes GD, Johnson MC: Clinical indica-tions for cervical spine radiographs in the traumatized patient.Am J Surg 153:473–478, 1987.
2. Bayless P, Ray VG: Incidence of cervical spine injuries in associ-ation with blunt head trauma. Am J Emerg Med 7:139–142, 1989.
3. Davis JW, Phreaner DL, Hoyt DB, Mackersie RC: The etiology ofmissed cervical spine injuries. J Trauma 34:342–346, 1993.
4. Gonzales RP, Fried PO, Bukhalo M, Holevar MR, Falimirski ME:Role of clinical examination in screening for blunt cervical spineinjury. J Am Coll Surg 189:152–157, 1999.
S34 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
5. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI:Validity of a set of clinical criteria to rule out injury to the cervicalspine in patients with blunt trauma: National EmergencyX-Radiography Utilization Study Group. N Engl J Med 343:94–99, 2000.
6. Hoffman JR, Schriger DL, Mower W, Luo JS, Zucker M: Low-riskcriteria for cervical spine radiography in blunt trauma: A pro-spective study. Ann Emerg Med 21:1454–1460, 1992.
7. Hoffman JR, Wolfson AB, Todd K, Mower WR: Selective cervicalspine radiography in blunt trauma: Methodology of the NationalEmergency X-Radiography Utilization Study (NEXUS). AnnEmerg Med 32:461–469, 1998.
8. Jacobs LM, Schwartz R: Prospective analysis of acute cervicalspine injury: A methodology to predict injury. Ann Emerg Med15:44–49, 1986.
9. Kreipke DL, Gillespie KR, McCarthy MC, Mail JT, Lappas JC,Broadie TA: Reliability of indications for cervical spine films intrauma patients. J Trauma 29:1438–1439, 1989.
10. Lindsey RW, Diliberti TC, Doherty BJ, Watson AB: Efficacy ofradiographic evaluation of the cervical spine in emergency situ-ations. South Med J 86:1253–1255, 1993.
11. Mace SE: Unstable occult cervical spine fracture. Ann Emerg Med20:1373–1375, 1991.
12. McNamara RM, Heine E, Esposito B: Cervical spine injury andradiography in alert, high risk patients. J Emerg Med 8:177–182, 1990.
13. Mirvis SE, Diaconis JN, Chirico PA, Reiner BI, Joslyn JN, MilitelloP: Protocol driven radiologic evaluation of suspected cervicalspine injury: Efficacy study. Radiology 170:831–834, 1989.
14. Neifeld GL, Keene JG, Hevesy G, Leikin J, Proust A, Thisted RA:Cervical injury in head trauma. J Emerg Med 6:203–207, 1988.
15. Pasquale M, Fabian TC: Practice Management Guidelines forTrauma from the Eastern Association for the Surgery of Trauma.J Trauma 44:945–957, 1998.
16. Roberge RJ, Wears RC, Kelly M, Evans TC, Kenny MA, DaffnerRD, Kremen R, Murray K, Cottington EC: Selective application ofcervical spine radiography in alert victims of blunt trauma: Aprospective study. J Trauma 28:784–788, 1988.
17. Ross SE, O’Malley KF, DeLong WG, Born CT, Schwab CW: Clin-ical predictors of unstable cervical spinal injury in multiply in-jured patients. Injury 23:317–319, 1992.
18. Roth BJ, Martin RR, Foley K, Barcia PJ, Kennedy P: Roentgeno-graphic evaluation of the cervical spine: A selective approach.Arch Surg 129:643–645, 1994.
19. Walters BC: Clinical practice parameter development in neuro-surgery, in Bean JR (ed): Neurosurgery in Transition: The Socioeco-nomic Transformation of Neurological Surgery. Baltimore, Williams& Wilkins, 1998, pp 99–111.
20. Woodring JH, Lee C: Limitations of cervical radiography in theevaluation of acute cervical trauma. J Trauma 34:32–39, 1993.
Frontispiece and plate from La Anatomia del corpo umano composta da M. Giovanni Valverde, Nuouamente Ristampata.Venice, Stamperia de Givnti, 1586. Courtesy, Rare Book Room, Norris Medical Library, Keck School of Medicine, Universityof Southern California, Los Angeles, California.
Radiographic Spinal Assessment in Asymptomatic Trauma Patients S35
CHAPTER 5
Radiographic Assessment of the Cervical Spine in SymptomaticTrauma Patients
RECOMMENDATIONSSTANDARDS: A three-view cervical spine series (anteroposterior, lateral, and odontoid views) is recom-mended for radiographic evaluation of the cervical spine in patients who are symptomatic after traumaticinjury. This should be supplemented with computed tomography (CT) to further define areas that aresuspicious or not well visualized on the plain cervical x-rays.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• It is recommended that cervical spine immobilization in awake patients with neck pain or tenderness and
normal cervical spine x-rays (including supplemental CT as necessary) be discontinued after either a) normaland adequate dynamic flexion/extension radiographs, or b) a normal magnetic resonance imaging study isobtained within 48 hours of injury.
• Cervical spine immobilization in obtunded patients with normal cervical spine x-rays (including supple-mental CT as necessary) may be discontinued a) after dynamic flexion/extension studies performed underfluoroscopic guidance, or b) after a normal magnetic resonance imaging study is obtained within 48 hoursof injury, or c) at the discretion of the treating physician.
RATIONALE
Trauma patients who are symptomatic, that is, complain ofneck pain, have cervical spine tenderness, or have symp-toms or signs of a neurological deficit associated with the
cervical spine, and trauma patients who cannot be assessed forsymptoms or signs (those who are unconscious, uncooperative,incoherent, or intoxicated, or who have associated traumaticinjuries that distract from their assessment) require radiographicstudy of the cervical spine before cervical spine immobilizationis discontinued. Many authors have proposed strategies andimaging techniques to accomplish x-ray clearance of the cervicalspine after trauma, particularly in the symptomatic or obtundedpatient. One-, three-, and five-view static cervical spine x-rays,computed tomography (CT), magnetic resonance imaging(MRI), bone scans, flexion/extension radiographs, dynamic flu-oroscopy with or without somatosensory evoked potential mon-itoring, and other studies have all been described as useful fordetermining spinal injury and potential spinal instability aftertraumatic injury (1–9, 11–17, 19–24, 26–28, 30–39, 41–43, 45–54,56, 57, 59–73). The purpose of this review is to determine theoptimal radiographic assessment strategy necessary and suffi-cient to exclude a significant cervical spine injury in the symp-tomatic trauma patient.
SEARCH CRITERIA
A computerized search of the database of the National Libraryof Medicine of the literature published from 1966 to 2001 wasperformed. MEDLINE medical subject headings and keywords
“spinal cord injury,” “spinal fractures,” or “spinal injuries” re-sulted in 7994 matches. Combination with the keyword “ cervi-cal” resulted in 1844 matches. These references were limited tohuman studies and the English language, resulting in 1268 arti-cles. Combination with the keywords “clearance,” “diagnosis,”or “radiography” yielded 184 matches. The titles and abstracts ofthese 184 articles were reviewed. All articles focusing on clinicaldecision-making in diagnosing cervical spine injuries in adultpatients with trauma injuries were included. Additional refer-ences were culled from the reference lists of the remaining arti-cles. The members of the author group were asked to contributearticles known to them on the subject matter that were not foundby other search means. The practice parameters and referencelist developed by the EAST (Eastern Association for the Surgeryof Trauma) (56) practice parameter workgroup for cervical spineclearance was reviewed, as was the reference list developed bythe NEXUS (National Emergency X-radiography UtilizationStudy) group (31, 33). A total of 73 references form the basis forthis guideline.
Twenty-one manuscripts were identified that provided ev-idence germane to the topic of this guideline. Four studiesprovided Class I evidence, six provided Class II evidence, and11 were individual case series and provided Class III evi-dence. These 21 manuscripts are summarized in Table 5.1.
SCIENTIFIC FOUNDATION
Patients who are asymptomatic with respect to a potentialcervical spinal injury after acute trauma do not require radio-
S36 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
graphic assessment to rule out a significant injury to thecervical spine (see Chapter 4). Radiographic studies do notincrease the sensitivity or specificity of the clinical examina-tion in this specifically defined population of patients (31, 33).There is, however, a 2 to 6% incidence of significant cervicalspine injury in the symptomatic patient population after acutetrauma (4, 31–33, 42, 53, 61, 62). These patients require radio-graphic assessment to exclude cervical spinal injury before thediscontinuation of cervical spine immobilization. The mostsignificant consequence of premature discontinuation of cer-vical spine immobilization is neurological injury. Prolongedimmobilization, however, is associated with morbidity aswell. Decubitus ulcers, increased cerebrospinal fluid pressure,pain, and pulmonary complications have all been describedwith prolonged immobilization of the cervical spine (18, 44,58). For these reasons, a diagnostic algorithm that is highlysensitive and specific for the occurrence of a significant cer-vical spine injury and that can be applied in an expeditiousfashion is desired.
The single most common cause of missed cervical spineinjury seems to be failure to adequately visualize the region ofinjury. This can be caused by failure to obtain radiographs, orby making judgments on technically suboptimal films. Thisoccurs most commonly at the extremes of the cervical spine,the occiput to C2 and at the C7–T1 levels (17, 25, 59). Davis etal. (17) described 32,117 acute trauma patients. Cervical spineinjuries were missed in 34 symptomatic patients; 23 of these34 symptomatic patients either did not have radiographs orhad inadequate radiographs that did not include the region ofinjury. Eight patients had adequate x-ray studies that weremisread by the treating physician. Only one patient had amissed injury that was undetectable on technically adequatefilms, even after retrospective review. The error in two pa-tients with missed injuries was not described. The reviews byDavis et al. (17) and other investigators (1, 6, 9, 16, 24, 43, 47)confirm that it is uncommon to miss cervical spine injurieswith adequate plain radiographic assessment of the occiputthrough T1.
The most prevalent initial x-ray assessment of the symp-tomatic or obtunded patient is the three-view cervical spineseries. When adequate visualization of the entire cervicalspine is achieved from occiput to T1, the negative predictivevalue of a normal three-view cervical spine series has beenreported to range from 93 to 98% in several Class I studies (1,6, 47), and from 85 to 100% in Class II and III studies (9, 16, 24,43). Although the negative predictive value of the three-viewcervical spine x-ray series is quite high, the sensitivity of thethree-view series is less impressive. The same Class I seriesreferenced above report sensitivity rates for the three-viewcervical spine series of 84, 62.5, and 83%, respectively (1, 6,47). In the best-case clinical scenario, assuming the highestvalues for negative predictive value and sensitivity, approxi-mately 98% of patients with a normal three-view cervicalspine x-ray series will have a truly normal cervical spine.
These same data suggest that the three-view cervical spineseries will also be normal in 15 to 17% of patients who havecervical spine injuries. If we assume a 6% incidence of spinalinjury in a high-risk population (the head-injured multi-trauma patient, for example), then an adequate three-viewcervical spine series alone would be expected to correctlyidentify 5 of 6 spinal injuries in a group of 100 patients, andcorrectly identify 94 of 94 patients without a spinal injury.One patient of the 100 with an injured spine would havecervical radiographs interpreted as normal. The addition ofoblique views (for a five-view series) does not seem to in-crease the overall sensitivity of the examination (24). Obliqueviews may be useful in lieu of a swimmer’s view to visualizeC7–T1 (36). Holliman et al. (34) have questioned the useful-ness of the anteroposterior cervical view, and they argue thatit is not an important addition to the assessment of the acutetrauma patient. The data presented by these authors are ClassIII evidence and have not been verified by others. Severalreports confirm that the lateral x-ray view alone will miss asubstantial portion of cervical spine injuries depicted in athree-view series (14, 26, 65).
To increase the sensitivity of the radiographic assessment ofthe cervical spine in trauma patients, many authors havedescribed experiences with CT and MRI in the acute setting.Several have reported greater sensitivity by using CT to viewareas not well visualized on plain films, typically the cranio-cervical and cervicothoracic junctions, or areas identified assuspicious on plain cervical spine x-rays (6, 9, 24, 48, 67, 68).In a small Class I study of 58 patients, Berne et al. (6) reportedthat helical CT of the entire cervical spine identified all clini-cally significant injuries in a series of patients assessed withplain films, CT, and MRI who were followed clinically forsubsequent events. Two injuries were missed; however, nei-ther required any treatment. The authors report a negativepredictive value of 95% for CT for all spinal injuries and anegative predictive value of 100% for unstable injuries. Otherauthors report 100% sensitivity for the detection of injurieswith CT limited to areas poorly visualized or identified assuspicious on plain films (24, 48, 67, 68). However, all studiescited provide Class II and III evidence, and most were im-paired by a common flaw: they treat CT as the “gold stan-dard” for the detection of injury. Although they suggest thatthe addition of CT increases diagnostic sensitivity, the use ofCT data as the gold standard represents a false end point forthe true variable of clinically relevant spinal injury.
Although the incidence of significant spinal injury with anormal cervical spine series supplemented with CT is ex-tremely low, missed injuries have been reported. Brohi andWilson-Macdonald (11) reported a missed C6–C7 facet dislo-cation in a patient with persistent neck pain who was studiedwith plain films and a CT occiput through C7–T1. Sweeney etal. (66) reported an autopsy series of three patients who diedof traumatic injuries and were found to have spinal injuriesundetected by plain films supplemented with CT through the
Radiographic Spinal Assessment in Symptomatic Trauma Patients
S37Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE5.
1.Su
mm
ary
ofR
epor
tson
Rad
iogr
aphi
cSp
inal
Ass
essm
ent
ofSy
mpt
omat
icTr
aum
aPa
tien
tsa
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Ban
itet
al.,
2000
(3)
Com
bine
dre
tros
pect
ive/
pros
pect
ive
stud
y.
4460
patie
nts
eval
uate
d.
2217
thou
ght
tore
quir
ex-
rays
.
6-m
ocl
inic
alfo
llow
-up
and
subs
eque
ntC
T/M
RI
used
as“g
old
stan
dard
”
for
plai
nx-
rays
(aut
hors
clai
mno
mis
sed
inju
ries
,cr
edib
lecl
aim
).
5-vi
ewse
ries
used
inal
lpa
tient
s.
IIIIn
sym
ptom
atic
patie
nts,
sens
itivi
tyof
plai
nx-
rays
was
84%
to88
%.
Inst
itutio
nof
acl
inic
ally
base
dim
agin
gpr
otoc
olre
sulte
din
ade
crea
sein
the
rate
ofm
isse
din
juri
esfr
om4%
to0%
.
Prot
ocol
had
sens
itivi
tyof
100%
and
incl
uded
use
ofde
laye
dex
amin
atio
nof
patie
nts
with
tend
erne
ss/p
ain
with
flexi
on/e
xten
sion
x-ra
ys(fa
lse-
posi
tives
not
give
n).
Ber
neet
al.,
1999
(6)
Pros
pect
ive
stud
yof
sele
ctpo
pula
tion
ofpa
tient
s(u
neva
luab
le,
mul
titra
uma,
havi
ngC
Tdo
nefo
ran
othe
rre
ason
).
58pa
tient
s,al
lun
derw
ent
3-vi
ewx-
ray
seri
esfo
llow
edby
helic
alC
Tof
entir
esp
ine.
“Sus
pici
ous
but
not
diag
nost
ic”
exam
inat
ions
wer
eev
alua
ted
with
MR
I,
flexi
on/e
xten
sion
view
s,or
repe
ated
clin
ical
exam
inat
ion.
I20
/58
(34%
)ha
din
juri
esde
tect
ed.
Plai
nx-
rays
iden
tifie
d12
for
ase
nsiti
vity
of60
%,
posi
tive
pred
ictiv
eva
lue
of10
0%,
nega
tive
pred
ictiv
eva
lue
of85
%.
CT
mis
sed
2in
juri
es(b
oth
“sta
ble”
).
Sens
itivi
ty:
90%
.
Spec
ifici
ty:
100%
.
Posi
tive
pred
ictiv
eva
lue:
100%
.
Neg
ativ
epr
edic
tive
valu
e:95
%.
D’A
lise
etal
.,19
99(1
5)12
1ob
tund
edpa
tient
sw
ithno
rmal
x-ra
ysst
udie
dw
ithM
RI.
CT
used
tost
udy
area
sof
MR
Iab
norm
ality
.
All
patie
nts
with
nega
tive
MR
Iun
derw
ent
flexi
on/e
xten
sion
imag
ing
imm
edia
tely
upon
“cle
aran
ce.”
III31
/121
(26%
)ha
din
juri
esde
tect
edon
MR
I.
90/1
21(7
4.4%
)ha
dno
inju
ryan
dw
ere
clea
red
(ver
ified
with
flexi
on/e
xten
sion
).
8pa
tient
sde
term
ined
toha
vesp
inal
inst
abili
ty(c
linic
al,
CT,
etc.
).
No
flexi
on/e
xten
sion
perf
orm
edon
patie
nts
with
abno
rmal
MR
I.
Can
not
dete
rmin
esi
gnifi
canc
eof
MR
Ifin
ding
sin
23/3
1pa
tient
s.
Aut
hors
indi
cate
that
nega
tive
MR
Ieq
uiva
lent
tone
gativ
efle
xion
/ext
ensi
on.
Kat
zber
get
al.,
1999
(39)
Pros
pect
ive
stud
yof
199
patie
nts
who
unde
rwen
tM
RI
inad
ditio
nto
stan
dard
radi
ogra
phic
stud
y.H
alf
ofpa
tient
sw
ere
sele
cted
beca
use
of
susp
ecte
dhi
ghpr
obab
ility
ofin
jury
.
IIIM
RI
dete
cted
inju
ries
ina
high
erfr
actio
nof
thes
epa
tient
sth
andi
dco
nven
tiona
lx-
rays
and
CT.
Sign
ifica
nce
ofth
ese
inju
ries
?
Gol
dst
anda
rd?
Kle
inet
al.,
1999
(41)
Ret
rosp
ectiv
ere
view
of32
patie
nts
with
75kn
own
spin
efr
actu
res.
Blin
ded
revi
ewof
MR
Isc
ans
byra
diol
ogis
ts.
II
Sele
ctpo
pula
tion
Post
erio
r/an
teri
orel
emen
tin
juri
es:
Sens
itivi
ty:
11.5
%/3
6.7%
.
Spec
ifici
ty:
97.0
%/9
8%.
Posi
tive
pred
ictiv
eva
lue:
83%
/91.
2%.
Neg
ativ
epr
edic
tive
valu
e:46
%/6
4%.
MR
Ino
tgo
odfo
rev
alua
ting
bony
path
olog
y.
Tan
etal
.,19
99(6
7)R
etro
spec
tive
revi
ewof
360
patie
nts
trea
ted
for
blun
tin
jury
who
unde
rwen
t3-
view
C-s
pine
film
ssu
pple
men
ted
with
CT
beca
use
of
nonv
isua
lizat
ion
ofC
7–T1
.
CT
findi
ngs
cons
ider
edgo
ldst
anda
rdfo
rde
tect
ion
offr
actu
re.
III11
inju
ries
dete
cted
byC
Tw
hich
wer
eno
tvi
sibl
eon
plai
nx-
rays
.
Sens
itivi
tyof
inad
equa
tepl
ain
x-ra
ysre
lativ
eto
CT
for
this
purp
ose:
97%
.
Whi
teet
al.,
1999
(72)
31pa
tient
sw
ithkn
own
orsu
spec
ted
spin
ein
jury
eval
uate
dw
ithM
RI.
IIIPr
ever
tebr
alhe
mat
oma
pick
edup
mor
eof
ten
byM
RI
than
bypl
ain
x-ra
ys(2
4/31
vers
us14
/30)
.
Sugg
ests
that
sens
itivi
tyof
plai
nx-
rays
for
prev
erte
bral
hem
atom
ais
66%
.
Aja
niet
al.,
1998
(1)
100
cons
ecut
ive
patie
nts
stud
ied
pros
pect
ivel
y.
All
radi
ogra
phed
(3vi
ews)
.
Follo
w-u
pcl
inic
alex
amin
atio
n,C
T,M
RI,
and
flexi
on/e
xten
sion
view
s
perf
orm
ed.
I1/
6in
juri
esm
isse
dby
x-ra
y(s
ensi
tivity
,84
%),
7/12
x-ra
yab
norm
aliti
esfo
und
tobe
insi
gnifi
cant
.
Posi
tive
pred
ictiv
eva
lue:
45%
.
Neg
ativ
epr
edic
tive
valu
e:98
.9%
.
1m
isse
din
jury
dete
cted
byfle
xion
/ext
ensi
onvi
ews.
Sees
etal
.,19
98(6
4)20
patie
nts
unde
rwen
tbe
dsid
efle
xion
/ext
ensi
onun
der
fluor
osco
pyaf
ter
3-vi
ewC
-spi
nex-
rays
norm
al.
III(fo
rflu
oros
copy
),II
for
3-vi
ewC
-spi
new
ith
fluor
osco
pyas
gold
stan
dard
One
patie
ntfo
und
toha
vesu
blux
atio
n.
No
gold
stan
dard
for
flexi
on/e
xten
sion
orflu
oros
copy
.
Sens
itivi
tyof
plai
nx-
rays
with
flexi
on/e
xten
sion
asgo
ldst
anda
rd:
95%
.
S38 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE5.
1.C
onti
nued
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Ben
zel
etal
.,19
96(5
)17
4pa
tient
ssu
spec
ted
ofha
ving
cerv
ical
spin
ein
jury
(equ
ivoc
alpl
ain
x-ra
ys/C
Tor
posi
tive
sym
ptom
s).
Und
erw
ent
MR
I.
CT
scan
sob
tain
edth
roug
har
eaof
inju
ryde
fined
byM
RI.
III36
%(6
2/17
4)ha
dM
RI
evid
ence
ofin
jury
.
61/6
2m
anag
edw
ithim
mob
iliza
tion
for
1–2
mo.
All
patie
nts
with
nega
tive
MR
Isc
ans
wer
ecl
eare
d,no
inst
ance
sof
late
inst
abili
ty.
Neg
ativ
epr
edic
tive
valu
eof
MR
I:10
0%.
Posi
tive
pred
ictiv
eva
lue?
Spec
ifici
ty?
Dav
iset
al.,
1995
(16)
116
patie
nts
with
GC
S�
13an
dno
rmal
x-ra
ysev
alua
ted
with
flexi
on/
exte
nsio
nvi
ews
unde
rflu
oros
copy
.
Ifo
rpl
ain
x-ra
ysve
rsus
flexi
on/e
xten
sion
asgo
ld
stan
dard
,III
(follo
w-u
p
ques
tiona
ble)
for
flexi
on/
exte
nsio
nru
ling
out
inju
ry.
113
patie
nts
had
noab
norm
ality
dete
cted
.
2pa
tient
sha
d“s
tabl
e”fa
cet
frac
ture
s.
1pa
tient
had
2m
mof
subl
uxat
ion
and
was
trea
ted
ina
colla
r(n
ofo
llow
-up
onth
ispa
tient
).
No
patie
ntha
dre
fera
ble
neur
olog
ical
inju
ryw
ithcl
inic
alfo
llow
-up.
Dec
ubitu
sul
cers
foun
dun
der
colla
rsin
44%
ofpa
tient
sw
ithm
ean
colla
rtim
eof
6d.
Neg
ativ
epr
edic
tive
valu
eof
flexi
on/e
xten
sion
fluor
osco
py:
100%
.
Hol
liman
,19
91(3
4)R
etro
spec
tive
seri
esof
148
patie
nts
with
know
nsp
ine
inju
ries
.
Late
ral
and
odon
toid
x-ra
ysre
tros
pect
ivel
yre
view
edse
para
tely
from
ante
ropo
ster
ior
film
.60
sets
offil
mav
aila
ble
for
revi
ew.
IIIIn
thes
e60
patie
nts,
all
inju
ries
note
don
ante
ropo
ster
ior
film
sw
ere
also
dete
ctab
leon
late
ral
orod
onto
idfil
ms.
Tehr
anza
deh
etal
.,19
94
(68)
Ret
rosp
ectiv
ere
view
of10
0pa
tient
sw
ithbl
unt
inju
ryan
dno
nvis
ualiz
ed
C7–
T1on
plai
nx-
rays
.
CT
findi
ngs
cons
ider
edgo
ldst
anda
rd.
III3
patie
nts
foun
dto
have
inju
ries
onC
Tno
tvi
sual
ized
bypl
ain
x-ra
ys.
Sens
itivi
tyof
inad
equa
tepl
ain
x-ra
ys:
97%
.
Bor
ock
etal
.,19
91(9
)U
sed
CT
toev
alua
tece
rvic
alsp
ine
in17
9pa
tient
sw
how
ere
sym
ptom
atic
with
norm
alx-
rays
(2),
who
seen
tire
cerv
ical
spin
eco
uld
not
bevi
sual
ized
(123
),or
who
had
equi
voca
l(1
3)or
abno
rmal
(41)
x-ra
ys.
Plai
nfil
mse
nsiti
vity
calc
ulat
edus
ing
CT
asgo
ldst
anda
rd;
auth
ors
clai
m
nom
isse
din
juri
es.
II
Poss
ible
fals
een
dpo
int
39/5
4x-
ray
abno
rmal
ities
wer
eve
rifie
dw
ithC
T(p
ositi
vepr
edic
tive
valu
e:72
%).
X-r
ays
mis
sed
both
inju
ries
insy
mpt
omat
icpa
tient
san
d1
C7
tran
sver
sepr
oces
sfr
actu
re
(neg
ativ
epr
edic
tive
valu
e:97
.6%
).
Coh
net
al.,
1991
(14)
60pa
tient
spr
ospe
ctiv
ely
stud
ied
with
late
ral
x-ra
ysin
emer
genc
y
depa
rtm
ent.
Full
radi
ogra
phic
wor
k-up
(3-
or5-
view
)fo
llow
ed.
Res
ults
ofla
tera
lvi
ewto
full
seri
esco
mpa
red.
II
Poss
ible
fals
een
dpo
int
(Iif
used
only
asco
mpa
riso
n)
Late
ral
view
mis
sed
3/7
tota
lin
juri
es.
Late
ral
view
posi
tive
pred
ictiv
eva
lue:
100%
.
Neg
ativ
epr
edic
tive
valu
e:94
%.
Sens
itivi
ty:
57%
.
Lew
iset
al.,
1991
(43)
Ret
rosp
ectiv
ere
view
of14
1pa
tient
sw
ithac
tive
flexi
on/e
xten
sion
x-ra
ys
perf
orm
edaf
ter
3-vi
ewse
ries
was
norm
al.
IIfo
rpl
ain
x-ra
ysve
rsus
flexi
on/e
xten
sion
asgo
ld
stan
dard
,II
for
flexi
on/
exte
nsio
nvi
ews
11/1
41fle
xion
/ext
ensi
onse
tsre
adas
unst
able
,4
ofw
hom
had
norm
al3-
view
seri
es(r
emai
nder
wer
equ
estio
nabl
e).
All
patie
nts
with
inst
abili
tyha
dpa
inor
wer
ein
toxi
cate
d.
1fa
lse-
nega
tive
flexi
on/e
xten
sion
stud
y.
For
plai
nfil
ms
vers
usfle
xion
/ext
ensi
onan
dpl
ain
film
s:
Sens
itivi
ty:
71%
and
99%
.
Spec
ifici
ty:
89%
and
89%
.
Neg
ativ
epr
edic
tive
valu
e:93
%an
d93
%.
Posi
tive
pred
ictiv
eva
lue:
67%
and
99%
.
Mac
dona
ldet
al.,
1990
(47)
775
patie
nts.
3vi
ews
com
pare
dag
ains
tgo
ldst
anda
rdof
all
othe
rst
udie
spe
rfor
med
and
clin
ical
outc
ome.
I3-
view
seri
es:
Sens
itivi
ty:
83%
.
Spec
ifici
ty:
97%
.
Posi
tive
pred
ictiv
eva
lue:
81%
.
Neg
ativ
epr
edic
tive
valu
e:98
%.
Radiographic Spinal Assessment in Symptomatic Trauma Patients S39
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
region of injury. Thin-cut computed tomographic imagesthrough the entire spine may increase sensitivity somewhat(6, 55), but no direct comparison between the two imagingstrategies in an appropriate patient population has been per-formed to date.
MRI has been used to evaluate patients at risk for acutespinal injury. Results have been mixed. Benzel et al. (5) stud-ied 174 symptomatic patients with low-field MRI within 48hours of injury. Soft tissue abnormalities were visualized onMRI scans in 62 patients. Two of these 62 were considered tohave unstable injuries. Both had plain film and CT abnormal-ities that revealed the injuries. The 60 patients with MRIabnormalities not thought to be significant were immobilizedfor 1 to 3 months and then studied with flexion/extensionradiographs. Not one was found to have an unstable injury.Patients with “negative” MRI studies were cleared of spinalprecautions, and no adverse events were reported. D’Alise etal. (15) reported their results of a Class III evidence study ofMRI in 121 obtunded patients. Ninety patients had normalstudies and were cleared. Follow-up flexion/extension radio-graphs did not reveal a single abnormality in this group.Thirty-one patients had injuries to soft tissues of the cervicalspine identified by MRI not detected by plain radiographs.Eight of these patients ultimately required surgery. Katzberget al. (39) and White et al. (72) have also described increasedsensitivity of MRI for the detection of soft tissue injuries of thecervical spine after trauma.
These studies demonstrate that MRI abnormalities are vi-sualized in a substantial number of cervical spine studiesperformed on patients after trauma. It is impossible to deter-mine the true incidence of clinically significant ligamentousinjury in this group examined with MRI because all patientswith MRI abnormalities were treated with immobilization.The incidence of significant cervical spine injury in previousstudies looking at similar patient populations is between 2and 6%, but the incidence of MRI abnormalities is reported tobe between 25 and 40%. MRI seems to “overcall” significantinjury. It should be noted that the optimal time frame for MRIassessment of the cervical spine is limited. MRI studies arepreferred within the first 48 hours after injury (5, 15, 21, 39,72). Even then, some injuries are poorly visualized. Emery etal. (21) used MRI to study 37 patients with known cervicalspine injury and found that MRI missed ligamentous injury in2 of 19 patients known to have ligamentous injury (abnormalflexion/extension films or surgical confirmation). These im-ages were obtained an average of 10.8 days after injury. Kleinet al. (41), comparing computed tomographic and MRI scansobtained from the same patients, demonstrated that MRI wasnot as effective as CT for recognizing bony abnormalities.MRI, when used early after trauma in conjunction with plainradiographs and CT, is exquisitely sensitive for detecting softtissue abnormalities of the cervical spine. The importance ofthese findings for most patients is, however, unknown.
Flexion/extension radiographs have been used to rule outligamentous injury of the cervical spine. In the awake patient,this maneuver is generally considered safe and effective. Nu-merous series have used flexion/extension films as the goldstandard for the exclusion of ligamentous injury in this pop-TA
BLE
5.1.
Con
tinu
ed
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Emer
yet
al.,
1989
(21)
MR
Ius
edto
stud
y37
patie
nts
with
know
nsp
ine
inju
ries
.
All
patie
nts
also
stud
ied
with
som
eco
mbi
natio
nof
plai
nx-
rays
/CT/
tom
ogra
ms/
surg
ical
expl
orat
ion/
clin
ical
follo
w-u
p.
II
Sele
ctpo
pula
tion
19pa
tient
sfo
und
toha
velig
amen
tous
inju
ry.
MR
Ide
tect
ed17
/19.
MR
Iw
asne
gativ
ein
18/1
8pa
tient
sw
ithin
tact
ligam
ents
.
Posi
tive
pred
ictiv
eva
lue:
100%
.
Neg
ativ
epr
edic
tive
valu
e:90
%.
Sens
itivi
ty:
89.5
%.
Free
mye
ret
al.,
1989
(24)
Pros
pect
ive
stud
yof
“hig
h-ri
sk”
sym
ptom
atic
patie
nts
imag
edw
ith5-
view
seri
es.
Rad
iogr
aphi
cre
view
of3
view
sco
mpa
red
with
5vi
ews.
CT
used
asgo
ldst
anda
rd.
II
Sele
ctpo
pula
tion
58pa
tient
sst
udie
d.
68in
juri
esde
tect
edin
33pa
tient
s.
Sens
itivi
tyof
3-vi
ewse
ries
was
83%
com
pare
dw
ithC
T.
Add
ition
ofob
lique
view
sha
dno
effe
cton
over
all
sens
itivi
ty.
Rei
det
al.,
1987
(59)
253
patie
nts
with
274
spin
alin
juri
esev
alua
ted.
38ha
dde
lay
indi
agno
sis.
Eval
uate
dca
use
for
dela
yin
diag
nosi
s.
III20
inju
ries
mis
sed
desp
itead
equa
tefil
ms.
Sens
itivi
ty:
92.3
%.
Shaf
fer
and
Dor
is,
1981
(65)
Ret
rosp
ectiv
ean
alys
isof
all
C-s
pine
inju
ries
dete
cted
inco
mm
unity
emer
genc
yro
oms.
Eval
uate
dre
adin
gof
late
ral
film
tore
adin
gof
3-vi
ewse
ries
(3-v
iew
seri
es
was
gold
stan
dard
).
Aut
hors
clai
mno
mis
sed
inju
ries
with
3-vi
ewse
ries
.
III35
inju
ries
dete
cted
.
9ca
ses
mis
sed
with
late
ral
view
alon
e.
Sens
itivi
tyof
late
ral
view
com
pare
dw
ith3
view
s:74
%.
aC
T,co
mpu
ted
tom
ogra
phy;
MR
I,m
agne
tic
reso
nanc
eim
agin
g;C
-spi
ne,
cerv
ical
spin
e;G
CS,
Gla
sgow
Com
aSc
ale
scor
e.
S40 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
ulation, and no serious adverse events have been reported (1,3, 5, 10, 11, 15). Brady et al. (10) used dynamic flexion/extension spine films to study 451 awake patients with blunttrauma evaluated in an urban emergency room. Flexion/extension views detected abnormalities in 5 of 372 patients inwhom static plain cervical spine films were thought to benormal. None of these patients required “invasive stabiliza-tion,” indicating that the abnormal examinations may havebeen false-positives. It should be noted, however, that false-negative examinations also occur, although infrequently.Lewis et al. (43) reported one false-negative examination in aseries of 141 patients studied with dynamic flexion/extensionfilms. These authors report the negative predictive value forthe combination of plain films and flexion/extension films tobe more than 99%.
The obtunded patient is not able to actively flex or extendthe neck for dynamic radiographic evaluation. Dynamic flu-oroscopy has been used to clear the cervical spine in thesepatients, and results of several series are available (16, 64).Ajani et al. (1) reported an unstable cervical spine injurydetected by flexion/extension radiographs in a patient withnormal plain films and CT (one of 100 patients studied). Daviset al. (16) used dynamic fluoroscopy to study 116 obtundedpatients who had normal cervical radiographs. Only one pa-tient was found to have an injury not visualized on plain filmsor CT. The significance of this injury, a 2-mm subluxation ina patient who was treated in a collar and subsequently lost tofollow-up, is questionable. Sees et al. (64) studied 20 obtundedpatients with normal three-view cervical spine series. Theyperformed bedside flexion/extension under fluoroscopy andfound one patient with C4–C5 subluxation caused by a facetinjury not appreciated on plain films but later confirmed withCT. It should be noted that 30% of the patients in the Sees etal. (64) series could not be cleared because of difficulty visu-alizing the lower cervical spine, whereas Davis et al. (16), byusing radiology staff in the fluoroscopy suite, were able tovisualize the entire spine in virtually all patients.
Because of the high negative predictive value of plain filmsand supplemental CT, application of MRI or flexion/extensionfluoroscopy for clearance of the cervical spine is probably notindicated for every obtunded patient. Use of these modalitiesshould be guided by clinical judgment based on patient historyand physical examination. Subgroups of obtunded trauma pa-tients exist with a low likelihood of cervical spine injury, andexhaustive study is not indicated for these patients. Hanson et al.(29) found that the incidence of cervical spine injury in a series of3684 patients without high-risk criteria was 0.2%, and that all ofthese injuries were detected by plain radiographs supplementedwith CT for poorly visualized or suspicious areas. The high-riskcriteria cited were: a high-speed motor vehicle accident (�35mph); an automobile crash with a death at the scene; a fall frommore than 10 feet; a significant traumatic closed-head injury ortraumatic intracranial hemorrhage; neurological signs or symp-toms referable to the cervical spine; or pelvis or multiple extrem-ity fractures. In support of this issue, Kaups and Davis (40) didnot identify a single cervical spine injury in a group of 215patients with gunshot wounds to the head. Similarly, Patton etal. (57) used MRI and flexion/extension fluoroscopy as a sup-
plement to x-rays to assess the cervical spines of a series ofpatients with isolated head injuries sustained as a result ofassault. They found no undiagnosed injuries.
SUMMARY
In summary, no single radiographic study can adequatelyrule out cervical spine injury in all symptomatic patients. Athree-view cervical spine series supplemented by CT throughareas difficult to visualize and “suspicious” areas will detectmost spinal injuries. This combination of studies representsthe minimum required for clearance of the cervical spine inthe symptomatic patient. The negative predictive value of thiscombination of studies is reported to be between 99 and 100%in several Class II and III evidence studies (9, 11, 24, 48, 67, 68).
In the awake patient, dynamic flexion/extension views(with at least 30-degree excursion in each direction) are safeand effective for detecting most “occult” cervical spine inju-ries not identified on plain x-rays. The negative predictivevalue of a normal three-view series and flexion/extensionviews exceeds 99% (43). Patients who are unable to cooperatewith active flexion/extension radiographs because of pain ormuscle spasm may be maintained in a cervical collar untilthey are able to cooperate, or they may be studied with MRI.A negative MRI study within the first 48 hours of injury, inaddition to normal radiographs and supplemental CT, seemsto be sufficient for clearing the cervical spine. The significanceof a positive MRI study is currently unclear. It is suggestedthat cervical immobilization be continued in these patientsuntil delayed flexion/extension views can be obtained.
In the obtunded patient with a normal three-view x-rayseries and appropriate CT of the cervical spine, the incidenceof significant spine injury is less than 1%. On the basis ofmechanism of injury and clinical judgment, the cervical spinein selected patients may be considered cleared without fur-ther study. In the remainder of cases, flexion/extension per-formed under fluoroscopic visualization seems to be safe andeffective for ruling out significant ligamentous injury, with areported negative predictive value of more than 99% (16).Because the incidence of occult injury diagnosed with dy-namic flexion/extension fluoroscopy in the setting of normalplain cervical spine x-rays and CT images is low, it is probablymost efficient for these procedures to be performed by staff inthe department of radiology, although variances in local ex-perience should be respected. MRI represents another optionfor clearance of the spine in this patient population, and anegative MRI within 48 hours of injury seems to effectivelyeliminate the likelihood of a significant ligamentous injury.However, MRI evaluation will result in a large number offalse-positive examinations, and the consequences of pro-longed unnecessary immobilization in the obtunded patientare not insignificant (18, 44, 58).
KEY ISSUES FOR FUTURE INVESTIGATION
The significance of positive MRI findings after cervicaltrauma should be evaluated by using flexion/extension ra-diographs and clinical follow-up as the gold standard. Theincidence of abnormal findings on flexion/extension fluoro-
Radiographic Spinal Assessment in Symptomatic Trauma Patients S41
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
scopic studies in obtunded patients should be evaluated in aprospective fashion with appropriate clinical follow-up. Aprospective comparison should be made between the three-view cervical spine series supplemented with selective CTthrough poorly visualized or suspicious areas and CT of theentire cervical spine.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Ajani AE, Cooper DJ, Scheinkestel CD, Laidlaw J, Tuxen DV:Optimal assessment of cervical spine trauma in critically ill pa-tients: A prospective evaluation. Anaesth Intensive Care 26:487–491, 1998.
2. Bachulis BL, Long WB, Hynes GD, Johnson MC: Clinical indica-tions for cervical spine radiographs in the traumatized patient.Am J Surg 153:473–478, 1987.
3. Banit DM, Grau G, Fisher JR: Evaluation of the acute cervicalspine: A management algorithm. J Trauma 49:450–456, 2000.
4. Bayless P, Ray VG: Incidence of cervical spine injuries in associ-ation with blunt head trauma. Am J Emerg Med 7:139–142, 1989.
5. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW, EspinosaMC: Magnetic resonance imaging for the evaluation of patientswith occult cervical spine injury. J Neurosurg 85:824–829, 1996.
6. Berne JD, Velmahos GC, El-Tawil Q, Demetriades D, Asensio JA,Murray JA, Cornwell EE, Belzberg H, Berne TV: Value of com-plete cervical helical computed tomographic scanning in identi-fying cervical spine injury in the unevaluable blunt trauma pa-tient with multiple injuries: A prospective study. J Trauma 47:896–903, 1999.
7. Blackmore CC, Emerson SS, Mann FA, Koepsell TD: Cervicalspine imaging in patients with trauma: Determination of fracturerisk to optimize use. Radiology 211:759–765, 1999.
8. Blackmore CC, Ramsey SD, Mann FA, Deyo RA: Cervical spinescreening with CT in trauma patients: A cost effective analysis.Radiology 212:117–125, 1999.
9. Borock EC, Gabram SG, Jacobs LM, Murphy MA: A prospectiveanalysis of a two-year experience using computed tomography asan adjunct for cervical spine clearance. J Trauma 31:1001–1006,1991.
10. Brady WJ, Moghtader J, Cutcher D, Exline C, Young J: ED use offlexion-extension cervical spine radiography in the evaluation ofblunt trauma. Am J Emerg Med 17:504–508, 1999.
11. Brohi K, Wilson-Macdonald J: Evaluation of unstable cervicalspine injury: A 6-year experience. J Trauma 49:76–80, 2000.
12. Chee SG: Review of the role of magnetic resonance imaging inacute cervical spine injuries. Ann Acad Med Singapore 22:757–761, 1993.
13. Clancy MJ: Clearing the cervical spine of adult victims of trauma.J Accid Emerg Med 16:208–214, 1999.
14. Cohn SM, Lyle WG, Linden CH, Lancey RA: Exclusion of cervicalspine injury: A prospective study. J Trauma 31:570–574, 1991.
15. D’Alise MD, Benzel EC, Hart BL: Magnetic resonance imagingevaluation of the cervical spine in the comatose or obtundedtrauma patient. J Neurosurg 91[Suppl 1]:54–59, 1999.
16. Davis JW, Parks SN, Detlefs CL, Williams GG, Williams JL, SmithRW: Clearing the cervical spine in obtunded patients: The use ofdynamic fluoroscopy. J Trauma 39:435–438, 1995.
17. Davis JW, Phreaner DL, Hoyt DB, Mackersie RC: The etiology ofmissed cervical spine injuries. J Trauma 34:342–346, 1993.
18. Dodd FM, Simon E, McKeown D, Patrick MR: The effect of acervical collar on the tidal volume of anaesthetised adult patients.Anaesthesia 50:961–963, 1995.
19. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ, FrederiksenSM: High-risk criteria for performing pre-hospital spinal immo-bilization in trauma. Ann Emerg Med 25:141–142, 1995.
20. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ, FrederiksenSM: Prospective validation of out-of-hospital spinal clearancecriteria: A preliminary report. Acad Emerg Med 4:643–646, 1997.
21. Emery SE, Pathria MN, Wilber RG, Masaryk T, Bohlman HH:Magnetic resonance imaging of posttraumatic spinal ligamentinjury. J Spinal Disord 2:229–233, 1989.
22. Fehlings MG, Rao SC, Tator CH, Skaf G, Arnold P, Benzel EC,Dickman C, Cuddy B, Green B, Hitchon P, Northrup B, SonntagVKH, Wagner F, Wilberger JE: The optimal method for assessingspinal canal compromise and cord compression in patients withcervical spinal cord injury. Spine 24:605–613, 1999.
23. Flanders AE, Schaefer DM, Doan HT, Mishkin MM, Gonzales CF,Northrup BE: Acute cervical spine trauma: Correlation of MRimaging findings with degree of neurologic deficit. Radiology177:25–33, 1990.
24. Freemyer B, Knopp R, Piche J, Wales L, Williams J: Comparison offive-view and three-vies cervical spine series in the evaluation ofpatients with cervical trauma. Ann Emerg Med 18:818–821, 1989.
25. Gerrelts BD, Petersen EU, Mabry J, Petersen SR: Delayed diagno-sis of cervical spine injuries. J Trauma 31:1622–1626, 1991.
26. Gonzales RP, Fried PO, Bukhalo M, Holevar MR, Falimirski ME:Role of clinical examination in screening for blunt injury. J AmColl Surg 189:152–157, 1999.
27. Grossman MD, Reilly PM, Gillett T, Gillett D: National survey ofthe incidence of cervical spine injury and approach to cervicalspine clearance in US trauma centers. J Trauma 47:684–690, 1999.
28. Hall AJ, Wagle VG, Raycroft J, Goldman RL, Butler AR: Magneticresonance imaging in cervical spine trauma. J Trauma 34:21–26,1993.
29. Hanson JA, Blackmore CC, Mann FA, Wilson AJ: Cervical spineinjury: A clinical decision rule to identify high risk patients forhelical CT screening. Am J Roentgenol 174:713–717, 2000.
30. Herr CH, Ball PA, Sargent SK, Quinton HB: Sensitivity ofprevertebral soft tissue measurement at C3 for detection of cer-vical spine fractures and dislocations. Am J Emerg Med 16:346–349, 1998.
31. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI:Validity of a set of clinical criteria to rule out injury to the cervicalspine in patients with blunt trauma. N Engl J Med 343:94–99,2000.
32. Hoffman JR, Schriger DL, Mower W, Luo JS, Zucker M: Low-riskcriteria for cervical spine radiography in blunt trauma: A pro-spective study. Ann Emerg Med 21:1454–1460, 1992.
33. Hoffman JR, Wolfson AB, Todd K, Mower WR: Selective cervicalspine radiography in blunt trauma: Methodology of the NationalEmergency X-Radiography Utilization Study (NEXUS). AnnEmerg Med 32:461–469, 1998.
34. Holliman CJ, Mayer JS, Cook RT Jr, Smith JS Jr: Is the anteropos-terior cervical spine radiograph necessary in initial traumascreening? Am J Emerg Med 9:421–425, 1991.
35. Iida H, Tachibana S, Kitahara T, Horiike S, Ohwada T, Fujii K:Association of head trauma with cervical spine injury, spinal cordinjury, or both. J Trauma 46:450–452, 1999.
36. Ireland AJ, Britton I, Forrester AW: Do supine oblique viewsprovide better imaging of the cervicothoracic junction than swim-mer’s views? J Accid Emerg Med 15:151–154, 1998.
S42 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
37. Jacobs LM, Schwartz R: Prospective analysis of acute cervicalspine injury: A methodology to predict injury. Ann Emerg Med15:44–49, 1986.
38. Kaneriya PP, Schweitzer ME, Spettell C, Cohen MJ, Karasick D:The cost effectiveness of oblique radiography in the exclusion ofC7–T1 injury in trauma patients. Am J Radiol 171:959–962, 1998.
39. Katzberg RW, Benedetti PF, Drake CM, Ivanovic M, Levine RA,Beatty CS, Nemzek WR, McFall RA, Ontell FK, Bishop DM,Poirier VC, Chong BW: Acute cervical spine injuries: ProspectiveMR imaging assessment at a level one trauma center. Radiology213:203–212, 1999.
40. Kaups K, Davis J: Patients with gunshot wounds to the head donot require cervical spine immobilization and evaluation.J Trauma 44:865–867, 1998.
41. Klein GR, Vaccaro AR, Albert TJ, Schweitzer M, Deely D, KarasickD, Cotler JM: Efficacy of magnetic resonance imaging in theevaluation of posterior cervical spine fractures. Spine 24:771–774,1999.
42. Kreipke DL, Gillespie KR, McCarthy MC, Mail JT, Lappas JC,Broadie TA: Reliability of indications for cervical spine films intrauma patients. J Trauma 29:1438–1439, 1989.
43. Lewis LM, Docherty M, Ruoff BE, Fortney JP, Keltner RA Jr,Britton P: Flexion-extension views in the evaluation of cervicalspine injuries. Ann Emerg Med 20:117–121, 1991.
44. Liew SC, Hill DA: Complications of hard cervical collars in multi-trauma patients. Aust N Z J Surg 64:139–140, 1994.
45. Lindsey RW, Diliberti TC, Doherty BJ, Watson AB: Efficacy ofradiographic evaluation of the cervical spine in emergency situ-ations. South Med J 86:1253–1255, 1993.
46. Lockey AS, Handley R, Willett K: “Clearance” of cervical spineinjury in the obtunded patient. Injury 29:493–497, 1998.
47. Macdonald RL, Schwartz ML, Mirich D, Sharkey PW, Nelson WR:Diagnosis of cervical spine injury in motor vehicle crash victims:How many x-rays are enough? J Trauma 30:392–397, 1990.
48. Mace SE: Emergency evaluation of cervical spine injuries: CTversus plain radiographs. Ann Emerg Med 14:973–975, 1985.
49. Mace SE: Unstable occult cervical spine fracture. Ann Emerg Med20:1373–1375, 1991.
50. Mahadevan S, Mower WR, Hoffman JR, Peeples N, Goldberg W,Sonner R: Interrater reliability of cervical spine injury criteria inpatients with blunt trauma. Ann Emerg Med 31:197–201, 1998.
51. McNamara RM, Heine E, Esposito B: Cervical spine injury andradiography in alert, high risk patients. J Emerg Med 8:177–182,1990.
52. Mirvis SE, Diaconis JN, Chirico PA, Reiner BI, Joslyn JN, MilitelloP: Protocol driven radiologic evaluation of suspected cervicalspine injury: Efficacy study. Radiology 170:831–834, 1989.
53. Neifeld GL Keene JG, Hevesy G, Leikin J, Proust A, Thisted RA:Cervical injury in head trauma. J Emerg Med 6:203–207, 1988.
54. Nunez DB Jr, Quencer RM: The role of helical CT in the assess-ment of cervical spine injuries. AAJR Am J Radiol 171:951–957,1998.
55. Nunez DB Jr, Zuluaga A, Fuentes-Bernardo DA, Rivas LA,Beccera JL: Cervical spine trauma: How much more do we learnby routinely using helical CT? Radiographics 16:1318–1321, 1996.
56. Pasquale M, Fabian TC: Practice management guidelines fortrauma from the Eastern Association for the Surgery of Trauma.J Trauma 44:945–946, 1998.
57. Patton JH, Kralovich KA, Cuschieri J, Gasparri M: Clearing thecervical spine in victims of blunt assault to the head and neck:What is necessary? Am Surg 66:326–330, 2000.
58. Raphael JH, Chotai R: Effects of the cervical collar of cerebrospi-nal fluid pressure. Anaesthesia 49:437–439, 1994.
59. Reid DC, Henderson R, Saboe L, Miller JD: Etiology and clinicalcourse of missed spine fractures. J Trauma 27:980–986, 1987.
60. Roberge RJ, Wears RC, Kelly M, Evans TC, Kenny MA, DaffnerRD, Kremen R, Murray K, Cottington EC: Selective application ofcervical spine radiography in alert victims of blunt trauma: Aprospective study. J Trauma 28:784–788, 1988.
61. Ross SE, O’Malley KF, DeLong WG, Born CT, Schwab CW: Clin-ical predictors of unstable cervical spine injury in multiply injuredpatients. Injury 23:317–319, 1992.
62. Roth BJ, Martin RR, Foley K, Barcia PJ, Kennedy P: Roentgeno-graphic evaluation of the cervical spine. Arch Surg 129:643–645,1994.
63. Scarrow AM, Levy EI, Resnick DK, Adelson PD, Sclabassi RJ:Cervical spine evaluation in obtunded or comatose pediatrictrauma patients: A pilot study. Pediatr Neurosurg 30:169–175,1999.
64. Sees DW, Cruz LR, Flaherty SF, Ciceri DP: The use of bedsidefluoroscopy to evaluate the cervical spine in obtunded patients.J Trauma 45:768–771, 1998.
65. Shaffer MA, Doris PE: Limitation of the cross table lateral view indetecting cervical spine injuries: A prospective analysis. AnnEmerg Med 10:508–513, 1981.
66. Sweeney JF, Rosemurgy AS, Gill S, Albrink MH: Is the cervicalspine clear? Undetected cervical fractures diagnosed only at au-topsy. Ann Emerg Med 21:1288–1290, 1992.
67. Tan E, Schweitzer ME, Vaccaro L, Spetell AC: Is computed to-mography of nonvisualized C7–T1 cost effective? J Spinal Disord12:472–476, 1999.
68. Tehranzadeh J, Bonk RT, Ansari A, Mesgarzadeh M: Efficacy oflimited CT for nonvisualized lower cervical spine in patients withblunt trauma. Skeletal Radiol 23:349–352, 1994.
69. Tracy PT, Wright RM, Hanigan WC: Magnetic resonance imagingof spinal injury. Spine 14:292–301, 1989.
70. Vaccaro AR, An HS, Lin S, Sun S, Balderston RA, Cotler JM:Noncontiguous injuries of the spine. J Spinal Disord 5:320–329,1992.
71. Walters BC: Clinical practice parameter development in neuro-surgery, in Bean JR (ed): Neurosurgery in Transition: The Socioeco-nomic Transformation of Neurological Surgery. Baltimore, Williams& Wilkins, 1998, pp 99–111.
72. White P, Seymour R, Powell N: MRI assessment of the pre-vertebral soft tissues in acute cervical spine trauma. Br J Radiol72:818–823, 1999.
73. Woodring JH, Lee C: Limitations of cervical radiography in theevaluation of acute cervical trauma. J Trauma 34:32–39, 1993.
Radiographic Spinal Assessment in Symptomatic Trauma Patients S43
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
CHAPTER 6
Initial Closed Reduction of Cervical SpineFracture-Dislocation Injuries
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Early closed reduction of cervical spine fracture-dislocation injuries with craniocervical traction is recom-
mended to restore anatomic alignment of the cervical spine in awake patients.• Closed reduction in patients with an additional rostral injury is not recommended.• Patients with cervical spine fracture-dislocation injuries who cannot be examined during attempted closed
reduction, or before open posterior reduction, should undergo magnetic resonance imaging (MRI) beforeattempted reduction. The presence of a significant disc herniation in this setting is a relative indication fora ventral decompression before reduction.
• MRI study of patients who fail attempts at closed reduction is recommended.• Prereduction MRI performed in patients with cervical fracture dislocation injury will demonstrate disrupted
or herniated intervertebral discs in one-third to one-half of patients with facet subluxation. These findingsdo not seem to significantly influence outcome after closed reduction in awake patients; therefore, theusefulness of prereduction MRI in this circumstance is uncertain.
RATIONALE
Spinal cord injury (SCI) is frequently associated withtraumatic cervical spine fractures and cervical facet dis-location injuries because the spinal canal can be nar-
rowed by displacement of fracture fragments or subluxationof one vertebra over another. Reduction of the deformityhelps to restore the diameter of the bony canal and eliminatesbony compression of the spinal cord caused by the vertebralfracture and/or subluxation. Theoretically, early decompres-sion of the spinal cord after injury may lead to improvedneurological outcome. Several large case series of patientsdescribe excellent results with closed reduction of cervicalfractures and facet subluxations. However, descriptive seriesusing prereduction magnetic resonance imaging (MRI) havereported a high incidence of cervical disc herniation in pa-tients with facet dislocation. Furthermore, there are case re-ports and small case series of patients who worsened neuro-logically after closed cervical spine reduction. Several of thesereports implicate ventral compression of the spinal cord bydisplaced disc material. The purpose of this qualitative reviewis to address the following issues:
1. Is closed reduction safe and effective for reducing cervicalspine deformity in patients with cervical fractures or uni-lateral or bilateral facet dislocation injuries?
2. Is a prereduction MRI study essential for managing thesepatients?
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasperformed. The following subject headings were combinedwith “spinal cord injury”: “spinal fracture,” “spinal injury,”and “human.” Approximately 12,300 citations were acquired.Non-English language citations were deleted. Searching thisset of publications with the keyword “cervical” resulted in2,154 matches. Further refining the search with the terms“reduction” or “fracture” yielded 606 articles; titles and ab-stracts of the articles were reviewed. Clinical series dealingwith adult patients in the acute setting were selected. Casereports and case collections were included in the review butare not listed in the table. Additional references were culledfrom the reference lists of the remaining articles. The membersof the author group were asked to contribute articles knownto them on the subject matter that were not found by othersearch means. Finally, the tables of contents of the journalSpine from 1993 through May 2000 were hand-searched. Atotal of 42 articles relevant to this topic were identified andform the basis of this guideline. The larger clinical series aresummarized in Table 6.1.
S44 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Among the articles reviewed, there were no randomizedclinical trials, no prospective cohort studies, and no case-control studies. One historical cohort study comparedtraction-reduction in awake patients with manipulation underanesthesia (MUA). The remainder of the articles consisted ofcase series of patients with acute or subacute unilateral orbilateral cervical facet dislocation injuries treated with open orclosed reduction. Several case reports and case series of pa-tients who deteriorated after closed reduction were identifiedand are included. Several studies included pre- and pos-treduction MRI findings.
SCIENTIFIC FOUNDATION
Walton (39), in 1893, first described closed reduction bymanipulation of cervical spine deformity caused by facet dis-location. Crutchfield (7) introduced tongs for in-line traction-reduction in 1933, and similar techniques have been success-fully used for traction-reduction of cervical deformity by alarge number of authors (1–4, 6, 11, 12, 16, 21, 23, 27, 29–31, 34,37, 38). Observations by Evans (10) and Kleyn (19) popular-ized reduction under anesthesia, although other authors con-demned the procedure as potentially dangerous comparedwith craniocervical traction-reduction. MUA is still a fre-quently practiced technique, usually used after failure oftraction-reduction but occasionally used as a primary meansof achieving reduction (9, 16, 41). Only one cohort study hasbeen performed comparing the two modalities. Lee et al. (20)found a higher rate of success and a lower complication ratewith traction-reduction as opposed to MUA. The significanceof their results is questionable because of the historical cohortdesign of the study. Lee et al. attributed the higher complica-tion rate in the MUA group to the effects of anesthesia onperfusion of the injured spinal cord. It is possible, however,that advances in the pharmacological and medical manage-ment of SCI patients during the 10-year period of data accrualaccounted for the improved results the authors noted in thetraction-reduction group. For this reason, the evidence pro-vided by this study is considered to be Class III medicalevidence.
Recent reports of neurological deterioration after closed oropen posterior reduction of cervical fracture-dislocation inju-ries has led some authors to recommend the use of prereduc-tion MRI to assess for ventral cord compromise caused bytraumatic disc disruption. Several investigators believe thatdisc disruption in association with facet fracture-dislocationincreases the risk of spinal cord injury by disc material afterreduction (7, 8, 20, 25). Other authors, however, have foundno relationship between findings on prereduction MRI, neu-rological outcome, or findings on postreduction MRI (33). Thenature of the injury predisposes a large percentage of patientswith cervical facet dislocation injuries to have MRI evidenceof disc material ventral to the spinal cord. Rizzolo et al. (28)found evidence of disc disruption/herniation in 42% of pa-
tients studied with prereduction MRI. The clinical relevanceof these findings has not been proven. The use of prereductionMRI may delay reduction of the spinal deformity and there-fore may delay decompression of the compromised spinalcord. Prereduction MRI assessment requires the transport of apatient with a highly unstable cervical spine fracture to theMRI suite. Recent laboratory work, as well as Class III evi-dence, suggests that early reduction of fracture-dislocationinjuries may improve neurological outcome (2, 10, 16, 20, 26).If stabilization of the unstable cervical spine protects againstadditional injury to the cervical spinal cord, then the informa-tion gained by prereduction MRI must be of sufficient value towarrant the delay in treatment and the associated potentialmorbidity of transport.
Most of the clinical series reviewed were based on patientdata accrued before the introduction of MRI. These combinedseries encompass more than 1200 patients treated with closedreduction in the acute or subacute period after injury. Thesuccess rate for restoring anatomic alignment by closed re-duction in these studies was approximately 80%. The reportedpermanent neurological complication rate was less than 1.0%(3, 6, 9, 11, 12, 15, 16, 20–21, 23, 27–31, 34, 36, 38, 42). Of the 11patients reported to develop new permanent neurologicaldeficits with attempted closed reduction, two had root inju-ries (11, 12), and two had ascending spinal cord deficits notedat the time of reduction (3, 30). Seven patients were noted tohave decreased American Spinal Injury Association (ASIA)scores postreduction, but neither the nature nor the cause ofthe new deficits in these patients was described (16).
Transient neurological deterioration after closed reductionhas been reported. In addition to the permanent deficits notedabove, temporary deficits have been described in 20 addi-tional patients of these 1200. These deficits reversed sponta-neously or improved after reduction of weight or after openreduction (3, 11, 12, 16, 20, 31). The causes of neurologicaldeterioration associated with closed reduction in these andother series included overdistraction (3, 21, 30), failure torecognize a more rostral noncontiguous lesion (30, 32), discherniation (11), epidural hematoma (17), and spinal cordedema (19, 20).
Several authors have provided general suggestions on howcraniocervical traction for closed reduction of cervical spinefracture-dislocation injuries is best accomplished (14, 15, 18,40). No study has been undertaken to determine the superi-ority of one method or technique over another. Tongs(Crutchfield or Gardner-Wells) or a halo ring are applied tothe patient’s head using sterile technique and local anesthesiaat the tong or pin insertion sites. Most contemporary descrip-tions incorporate the use of an MRI-compatible halo ring asthe cranial fixation device for four-point fixation of the cra-nium. This device gives better control of the head and neck ifpositioning and directional traction are needed (e.g., passivedirectional traction in positions of flexion or extension of theneck depending on the injury type) and facilitates halo-ring
Initial Closed Reduction of Fracture-Dislocation Injuries
S45Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE6.
1.Su
mm
ary
ofR
epor
tson
Clo
sed
Red
ucti
onof
Cer
vica
lSp
inal
Frac
ture
-Dis
loca
tion
Inju
ries
a
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yR
esul
tsEv
iden
ce
Cla
ssC
oncl
usio
ns
Gra
ntet
al.,
1999
(13)
82pa
tient
s.
Ret
rosp
ectiv
ese
ries
.
All
clos
edC
-spi
nein
juri
esw
ithm
alal
ignm
ent
incl
uded
.
Uni
late
ral
and
bila
tera
llo
cked
face
ts.
Earl
yra
pid
clos
edre
duct
ion
atte
mpt
edin
all
patie
nts.
MR
Isc
ans
obta
ined
afte
rre
duct
ion.
ASI
Aan
dFr
anke
lgr
ades
dete
rmin
edon
adm
issi
onan
dat
6an
d24
h.
Wei
ght
upto
80%
ofpa
tient
’sbo
dyw
eigh
t.
Succ
essf
ulre
duct
ion
in97
.6%
.
Ave
rage
time
tore
duct
ion
2.1
�0.
24h.
Ove
rall
ASI
Asc
ores
impr
oved
by24
haf
ter
redu
ctio
n.
1pa
tient
dete
rior
ated
6h
post
redu
ctio
n(p
roba
ble
root
lesi
on).
46%
had
disc
inju
ryon
MR
I,22
%ha
dhe
rnia
tion.
Dis
cin
jury
onM
RI
did
corr
elat
ew
ithco
rded
ema
onM
RI.
IIIC
lose
dre
duct
ion
isef
fect
ive
and
safe
,de
spite
high
inci
denc
eof
MR
I-de
mon
stra
ble
disc
inju
ries
/her
niat
ions
.
Vita
let
al.,
1998
(38)
168
patie
nts.
Ret
rosp
ectiv
ese
ries
.
Uni
late
ral
and
bila
tera
lfa
cet
inju
ries
.
Empl
oyed
man
ipul
atio
nun
der
gene
ral
anes
thes
iain
min
ority
ofca
ses.
Use
dre
lativ
ely
light
wei
ghts
(max
imum
8.8
lbpl
us2.
2pe
rle
vel
for
max
imum
of40
lb).
All
patie
nts
oper
ated
onim
med
iate
lypo
stre
duct
ion
orpo
st-f
ailu
reof
redu
ctio
n.
MR
Isc
ans
not
done
prer
educ
tion
(alth
ough
disc
sno
ted
in7
patie
nts?
).
43%
redu
ced
bytr
actio
nw
ithou
tan
esth
esia
(tim
e�
2h)
.
30%
redu
ced
byM
UA
.
27%
redu
ced
intr
aope
rativ
ely.
5pa
tient
sdi
dno
tre
duce
(del
ayed
refe
rral
,su
rgic
aler
ror)
.
Aut
hors
obse
rved
noca
ses
ofne
urol
ogic
alde
teri
orat
ion.
IIIA
utho
rspr
omot
eth
eir
prot
ocol
asa
safe
and
effe
ctiv
em
eans
for
redu
ctio
nan
dst
abili
zatio
nof
frac
ture
s.
Lee
etal
.,19
94(2
0)21
0pa
tient
s.
RT
in11
9,M
UA
in91
.
Ret
rosp
ectiv
ehi
stor
ical
coho
rtst
udy.
Gro
ups
sim
ilar
exce
pttr
actio
ngr
oup
had
long
erde
lay
totr
eatm
ent.
Wei
ghts
upto
150
lbus
ed.
No
MR
Ido
nepr
ered
uctio
n.
Red
uctio
nsu
cces
sful
:
MU
A:
66/9
1(7
3%).
RT:
105/
119
(88%
).
All
failu
res
inR
Tgr
oup
wer
edu
eto
asso
ciat
edfr
actu
res
or
dela
yed
refe
rral
.
Tim
eto
redu
ctio
n:
RT
21m
in.
MU
A:
not
repo
rted
.
No
loss
ofFr
anke
lgr
ade
inei
ther
grou
p.
6M
UA
and
1R
Tha
dde
teri
orat
ion
onA
SIA
scor
e.
IIIR
Tsu
peri
orto
MU
A,
both
proc
edur
essa
fean
d
effe
ctiv
e,M
RI
not
done
.
Cot
ler
etal
.,19
93(5
)24
patie
nts
(all
awak
e).
Pros
pect
ive
stud
y.
No
frac
ture
dfa
cets
.
All
acut
ein
juri
es(1
5-d
patie
nt).
Wei
ghts
upto
140
lbus
ed.
No
CT
orM
RI
done
.
All
24re
duce
d.
No
inci
denc
eof
neur
olog
ical
dete
rior
atio
n.
Man
ipul
atio
nus
edin
addi
tion
tow
eigh
tsin
9pa
tient
s(w
hen
face
tspe
rche
d).
Tim
ere
quir
edra
nged
from
8to
187
min
.
IIIR
educ
tion
with
wei
ghts
upto
140
lbis
safe
and
effe
ctiv
ein
mon
itore
dse
tting
with
expe
rien
ced
phys
icia
ns.
Mah
ale
etal
.,19
93(2
4)34
1pa
tient
str
eate
dfo
rtr
aum
atic
disl
ocat
ions
ofce
rvic
alsp
ine.
15ha
dne
urol
ogic
alde
teri
orat
ion.
Var
iety
oftr
eatm
ents
used
tore
duce
defo
rmity
,in
clud
ing
oper
ativ
e
redu
ctio
n.
Com
plet
ein
juri
es:
6af
ter
OR
,1
afte
rm
anip
ulat
ion.
Inco
mpl
ete
inju
ries
:1
afte
rO
R,
3af
ter
man
ipul
atio
n,2
afte
r
trac
tion,
1du
ring
appl
icat
ion
ofca
st.
Rad
icul
opat
hy:
1(o
ccur
red
whe
nto
ngs
slip
ped
duri
ng
trac
tion)
.
Det
erio
ratio
nde
laye
din
11pa
tient
s.
IIIN
umbe
rsof
patie
nts
subj
ect
toea
chtr
eatm
ent
arm
not
give
n.Pu
rely
ade
scri
ptiv
epa
per.
Onl
y
conc
lusi
onis
that
neur
olog
ical
dete
rior
atio
nca
n
happ
en.
Had
ley
etal
.,19
92(1
5)68
patie
nts.
Ret
rosp
ectiv
ese
ries
.
Face
tfr
actu
redi
sloc
atio
nson
ly.
Uni
late
ral
and
bila
tera
llo
cked
face
ts.
66tr
eate
dw
ithea
rly
atte
mpt
edcl
osed
redu
ctio
n(2
late
refe
rral
s).
Ave
rage
wei
ghts
used
for
succ
essf
ulre
duct
ion
wer
ebe
twee
n9
and
10
lb/c
rani
alle
vel.
58%
ofpa
tient
sha
dsu
cces
sful
redu
ctio
n.
Ove
rall,
mos
tpa
tient
s(7
8%)
dem
onst
rate
dne
urol
ogic
al
reco
very
byla
stfo
llow
-up
(not
quan
tifie
d).
7pa
tient
sde
teri
orat
eddu
ring
“tre
atm
ent”
(6im
prov
edaf
ter
oper
ativ
ere
duct
ion,
1pe
rman
ent
root
defic
itaf
ter
trac
tion)
.
No
MR
Ida
tare
port
ed.
IIIEa
rly
deco
mpr
essi
onby
redu
ctio
nle
dto
impr
oved
outc
omes
,ba
sed
onfa
ctth
atpa
tient
s
who
did
best
wer
ere
duce
dea
rly
(�5–
8h)
.
No
com
pari
son
poss
ible
betw
een
CR
and
oper
ativ
ere
duct
ion
due
tosm
all
num
bers
.
1.2%
perm
anen
tde
ficit
(roo
t)re
late
dto
trac
tion.
S46 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE6.
1.C
onti
nued
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yR
esul
tsEv
iden
ce
Cla
ssC
oncl
usio
ns
Star
etal
.,19
90(3
5)57
patie
nts.
Ret
rosp
ectiv
ese
ries
.
Uni
late
ral
and
bila
tera
lin
juri
es.
Earl
yra
pid
redu
ctio
nat
tem
pted
inal
lpa
tient
s.
No
MR
Ido
nepr
ered
uctio
n.
1pa
tient
was
ade
laye
dtr
ansf
er.
Wei
ghts
upto
160
lb(b
egan
at10
lb).
Fran
kel
grad
esre
cord
edat
adm
issi
onan
ddi
scha
rge.
53/5
7(9
3%)
redu
ced
mea
ntim
eto
redu
ctio
nw
as8
h.
No
patie
ntde
teri
orat
eda
Fran
kel
grad
e.
2pa
tient
slo
stro
otfu
nctio
n,1
tran
sien
tly.
45%
impr
oved
one
Fran
kel
grad
eby
time
ofdi
scha
rge,
23%
impr
oved
less
subs
tant
ially
.
75%
ofpa
tient
sre
quir
ed�
50lb
wei
ght.
IIIC
lose
dre
duct
ion
issa
fean
def
fect
ive
for
deco
mpr
essi
ngco
rdan
des
tabl
ishi
ngsp
inal
alig
nmen
t.
Sabi
ston
etal
.,19
88(3
1)39
patie
nts.
Ret
rosp
ectiv
ese
ries
.
Uni
late
ral
and
bila
tera
lin
juri
es.
Up
to70
%of
body
wei
ght
used
.
All
acut
ein
juri
es.
No
MR
I.
35/3
9(9
0%)
ofin
juri
essu
cces
sful
lyre
duce
d,av
erag
e
wei
ght
used
62.5
lb.
No
neur
olog
ical
dete
rior
atio
n.
Failu
res
due
tosu
rgeo
nun
will
ingn
ess
tous
em
ore
wei
ght.
IIIC
lose
dre
duct
ion
with
upto
70%
ofbo
dyw
eigh
tis
safe
and
effe
ctiv
efo
rre
duci
nglo
cked
face
ts.
Aut
hors
stat
eth
atpa
tient
sse
enin
dela
yed
fash
ion
(�10
d)ar
eun
likel
yto
redu
ce(n
oev
iden
cepr
esen
ted
here
).
Mai
man
etal
.,19
86(2
5)28
patie
nts.
Var
iety
oftr
eatm
ents
offe
red.
No
MR
I.
18pa
tient
sha
dat
tem
pted
clos
edre
duct
ion
(max
imum
wei
ght
50lb
).
10/1
8re
duce
dw
ithtr
actio
n.
No
patie
nttr
eate
dby
auth
ors
dete
rior
ated
.
One
refe
rred
patie
ntha
dan
over
dist
ract
ion
inju
ry.
IIIM
ixed
grou
pof
patie
nts
and
trea
tmen
ts.
Inge
nera
l,
trac
tion
seem
edto
besa
fe.
Kle
yn,
1984
(19)
101
patie
nts.
Uni
late
ral
and
bila
tera
lin
juri
es,
all
with
neur
olog
ical
invo
lvem
ent.
All
trea
ted
with
trac
tion.
Ifin
jury
�24
h,M
UA
atte
mpt
edin
itial
ly;
iftr
actio
nre
duct
ion
fails
with
max
imum
18kg
wei
ght,
MU
Ape
rfor
med
.
Pre-
MR
Iut
ilize
d.
82/1
01su
cces
sful
lyre
duce
d(4
open
redu
ctio
n,6
part
ial
redu
ctio
nac
cept
ed,
9no
furt
her
atte
mpt
due
topo
or
cond
ition
ofpa
tient
).
37/4
5in
com
plet
ele
sion
sim
prov
ed.
7/56
com
plet
ele
sion
sim
prov
ed.
No
neur
olog
ical
dete
rior
atio
n.
IIITr
actio
nfo
llow
edby
MU
Ais
safe
,us
ually
(80%
)
effe
ctiv
e,an
dm
ayre
sult
inim
prov
edne
urol
ogic
al
func
tion.
Sonn
tag,
1981
(34)
15pa
tient
s.
Ret
rosp
ectiv
ean
alys
is.
All
bila
tera
llo
cked
face
ts.
All
acut
ein
juri
es.
Man
ual
trac
tion,
tong
trac
tion,
and
open
redu
ctio
nus
ed.
Wei
ght
used
rang
edfr
om30
to75
lb.
No
MR
Ido
ne.
Red
uctio
nw
ithtr
actio
nsu
cces
sful
in10
patie
nts.
5fa
iled:
1w
ithC
1fr
actu
rew
hich
did
not
allo
wtr
actio
n,
2w
ithfr
actu
red
face
ts,
1w
ithra
dicu
lar
sym
ptom
s
wor
sene
dby
trac
tion
(tran
sien
t),1
with
anas
cend
ing
spin
alco
rdin
jury
(pat
ient
died
ofpu
lmon
ary
com
plic
atio
ns2
wk
late
r).
IIISt
epw
ise
algo
rith
m(tr
actio
n,m
anua
lm
anip
ulat
ion,
oper
ativ
ere
duct
ion)
isin
dica
ted.
Clo
sed
redu
ctio
nby
wei
ght
appl
icat
ion
isth
e
pref
erre
dm
etho
dfo
rre
duct
ion
ofde
form
ity.
Shro
sbre
e,19
79(3
3)21
6pa
tient
sid
entif
ied
with
lock
edfa
cets
(86
died
with
in3
mo,
excl
uded
from
seri
es).
Use
dtr
actio
n(n
ow
eigh
tsp
ecifi
ed)
follo
wed
bym
anip
ulat
ion
unde
r
anes
thes
iaif
trac
tion
faile
d.
Pre-
MR
I.
70/9
5un
ilate
ral
ijuri
esre
duce
d(7
4%).
77/1
21bi
late
ral
inju
ries
redu
ced
(64%
).
No
neur
olog
ical
mor
bidi
tyre
port
ed.
Patie
nts
who
wer
esu
cces
sful
lyre
duce
dim
prov
edm
ore
ofte
nth
anpa
tient
sw
how
ere
not
succ
essf
ully
redu
ced
(41%
vers
us32
%un
ilate
ral,
16%
vers
us0%
bila
tera
l).
IIID
isca
rded
patie
nts
and
lack
ofst
atis
tical
anal
ysis
prec
lude
firm
stat
emen
ts.
Hig
hly
sugg
estiv
epa
per.
Con
clus
ions
:tr
actio
nfo
llow
edby
man
ipul
atio
nis
safe
and
usua
llyef
fect
ive,
and
redu
ctio
nse
ems
toim
prov
e
outc
ome
(or
patie
nts
who
are
redu
cibl
edo
bette
r).
Bur
kean
dB
erry
man
,19
71
(3)
41pa
tient
str
eate
dby
MU
A.
Ligh
ttr
actio
nan
dth
enin
duct
ion
ofan
esth
esia
and
intu
batio
n,th
en
MU
Aif
nece
ssar
y(s
ame
asEv
ans
[10]
).
32pa
tient
str
eate
dw
ithtr
actio
nal
one.
3tr
eate
dby
trac
tion
afte
rM
UA
faile
d.
C7–
T1no
tat
tem
pted
.Pr
e-M
RI.
37/4
1su
cces
sful
lyre
duce
dby
MU
A.
21/2
5re
duce
dw
ithtr
actio
nbe
fore
anes
thet
ic.
7pa
tient
sw
ere
judg
edto
osi
ckfo
ran
esth
esia
and
unde
rwen
ttr
actio
nfo
rst
abili
zatio
n,no
tre
duce
d.
2ca
ses
ofne
urol
ogic
alde
teri
orat
ion:
1ov
erdi
stra
ctio
n,1
unre
cogn
ized
inju
ry.
IIIM
UA
and
trac
tion
both
safe
ifpr
oper
diag
nosi
san
d
care
ful
atte
ntio
npa
idto
radi
ogra
phs.
Evan
s,19
61(1
0)17
patie
nts.
Trea
ted
byin
duct
ion
ofan
esth
esia
and
intu
batio
n,
som
etim
esw
ithm
anip
ulat
ion
unde
ran
esth
esia
.Pr
e-M
RI.
No
neur
olog
ical
dete
rior
atio
nno
ted.
All
succ
essf
ully
redu
ced.
2un
chan
ged,
2di
ed,
13
impr
oved
.
IIIR
educ
tion
unde
ran
esth
esia
safe
and
effe
ctiv
e.
Smal
lse
ries
.
aM
RI,
mag
neti
cre
sona
nce
imag
ing;
ASI
A,A
mer
ican
Spin
alIn
jury
Ass
ocia
tion
;MU
A,m
anip
ulat
ion
unde
ran
esth
esia
;RT,
rapi
dtr
acti
on-r
educ
tion
;CT,
com
pute
dto
mog
raph
y;O
R,
open
redu
ctio
n;O
RIF
,op
enre
duct
ion
wit
hin
tern
alfix
atio
n;C
R,
clos
edre
duct
ion;
C-s
pine
,ce
rvic
alsp
ine.
Initial Closed Reduction of Fracture-Dislocation Injuries S47
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
vest application after closed reduction has been accomplished(14, 15, 18, 40).
Hadley et al. (14) suggest that closed reduction of acutecervical spine fracture-dislocation injuries is best accom-plished as part of the early overall medical management of thepotential SCI patient in the intensive care unit setting usingbedside fluoroscopy, with close monitoring of each patient’sclinical and neurological status, as well as cardiac, respiratory,and hemodynamic parameters. Pain control and modestsedation-relaxation is provided using short-acting intrave-nous agents that do not impair the patient’s level of conscious-ness or alter hemodynamic performance parameters (14, 18).
Craniocervical traction is typically arbitrarily initiated withthe application of weight beginning with 3 pounds per supe-rior injury level. By this scheme, a patient with an isolatedC5–C6 fracture-dislocation injury would begin treatment withan initial weight of 15 pounds (3 lb � 5 rostral vertebrallevels). Caveats to the use of this initial weight suggestioninclude patients with fracture-dislocations involving C2 andabove and patients with ankylosing spondylitis in whom verylittle weight, if any, may be needed to accomplish reduction.Judgment and experience must be used in this setting becausesome more-proximal cervical spine injuries may be distractioninjuries and will not require traction. These injuries are per-haps best managed with realignment and compression.
Weight may be added as often as every 10 to 15 minutes aslong as close clinical, neurological, and radiographic monitor-ing is reassessed throughout the process. No upper limit ofweight has been described in the literature. In general, in-creasing weight is applied until closed reduction and realign-ment occurs or until patient complaints are great (increasingintractable pain or a subjective worsening of neurologicalstatus), the patient’s neurological examination worsens,overdistraction occurs as noted on fluoroscopy or lateral cer-vical spine x-rays, it is impractical to add further weight (noadditional weight can be applied to the distraction device,no additional weight available, patient sliding upward inbed), or the clinician judges that closed reduction has failed.
After closed reduction has been accomplished, or once thedetermination has been made that closed reduction has failed,it is recommended that the patient be immobilized to providestabilization of the cervical spine injury before transport toradiology for further assessment or to the operating room forsurgical management (14, 18).
Herniated disc fragments causing compression of the cer-vical spinal cord at the level of facet dislocation have beendescribed by several authors (3, 8, 11, 16, 20, 22, 25, 28, 37).Eismont et al. (9) reported a series of 63 patients managedwith closed traction-reduction and then by open reduction ifclosed reduction was unsuccessful. One of these patientsworsened after posterior open reduction and fusion. A herni-ated disc was found ventral to the cord on postproceduremyelography. Herniated discs were found in three other pa-tients in whom closed reduction failed and in two patientswith static neurological deficits after fracture-dislocation re-duction (one open, one closed). One of these patients deteri-orated after subsequent anterior cervical discectomy and fu-sion. Eismont et al. (9) did not report their overall results with
closed reduction. However, it is clear from their case descrip-tions that no awake patient experienced neurological deteri-oration as a result of attempted closed reduction. Olerud andJohnson (26) described two patients found to have disc her-niations on postreduction MRI or CT myelography. Both pa-tients deteriorated after open reduction after failure of at-tempted closed reduction. Robertson and Ryan (29) reportedthree patients who deteriorated during management of cervi-cal subluxation injuries. One of their patients worsened dur-ing transport to the hospital. That patient’s vertebral injurywas found to have spontaneously partially reduced. MRIrevealed a disc fragment compressing the cord. A secondpatient deteriorated after posterior open reduction. MRI re-vealed disc fragments compressing the cord. The patient un-derwent subsequent anterior decompression. The third pa-tient deteriorated 3 days after successful closed reduction. Asubsequent MRI study demonstrated disc material compress-ing the ventral cervical spinal cord. Five days after deteriora-tion, the patient underwent anterior decompression. All threepatients recovered to at least their pre-deterioration neurolog-ical examination. Grant et al. (13) reported a single case ofneurological deterioration in their series associated withclosed reduction that also occurred in a delayed fashion (6 hafter reduction). This occurred in a patient subsequentlyfound to have a herniated disc at the level of injury.
Mahale et al. (24) reviewed 16 cases of neurological deteri-oration in patients with cervical spinal cord injuries afterreduction of facet dislocations. Seven of the 16 patients devel-oped complete cord injuries, 6 after open reduction, and 1after manipulation under anesthesia. Five patients developedpartial injuries, three after manipulation under anesthesia andtwo after closed traction-reduction. Of the two patients whodeteriorated after closed reduction, one was found to beoverdistracted. Minor injuries were sustained by the remain-ing four patients, including one who deteriorated when thecranial traction pins slipped, one who deteriorated in a plasterbrace, one who lost reduction and had neurological worsen-ing, and one patient who underwent open reduction. Sixpatients were investigated with myelography after deteriora-tion, two with MRI, and one with CT. A disc protrusion wasnoted in one patient, and a “disc prolapse with hematoma”was noted in another. Both of these patients were treatedconservatively. The most common imaging finding in thesenine patients was cord edema (20).
The prevalence of MRI-documented disc herniation in as-sociation with cervical facet injury with subluxation hascaused a number of authors to recommend prereduction MRIin patients with these injuries. Harrington et al. (16) reporteda series of 37 patients managed with closed reduction, inwhom a 97% rate of successful reduction was achieved withno neurological morbidity. Postreduction imaging revealeddisc herniations in nine patients, four of whom underwentlater anterior decompression. Doran et al. (8) reported a seriesof 13 patients drawn from four institutions during an unspec-ified time period. Nine patients were treated with attemptedearly closed reduction. Subluxations in three patients werereduced without incident; failure to reduce was noted inthree patients. Closed reduction was abandoned in an-
S48 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
other three patients because of worsening pain (one patient)or arm weakness (two patients). All patients underwent MRIevaluation (four prereduction). Herniated discs were visual-ized in 10 patients, bulging discs were imaged in 3. No treatedpatient developed a permanent neurological deficit as a resultof attempted closed reduction. No patient who underwentsuccessful closed reduction deteriorated. All three who hadinjuries that were successfully reduced showed significantneurological improvement despite the MRI appearance of adisc herniation in two and a disc bulge in the third. Theauthors concluded that prereduction MRI studies should beobtained before closed reduction in patients with cervicalspine facet dislocation injuries. Because there was no correla-tion between the presence of disc herniation and neurologicaldeterioration, the recommendation for a prereduction MRI inpatients with cervical facet dislocation injuries cannot besupported.
Vaccaro et al. (37) studied 11 consecutive patients with MRIbefore and after reduction. The authors found herniated discsin two patients in the prereduction group and in five of ninepatients who underwent successful closed reduction. Thepresence of a herniated disc on prereduction MRI or pos-treduction MRI did not predict neurological deterioration. Nocase of deterioration after successful reduction occurred (37).Grant et al. (13) obtained postreduction MRI studies on 80patients treated with closed reduction and found herniated orbulging discs in 46%. They found no correlation between MRIresults and neurological outcome. Rizzolo et al. (28) per-formed MRI prereduction on 55 patients with cervical frac-tures and dislocation injuries. They found evidence of discherniation in 54% of these patients. Awake and alert patientsunderwent closed traction-reduction. There were no instancesof neurological deterioration in this group. The authors didnot attempt closed reduction in patients who were not awake.
In summary, a review of the literature reveals only twodocumented cases (11, 25) of neurological deterioration asso-ciated with attempted closed reduction of cervical spinefracture-dislocation injuries caused by cord compression fromdisc herniation. Both of these cases were characterized bydeterioration hours to days after closed reduction. A numberof large clinical series have failed to establish a relationshipbetween the presence of a prereduction herniated disc andsubsequent neurological deterioration with attempted closedtraction-reduction in awake patients.
SUMMARY
Closed reduction of fracture-dislocation injuries of the cer-vical spine by traction-reduction seems to be safe and effectivefor the reduction of spinal deformity in awake patients. Ap-proximately 80% of patients will have their injuries reducedwith this technique. The overall permanent neurological com-plication rate of closed reduction is approximately 1%. Theassociated risk of a transient injury with closed reductionseems to be 2 to 4%. Closed traction-reduction seems to besafer than manipulation under anesthesia.
There are numerous causes of neurological deterioration inpatients with unstable cervical spine injuries. These include
inadequate immobilization, unrecognized rostral injuries,overdistraction, loss of reduction, and cardiac, respiratory,and hemodynamic instability. Therefore, the treatment of cer-vical spine fracture-dislocation injuries must be supervised byan appropriately trained specialist.
Although prereduction MRI will demonstrate disc herniationin up to half of patients with facet subluxation injuries, theclinical importance of these findings is questionable. Only twocase reports exist that document neurological deteriorationcaused by disc herniation after successful closed traction-reduction in awake patients. Both occurred in delayed fashionafter closed reduction. In addition, several investigators havedemonstrated the lack of correlation between the MRI finding ofdisc herniation and neurological deterioration in this patientpopulation. The use of prereduction MRI has therefore not beenshown to improve the safety or efficacy of closed traction-reduction in awake patients. MRI before fracture-dislocationreduction may result in unnecessary delays in accomplishingfracture realignment and decompression of the spinal cord. Be-cause Class III evidence exists in support of early closed reduc-tion of cervical fracture-dislocation injuries with respect to neu-rological function, prereduction MRI in this setting is notnecessary. The ideal timing of reduction is unknown, but manyinvestigators favor reduction as rapidly as possible after injury tomaximize the potential for neurological recovery.
Patients in whom attempted closed reduction of cervical frac-ture injuries fails have a higher incidence of anatomic obstaclesto reduction, including facet fractures and disc herniation. Pa-tients in whom closed reduction fails should undergo moredetailed radiographic study before attempts at open reduction.The presence of a significant disc herniation in this setting is arelative indication for an anterior decompression procedure, ei-ther in lieu of or preceding a posterior procedure.
Patients with cervical fracture-dislocation who cannot be ex-amined, because of head injury or intoxication, cannot be as-sessed for neurological deterioration during attempted closedtraction-reduction. For this reason, an MRI study before at-tempted reduction is recommended as a treatment option.
KEY ISSUES FOR FUTURE INVESTIGATION
A prospective cohort study of patients with cervical spinalcord injuries caused by facet fracture-subluxation injuries treatedwith or without prereduction MRI would provide Class II med-ical evidence in support of a treatment recommendation on thisissue. This could address issues of timing. A randomized con-trolled trial may provide Class I medical evidence.
No prospective comparative study has been performed ofclosed reduction versus anterior decompression and stabili-zation for patients with MRI-documented herniated discs inassociation with unreduced cervical fracture-dislocation inju-ries. A prospective study of this issue would provide Class IImedical evidence in support of a treatment recommendation.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
Initial Closed Reduction of Fracture-Dislocation Injuries S49
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
REFERENCES
1. Alexander E Jr, Davis CH Jr, Forsyth HF: Reduction and fusion offracture dislocation of the cervical spine. J Neurosurg 27:588–591,1967.
2. Brunette DD, Rockswold GL: Neurologic recovery followingrapid spinal realignment for complete cervical spinal cord injury.J Trauma 27:445–447, 1987.
3. Burke DC, Berryman D: The place of closed manipulation in themanagement of flexion-rotation dislocations of the cervical spine.J Bone Joint Surg Br 53B:165–182, 1971.
4. Cloward RB: Reduction of traumatic dislocation of the cervicalspine with locked facets: Technical note. J Neurosurg 38:527–531,1973.
5. Cotler JM, Herbison GJ, Nasuti JF, Ditunno JF Jr, An H, Wolff BE:Closed reduction of traumatic cervical spine dislocation usingtraction weights up to 140 pounds. Spine 18:386–390, 1993.
6. Cotler HB, Miller LS, DeLucia FA, Cotler JM, Dayne SH: Closedreduction of cervical spine dislocations. Clin Orthop 214:185–199,1987.
7. Crutchfield W: Skeletal traction in treatment of injuries to thecervical spine. JAMA 155:29–32, 1954.
8. Doran SE, Papadopoulos SM, Ducker TB, Lillehei KO: Magneticresonance imaging documentation of coexistent traumatic lockedfacets of the cervical spine and disc herniation. J Neurosurg79:341–345, 1993.
9. Eismont FJ, Arena MJ, Green BA: Extrusion of an intervertebraldisc associated with traumatic subluxation or dislocation of cer-vical facets. J Bone Joint Surg Am 73A:1555–1560, 1991.
10. Evans D: Reduction of cervical dislocations. J Bone Joint Surg Br43B:552–555, 1961.
11. Farmer J, Vaccaro A, Albert TJ, Malone S, Balderston RA, CotlerJM: Neurologic deterioration after spinal cord injury. J SpinalDisord 11:192–196, 1998.
12. Fehlings MG, Tator CH: An evidence based review of decompres-sive surgery in acute spinal cord injury: Rationale, indications,and timing based on experimental and clinical studies.J Neurosurg 91[Suppl 1]:1–11, 1999.
13. Grant GA, Mirza SK, Chapman JR, Winn HR, Newell DW, JonesDT, Grady MS: Risk of early closed reduction in cervical spinesubluxation injuries. J Neurosurg 90[Suppl 1]:13–18, 1999.
14. Hadley MN, Argires PJ: The acute/emergent management ofvertebral column fracture dislocation injuries, in NeurologicalEmergencies. Park Ridge, AANS, 1994, vol 2, pp 249–262.
15. Hadley MN, Fitzpatrick BC, Sonntag VKH, Browner CM: Facetfracture-dislocation injuries of the cervical spine. Neurosurgery30:661–666, 1992.
16. Harrington JF, Likavec MJ, Smith AS: Disc herniation in cervicalfracture subluxation. Neurosurgery 29:374–379, 1991.
17. Key A: Cervical spine dislocations with unilateral facet interlock-ing. Paraplegia 13:208–215, 1975.
18. Kilburn MP, Smith DP, Hadley MN: The initial evaluation andtreatment of the patient with spinal trauma, in Batjer HH, LoftusCM (eds): Textbook of Neurological Surgery: Principles and Practice.Philadelphia, Lippincott Williams & Wilkins (in press).
19. Kleyn PJ: Dislocations of the cervical spine: Closed reductionunder anaesthesia. Paraplegia 22:271–281, 1984.
20. Lee AS, MacLean JC, Newton DA: Rapid traction for reduction ofcervical spine dislocations. J Bone Joint Surg Br 76B:352–356, 1994.
21. Lu K, Lee TC, Chen HJ: Closed reduction of bilateral locked facetsof the cervical spine under general anaesthesia. Acta Neurochir(Wien) 40:1055–1061, 1998.
22. Ludwig SC, Vaccaro AR, Balderston RA, Cotler JM: Immediatequadriparesis after manipulation for bilateral cervical facet sub-luxation. J Bone Joint Surg Am 79A:587–590, 1997.
23. Mahale YJ, Silver JR: Progressive paralysis after bilateral facetdislocation of the cervical spine. J Bone Joint Surg Br 74B:219–223, 1992.
24. Mahale YJ, Silver JR, Henderson NJ: Neurological complicationsof the reduction of cervical spine dislocations. J Bone Joint SurgBr 75B:403–409, 1993.
25. Maiman DJ, Barolat G, Larson SJ: Management of bilateral lockedfacets of the cervical spine. Neurosurgery 18:542–547, 1986.
26. Olerud C, Jonsson H Jr: Compression of the cervical spinal cordafter reduction of fracture dislocations: Report of 2 cases. ActaOrthop Scand 62:599–601, 1991.
27. Osti OL, Fraser RD, Griffiths ER: Reduction and stabilisation ofcervical dislocations: An analysis of 167 cases. J Bone Joint SurgBr 71B:275–282, 1989.
28. Rizzolo SJ, Piazza MR, Cotler JM, Balderston RA, Schaefer D,Flanders A: Intervertebral disc injury complicating cervical spinetrauma. Spine 16(6 Suppl):S187–S189, 1991.
29. Robertson PA, Ryan MD: Neurological deterioration after reduc-tion of cervical subluxation: Mechanical compression by disctissue. J Bone Joint Surg Br 74B:224–227, 1992.
30. Rosenfeld JF, Vaccaro AR, Albert TJ, Klein GR, Cotler JM: Thebenefits of early decompression in cervical spinal cord injury.Am J Orthop 27:23–28, 1998.
31. Sabiston CP, Wing PC, Schweigel JF, Van Peteghem PK, Yu W:Closed reduction of dislocations of the lower cervical spine.J Trauma 28:832–835, 1988.
32. Schaefer DM, Flanders A, Northrup BE, Doan HT, Osterholm JL:Magnetic resonance imaging of acute cervical spine trauma.Spine 14:1090–1095, 1989.
33. Shrosbree RD: Neurological sequelae of reduction of fracturedislocations of the cervical spine. Paraplegia 17:212–221, 1979.
34. Sonntag VKH: Management of bilateral locked facets of the cer-vical spine. Neurosurgery 8:150–152, 1981.
35. Star AM, Jones AA, Cotler JM, Balderston RA, Sinha R: Immedi-ate closed reduction of cervical spine dislocations using traction.Spine 15:1068–1072, 1990.
36. Vaccaro AR, An HS, Lin S, Sun S, Balderston RA, Cotler JM:Noncontiguous injuries of the spine. J Spinal Disord 5:320–329,1992.
37. Vaccaro AR, Falatyn SP, Flanders AE, Balderston RA, NorthrupBE, Cotler JM: Magnetic resonance evaluation of the interverte-bral disc, spinal ligaments, and spinal cord before and after closedtraction-reduction of cervical spine dislocations. Spine 24:1210–1217, 1999.
38. Vital JM, Gille O, Senegas J, Pointillart V: Reduction technique foruni- and biarticular dislocations of the cervical spine. Spine 23:949–955, 1998.
39. Walton G: A new method of reducing dislocation of cervicalvertebrae. J Nerv Ment Dis 20:609, 1893.
40. Wilberger JE: Immobilization and traction, in Tator CH, BenzelEC (eds): Contemporary Management of Spinal Cord Injury: FromImpact to Rehabilitation. Park Ridge, AANS, 2000, pp 91–98.
41. Xiong XH, Bean A, Anthony A, Inglis G, Walton D: Manipulationfor cervical spinal dislocation under general anaesthesia: Serialreview for 4 years. Spinal Cord 36:21–24, 1998.
42. Yashon D, Tyson G, Vise WM: Rapid closed reduction of cervicalfracture dislocations. Surg Neurol 4:513–514, 1975.
S50 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
CHAPTER 7
Management of Acute Spinal Cord Injuries in an IntensiveCare Unit or Other Monitored Setting
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Management of patients with acute spinal cord injury, particularly patients with severe cervical level
injuries, in an intensive care unit or similar monitored setting is recommended.• Use of cardiac, hemodynamic, and respiratory monitoring devices to detect cardiovascular dysfunction and
respiratory insufficiency in patients after acute cervical spinal cord injury is recommended.
RATIONALE
The intensive care unit (ICU) setting has traditionally beenreserved for critically ill patients who require aggressivemedical care and exceptional medical attention. Most con-
temporary medical centers have multiple critical care units; eachunit is designed to provide discipline-specific observation andintensive care to patients in need. Select institutions have createdacute spinal cord injury centers and offer multidisciplinary care,including ICU care, to patients who have sustained acute spinalcord injuries (ASCIs) (2, 11, 12, 16, 21, 22, 25, 30–32, 34). Severalreports describe improved patient management and lower mor-bidity and mortality after ASCIs with ICU monitoring and ag-gressive medical management (11, 12, 16, 22, 25, 31, 32, 34).Despite this interest in and commitment to more comprehensivecare for the patient with an ASCI over the last 30 years by someindividuals and centers, many patients who sustain an ASCI arenot managed in an ICU setting, nor are they routinely monitoredfor cardiac or respiratory dysfunction. There exist divergentmanagement strategies for ASCI patients within regions, com-munities, even within institutions, depending on the trainingand experiences of the clinicians providing care. Recently com-pleted randomized clinical trials to investigate pharmacologicalagents in the treatment of ASCI patients did not suggest aspecific, common medical management paradigm to guide pa-tient care provided by participating investigators, other than thetiming and dosage of the pharmacological agents being tested(4–7, 9, 10). These studies included large numbers of seriouslyinjured ASCI patients managed outside the ICU setting, mostwithout continuous cardiac or respiratory monitoring.
QUESTIONS
1. Do patients with acute spinal cord injuries benefit fromcare in the ICU setting?
2. Is monitoring of cardiac, hemodynamic, and pulmonaryperformance of benefit to patients who have sustained anASCI?
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The following medical subject headings wereused in combination with “spinal cord injury”: medical man-agement, nonoperative management, surgical management,hypotension, respiratory insufficiency, pulmonary complica-tions, and ICU. Approximately 3400 citations were acquired.Non-English language citations were deleted. Titles and ab-stracts of the remaining publications were reviewed, and rel-evant articles were selected to develop this guideline. Wefocused on four specific topics concerning human patientswith ASCIs: management in an ICU (18 articles reviewed),cardiac instability (8 articles reviewed), hypotension (22 arti-cles reviewed), and respiratory/pulmonary dysfunction (20articles reviewed). Additional references were culled from thereference lists of the remaining papers. Finally, members ofthe author group were asked to contribute articles known tothem on the subject matter that were not found by othersearch means. Articles describing nonhuman laboratory in-vestigations germane to the topic and related general reviewarticles referenced in the Scientific Foundation are includedamong the 34 references in the reference list. Articles describ-ing economics, epidemiology, anesthesia, monitoring tech-niques, penetrating cord injuries, nursing care, infectious orurological complications, chronic complications, or patientswith remote spinal cord injuries (SCIs) were not included.These efforts resulted in 17 manuscripts, all of which arereports of case series (Class III medical evidence), which form
S51Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
the foundation for this guideline. These articles are summa-rized in Table 7.1.
SCIENTIFIC FOUNDATIONThe pathophysiology of ASCI is complex and multifaceted.
It involves a primary mechanical injury by way of compres-sion, penetration, laceration, shear, and/or distraction. Theprimary injury seems to initiate a host of secondary injury
mechanisms, including: 1) vascular compromise leading toreduced blood flow, loss of autoregulation, loss of microcir-culation, vasospasm, thrombosis, and hemorrhage; 2) electro-lyte shifts, permeability changes, loss of cellular membraneintegrity, edema, and loss of energy metabolism; and 3) bio-chemical changes including neurotransmitter accumulation,arachidonic acid release, free radical and prostaglandin pro-duction, and lipid peroxidation (1, 13, 26–29). These mecha-
TABLE 7.1. Summary of Reports on Management of Patients with Acute Spinal Cord Injuries in an Intensive CareUnit Settinga
Series (Ref. No.) Description of Study Evidence Class Conclusions
Lu et al., 2000 (18) Retrospective review of apnea in 36 ASCI patients. III Delayed apnea most likely in ASCI patients with severe,diffuse ASCI. Apnea most likely within first 7–10 d.
Botel et al., 1997 (3) 225 ASCI patients treated in ICU. Only 87 admittedwithin 24 h of injury.
III Significant numbers of multiply injured and head-injuredpatients. No complete injury recorded. Improved outcomewhen admitted to ICU early after injury.
Vale et al., 1997 (32) Prospective assessment of 77 ASCI patients treated inICU, aggressive hemodynamic support, MAP �85.
III Improved outcome with aggressive medical care, distinctfrom potential benefit from surgery at 1-yr follow-up.
Levi et al., 1993 (16) 50 patients treated in ICU, aggressive medicaltreatment, MAP �90.
III Improved outcome with aggressive hemodynamic support at6 wk postinjury.
Tator et al., 1993 (30) 201 ASCI patients, ICU care, hemodynamic supportcompared with 351 earlier patients.
III Less severe cord injuries due to immobilization,resuscitation, and early transfer to ICU setting.
Levi et al., 1991 (17) 103 ASCI patients: 50 incomplete (Group A), 53complete (Group B). ICU care, hemodynamicsupport, MAP �85.
III Improved neurological outcome, no significant differencebetween early and late surgery in either group.
Wolf et al., 1991 (33) 52 patients with locked facets reduced within 4 h,ICU care, MAP �85. 49 operated on, 23 Day 1, 26delayed (8.7 d mean).
III Closed reduction, 61%.Closed (a), 15%.52% follow-up at 1 y; in general, improved neurologicaloutcome.
Lehmann et al., 1987 (15) 71 consecutive ASCI patients, ICU care, monitoringof cardiac/hemodynamic parameters.
III Bradycardia, 100%; hypotension (�90 systolic); 68% life-threatening bradyarrhythmias; 16% incidence related toseverity of SCI.
Reines and Harris, 1987 (22) 123 cases. ASCI patients in ICU, aggressivepulmonary treatment.
III Respiratory insufficiency major cause of morbidity andmortality after ASCI. Aggressive ICU care, pulmonarytreatment reduces incidence.
Piepmeier et al., 1985 (21) 45 ASCI patients, all managed in ICU setting withcardiac, hemodynamic monitoring.
III Cardiac dysrhythmia, hypotension, and hypoxia common infirst 2 wk after ASCI. Incidence related to severity of injury.
Bose et al., 1984 (2) 28 ASCI patients, 22 managed in ICU setting.Group I: medical treatment.Group II: medical/surgical treatment.
III Improved neurological outcome at discharge for Group IIbut better scores initially. Group I with intrinsic cord injuryversus Group II compression on myelography and/orinstability.
Tator et al., 1984 (31) 144 ASCI patients. ICU care, hemodynamic support,compared with prior series.
III Improved neurological outcome, less mortality with earlytransfer, avoidance of hypotension, and ICU care.
Ledsome and Sharp, 1981 (14) Reassessment of pulmonary function in ASCI patients,comparison over time.
III Reduced vital capacity, flow rates, and hypoxia after ASCI.Incidence related to severity of SCI. Marked improvement inpulmonary functions 3 mo postinjury.
McMichan et al., 1980 (20) Prospective study of pulmonary complications in 22ASCI patients compared with 22 earlier patientsmanaged with aggressive ICU care.
III No deaths in series versus 9 of 22 deaths in earlier group.ICU care and vigorous pulmonary therapy improvessurvival, reduces complications.
Gschaedler et al., 1979 (11) 51 ASCI patients managed in ICU, aggressive medicaltreatment, avoidance of hypotension.
III Improved morbidity and mortality with early transfer,avoidance of hypotension, respiratory insufficiency.
Hachen, 1977 (12) 188 ASCI patients managed in ICU, aggressivetreatment of hypotension, respiratory insufficiency.
III Reduced morbidity and mortality with early transfer,attentive ICU care and monitoring, and aggressive treatmentof hypotension and respiratory failure.
Zach et al., 1976 (34) 117 ASCI patients at Swiss Center, ICU setting,aggressive blood pressure and volume therapy.Rheomacrodex � 5 d.Dexamethasone � 10 d.
III Improved neurological outcome with aggressive medicaltreatment. Better outcome for early referrals.
a ASCI, acute spinal cord injury; ICU, intensive care unit; MAP, mean arterial pressure; closed (a), closed reduction under anesthesia.
S52 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
nisms, if unchecked, result in axonal disruption and cellulardeath. A number of contemporary reviews describe thesetheories and provide experimental evidence in their support(1, 26–28).
Animal models of SCI suggest that ischemia of the spinalcord underlies much of the mechanism of posttraumatic SCIand is the important common denominator resulting in neu-rological deficit after primary injury (1, 26, 27). Ischemiaseems to be related to both local and systemic vascular alter-ations after severe injury. Local vascular alterations are due tothe direct SCI and focal, postinjury vasospasm, both of whichlead to loss of autoregulation of spinal cord blood flow (1, 8,13, 23, 24, 26, 27, 29). Systemic vascular alterations of bloodflow to the spinal cord after ASCI observed in both animalstudies and in human SCI patients include reduced heart rate,cardiac rhythm irregularities, reduced mean arterial bloodpressure, reduced peripheral vascular resistance, and com-promised cardiac output (1, 8, 15, 17, 18, 21, 24, 26, 27, 29, 32).Any of these untoward hemodynamic occurrences can con-tribute to systemic hypotension after severe injury (15–17, 21,25–27, 32). Systemic hypotension in the setting of ASCI, withcoincident loss of spinal cord autoregulatory function, com-pounds local spinal cord ischemia by further reducing spinalcord blood flow and perfusion (1, 26, 27, 29).
Respiratory insufficiency and pulmonary dysfunction iscommon after traumatic SCI, particularly when the injuryoccurs at cervical spinal cord levels (11, 12, 14, 18–20, 22, 25).Severely injured patients demonstrate marked reductions inexpected vital capacity and inspiratory capacity and mayexperience relative hypoxemia, all of which contribute toglobal hypoxemia and can exacerbate spinal cord ischemiaafter acute injury (14, 18–20, 22). It seems that the earliercardiac and/or ventilatory/pulmonary dysfunction is de-tected, the more likely effective, often life-saving treatmentcan be initiated. It is for these reasons that the issues of earlyICU care and cardiac and pulmonary monitoring for humanpatients after ASCI have been raised.
Several clinical series have been reported in which humanpatients with ASCIs have been managed in ICU environmentswith attention to heart rate, cardiac function, pulmonary per-formance, and mean arterial blood pressure (2, 3, 11, 12,14–22, 25, 31–34). Zach et al. (34), in 1976, provided a prelim-inary report on their prospective medical management para-digm in the treatment of 117 consecutive ASCI patients in theSwiss Paraplegic Center of Basel, Switzerland. All patientswere treated in the ICU with central venous pressure moni-toring and were given dexamethasone, 0.5 mg/kg for 4 days,with a tapering dose through 10 days, and volume expansionwith Rheomacrodex 40 (Medisan, Parsippany, NJ), 500 ml/dfor 7 days. Patients were stratified by injury level, degree ofdeficit (Frankel grade), and time of admission after injury. Theauthors reported that 62% of cervical level SCI patients theymanaged in this way improved at last follow-up, including 8of 18 Frankel Grade A patients, two patients by two gradesand one patient by three grades. No patient with a cervicalinjury worsened; 38% were unchanged from admission. Pa-tients with thoracic T1–T10 level SCIs fared less well; 38%improved, none worsened, and 62% were without change,
including 22 of 26 Frankel Grade A patients. Two FrankelGrade A patients experienced a complete recovery. Seventypercent of acute T11–L1 level SCIs improved with this treat-ment paradigm, none worsened, and 30% were unchangedfrom admission. Of patients who arrived within 12 hours ofinjury, 67% improved compared with their admission neuro-logical examination. Of patients admitted between 12 and 48hours after injury, only 59% improved. When admission oc-curred 48 hours after injury, improvement was seen in only50% of patients. The authors concluded that early transfer andimmediate medical specific treatment of the spinal injuryseemed to improve neurological recovery (34).
Hachen (12), in 1977, reported the 10-year experience withacute traumatic tetraplegia from the National Spinal InjuriesCenter in Geneva, Switzerland. He described 188 ASCI pa-tients treated in a 10-year period in the ICU setting afterimmediate transfer from the scene of the injury. The centerreported a marked reduction in mortality rates after acutecervical SCI compared with annual statistics from 1966. Mor-tality for complete tetraplegia was reduced from 32.5 to 6.8%over the 10-year period. Mortality for patients with incom-plete tetraplegia dropped from 9.9% in 1966 to 1.4% in 1976.Most early deaths in the center’s experience were related topulmonary complications. The likelihood of severe respira-tory insufficiency was related to the severity of the cervicalSCI. Seventy percent of patients with complete lesions expe-rienced severe respiratory insufficiency in the center’sexperience, compared with 27% of patients with incompletelesions. The improvement in mortality rates described wasdirectly related to early monitoring and treatment of respira-tory insufficiency in the ICU setting. Hachen stressed thatfacilities for continuous monitoring of central venous pres-sure, arterial pressure, pulse, respiration rate and pattern, andoxygenation-perfusion parameters must be available for allpatients with neurological injuries after ASCI, particularlythose injuries above the C6 level.
In 1979, Gschaedler et al. (11) described the comprehensivemanagement of 51 patients with acute cervical SCI in the ICUsetting in Colmar, France. Forty percent of the patients theymanaged had multiple organ system injuries. The authorsreported a low mortality rate (7.8%) and described severalseverely injured patients who made important neurologicalimprovements, including one patient improving from FrankelGrade A to Grade D and two patients improving fromFrankel Grade B to Grade D. The authors cited early transportafter injury and comprehensive intensive medical care withattention to and avoidance of hypotension and respiratoryinsufficiency as essential to the improved management out-come they experienced.
McMichan et al. (20) reported in 1980 their prospectiveassessment of pulmonary complications identified in 22 pa-tients with cervical level ASCI managed in the ICU setting.They compared their results with 22 retrospective patientswith similar injuries. Use of a new, aggressive pulmonarytreatment paradigm resulted in no deaths and fewer respira-tory complications compared with those experienced by theretrospective group (nine deaths). The authors concluded thatvigorous pulmonary therapy initiated early after ASCI was
Management of Acute Spinal Cord Injuries in an ICU Setting S53
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
associated with increased survival, a reduced incidence ofpulmonary complications, and a decreased need for ventila-tory support.
Ledsome and Sharp (14) measured pulmonary function in16 cervical level complete ASCI patients and compared initialvalues with those obtained in the same patients at 1, 3, and 5weeks and 3 and 5 months after injury. In their 1981 report,they noted profound reductions in forced vital capacity (FVC)and expiratory flow rates immediately after injury. Patientswith an FVC less than 25% of expected had a high incidenceof respiratory failure requiring ventilator support. This wasespecially true of patients with injuries at C4 or above. Theauthors found a significant increase in FVC at 5 weeks postin-jury and an approximate doubling of FVC at 3 months, irre-spective of the level of cervical cord injury. Importantly, theyidentified hypoxemia (PO2 � 80 mm Hg) in most of theirpatients (74% of those who did not require ventilator sup-port), despite adequate alveolar ventilation (PCO2 normaldespite low FVC). Ledsome and Sharp attributed this to aventilation perfusion imbalance that occurs immediately afterASCI. Systemic hypoxemia was identified by blood gas mea-surements and was effectively treated with the addition ofsupplemental oxygen in most patients.
Piepmeier et al. (21) identified cardiovascular instabilityafter acute cervical SCI in 45 patients they managed in the ICUsetting in New Haven, CT. Twenty-three patients had FrankelGrade A injuries, eight were Grade B, seven Grade C, andseven Grade D. The authors discovered a high incidence ofcardiovascular irregularities in these patients and identified adirect correlation between the severity of the cord injury andthe incidence and severity of cardiovascular problems. Threepatients returned to the ICU setting during the 2-week obser-vation period of the study because of cardiac dysfunction,despite a period of initial stability. Of the 45 patients, 29 hadan average daily pulse rate of less than 55 beats per minute; 32had episodes during which their pulse rate was below 50beats per minute for prolonged periods. Hypotension wascommon after ASCI in their series, but most patients re-sponded well to volume replacement. Nine patients requiredvasopressors to maintain a systolic pressure above 100mm Hg, therapy that ranged from hours to 5 days duration.Cardiac arrest occurred in five (11%) patients. All had FrankelGrade A injuries. Three arrests occurred during endotrachealsuctioning. The authors found that the first week after injurywas the timeframe during which patients were most vulner-able to cardiovascular instability. Patients with the most se-vere neurological injuries were most likely to experience car-diovascular instability after ASCI. These events occurreddespite the absence of complete autonomic disruption. Hyp-oxia and endotracheal suctioning were associated with car-diac arrest in most instances. Piepmeier et al. concluded thatcareful monitoring of severely injured ASCI patients in theICU setting reduces the risk of life-threatening emergencies.
In 1984, Tator et al. (31) described their experience with 144patients with ASCIs managed between 1974 and 1979 at adedicated SCI unit at Sunnybrook Medical Centre in Toronto,Ontario, Canada. The authors compared their results with acohort of 358 SCI patients managed between 1948 and 1973,
before the development of the acute care SCI facility. All 144patients managed from 1974 to 1979 were treated in an ICUsetting with strict attention to the treatment of hypotensionand respiratory failure. Their medical paradigm was devel-oped on the principle that avoiding hypotension is one of themost important aspects of the immediate management ofacute cord injury (31). Hypotension was treated vigorouslywith crystalloid and transfusion of whole blood or plasma forvolume expansion. Patients with respiratory dysfunctionwere treated with ventilatory support as indicated. They re-ported that mean time from injury to admission and treatmentwas 4.9 hours during the period 1974 to 1979 compared withmore than 12 hours during the period 1948 to 1973. Neuro-logical improvement was observed in 41 (43%) of 95 patientsmanaged under the aggressive ICU medical paradigm; 52patients (55%) demonstrated no improvement; only 2 patients(2%) deteriorated. The authors reported lower mortality, re-duced morbidity, shorter length of stay, and lower cost oftreatment with their contemporary comprehensive manage-ment paradigm compared with the 1948 to 1973 experience.They cited improved respiratory management in their ICU asone of the principal factors responsible for reduced mortalityand credited the avoidance of hypotension, sepsis, and uro-logical complications for reduced morbidity after injury.These improved management results were realized despitethe fact that 28% of the ASCI patients they treated had addi-tional injuries that increased their risk of morbidity andmortality.
Lehmann et al. (15), in a follow-up study in 1987, reportedon 71 ASCI patients they managed in the ICU at the Yale/New Haven Medical Center. Patients were admitted within 12hours of injury and were stratified by level and severity ofneurological injury (Frankel scale). No patient had an associ-ated head injury, a history of diabetes mellitus, a preexistingcardiac disorder, or a history of cardiac medication use. Allwere monitored and aggressively treated to avoid hypoten-sion. The authors found that all patients with severe cervicalSCIs, Frankel Grades A and B, had persistent bradycardia,defined as a heart rate below 60 beats per minute for at least1 day. Thirty-five percent of Frankel Grade C and D patientswere identified to have persistent bradycardia. Only 13% ofthoracic and lumbar SCIs had this finding. Similarly, markedbradycardia, below 45 beats per minute, was frequent in thesevere cervical injury group (71%) and less common in themilder cervical injury (12%) and thoracolumbar injury pa-tients (4%). Sinus node slowing was often profound enough toproduce hemodynamic compromise and systemic hypoten-sion. Bolus injections of atropine or placement of a temporarypacemaker was often performed. This therapy was requiredby 29% of the severe cervical injury patients and by none inthe two other injury groups. Episodic hypotension unrelatedto hypovolemia was identified in 68% of the severe cervicalinjury group and in none of the other two injury groups.Thirty-five percent of the severe cervical injury group patientsrequired the use of intravenous pressors to maintain an ac-ceptable blood pressure. Five (16%) of 31 patients in thesevere injury group experienced a primary cardiac arrest,three of which were fatal. All five patients had Frankel Grade
S54 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
A SCIs. No patient in the study experienced a significantcardiac rate disturbance or spontaneous episode of hypoten-sion beyond 14 days of injury. The authors concluded thatpotentially life-threatening cardiac arrhythmias and episodesof hypotension regularly accompanied acute severe injury tothe cervical spinal cord within the first 14 days of injury.These events were not solely attributable to disruption of theautonomic nervous system. Detection and treatment was bestaccomplished in the ICU setting.
Wolf et al. (33), in 1991, described their experience withbilateral facet dislocation injuries of the cervical spine at theUniversity of Maryland in Baltimore. The authors described52 patients with acute cervical trauma managed with an ag-gressive treatment paradigm that included ICU care, aggres-sive resuscitation, invasive monitoring, and hemodynamicmanipulation to maintain mean blood pressure above 85mm Hg for 5 days. Thirty-four patients had complete neuro-logical injuries, 13 had incomplete injuries, and 5 patientswere intact. The authors attempted closed reduction within 4hours of patient arrival to their center and performed earlyopen reduction on patients who could not be reduced byclosed means, including closed reduction under anesthesia.All but three patients underwent surgery for stabilization andfusion. The authors reported neurological improvement atdischarge in 21% of complete SCI patients and in 62% ofpatients with incomplete cervical SCI at admission. No intactpatient deteriorated. Only 52% 1-year follow-up was pro-vided. The authors concluded that their protocol of aggres-sive, early medical and surgical management of patients withASCIs improved outcome after injury. Treatment in the ICUsetting, hemodynamic monitoring with maintenance of meanarterial pressure, and early decompression of the spinal cordby open or closed means seemed to reduce secondary com-plications after ASCIs in their study.
Levi et al. (16) treated 50 acute cervical SCI patients in theICU at the University of Maryland in Baltimore according toan aggressive management protocol that included invasivehemodynamic monitoring and volume and pressor support tomaintain a hemodynamic profile with adequate cardiac out-put and mean blood pressure above 90 mm Hg. Their 1993report described 31 patients with Frankel Grade A injuries atadmission, 8 patients with Frankel Grade B injuries, and 11patients in Frankel Grades C and D. Eight patients had shockat the time of admission (systolic blood pressure �90mm Hg), and 82% of patients had volume-resistant hypoten-sion requiring pressors within the first 7 days of treatment.This was 5.5 times more common among patients with com-plete motor injuries. The authors reported that the overallmean peripheral vascular resistance index for the 50 patientsthey studied was less than the normal range, and it was lessthan the normal value in 58% of patients. Half of their ASCIpatients had a lower than normal systemic vascular resistanceindex. No patient with a complete motor deficit (FrankelGrades A and B) and marked deficits in indexes of peripheraland/or systemic vascular resistance experienced neurologicalrecovery at 6 weeks. At 6 weeks after injury, 40% of patientsmanaged by protocol improved, including several with com-plete injuries, 42% remained unchanged, and 18% (9 patients)
died. Minimal morbidity was associated with invasive hemo-dynamic monitoring. The authors concluded that hemody-namic monitoring in the ICU allows early identification andprompt treatment of cardiac dysfunction and hemodynamicinstability and can reduce the potential morbidity and mor-tality after ASCI.
Vale et al. (32), in 1997, reported their experience with anonrandomized, prospective pilot study in the assessment ofaggressive medical resuscitation and blood pressure manage-ment in 77 consecutive ASCI patients treated at the Universityof Alabama in Birmingham. There was no control group. Allpatients were managed in the ICU with invasive monitoring(Swan Ganz catheters and arterial lines) and blood pressureaugmentation to maintain mean arterial pressure (MAP)above 85 mm Hg for 7 days postinjury. The authors reported10 patients with complete cervical SCIs (American SpinalCord Injury Association [ASIA] Grade A), 25 patients withincomplete cervical injuries (ASIA Grades B, C, and D), 21patients with complete thoracic SCIs, and 8 patients withincomplete thoracic level SCIs (Grades B, C, and D). Theaverage admission MAP for Grade A cervical SCI patientswas 66 mm Hg. Nine of 10 patients required pressors aftervolume replacement to maintain MAP of 85 mm Hg. Fifty-two percent of incomplete cervical SCI patients required pres-sors to maintain MAP at 85 mm Hg. Only 9 of 29 patients withthoracic level SCIs required the use of pressors. The authorsreported minimal morbidity with the use of invasive moni-toring or with pharmacological therapy to augment MAP. At1-year follow-up (mean, 17 mo), neurological recovery wasvariable and typically incomplete. Three of 10 ASIA Grade Acervical SCI patients regained ambulatory capacity, and tworegained bladder function. Incomplete cervical SCI patientsfared better. Twenty-three of these patients regained ambula-tory function at 12 months follow-up, only four of whom hadinitial examination scores consistent with walking. Twenty-two (88%) of 25 patients regained bladder control. Thirty-oneof 35 cervical SCI patients and 27 of 29 thoracic level SCIpatients were treated surgically. The authors statistically com-pared selection for and timing of surgery with admissionneurological function and compared surgical treatment, earlyand late, with neurological outcome and found no statisticalcorrelation. They concluded that the enhanced neurologicaloutcome identified in their series after ASCI was optimized byearly and aggressive volume resuscitation and blood pressureaugmentation and was in addition to and/or distinct fromany potential benefit provided by surgery.
SUMMARY
Patients with severe ASCIs, particularly cervical level inju-ries, or patients with multisystem traumatic injury, frequentlyexperience hypotension, hypoxemia, and pulmonary dys-function, and many exhibit cardiovascular instability, despiteearly acceptable cardiac and pulmonary function after initialresuscitation. These occurrences are not limited to ASCI pa-tients with complete autonomic disruption. Life-threateningcardiovascular instability and respiratory insufficiency maybe transient and episodic and may occur in patients who seem
Management of Acute Spinal Cord Injuries in an ICU Setting S55
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
to have stable cardiac and respiratory function early in theirpostinjury course. Patients with the most severe neurologicalinjuries after ASCI seem to have the greatest risk of theselife-threatening events. Monitoring allows the early detectionof hemodynamic instability, cardiac rate disturbances, pulmo-nary dysfunction, and hypoxemia. Identification and treat-ment of these events seems to reduce cardiac- and respiratory-related morbidity and mortality. Management in an ICU orsimilar setting with cardiovascular and pulmonary monitor-ing has an effect on neurological outcome after ASCI. Patientswith ASCIs seem to be best managed in the ICU setting for thefirst 7 to 14 days after injury, the timeframe during which theyseem most susceptible to significant fluctuations in cardiacand pulmonary performance. This seems to be particularlytrue for severe cervical SCI patients, specifically acute ASIAGrades A and B.
KEY ISSUES FOR FUTURE INVESTIGATION
The length of stay in the ICU setting necessary to provideoptimal management of patients with ASCI is unknown. Theavailable evidence suggests that most untoward and poten-tially life-threatening cardiac and respiratory events occurwithin the first week or two after injury. Patients with lesssevere ASCIs may require less time in a monitored settingthan patients with more severe injuries. These issues could beaddressed in a prospective cohort study or, potentially, aretrospective case-control study.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Amar AP, Levy ML: Pathogenesis and pharmacological strategiesfor mitigating secondary damage in acute spinal cord injury.Neurosurgery 44:1027–1040, 1999.
2. Bose B, Northrup BE, Osterholm JL, Cotler JM, DiTunno JF:Reanalysis of central cervical cord injury management. Neuro-surgery 15:367–372, 1984.
3. Botel U, Gläser E, Niedeggen A: The surgical treatment of acutespinal paralysed patients. Spinal Cord 35:420–428, 1997.
4. Bracken MB, Collins WF, Freeman DF, Shepard MJ, Wagner FW, SiltenRM, Hellenbrand KG, Ransohoff J, Hunt WE, Perot PL Jr, GrossmanRG, Green BA, Eisenberg HM, Rifkinson N, Goodman JH, Meagher JN,Fischer B, Clifton GL, Flamm ES, Rawe SE: Efficacy of methylpred-nisolone in acute spinal cord injury. JAMA 251:45–52, 1984.
5. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC, MarshallLF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr,Wilberger JL, Winn HR, Young W: Methylprednisolone or nalox-one treatment after acute spinal cord injury: 1-year follow updata—Results of the Second National Acute Spinal Cord InjuryStudy. J Neurosurg 76:23–31, 1992.
6. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized trial of methylpred-nisolone or naloxone in the treatment of acute spinal cord injury:Results of the Second National Acute Spinal Cord Injury Study(NASCIS II). N Engl J Med 322:1405–1411, 1990.
7. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings M, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, Sonntag VKH,Wagner F, Wilberger JE, Winn HR, Young W: Administration ofmethylprednisolone for 24 or 48 hours or tirilazad mesylate for 48hours in the treatment of acute spinal cord injury: Results of theThird National Acute Spinal Cord Injury Randomized ControlledTrial—National Acute Spinal Cord Injury Study. JAMA 277:1597–1604, 1997.
8. Dolan EJ, Tator CH: The effect of blood transfusion, dopamine,and gamma hydroxybutyrate on posttraumatic ischemia of thespinal cord. J Neurosurg 56:350–358, 1982.
9. Geisler FH: GM-1 ganglioside and motor recovery following hu-man spinal cord injury. J Emerg Med 11[Suppl 1]:49–55, 1993.
10. Geisler FH, Dorsey FC, Coleman WP: Recovery of motor functionafter spinal-cord injury: A randomized, placebo-controlled trialwith GM-1 ganglioside. N Engl J Med 324:1829–1838, 1991.
11. Gschaedler R, Dollfus P, Molé JP, Molé L, Loeb JP: Reflections onthe intensive care of acute cervical spinal cord injuries in a generaltraumatology centre. Paraplegia 17:58–61, 1979.
12. Hachen HJ: Idealized care of the acutely injured spinal cord inSwitzerland. J Trauma 17:931–936, 1977.
13. Hall ED, Wolf DL: A pharmacological analysis of the pathophys-iological mechanisms of posttraumatic spinal cord ischemia.J Neurosurg 64:951–961, 1986.
14. Ledsome JR, Sharp JM: Pulmonary function in acute cervical cordinjury. Am Rev Respir Dis 124:41–44, 1981.
15. Lehmann KG, Lane JG, Piepmeier JM, Batsford WP: Cardiovas-cular abnormalities accompanying acute spinal cord injury inhumans: Incidence, time course and severity. J Am Coll Cardiol10:46–52, 1987.
16. Levi L, Wolf A, Belzberg H: Hemodynamic parameters in patientswith acute cervical cord trauma: Description, intervention, andprediction of outcome. Neurosurgery 33:1007–1017, 1993.
17. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does the timingof surgery affect the outcome? Neurosurgery 29:216–222, 1991.
18. Lu K, Lee TC, Liang CL, Chen HJ: Delayed apnea in patients withmid- to lower cervical spinal cord injury. Spine 25:1332–1338,2000.
19. Mansel JK, Norman JR: Respiratory complications and manage-ment of spinal cord injuries. Chest 97:1446–1452, 1990.
20. McMichan JC, Michel L, Westbrook PR: Pulmonary dysfunctionfollowing traumatic quadriplegia: Recognition, prevention, andtreatment. JAMA 243:528–531, 1980.
21. Piepmeier JM, Lehmann KB, Lane JG: Cardiovascular instabilityfollowing acute cervical spinal cord trauma. Cent Nerv SystTrauma 2:153–160, 1985.
22. Reines HD, Harris RC: Pulmonary complications of acute spinalcord injuries. Neurosurgery 21:193–196, 1987.
23. Sandler AN, Tator CH: Effect of acute spinal cord compressioninjury on regional spinal cord blood flow in primates.J Neurosurg 45:660–676, 1976.
24. Sandler AN, Tator CH: Review of the effect of spinal cord traumaon the vessels and blood flow in the spinal cord. J Neurosurg45:638–646, 1976.
25. Tator CH: Vascular effects and blood flow in acute spinal cordinjuries. J Neurosurg Sci 28:115–119, 1984.
26. Tator CH: Ischemia as a secondary neural injury, in Salzman SK,Faden AI (eds): Neurobiology of Central Nervous System Trauma.New York, Oxford University Press, 1994, pp 209–215.
S56 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
27. Tator CH: Experimental and clinical studies of the pathophysiol-ogy and management of acute spinal cord injury. J Spinal CordMed 19:206–214, 1996.
28. Tator CH: Biology of neurological recovery and functionalrestoration after spinal cord injury. Neurosurgery 42:696–708,1998.
29. Tator CH, Fehlings MG: Review of the secondary injury theory ofacute spinal cord trauma with emphasis on vascular mechanisms.J Neurosurg 75:15–26, 1991.
30. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Changes in epidemiology of acute spinal cord injury from 1947 to1981. Surg Neurol 40:207–215, 1993.
31. Tator CH, Rowed DW, Schwartz MI, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–293, 296, 1984.
32. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medical andsurgical treatment after acute spinal cord injury: Results of a prospectivepilot study to assess the merits of aggressive medical resuscitation andblood pressure management. J Neurosurg 87:239–246, 1997.
33. Wolf A, Levi L, Mirvis S, Ragheb J, Huhn S, Rigamonti D,Robinson WL: Operative management of bilateral facet disloca-tion. J Neurosurg 75:883–890, 1991.
34. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cord injuriesin the Swiss Paraplegic Centre of Basel. Paraplegia 14:58–65, 1976.
Drawings by Leonardo da Vinci of the human spine. His representation of the spinal column was perhaps the first to showthe correct curvatures, articulations, and number of vertebrae. Courtesy, Dr. Edwin Todd, Pasadena, California.
Management of Acute Spinal Cord Injuries in an ICU Setting S57
CHAPTER 8
Blood Pressure Management after Acute Spinal Cord Injury
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Hypotension (systolic blood pressure <90 mm Hg) should be avoided if possible or corrected as soon as
possible after acute spinal cord injury.• Maintenance of mean arterial blood pressure at 85 to 90 mm Hg for the first 7 days after acute spinal cord
injury to improve spinal cord perfusion is recommended.
RATIONALE
Acute traumatic spinal cord injury is frequently associ-ated with systemic hypotension. Hypotension may beattributable to associated traumatic injuries with hypo-
volemia, direct severe spinal cord trauma itself, or a combi-nation. The occurrence of hypotension has been shown to beassociated with worse outcomes after traumatic injury, in-cluding severe head injury (1, 2, 8, 16, 21, 25). Although aprospective controlled assessment of the effects of hypoten-sion on acute spinal cord injury (ASCI) in humans has notbeen performed, laboratory evidence suggests that hypoten-sion contributes to secondary injury after ASCI by furtherreducing spinal cord blood flow and perfusion (1, 3, 4, 8, 16,18–22, 25). Hypotension in animal models of spinal cord in-jury (SCI) results in worse neurological outcome (13, 14, 23,26, 28, 29). Several clinical series of human patients with ASCImanaged in an aggressive fashion with attention to bloodpressure, oxygenation, and hemodynamic performance reportno deleterious effects of therapy and suggest improved neu-rological outcome (13, 14, 23, 26, 28, 29). Despite these obser-vations, most patients with ASCI treated in contemporarypractice are not routinely monitored or treated with bloodpressure augmentation after injury. For these reasons, theissues of routine blood pressure support and threshold levelsof mean arterial pressure maintenance after ASCI have beenraised.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature from 1966 to 2001 was undertaken.The following medical subject headings were used in combi-nation with “spinal cord injury”: medical management, non-operative management, hypotension, and spinal cord bloodflow. Approximately 3000 citations were acquired. Non-English language citations were deleted. Titles and abstractsof the remaining publications were reviewed, and relevantarticles were selected to develop the guidelines. We focused
on two specific topics concerning human patients with ASCI:hypotension (22 articles reviewed) and spinal cord blood flow(no articles identified). Additional references were culledfrom the reference lists of the remaining papers. Finally, mem-bers of the author group were asked to contribute articlesknown to them on the subject matter that were not found byother search means. Articles describing nonhuman laboratoryinvestigations germane to the topic, related general reviewarticles, and relevant studies of hypotension and human trau-matic brain injury referenced in the Scientific Foundation areincluded among the 29 citations in the references. These ef-forts resulted in six articles describing clinical case series(Class III medical evidence), which form the foundation forthis guideline. They are summarized in Table 8.1.
SCIENTIFIC FOUNDATION
Ischemia of the spinal cord is thought to be one of the mostimportant contributors to neuronal injury and neurologicaldeficit after ASCI. Both local and systemic vascular alterationscan contribute to ischemia after ASCI by further reducingspinal cord blood flow that can exacerbate and extend theprincipal spinal cord insult (1, 6, 8, 16, 21, 25).
In the normal, noninjured spinal cord, arterial blood supplyis diffuse, primarily delivered via a single anterior spinalartery and two posterior spinal arteries. A variable number ofanterior and posterior radicular arteries provide segmentalcontributions over the length of the cord (24, 25). They feedanastomotic arterial channels over the pial surface that supplythe outer half of the cord and penetrating central arteries fromthe anterior spinal artery, which supply the central portion ofthe cord. Terminal branches of the central arteries extendrostral and caudal to overlap with adjacent terminal arteries,but the terminal arterioles that originate from the terminalarteries do not interconnect within the cord. They in turn giverise to an extensive capillary network, which does intercon-nect within the deep gray and white matter of the cord.Capillaries are much more numerous and extensive in the
S58 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
gray matter than in the white matter, reflecting the increasedmetabolic needs of cell bodies compared with axons (24, 25).Perfusion of the spinal cord under normal physiological cir-cumstances is maintained over a wide range of systemic bloodpressure by autoregulatory mechanisms that seem identical tothose that regulate cerebral blood flow (1, 3–5, 7, 9, 10, 15, 16,18, 20–22, 25).
Local vascular alterations after ASCI are multiple, and theprecise mechanisms of injury-induced ischemia of the cordhave yet to be elucidated. Most investigators cite direct vas-cular injury at the site of the primary trauma as the earliestcomponent of the ischemic injury process (1, 6, 8, 19–21). Theprincipal SCI not only leads to white and gray matter injury atthe insult site but, because of sulcal vessels and collateralterminal arteries that pass through the primary injury site,creates white matter ischemia distal to the direct injury site (8,21, 22, 25). In addition, the primary SCI creates intraluminalthrombosis and vasospasm and initiates a variety of second-ary injury biochemical phenomena that further reduce bloodflow, injure endothelium, increase edema and microvascularcompression, and contribute to microvascular collapse (1, 8,19, 21, 22, 27). Posttraumatic spinal cord ischemia has beenshown to become progressively worse over the first severalhours after injury in animals (1, 4, 6, 7, 16, 21). Laboratory
models of SCI have convincingly demonstrated that autoreg-ulation of spinal cord blood flow is lost after injury, exacer-bating local spinal cord ischemia and rendering the spinalcord vulnerable to systemic hypotension (1, 3–5, 7, 8, 16, 18,21, 27). This is analogous to what often occurs in regionalcerebrovasculature after acute traumatic brain injury (1, 4, 5,7–9, 15, 16, 19, 21, 25, 27).
Systemic hemodynamic alterations after ASCI have beenwell documented and include hypotension, cardiac dysrhyth-mias, reduced peripheral vascular resistance, and reducedcardiac output (1, 12–14, 17, 21, 26). Patients with the mostsevere injuries, particularly those with severe cervical SCIs,are at greatest risk for cardiac, hemodynamic, and respiratorydisturbances in the first week after ASCI (11, 12, 17). Theseuntoward occurrences, which may be episodic in nature, canresult in hypotension and hypoxia. If, as many investigatorssuspect, ASCI with loss of spinal cord autoregulation is anal-ogous to acute traumatic brain injury, hypotension and hyp-oxia can worsen the severity of the original insult and can bedisastrous for potential neurological recovery (1, 8, 20, 21).Although the relationship between systemic hypotension andoutcome after ASCI has not been directly studied in humanpatients, inference from studies of patients with traumaticbrain injury seems appropriate (2, 8, 21). Prospectively col-
TABLE 8.1. Summary of Reports on Blood Pressure Management after Acute Spinal Cord Injurya
Series (Ref. No.) Description of Study Evidence Class Conclusions
Vale et al., 1997 (26) Prospective assessment of 77 ASCI patientstreated in ICU, aggressive hemodynamicsupport, MAP �85.No control group.
III Improved outcome with aggressivemedical care, distinct from potentialbenefit from surgery at 1-yr follow-up.
Levi et al., 1993 (13) 50 patients treated in ICU, aggressivemedical treatment, MAP �90.
III Improved outcome with aggressivehemodynamic support at 6 wk postinjury.
Levi et al., 1991 (14) 103 ACSI patients, 50 incomplete (GroupA), 53 complete (Group B), ICU care,hemodynamic support, MAP �85.
III Improved neurological outcome, nosignificant difference between early andlate surgery in either group.
Wolf et al., 1991 (28) 52 patients with locked facets reducedwithin 4 h, ICU care, MAP �85. 49operated on, 23 day 1, 26 delayed.
III Closed reduction 61%. 52% 1-yr follow-up. In general, improved neurologicaloutcome with hemodynamic therapy.
Tator et al., 1984 (23) 144 ASCI patients managed per protocolof ICU care, hemodynamic support.Compared with earlier cohort.
III Improved neurological outcome, lessmortality with early transfer and ICU care.
Zach et al., 1976 (29) Prospective assessment of 117 ACSIpatients at Swiss center, ICU setting.Aggressive medical therapy and bloodpressure support (Rheomacrodexb � 7 d;dexamethasone � 10 d).No comparison or control group.
III Improved neurological outcome withaggressive medical treatment and bloodpressure management. Better outcome forearly referrals.
a ASCI, acute spinal cord injury; MAP, mean arterial pressure; ICU, intensive care unit.b Rheomacrodex, dextran 40 (Medisan, Parsippany, NJ).
Blood Pressure Management after Acute Cervical Spinal Cord Injury
S59Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
lected data from the Traumatic Coma Data Bank (Class IIevidence) demonstrate that hypotension (systolic blood pres-sure �90 mm Hg) or hypoxia (paO2 �60 mm Hg) was inde-pendently associated with significant increases of morbidityand mortality after severe traumatic brain injury (2). A singleepisode of hypotension was associated with a 150% increasein mortality. It is in this very setting that therapeutic inter-vention aimed at correcting hypotension and maintainingthreshold levels of mean arterial pressure (MAP) to improvecerebral or spinal cord perfusion has its greatest potential.Several reports of case series suggest that treatment of hypo-tension and resuscitation to maintain MAP at high-normallevels, 85 to 90 mm Hg, may enhance neurological outcomeafter acute traumatic SCI (13, 14, 23, 26, 28, 29).
Zach et al. (29) used a prospective aggressive medical man-agement paradigm in the treatment of 117 consecutive ASCIpatients. All patients were treated in the intensive care unit(ICU) with central venous pressure monitoring and weretreated with volume expansion (Rheomacrodex 40 [dextran40; Medisan, Parsippany, NJ], 500 ml/d) for maintenance ofsystemic blood pressure for 7 days. Patients were stratified byinjury level, degree of deficit (Frankel grade), and time ofadmission after injury. The authors reported that 62% ofcervical level SCI patients they managed in this way im-proved at last follow-up, including 8 of 18 Frankel Grade Apatients, with 2 patients improving by two grades, and 1patient improving by three grades. No patient with a cervicalinjury worsened; 38% were unchanged from admission. Ofpatients who arrived within 12 hours of injury, 67% improvedcompared with their admission neurological examination. Ofpatients admitted between 12 and 48 hours after injury, only59% improved. When admission occurred 48 hours after in-jury, improvement was seen in only 50% of patients. Theauthors concluded that early transfer and immediate medicalspecific treatment of the spinal injury with attention to main-tenance of acceptable blood pressure seemed to improve neu-rological recovery (29).
Tator et al. (23) in 1984 described their experience with 144patients with ASCI managed between 1974 and 1979 at adedicated SCI unit in Toronto, Ontario, Canada. The authorscompared their results with a cohort of 358 SCI patientsmanaged between 1948 and 1973, before the development ofthe acute care SCI facility. All 144 patients managed from 1974to 1979 were treated in an ICU setting with strict attention tothe treatment of hypotension and respiratory failure. Hypo-tension was “treated vigorously” with crystalloid and trans-fusion of whole blood or plasma for volume expansion. Pa-tients with respiratory dysfunction were treated withventilatory support as indicated. Tator et al. reported thatmean time from injury to admission and treatment was 4.9hours, compared with more than 12 hours from 1948 to 1973.Neurological improvement was observed in 41 (43%) of 95patients managed under the aggressive ICU medical para-digm. Fifty-two patients (55%) demonstrated no improve-ment. Only two patients (2%) deteriorated. The authors re-ported lower mortality, reduced morbidity, shorter length ofstay, and lower cost of treatment with their contemporarycomprehensive management paradigm compared with the
1948 to 1973 experience. They cited improved respiratorymanagement in their ICU as one of the principal factorsresponsible for reduced mortality and credited the avoidanceof hypotension, sepsis, and urological complications for re-duced morbidity after injury. These improved managementresults were realized despite the fact that 28% of the ASCIpatients they treated had additional injuries that increasedtheir risk of morbidity and mortality.
Wolf et al. (28), in 1991, reported their experience with 52patients with acute cervical bilateral facet dislocation injuriesmanaged with an aggressive treatment paradigm that in-cluded ICU care, aggressive resuscitation, invasive monitor-ing, and hemodynamic manipulation to maintain mean bloodpressure above 85 mm Hg for 5 days. Thirty-four patients hadcomplete neurological injuries, 13 patients had incompleteinjuries, and 5 patients were intact. The authors attemptedclosed reduction within 4 hours of patient arrival to theircenter and performed early open reduction on patients whocould not be reduced by closed means. The authors describedneurological improvement at discharge in 21% of completeSCI patients and in 62% of patients with incomplete cervicalSCIs at admission. No intact patient deteriorated. The authorsconcluded that their protocol of aggressive, early medical andsurgical management of patients with ASCI improved out-come after injury. Treatment in the ICU setting, hemodynamicmonitoring with maintenance of MAP above 85 mm Hg, andearly decompression of the spinal cord by open or closedmeans seemed to reduce secondary complications after ASCIin their study.
Levi et al. (13) treated 50 acute cervical SCI patients in theICU setting according to an aggressive management protocolthat included invasive hemodynamic monitoring and volumeand pressor support to maintain a hemodynamic profile withadequate cardiac output and mean blood pressure above 90mm Hg. Their 1993 report described 31 patients with FrankelGrade A injuries at admission, 8 patients with Frankel GradeB injuries, and 11 patients with Frankel Grades C and D. Eightpatients had shock at the time of admission (systolic bloodpressure �90 mm Hg), and 82% of patients had volume-resistant hypotension requiring pressors within the first 7days of treatment. Volume-resistant hypotension was 5.5times more common among patients with complete motorinjuries. Forty percent of patients managed by protocol im-proved, including several with complete injuries; 42% re-mained unchanged; and 18% (9 patients) died. There wasminimal morbidity associated with invasive hemodynamicmonitoring. The authors concluded that hemodynamic mon-itoring in the ICU allows early identification and prompttreatment of cardiac dysfunction and hemodynamic instabil-ity and can reduce the potential morbidity and mortality afterASCI.
Vale et al. (26), in 1997, reported their experience with anonrandomized, prospective pilot study in the assessment ofaggressive medical resuscitation and blood pressure manage-ment in 77 consecutive ASCI patients. All patients were man-aged in the ICU with invasive monitoring (Swan Ganz cath-eters and arterial lines) and blood pressure augmentation tomaintain MAP above 85 mm Hg for 7 days after injury. The
S60 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
authors reported 10 patients with complete cervical SCIs, 25patients with incomplete cervical injuries, 21 patients withcomplete thoracic SCIs, and 8 patients with incomplete tho-racic level SCIs. The average admission MAP for completecervical SCI patients was 66 mm Hg. Nine of 10 completecervical SCI patients required pressors after volume replace-ment to maintain MAP at 85 mm Hg. Fifty-two percent ofincomplete cervical SCI patients required pressors to maintainMAP at 85 mm Hg. Only 9 of 29 patients with thoracic levelSCIs required the use of pressors. The authors reported min-imal morbidity with the use of invasive monitoring or withpharmacological therapy to augment MAP. At 1-yearfollow-up (mean, 17 mo), 3 of 10 complete cervical SCI pa-tients regained ambulatory capacity and 2 patients regainedbladder function. Incomplete cervical SCI patients fared bet-ter. Twenty-three of these patients regained ambulatory func-tion at 12 months follow-up, only four of whom had initialexamination scores consistent with walking. Twenty-two(88%) of 25 patients regained bladder control. Thirty-one of 35cervical SCI patients and 27 of 29 thoracic level SCI patientswere treated surgically. The authors statistically comparedselection for and timing of surgery with admission neurolog-ical function and compared surgical treatment, early and late,with neurological outcome and found no statistical correla-tion. They concluded that the enhanced neurological outcomeidentified in their series after ASCI was optimized by earlyand aggressive volume resuscitation and blood pressure aug-mentation and was in addition to and/or distinct from anypotential benefit provided by surgery.
The collective experience described in these case series (ClassIII evidence) strongly suggests that maintenance of MAP at 85 to90 mm Hg improves spinal cord perfusion or affects neurologi-cal outcome (13, 14, 23, 26, 28, 29). Prompt treatment of hypo-tension and resuscitation to MAP levels of 85 to 90 mm Hg issafe and suggests that elevation of MAP to threshold levels maybe beneficial to patients with ASCIs. The 7-day duration oftreatment and the threshold levels of MAP maintenance seem tohave been chosen arbitrarily by the individual clinical investiga-tors (13, 26, 28). They are thought to be analogous to initialduration and threshold MAP level recommendations for man-agement of patients after acute traumatic brain injury. None ofthe authors provides a specific recipe or an algorithm to guideblood pressure augmentation. All of the articles describe acutelyinjured patients who have arterial lines and central venous orSwan Ganz catheters in place to monitor pressures and volumestatus (13, 14, 23, 26, 28, 29). Initially, crystalloid is given intra-venously in response to MAP levels below 85 mm Hg. Colloid isadministered if the hematocrit is low (blood) or as a volumeexpander (albumin). If the patient’s volume status is optimal butthe MAP remains below threshold, the authors describe the useof pressors, typically (although not exclusively) a �-agonist (do-pamine), before the addition of an �-agonist (Neo-Synephrine[Sanofi Winthrop Pharmaceuticals, New York, NY]), to elevatethe MAP. These agents are titrated to the appropriate dose levelto achieve the threshold MAP using volume, pressure, and car-diac performance data provided by the invasive monitoringdevices.
SUMMARY
Hypotension is common after acute traumatic SCI in hu-mans. Hypotension contributes to spinal cord ischemia afterinjury in animal models and can worsen the initial insult andreduce the potential for neurological recovery. Although un-proven by Class I medical evidence studies, it is likely thatthis occurs in human SCI patients as well. Because the correc-tion of hypotension and maintenance of homeostasis is a basicprinciple of ethical medical practice in the treatment of pa-tients with traumatic neurological injuries, depriving ASCIpatients of this treatment would be untenable. For this reason,Class I evidence about the effects of hypotension on outcomeafter acute human SCI will never be obtained. However,correction of hypotension has been shown to reduce morbid-ity and mortality after acute human traumatic brain injuryand is a guideline level recommendation for the managementof traumatic brain injury. Although a similar treatment guide-line cannot be supported by the existing SCI literature, cor-rection of hypotension in the setting of acute human SCI isoffered as a strong treatment option. Class III evidence fromthe literature suggests that maintenance of MAP at 85 to 90mm Hg after ASCI for 7 days is safe and may improve spinalcord perfusion and, ultimately, neurological outcome.
KEY ISSUES FOR FUTURE RESEARCH
The issue of whether blood pressure augmentation affectsoutcome after human SCI is important and deserves furtherstudy. If augmentation of MAP is determined to be of poten-tial benefit, the threshold levels of MAP most appropriate andthe length of augmentation therapy need definition. Theseissues are best analyzed in a multi-institutional prospectivecohort study or a properly designed multi-institutional retro-spective case-control study.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Amar AP, Levy ML: Pathogenesis and pharmacological strategiesfor mitigating secondary damage in acute spinal cord injury.Neurosurgery 44:1027–1040, 1999.
2. Chesnut RM, Marshall LF, Klauber MR, Blunt BA, Baldwin N,Eisenberg HM, Jane JA, Marmarou A, Foulkes MA: The role ofsecondary brain injury in determining outcome from severe headinjury. J Trauma 34:216–222, 1993.
3. Dolan EJ, Tator CH: The effect of blood transfusion, dopamine,and gamma hydroxybutyrate on posttraumatic ischemia of thespinal cord. J Neurosurg 56:350–358, 1982.
4. Ducker TB, Kindt GW, Kempf LG: Pathological findings in acuteexperimental spinal cord trauma. J Neurosurg 35:700–708, 1971.
5. Flohr H, Pöll W, Brock M: Regulation of spinal cord blood flow,in Russell RWR (ed): Brain and Blood Flow: Proceedings of the 4thInternational Symposium on the Regulation of Cerebral Blood Flow.London, Pitman Medical, 1971, pp 406–409.
6. Hall ED, Wolf DL: A pharmacological analysis of the pathophys-iological mechanisms of posttraumatic spinal cord ischemia.J Neurosurg 64:951–961, 1986.
Blood Pressure Management after Acute Cervical Spinal Cord Injury S61
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
7. Kindt GW, Ducker TB, Huddlestone J: Regulation of spinal cordblood flow, in Russell RWR (ed): Brain and Blood Flow: Proceedingsof the 4th International Symposium on the Regulation of Cerebral BloodFlow. London, Pitman Medical, 1971, pp 401–405.
8. King BS, Gupta R, Narayan RK: The early assessment and inten-sive care unit management of patients with severe traumatic brainand spinal cord injuries. Surg Clin North Am 80:855–870, 2000.
9. Kobrine AI, Doyle TF, Martins AN: Autoregulation of spinal cordblood flow. Clin Neurosurg 22:573–581, 1975.
10. Kobrine AI, Doyle TF, Rizzoli HV: Spinal cord blood flow asaffected by changes in systemic arterial blood pressure.J Neurosurg 44:12–15, 1976.
11. Ledsome JR, Sharp JM: Pulmonary function in acute cervical cordinjury. Am Rev Respir Dis 124:41–44, 1981.
12. Lehmann KG, Lane JG, Piepmeier JM, Batsford WP: Cardiovas-cular abnormalities accompanying acute spinal cord injury inhumans: Incidence, time course and severity. J Am Coll Cardiol10:46–52, 1987.
13. Levi L, Wolf A, Belzberg H: Hemodynamic parameters in patientswith acute cervical cord trauma: Description, intervention, andprediction of outcome. Neurosurgery 33:1007–1017, 1993.
14. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does the timingof surgery affect the outcome? Neurosurgery 29:216–222, 1991.
15. Lewelt W, Jenkins LW, Miller JD: Autoregulation of cerebralblood flow after experimental fluid percussion injury of the brain.J Neurosurg 53:500–511, 1980.
16. Osterholm JL: The pathophysiological response to spinal cordinjury: The current status of related research. J Neurosurg 40:5–33, 1974.
17. Piepmeier JM, Lehmann KB, Lane JG: Cardiovascular instabilityfollowing acute cervical spinal cord trauma. Cent Nerv SystTrauma 2:153–160, 1985.
18. Senter HJ, Venes JL: Loss of autoregulation and posttraumaticischemia following experimental spinal cord trauma. J Neurosurg50:198–206, 1979.
19. Tator CH: Vascular effects and blood flow in acute spinal cordinjuries. J Neurosurg Sci 28:115–119, 1984.
20. Tator CH: Hemodynamic issues and vascular factors in acuteexperimental spinal cord injury. J Neurotrauma 9:139–141, 1992.
21. Tator CH: Ischemia as a secondary neural injury, in Salzman SK,Faden AI (eds): Neurobiology of Central Nervous System Trauma.New York, Oxford University Press, 1994, pp 209–215.
22. Tator CH: Experimental and clinical studies of the pathophysiol-ogy and management of acute spinal cord injury. J Spinal CordMed 19:206–214, 1996.
23. Tator CH, Rowed DW, Schwartz MI, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–293, 296, 1984.
24. Turnbull IM: Microvasculature of the human spinal cord.J Neurosurg 35:141–147, 1971.
25. Turnbull IM: Blood supply of the spinal cord: Normal and patho-logical considerations. Clin Neurosurg 20:56–84, 1973.
26. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medical andsurgical treatment after acute spinal cord injury: Results of aprospective pilot study to assess the merits of aggressive medicalresuscitation and blood pressure management. J Neurosurg 87:239–246, 1997.
27. Wallace MC, Tator CH: Successful improvement of blood pres-sure, cardiac output, and spinal cord blood flow after experimen-tal spinal cord injury. Neurosurgery 20:710–715, 1987.
28. Wolf A, Levi L, Mirvis S, Ragheb J, Huhn S, Rigamonti D,Robinson WL: Operative management of bilateral facet disloca-tion. J Neurosurg 75:883–890, 1991.
29. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cordinjuries in the Swiss Paraplegic Centre of Basel. Paraplegia 14:58–65, 1976.
Images of the human spine. From, Tilney F, Alsop Riley H: The Form and Functions of the Central Nervous System: An Intro-duction to the Study of Nervous Diseases. New York, Harper & Brothers, 1938, 3rd ed.
S62 Guidelines for Management of Acute Cervical Spinal Injuries
CHAPTER 9
Pharmacological Therapy after Acute Cervical SpinalCord Injury
RECOMMENDATIONSCORTICOSTEROIDS:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment with methylprednisolone for either 24 or 48 hours is recommended as an option in the
treatment of patients with acute spinal cord injuries that should be undertaken only with the knowledgethat the evidence suggesting harmful side effects is more consistent than any suggestion of clinical benefit.
GM-1 GANGLIOSIDE:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment of patients with acute spinal cord injuries with GM-1 ganglioside is recommended as an
option without demonstrated clinical benefit.
RATIONALE
The hope that administration of a pharmacological agentdelivered shortly after acute spinal cord injury (ASCI)might improve neurological function and/or assist neu-
rological recovery has long been held. A variety of promisingsubstances have been tested in animal models of ASCI, butfew have had potential application to human spinal cordinjury (SCI) patients. Four pharmacological substances havemet rigorous criteria in laboratory testing and initial humaninvestigations: two corticosteroids (methylprednisolone andtirilazad mesylate), naloxone, and GM-1 ganglioside. All fourpharmacological agents have been evaluated in controlled,randomized, blinded clinical trials of human patients withASCIs. Two of these substances, tirilazad and naloxone, havebeen studied less extensively and as yet have unclear efficacyin the management of acute human SCI. The purpose of thismedical evidence-based review is to define the usefulness ofadministration of methylprednisolone with or without GM-1ganglioside in the contemporary management of ASCI patients.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of literature published from 1966 to 2001 was un-dertaken. The following medical subject headings were usedin combination with “spinal cord injury” and “neurologicaldeficit”: steroids, methylprednisolone, and GM-1 ganglioside.
Approximately 2400 citations were acquired. Non-Englishlanguage citations and nonhuman experimental studies weredeleted. Titles and abstracts of 652 manuscripts were re-viewed, 639 on the topic of corticosteroids and human SCIand 13 on the topic of GM-1 ganglioside and human SCI. Ad-ditional references were culled from the reference lists of theremaining papers. Finally, the members of the author groupwere asked to contribute articles known to them on the subjectmatter that were not found by other search means. Duplica-tions, case reports, pharmacokinetic reports, general reviews,and articles with mention of one agent or another but withoutscientific assessment were eliminated. Several editorials, cri-tiques, and responses to published reports and studies wereincluded. Forty-six published references on the topic of meth-ylprednisolone in the treatment of patients with ASCI andseven published references for GM-1 ganglioside provide thebasis for this guideline. Thirteen studies on methylpred-nisolone and two studies on GM-1 ganglioside are summa-rized in Tables 9.1 and 9.2.
SCIENTIFIC FOUNDATION
Methylprednisolone
Corticosteroids, particularly methylprednisolone, havebeen studied extensively in animal models of SCI (2, 19, 47, 48,50, 51). Although their precise mechanisms of action are not
This chapter remains the most controversial of the Guidelines. The readers are advised to carefully review the availabledata and Comments provided within this Supplement to establish their own perspective on this evolving matter.
Michael L.J. Apuzzo
S63Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TABLE 9.1. Summary of Reports on Treatment with Methyprednisolone after Acute Cervical Spinal Cord Injurya
Series (Ref. No.) Description of Study Evidence Class Conclusions
Bracken et al., 1984
(11)
Multicenter, double-blind randomized trial
comparing MP (1000 mg/d versus 100 mg/d for
11 d) in treatment of 330 ASCI patients (NASCIS
I study).
III
(Study design, data presentation,
interpretation, and analysis
flaws)
No treatment effect at 6 wk and 6 mo postinjury. No control group.
Bracken et al., 1985
(15)
1-yr follow-up of NASCIS I study. III
(Study design, data presentation,
interpretation, and analysis
flaws)
No significant difference in neurological recovery of motor or sensory
function 1-yr postinjury.
Bracken et al., 1990
(14)
Multicenter, randomized, double-blind, placebo-
controlled trial comparing MP with naloxone
and placebo in treatment of 487 ASCI patients
(NASCIS II study).
III
(Study design, data presentation,
interpretation, and analysis
flaws)
Significant improvement in motor change scores (P � 0.03), and
sensation change scores (P � 0.02) at 6 mo postinjury for patients treated
with MP within 8 h of injury.
Bracken et al., 1992
(13)
1-yr follow-up of NASCIS II study. III
(Study design, data presentation,
interpretation, and analysis
flaws)
Significant improvement in motor change scores 1 year postinjury for
patients treated with MP within 8 h of injury (P � 0.03). Administration of
MP detrimental if given more than 8 h after injury.
Galandiuk et al.,
1993 (21)
Prospective assessment of 15 patients from 1990
to 1993 with retrospective review of 17 patients
from 1987 to 1990 to assess differences in
treatment outcome with MP compared with
treatment without corticosteroids.
III No difference in neurological outcome between two sets of patients. MP
patients had immune response alterations, higher rate of pneumonia, and
longer hospital stays than patients who did not receive corticosteroids.
Gerhart et al., 1995
(29)
Concurrent cohort comparison study
(population-based) of 363 ASCI patients
managed from 1990 to 1991 and 1993. 188
patients managed with NASCIS II MP compared
with 90 patients with no MP.
III
(Inadequate statistical power)
No differences in neurological outcome using Frankel classification
between MP and No-MP patients. However, may be insufficient numbers
of patients to show significant differences.
George et al., 1995
(28)
Retrospective review of 145 ASCI patients, 80
treated with MP compared with 65 who did not
receive MP.
III No difference in mortality or neurological outcome between groups
despite younger age, less severe injury in MP-treated patients.
Gerndt et al., 1997
(30)
Retrospective review with historical control of
231 ASCI patients. 91 excluded. Comparison of
medical complications among 93 MP patients
compared with 47 who received no
corticosteroid.
III MP-treated patients had significant increases in pneumonia (P � 0.02),
acute pneumonia (P � 0.03), ventilated days (P � 0.04), and ICU stay
(P � 0.45), but no adverse effect on long-term outcome.
Poynton et al., 1997
(39)
Case-control analysis of 71 consecutive ASCI
admissions. 63 available for 13 mo to 57 mo
follow-up. 38 patients treated with MP
compared with 25 referred �8 h after injury
who received no MP.
III Multiple factors influence recovery after SCI. No effect of MP or surgery
on outcome.
Bracken et al., 1997
(16)
Multicenter, randomized, double-blind trial
comparing MP administered for 24 hr to MP
administered 48 hr and TM in the treatment of
499 ASCI patients (NASCIS III study).
III
(Study design, data presentation,
interpretation, and analysis
flaws)
48 MP patients had improved motor recovery at 6 wk and at 6 mo
compared with 24 MP and 48 TM groups NS. When treatment initiated
between 3 h and 8 h after injury, 48 MP had significant improvement of
motor scores at 6 wk (P � 0.04) and 6 mo (P � 0.01). 48 MP was
associated with high rates of sepsis and pneumonia. No control group.
Bracken et al., 1998
(17)
1-yr follow-up of NASCIS III study. III
(Study design, data presentation,
interpretation, and analysis
flaws)
Recovery rates equal in all 3 groups when treatment initiated within 3 h
of injury. When treatment initiated between 3 h and 8 h, 24 MP patients
had diminished recovery, 48 MP patients had increased motor recovery
(P � 0.053).
Pointillart et al., 2000
(38)
Multicenter, prospective, randomized clinical
trial of 106 ASCI patients treated with MP,
nimodipine, neither, or both.
III
(Inadequate statistical power)
No significant difference in neurological outcome at 1-yr follow-up
between groups. Incomplete ASCI had significant improvement below
level of injury compared to complete patients (P � 0.0001). Higher
incidence of infectious complications among patients receiving
corticosteroids (NS).
Matsumoto et al.,
2001 (36)
Prospective, randomized, double-blind study
comparing incidence of medical complications
among 46 ASCI patients, 23 treated with MP, 23
with placebo.
I MP patients had higher incidence of complications (56.5% versus 34.8%).
Respiratory complications (P � 0.009) and gastrointestinal bleed
(P � 0.036) were most significant between groups. No data on
neurological improvement.
a ASCI, acute spinal cord injury; NASCIS, National Acute Spinal Cord Injury Study; MP, methylprednisolone; ICU, intensive care unit; SCI,spinal cord injury; TM, tirilazad mesylate; NS, not significant.
S64 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
completely known, they have the potential to stabilize mem-brane structures, maintain the blood-spinal cord barrier po-tentially reducing vasogenic edema, enhance spinal cordblood flow, alter electrolyte concentrations at the site of in-jury, inhibit endorphin release, scavenge damaging free rad-icals, and limit the inflammatory response after injury (2, 47,48, 50, 51). After considerable positive study in the laboratory,methylprednisolone was studied in human SCI patients in amulticenter, randomized, double-blinded clinical trial initi-ated in 1979. The first National Acute Spinal Cord InjuryStudy (NASCIS I) (11), reported in 1984, compared the effi-cacy of administration of a 100-mg bolus of methylpred-nisolone and then 100 mg daily thereafter for 10 days withadministration of a 1000-mg bolus and then 1000 mg daily for10 days in 330 acute injury patients assessed 6 weeks and 6months after injury. There was no control group. The studyrevealed no difference in neurological recovery (motor orsensory function) between the treatment groups at either 6weeks or 6 months after injury. Motor scores were determinedfrom the examination of seven muscle groups on each side ofthe body scored on a 6-point scale. Sensory function wasassessed using a 3-point scale of dermatomal light touch andpinprick sensation. The authors reported the motor and sen-sory scores from the right side of the body only. There was noanatomic level injury limit (superior to T12 vertebral level, forexample) in the study to include only SCI patients and ex-clude primary cauda equina injuries or “mixed” central andcauda equina injuries that might occur with a lower fractureinjury (e.g., T12–L1 or L1–L2 injuries). The study did notrequire a minimum motor impairment for inclusion; hence,patients with normal motor examinations and those withminimal neurological deficits were included in the study if theattending physician determined that the patient had an SCI ofany severity. In 1985, the same group of investigators reportedon the 1-year follow-up of these study patients (15). No dif-ferences in motor or sensory outcome were identified betweenthe two treatment groups.
Animal studies of the efficacy of methylprednisolone afterexperimental SCI suggested that the doses of methylpred-nisolone used in the NASCIS I investigation were too low to
demonstrate a significant difference in outcome (2, 14, 19, 50,51). A multicenter NASCIS II trial was initiated in 1985 usinga much higher dose of methylprednisolone (30 mg/kg as abolus and then 5.4 mg/kg/h infusion for 23 h). These patientswere compared with similarly injured patients who receivedeither naloxone (5.4 mg/kg bolus and then an infusion of 4.0mg/kg/h for 23 h) or placebo. Patients had to be randomizedto one of three treatment arms within 12 hours of ASCI. Theresults of NASCIS II were reported in 1990 (14). Four hundredeighty-seven patients were entered into the study; 162 re-ceived methylprednisolone, 154 were given naloxone, and 171patients were in the placebo control group. The authors re-ported that the administration of methylprednisolone within8 hours of injury was associated with a significant improve-ment in motor function (neurological change scores, right sideof body only, P � 0.03), and in sensation (pinprick, P � 0.02;light touch, P � 0.03) at the 6-month follow-up, comparedwith patients receiving methylprednisolone more than 8hours after injury and patients receiving naloxone or placebo.No similar significant improvements were noted at the6-week follow-up, either motor or sensory. Motor scores weredetermined from the examination of seven muscle groups oneach side of the body scored on a scale of 0 to 5 points.Sensory function was assessed using a 3-point scale of der-matomal light touch and pinprick sensation. The NASCIS IIstudy reported on the motor scores from the right side of thebody only. Bilateral sensory scores were provided. Like theNASCIS I study, there was no anatomic level injury limit inthe study (superior to T12 vertebral level, for example), toensure that only SCI patients were included for study (11, 15).Similarly, NASCIS II did not require a minimum motor im-pairment for inclusion; hence, patients with normal motorexaminations and those with minimal neurological deficitswere included. No outcome measures involving patient func-tion were used in this study. In 1992, NASCIS investigatorsreported on the 1-year follow-up of NASCIS II study patients(13). They reported statistically significant improvement inmotor scores on the right side of the body for 62 of 487 studypatients (P � 0.03). These 62 patients received methylpred-nisolone within 8 hours of injury. Significant right body motor
TABLE 9.2. Summary of Reports on Treatment with GM-1 Ganglioside after Acute Spinal Cord Injurya
Series (Ref. No.) Description of Study Evidence Class Conclusions
Geisler et al., 1991 (25) Prospective, randomized, double-blindtrial of GM-1 ganglioside in 37 humanASCI patients. All received 250-mg MPbolus followed by 125 mg every 6 h �72 h before randomization (placebogroup).
I GM-1 ganglioside enhances recovery ofneurological function, significantdifference in recovery compared withMP group (P � 0.047). Insufficientnumbers of patients to draw meaningfulconclusions. No true placebo group.
Geisler et al., 2001 (23) Prospective, randomized, double-blind,stratified multicenter trial of GM-1ganglioside in 760 ASCI patients. Allreceived MP per NASCIS II protocol(placebo group).
I No significant differences inneurological recovery identifiedbetween GM-1-treated patients andMP-treated patients at 26-wk follow-up.Trend for earlier recovery in GM-1-treated patients. No true placebo group.
a ASCI, acute spinal cord injury; NASCIS, National Acute Spinal Cord Injury Study; MP, methylprednisolone.
Pharmacological Therapy in Acute Cervical Spinal Cord Injury S65
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
score improvement was identified in two of three categoriesof patients, plegic patients with total sensory loss (P � 0.019)and paretic patients with variable sensory loss (P � 0.024), butnot among plegic patients with partial sensory loss (P �0.481). There were no significant improvements in motorchange scores described among the remaining 421 patientsentered in the study. There were no significant differences insensory scores for any treatment group or categories of pa-tients despite the differences reported at the 6-monthfollow-up for patients receiving methylprednisolone within 8hours of injury. Patients treated more than 8 hours after injurywith methylprednisolone or naloxone experienced less recov-ery of motor function compared with placebo treatment pa-tients. The authors concluded that treatment with the studydose of methylprednisolone administered within 8 hours ofinjury improves neurological outcome and is therefore indi-cated in the treatment of patients with ASCI. The use of studydose methylprednisolone in patients was not associated withharmful side effects compared with patients in the othertreatment groups, although the authors reported an increasedincidence of wound infection and gastrointestinal bleedingamong corticosteroid-treated patients. Treatment with meth-ylprednisolone beyond 8 hours after injury was notrecommended.
There are several flaws in the NASCIS II study, and criti-cism has been offered on several methodological, scientific,and statistical issues (18, 19, 22, 31, 32, 35, 37, 40–42, 44–46, 51).The investigators described two a priori hypotheses: thattreatment effect would be influenced by how soon the drugwas given after injury and by the severity of injury. Patientswere considered eligible for inclusion if they were admitted tothe study and randomized to treatment within 12 hours ofinjury. At some point, patient outcome was stratified accord-ing to the timing of methylprednisolone administration (�8 h,�8 h). Some reviewers have requested examination of the rawdata to look for time-related diminishing effects of methyl-prednisolone administration relative to injury rather than as-signment of an “all or nothing” time cutoff (18, 32, 37, 40, 42,51). Analysis of results of the entire population of patientsaccording to the second a priori hypothesis was not providedby the authors (18, 31, 37, 40, 42, 51). Analysis using thesecond hypothesis was accomplished on the group of patientspreviously stratified according to the first hypothesis. It maybe that the two hypotheses are fully independent, yet nojustification for this assumption was offered (31, 40). Thestudy did not offer a standardized medical treatment regimenfor all ASCI patients in this study. The medical managementof study patients including monitoring, blood pressure aug-mentation, respiratory care, deep venous thrombosis prophy-laxis, nutritional support, and initiation of rehabilitation ac-tivities was neither consistent within centers nor consistentfrom center to center (18, 22, 31, 37). Similarly, surgical treat-ment offered to patients in the NASCIS II study was notconsistent from center to center (19, 31, 35, 51). There was nodescription of surgical approaches used for specific pathologyor documentation of the timing of surgical intervention forindividual patients. There was no consideration given to theindependent effect that either aggressive medical manage-
ment or surgery had, or may have had, on outcome (18, 19, 22,31, 35, 37, 51).
The most important and significant criticism of the NASCISII study is the failure to measure patient functional recovery(e.g., functional independence measure [FIM]) to determinewhether the modest improvement reported in neurological ex-amination (change in motor scores) in the methylprednisolone-treated patients had meaningful clinical significance (18, 32, 35,37, 44). It is unclear from the change in score data providedwhether the improvement had any clinical significance to theinjured patients (1, 18, 32, 35, 37, 44–46). One of the most frequentcriticisms of the reported NASCIS II results is the failure toprovide scientific data on which statistical comparisons weremade (18, 19, 31, 32, 37, 40–42, 46, 51). As with the NASCIS Istudy, only right-sided motor scores were reported in NASCISII, but bilateral sensory scores were reported. Change in motorscore (improvement) on the right side only of ASCI patients hasbeen cited by the study authors as a significant neurologicalbenefit associated with methylprednisolone administrationgiven at study doses within 8 hours of injury and assessed at6-month and 1-year follow-up (P � 0.03) (13, 14). These findingswere observed in only a small subset of study patients (18, 31, 37,41). Was this an a priori hypothesis of the investigators, and wasthe result significant for the whole population of patients? If so,then the finding stands and the post hoc subgroup analysissuggests which subgroup receives the benefit. If, however, theentire result is from a post hoc hypothesis and analysis and issignificant only for the subgroup and not for all of the patientsanalyzed together, then it is a weak suggestive finding. This isnot made clear by the authors. Reviewers have argued againstthe use of right-side only motor scores, and particularly thechange of score results in NASCIS II publications (18, 22, 31, 32,40, 41). The lack of evidence describing left-sided motor scoresand total body motor scores in NASCIS II is confusing (4, 8–10,12, 50).
Also confusing is the reported difference in change of motorscore outcome for patients with incomplete SCI who were inthe placebo treatment arm. Patients with incomplete SCIs inthe NASCIS II study who received placebo more than 8 hoursafter injury had significantly better neurological recovery thandid patients who received placebo within 8 hours of injury(13, 18, 32, 42). Additionally, the neurological recovery curvegenerated for patients with incomplete SCIs treated withmethylprednisolone within 8 hours of injury is virtually iden-tical to that of patients with incomplete SCIs treated withplacebo beyond 8 hours after injury. The benefit of treatmentwith respect to neurological recovery (motor change score)with methylprednisolone given within 8 hours of injuryseems equal to treatment with placebo more than 8 hours afterinjury (18, 37, 42).
Statistical criticisms of the NASCIS II results are many (18,19, 22, 31, 32, 40–42, 45, 46, 51). They include potential inter-pretive errors, problematic statistical comparisons, simplifica-tion of subgroup analysis from the pre-planned 15 categoriesto 3 seemingly arbitrarily determined categories, an improperand incomplete presentation of odds ratios, and a post hocanalysis of study data including only 127 patients (62 meth-ylprednisolone, 65 placebo) treated within 8 hours of injury,
S66 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
rather than the entire study population of 487 patients (18, 19,22, 31, 32, 40–42, 45, 46, 51). NASCIS II was designed andimplemented to be a randomized, controlled, double-blindedclinical study in an attempt to generate Class I evidence on theefficacy of methylprednisolone and naloxone after ASCI inhuman subjects. The lack of a measure of functional signifi-cance, the dependence on post hoc analyses, and the absenceof an analysis of surgical treatment diminish the quality andusefulness of the evidence provided by these studies.
In 1993, Galandiuk et al. (21) described 32 patients withcervical or upper thoracic ASCIs managed in an urban traumacenter. Fourteen patients who received NASCIS II doses ofmethylprednisolone within 8 hours of injury were comparedwith 18 ASCI patients with similar injuries managed withoutcorticosteroids. The authors reported no difference in neurolog-ical outcome between the two sets of patients but noted thatmethylprednisolone-treated patients had immune response al-terations (lower percentage and density of monocyte Class IIantigen expression and lower T-cell helper/suppressor cell ra-tios), a higher rate of pneumonia (79% versus 50%), and longerhospital stays (44.4 d versus 27.7 d) than similar ASCI patientsthey managed without administration of corticosteroids. Al-though the conclusions drawn by the authors are interesting,they have little scientific power. The mix of historical patientswith contemporary patients, the lack of a prospective design,and the haphazard assignment and assessment of patients dilutethe quality of the evidence provided.
Bracken and Holford (8) described the effect of timing ofmethylprednisolone on neurological recovery in NASCIS IIstudy patients in 1993. They concluded from post hoc analysisof the NASCIS II data that methylprednisolone administeredto patients within 8 hours of ASCI improves neurologicalfunction below the level of the spinal cord lesion in patientsinitially diagnosed as having complete or incomplete injuries.The majority of the improvements they reported were amongpatients with incomplete SCIs at admission. Complete injurypatients demonstrated very little recovery below the level ofinjury irrespective of treatment. Their post hoc analysis alsoconfirmed that methylprednisolone administered more than 8hours after injury may be associated with a worse neurolog-ical outcome.
This 1993 article (8) refers to and references the 1-yearfollow-up NASCIS II study data but only describes patientgroups and offers percentages (18, 42). It provides neither newevidence nor the numbers of patients on whom Bracken andHolford based their conclusions. Although the result that theauthors describe is positive (methylprednisolone adminis-tered within 8 h of injury improves spinal cord function inpatients with SCI), it was identified in a very small subgroupof patients, which raises questions as to its true weight andvalidity. The manner in which the data and conclusions werepresented is ambiguous and suggests that this was a positiveresult reflected by analysis of the entire NASCIS II studypopulation (n � 487) (18). In fact, it was only a subgroupanalysis of the population of patients who received methyl-prednisolone within 8 hours of injury (n � 62), compared withthose who received placebo within 8 hours of injury (n � 65).Forty-five methylprednisolone-treated patients had complete
injuries and demonstrated very little change in function belowthe level of injury. The same is true for 43 similar (complete)patients who received placebo (no significant difference). Theactual differences described by the authors are based on 17methylprednisolone patients compared with 22 placebo-treated patients, all of whom had incomplete SCI and hadtherapy initiated within 8 hours of injury (18).
Their report (8) does help to clarify the issue of recovery offunction (motor score change) in NASCIS II patients withcomplete injuries at admission who received methylpred-nisolone within 8 hours of injury. The NASCIS II results at 1year cite a significant improvement in motor function forpatients who received methylprednisolone at study doseswithin 8 hours of injury compared with placebo-treated pa-tients (P � 0.03) (13). For the patients who had completeinjuries who met the early treatment criteria (n � 45), thesignificance of improvement (change in motor score) was P �0.019, compared with similar patients who received placebo.Bracken and Holford’s (8) post hoc analysis revealed no sig-nificant difference in recovery below the level of the lesion inthese patients compared with placebo-treated patients. Thissuggests that the primary improvements in function identi-fied in the NASCIS II study for patients with complete spinalinjuries treated within 8 hours were at the level of injury,likely root recovery, rather than a significant gain in spinalcord function (18). Again, the relationship between any suchrecovery and an improvement in patient function is un-known, irrespective of the sample size, because the study didnot use functional outcome assessments (18, 35, 37).
In 1994, Duh et al. (20) reported on the effect of surgery onoutcome among NASCIS II study patients. In all, 298 of 487study patients underwent 303 operative procedures, 56 viathe anterior approach and 247 via the posterior approach. Theauthors examined the influence of surgery on neurologicaloutcome across all study groups of patients at time periods ofless than 25 hours, 26 to 50 hours, 51 to 100 hours, 101 to 200hours, and more than 200 hours. They found that the mostseverely injured patients were less likely to be treated surgi-cally. The authors did not identify significant differences inoutcome, motor or sensory, with surgical treatment, eitherearly or late. Functional recovery was not measured.
Gerhart et al. (29), in 1995, reported a population-based,concurrent cohort comparison study of 363 ASCI survivorstreated in Colorado. Two hundred eighteen patients weremanaged between May 1990 and December 1991, and 145injury patients were managed 2 years later in 1993. Of 218patients managed in 1990 to 1991, 100 (46%) were treatedaccording to the NASCIS II protocol. Fifty-one patients (23%)received no methylprednisolone, and 67 patients (31%) re-ceived another corticosteroid, were given an incorrect dose, orhad insufficient data. In the 1993 study population, 61% ofASCI patients (n � 88) received methylprednisolone accord-ing to NASCIS II protocol. Thirty-nine patients (27%) receivedno methylprednisolone and 18 patients (13%) were givenanother corticosteroid, received an incorrect dose, or hadinsufficient data. The authors reported no significant differ-ences in outcome as assessed by the Frankel scale at the timeof hospital discharge when 188 patients who received proto-
Pharmacological Therapy in Acute Cervical Spinal Cord Injury S67
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
col methylprednisolone (appropriate dose and timing) werecompared with those (n � 90) who did not receive any meth-ylprednisolone during treatment. This was true for the com-bined population of patients and for both the 1990 to 1991 andthe 1993 patient populations. It does not seem, however, thatadequate numbers of patients were analyzed by the authors,substantially diluting the statistical power of their findings.
In 1995, George et al. (28) reported their experience withASCI patients at Michigan State University from 1989 through1992. One hundred forty-five patients were described, 80 ofwhom were treated with methylprednisolone per the NASCISII protocol (MP group) and 65 of whom did not receivemethylprednisolone (No-MP group). Admission, discharge,and follow-up neurological assessments were accomplishedaccording to the FIM instrument. Fifteen patients were ex-cluded from review, leaving 130 patients (75 MP, 55 No-MP).The MP group was significantly younger than the No-MPgroup (30 yr versus 38 yr, P � 0.05). Although the meantrauma scores were similar between the two groups, the MPpatients had a significantly lower injury severity score (ISS)than the No-MP patients (P � 0.05). The authors found nodifferences in mortality or neurological outcome between pa-tients treated with methylprednisolone and those who werenot. Despite older age and higher injury severity score, theNo-MP group had better mobility at the time of hospitaldischarge. Admission mobility scores were similar (MP � 5.99versus No-MP � 5.90), but the mobility scores differed sig-nificantly on hospital discharge (MP � 5.16 versus No-MP �4.67, P � 0.05). The authors argued that the MP patient grouphad a more favorable opportunity for improvement than theNo-MP patient group owing to younger age and lower ISSscores; however, neurological improvements in the MP groupcompared with the No-MP group were not observed. It isunclear from the study why most patients did not receivecorticosteroid therapy, and this is the weakness of a nonran-domized study in which patient assignment to treatment mayintroduce bias. For example, an examination of the data indi-cates that the worst neurologically injured patients at admis-sion were more likely to have received methylprednisolone.The findings of no difference in neurological examinationimprovement or functional recovery in this group seem torefute the finding of neurological improvement in NASCIS IIpatients who received methylprednisolone less than 8 hoursafter injury compared with those who did not receive thedrug.
Gerndt et al. (30), in 1997, reported a retrospective review of231 patients with ASCI for the purpose of examining medicalcomplications. Ninety-one patients were excluded becausethey received corticosteroids outside the NASCIS II protocol.One hundred forty patients were reviewed, comparing 93patients who received methylprednisolone per the NASCIS IIprotocol with a historical control group of 47 patients whoreceived no corticosteroid during treatment. The patientgroups were similar with respect to age and injury severity.The authors found significant differences (increases) in theincidence of pneumonia (P � 0.02, 2.6-fold increase), particu-larly acute pneumonia (P � 0.03, 4-fold increase), ventilateddays (P � 0.04), and ICU length of stay (P � 0.045) in
methylprednisolone-treated patients compared with thosewho did not receive corticosteroids during treatment. Non-corticosteroid-treated patients had a higher incidence of uri-nary tract infections (P � 0.01). Methylprednisolone-treatedpatients had decreased general care floor length of stay (P �0.02) and rehabilitation length of stay (P � 0.035). The authorsconcluded that methylprednisolone may increase the inci-dence of early infection, particularly pneumonia, in ASCIpatients but has no adverse effect on long-term outcome. In1997, Poynton et al. (39) described 71 consecutive ASCI pa-tients managed at the National Spinal Trauma Unit in Dublin,Ireland. They attempted a case-control analysis of ASCI pa-tients treated with methylprednisolone (n � 38) comparedwith patients who did not receive methylprednisolone (n �25) and provided follow-up from 13 months to 57 monthsafter injury. Patients who did not receive methylprednisolonewere referred more than 8 hours after injury. The authorsconcluded that multiple factors influenced outcome afterASCI. They found no difference in neurological outcomewhen they compared patients who received methylpred-nisolone with those who did not.
The results of the third NASCIS study (NASCIS III) werepublished in 1997 (16). NASCIS III was a double-blind,randomized clinical trial comparing the efficacy of methyl-prednisolone administered for 24 hours with that ofmethylprednisolone administered for 48 hours and tirilazadmesylate administered for 48 hours. There was no placebogroup. Entry criteria were similar to those described for NAS-CIS II study patients. Patients were assessed neurologicallyaccording to NASCIS I and II (change in motor and sensoryscores) and by change in FIM at 6 weeks and 6 months. Fourhundred ninety-nine patients were entered into the study, 166in the 24-hour methylprednisolone group (24 MP), 167 in the48-hour tirilazad mesylate group (48 TM), and 166 in the48-hour methylprednisolone group (48 MP). The authors re-ported that patients in the 48 MP group showed improvedmotor recovery at 6 weeks (P � 0.09) and at 6 months (P �0.07) follow-up compared with 24 MP patients and 48 TMpatients. When therapy was initiated between 3 and 8 hoursafter injury, the effect of the 48 MP regimen on change inmotor score was significant at 6 weeks (P � 0.04) and at 6months (P � 0.01) follow-up compared with patients in the 24MP and 48 TS treatment groups. 48 MP patients had moreimprovement in FIM at the 6-month follow-up (P � 0.08)compared with patients in the other two treatment groups. 48MP treatment patients also had higher rates of severe sepsis (P� 0.07) and severe pneumonia (P � 0.02). When treatmentwas initiated within 3 hours of injury, the same recoverypattern was observed in all three treatment groups. The au-thors concluded that patients with ASCI who receive methyl-prednisolone within 3 hours of injury should be maintainedon the 24 MP regimen. When methylprednisolone is admin-istered 3 to 8 hours after injury, they recommended the 48 MPregimen.
In 1998, the 1-year follow-up results of the NASCIS III trialwere reported (17). The authors reported that for patientstreated within 3 hours of injury, recovery rates at 1 year wereequal in all three treatment groups. For patients treated be-
S68 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
tween 3 and 8 hours after injury, 24 MP patients had dimin-ished motor recovery and 48 MP patients had increased motorrecovery at 1 year (P � 0.053). They noted no significantdifference in functional outcome as measured by FIM in anytreatment group. The authors concluded that if methylpred-nisolone is administered to patients with ASCI within 3 hoursof injury, 24-hour maintenance is recommended. If methyl-prednisolone is administered 3 to 8 hours after injury, theyrecommended that a 48-hour maintenance regimen be fol-lowed. These final recommendations seem to be based onmotor recovery score improvement alone (P � 0.053).
Predominant criticisms of the NASCIS III study and thereported results focus on three major issues: determination ofoptimum timing of therapy, method of motor assessment ofSCI patients, and insignificant differences in motor recoveryscores and functional outcome measures among study pa-tients (18, 19, 32, 33, 37, 51). For optimum timing of therapy,time-to-treatment data were not offered or explained. Like the8-hour time for treatment cutoff “result” that came from theNASCIS II study, the “within 3 hours of injury” versus the “3to 8 hours after injury” timeframes reported in NASCIS IIIseem arbitrary (18, 32, 37). It is not intuitive or likely that the3-hour treatment time is an “all or nothing” time periodsupported by physiological evidence. With respect to themethod of motor assessment and reporting, like the NASCISII study, NASCIS III motor scores were reported as change inmotor scores from the right side of the body. Left-side motorscores and total body motor scores were not provided. Thefailure to provide this study’s scientific evidence (particularlyin light of the NASCIS I and II criticisms) suggests that thechanges in right-side only motor scores are the only findingsthat approach significance at 1 year (P � 0.053) and argueagainst the meaningful nature of the data as interpreted andprovided by the authors (18, 32, 37). Finally, the clinical sig-nificance of the changes in motor scores between groups, inlight of the nonsignificant differences in patient function asdetermined by FIM scores, is not evident. NASCIS III patientswho received 48 MP treatment had a 2-fold higher incidenceof severe pneumonia, a 4-fold higher incidence of severesepsis, and a 6-fold higher incidence of death due to respira-tory complications than patients in the 24 MP treatment group(8, 32). These differences, although not statistically significant,raise questions about the safety of the 48-hour treatmentstrategy proposed for patients with ASCI treated within 3 to 8hours of injury. Additional important criticisms of theNASCIS III trial include those levied against both the NASCISI and II studies (i.e., lack of standardized medical treatment,lack of a minimum motor impairment for inclusion [hence,normal motor function patients admitted to the study], novertebral level of injury cutoff, and unclear statistical meth-odology, analysis, and data interpretation) (18, 32, 37). NAS-CIS III was designed and implemented to be a randomized,double-blind clinical study in an attempt to generate Class Ievidence on the efficacy of methylprednisolone, offered intwo different treatment regimens, and tirilazad mesylate afterASCI in human subjects. The absence of evidence for func-tional improvement in any group argues against the clinicalrelevance of any of these regimens.
Wing et al. (49) examined the effect of methylprednisoloneadministered per the NASCIS II protocol on avascular necro-sis (AVN) of the femoral heads of 91 ASCI patients, 59 whoreceived the corticosteroid, and 32 who did not. The authorsfound no case of AVN in their study population and estimatethe relative risk of AVN with high-dose 24-hour methylpred-nisolone therapy to be less than 5%.
In 2000, Pointillart et al. (38) reported the results of a pro-spective, randomized clinical trial designed to evaluate thesafety and effect of nimodipine, methylprednisolone, or bothversus no pharmacological therapy in 106 ASCI patients. Pa-tients were randomly assigned to one of four treatmentgroups, methylprednisolone per NASCIS II protocol (M), ni-modipine (N), both methylprednisolone and nimodipine(MN), and neither medication (P). Blinded neurological as-sessment was accomplished via the American Spinal CordInjury Association (ASIA) score at initiation of treatment andat 1-year follow-up. The authors performed early spinal de-compression and stabilization as indicated. One hundred pa-tients were available at 1-year follow-up. There was no sig-nificant difference in outcome among the four treatmentgroups for any of the ASIA scores recorded. Patients in allfour treatment groups demonstrated significant neurologicalimprovement at the 1-year follow-up compared with admis-sion (P � 0.0001). Two-way analysis of variance revealed nointeraction between methylprednisolone and nimodipine.There was a significant difference in recovery below the levelof injury among patients with complete SCIs compared withthose with incomplete injuries (P � 0.0001). Improvementamong complete injury patients, when present, involved thelevel of the lesion and the two adjacent caudal levels. Thegreatest neurological improvements were identified in incom-plete injury patients. There was no significant difference inneurological outcome for patients who underwent surgerywithin 8 hours of injury, patients treated surgically between 8and 24 hours after injury, and those managed without sur-gery. The incidence of infectious complications was higheramong patients treated with methylprednisolone comparedwith those who did not receive corticosteroids (66% versus45%), but this difference was not significant. The authorsconcluded that pharmacological therapy offered no addedbenefit to patients with ASCIs. Unfortunately, sample sizecalculations are not provided by the authors, and thereforethe statistical power of the study to show a significant benefitof the treatment(s) is unknown. In addition, indications forsurgery and for timing of surgery were not provided, poten-tially adding bias. The failure to show a difference betweengroups in this study may be explained by these potentialstudy design flaws.
A number of published critiques of the NASCIS data andtheir presentation in support of the use of methylprednisolonein the management of patients with ASCI have been offered(1, 18, 19, 22, 31–33a, 35, 37, 40–42, 44–46, 51). A recent medicalevidence-based review is provided by Short et al. (46). Theseauthors conclude, after review of the medical literature on theuse of methylprednisolone for ASCI (animal and human ex-perimental studies, including randomized human clinical tri-als), that the available evidence does not support the use of
Pharmacological Therapy in Acute Cervical Spinal Cord Injury S69
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
methylprednisolone in the treatment of ASCI. A number ofreviews that support the use of methylprednisolone afterASCI have also been published, including a Cochrane Data-base of Systematic Reviews (3–5, 6a, 7, 9, 10, 50).
In 2001, Matsumoto et al. (36) reported their results of aprospective, randomized double-blind clinical trial compar-ing methylprednisolone with placebo in the treatment of pa-tients with acute cervical SCI. The authors focused on poten-tial medical complications after ASCI. Forty-six patients wereincluded in the study: 23 treated with methylprednisolone perthe NASCIS II protocol were compared with 23 patients in aplacebo treatment group. Complications associated with ther-apy were noted at 2-month follow-up. Patients treated withmethylprednisolone had a higher incidence of complicationscompared with placebo-treated patients (56.5% versus 34.8%).Respiratory complications (P � 0.009) and gastrointestinalcomplications (P � 0.036) were the most significant betweenthe two treatment populations. The authors concluded thatpatients with ASCI treated with methylprednisolone (partic-ularly older patients) are at increased risk for pulmonary andgastrointestinal complications and deserve special care. Thisincidence of medical complications using methylprednisolonefor 24 hours seems clinically important. The NASCIS III studydemonstrated that these complications are even higher for48-hour methylprednisolone administration as describedabove (17). This calls into question the use of corticosteroidsfor any timeframe, but especially for the 48-hour duration.
Finally, a review of the data in a large number of patients inthe most recent GM-1 ganglioside trial who had methylpred-nisolone alone according to NASCIS II and III protocols didnot confirm the findings of the NASCIS II and III trials (23).This is described in detail in the section below on the GM-1ganglioside trials.
In summary, the available medical evidence does not supporta significant clinical benefit from the administration of methyl-prednisolone in the treatment of patients after ASCI for either 24or 48 hours duration. Three North American, multicenter ran-domized clinical trials have been completed and several otherstudies have been accomplished addressing this issue (11, 13–17,21, 28, 29, 38, 39). The neurological recovery benefit of methyl-prednisolone when administered within 8 hours of ASCI hasbeen suggested but not convincingly proven. The administrationof methylprednisolone for 24 hours has been associated with asignificant increase in severe medical complications. This is evenmore striking for methylprednisolone administered for 48 hours.In light of the failure of clinical trials to convincingly demon-strate a significant clinical benefit of administration of methyl-prednisolone, in conjunction with the increased risks of medicalcomplications associated with its use, methylprednisolone in thetreatment of acute human SCI is recommended as an option thatshould only be undertaken with the knowledge that the evi-dence suggesting harmful side effects is more consistent than thesuggestion of clinical benefit.
GM-1 ganglioside
GM-1 ganglioside has been evaluated in both animal andhuman studies of ASCI (2, 26, 27, 47, 48). In 1991, Geisler et al.
(25) described the results of a prospective, randomized,placebo-controlled, double-blind trial of GM-1 ganglioside inthe treatment of human patients with ASCI. Of 37 patientsentered into the study, 34 were available for 1-year follow-up(16 GM-1 patients, 18 placebo). All patients received a 250-mgbolus of methylprednisolone and then 125 mg every 6 hoursfor 72 hours. GM-1 patients were administered 100 mg ofGM-1 ganglioside per day for 18 to 32 days, with the first doseprovided within 72 hours of injury. Neurological evaluationwas accomplished with Frankel scale and ASIA motor scoreassessments. The authors reported that GM-1 ganglioside-treated patients had significant improvements in the distribu-tion of Frankel grades from baseline to 1-year follow-up (P �0.034) and significantly improved ASIA motor scores com-pared with placebo-treated patients (P � 0.047) (26, 27). Therecovery of motor function in GM-1 ganglioside-treated pa-tients was thought to caused by recovery of strength in par-alyzed muscles rather than strengthening of paretic muscles.There were no adverse effects attributed to the administrationof the study drug. The authors concluded that GM-1 gangli-oside enhances neurological recovery in human patients afterSCI and deserves further study.
In 1992, a multicenter GM-1 ganglioside ASCI study wasinitiated. It was a prospective, double-blind, randomized andstratified trial that enrolled 797 patients by study end in early1997 (23). All patients received methylprednisolone per theNASCIS II protocol. Patients were randomized into threeinitial study groups: placebo, low-dose GM-1 (300-mg loadingdose and then 100 mg/d for 56 d), and high-dose GM-1(600-mg loading dose and then 200 mg/d for 56 d). Placebo orGM-1 was administered at the conclusion of the 23-hourmethylprednisolone infusion. Patients were assessed usingthe modified Benzel Classification and the ASIA motor andsensory examinations at 4, 8, 16, 26, and 52 weeks after injury.Aggressive medical and surgical management paradigmswere used. Patients had to have an acute, nonpenetrating SCI(anatomic vertebral level C2 through T11) of at least moderateseverity (no neurologically normal or nearly normal patients).The primary efficacy assessment was the proportion of pa-tients who improved at least two grades from baseline exam-ination (defined as “marked recovery”), at Week 26 of thestudy. Secondary efficacy assessments included the timecourse of marked recovery, the ASIA motor score, and ASIAsensory evaluations, relative and absolute sensory levels ofimpairment, and assessments of bladder and bowel function.A planned interim analysis of the first 180 patients resulted inthe addition of stratification by patient age and discontinua-tion of the high-dose GM-1 treatment strategy because of anearly trend for higher mortality. At the study conclusion, 37patients were judged ineligible, leaving 760 patients for pri-mary efficacy analysis. The authors found no significant dif-ference in mortality between treatment groups (23). The au-thors did not identify a higher proportion of patients withmarked recovery in motor function at 26 weeks when theycompared GM-1-treated patients to the placebo treatmentgroup in their primary efficacy analysis. The time course ofrecovery indicated earlier attainment of marked recovery inGM-1-treated patients. The authors concluded that, despite
S70 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
the lack of statistical significance in the primary analysis,numerous positive secondary analyses indicate that GM-1ganglioside is a useful drug in the management of ASCI (23).The placebo group within this study of GM-1 represents agroup of 322 patients who received methylprednisolonewithin 8 hours of injury. Of interest, these 322 patients (mea-sured in a similar, albeit, more detailed manner as NASCIS IIpatients) did not demonstrate the previously published neu-rological examination improvement found in 62 NASCIS IIpatients treated within the same timeframe (13, 14). Similarly,218 of these patients received 24 hours methylprednisolonetreatment within 3 hours of injury, as suggested in NASCISIII, and did not show the same neurological examinationmotor improvement as the 75 NASCIS III patients who re-ceived the same regimen (16, 17). The authors could notconfirm the NASCIS findings that timing of methylpred-nisolone therapy had an impact on spinal cord recovery. Thisfurther brings into question the conclusions of the NASCIS IIand III methylprednisolone trials.
In summary, the available medical evidence does not sup-port a significant clinical benefit from the administration ofGM-1 ganglioside in the treatment of patients after ASCI. TwoNorth American multicenter, randomized clinical trials havebeen completed addressing this issue (23, 29). The neurolog-ical recovery benefit of GM-1 ganglioside when administeredfor 56 days after the administration of methylprednisolonewithin 8 hours of ASCI has been suggested but not convinc-ingly proven. At present, GM-1 ganglioside (a 300-mg loadingdose and then 100 mg/d for 56 d), when initiated after theadministration of methylprednisolone given within 8 hours ofinjury (NASCIS II protocol), is recommended as an option inthe treatment of adult patients with ASCI.
KEY ISSUES FOR FUTURE INVESTIGATION
Given the problems associated with the many trials at-tempting to answer the questions surrounding the use ofpharmacological agents in acute spinal cord-injured patients,it is clear that more research is required. Issues such as ade-quate numbers of patients to achieve statistical power, a pla-cebo group as one of the treatment arms, standardized med-ical and surgical protocols to diminish bias, careful collectionof relevant outcome data, especially functional outcomes, andappropriate statistical analyses need to be further addressed apriori. Research into all potentially promising pharmacologicalagents, including, but not limited to, tirilazad mesylate, nalox-one, methylprednisolone, and GM-1, should be undertaken.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Acland RH, Anthony A, Inglis GS, Walton DI, Xiong X: Methylpred-nisolone use in acute spinal cord injury. N Z Med J 9:99, 2001 (letter).
2. Amar PA, Levy ML: Pathogenesis and pharmacological strate-gies for mitigating secondary damage in acute spinal cord injury.Neurosurgery 44:1027–1040, 1999.
3. Bracken MB: National Acute Spinal Cord Injury Study of methyl-prednisolone or naloxone. Neurosurgery 28:628–629, 1991 (letter).
4. Bracken MB: Methylprednisolone and spinal cord injury.J Neurosurg 93:175–177, 2000 (letter).
5. Bracken MB: The use of methylprednisolone. J Neurosurg 93:340–341, 2000 (letter).
6. Bracken MB: High dose methylprednisolone must be given for 24 or48 hours after acute spinal cord injury. BMJ 322:862–863, 2001 (letter).
6a. Bracken MB: Methylprednisolone and acute spinal cord injury: Anupdate of the randomized evidence. Spine 26(24 Suppl):S47–S54, 2001.
7. Bracken MB: Pharmacological interventions for acute spinal cordinjury. Cochrane Database Syst Rev 1:1–32, 2001.
8. Bracken MB, Holford TR: Effects of timing of methylprednisoloneor naloxone administration on recovery of segmental and long-tractneurological function in NASCIS 2. J Neurosurg 79:500–507, 1993.
9. Bracken MB, Holford TR: Response: Treatment of spinal cordinjury. J Neurosurg 80:954–955, 1994 (letter).
10. Bracken MB, Aldrich EF, Herr DL, Hitchon PW, Holford TR,Marshall LF, Nockels RP, Pascale V, Shepard MJ, Sonntag VKH,Winn HR, Young W: Clinical measurement, statistical analysis,and risk-benefit: Controversies from trials of spinal injury.J Trauma 48:558–561, 2000.
11. Bracken MB, Collins WF, Freeman DF, Shepard MJ, Wagner FW,Silten RM, Hellenbrand KG, Ransohoff J, Hunt WE, Perot PL Jr,Grossman RG, Green BA, Eisenberg HM, Rifkinson N, GoodmanJH, Meagher JN, Fischer B, Clifton GL, Flamm ES, Rawe SE:Efficacy of methylprednisolone in acute spinal cord injury.JAMA 251:45–52, 1984.
12. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Flamm E, Eisenberg HM, Leo-Summers L, Maroon JC, Marshall LF,Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr, Wilberger JL,Winn HR, Young W: Response: Methylprednisolone for spinal cordinjury. J Neurosurg 77:325–327, 1992 (letter).
13. Bracken MB, Shepard MJ, Collins WF Jr, Holford TR, Baskin DS,Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC, Marshall LF,Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr, Wilberger JE,Winn HR, Young W: Methylprednisolone or naloxone treatmentafter acute spinal cord injury: 1-year follow-up data—Results of theSecond National Acute Spinal Cord Injury Study. J Neurosurg76:23–31, 1992.
14. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized, controlled trial of meth-ylprednisolone or naloxone in the treatment of acute spinal cordinjury: Results of the Second National Acute Spinal Cord InjuryStudy (NASCIS-2). N Engl J Med 322:1405–1411, 1990.
15. Bracken MB, Shepard MJ, Hellenbrand KG, Collins WF, Leo-Summers L, Freeman DF, Wagner FC, Flamm ES, Eisenberg HM,Goodman JH, Perot PL Jr, Green BA, Grossman RG, Meagher JN,Young W, Fischer B, Clifton GL, Hunt WE, Rifkinson N: Meth-ylprednisolone and neurological function 1 year after spinal cordinjury: Results of the National Acute Spinal Cord Injury Study.J Neurosurg 63:704–713, 1985.
16. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings M, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, Sonntag VKH,Wagner F, Wilberger JE, Winn HR, Young W: Administration ofmethylprednisolone for 24 or 48 hours or tirilazad mesylate for48 hours in the treatment of acute spinal cord injury: Results ofthe Third National Acute Spinal Cord Injury Randomized Con-trolled Trial—National Acute Spinal Cord Injury Study. JAMA277:1597–1604, 1997.
Pharmacological Therapy in Acute Cervical Spinal Cord Injury S71
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
17. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings MG, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, Sonntag VKH,Wagner F, Wilberger JE, Winn HR, Young W: Methylpred-nisolone or tirilazad mesylate administration after acute spinalcord injury: 1-year follow-up—Results of the Third NationalAcute Spinal Cord Injury Randomized Controlled Trial.J Neurosurg 89:699–706, 1998.
18. Coleman WP, Benzel D, Cahill DW, Ducker T, Geisler F, GreenB, Gropper MR, Goffin J, Madsen PW III, Maiman DJ, Ondra SL,Rosner M, Sasso RC, Trost GR, Zeidman S: A critical appraisal ofthe reporting of the National Acute Spinal Cord Injury Studies(II and III) of methylprednisolone in acute spinal cord injury.J Spinal Disord 13:185–199, 2000.
19. Ducker TB, Zeidman SM: Spinal cord injury: Role of steroidtherapy. Spine 19:2281–2287, 1994.
20. Duh MS, Shepard MJ, Wilberger JE, Bracken MB: The effective-ness of surgery on the treatment of acute spinal cord injury andits relation to pharmacological treatment. Neurosurgery 35:240–249, 1994.
21. Galandiuk S, Raque G, Appel S, Polk HC Jr: The two-edgedsword of large-dose steroids for spinal cord trauma. Ann Surg218:419–427, 1993.
22. Geisler FH: Commentary on NASCIS-2. J Spinal Disord 5:132–133,1992 (comment).
23. Geisler FH, Coleman WP, Grieco G, Dorsey FC, Poonian D, TheSygen Study Group: The GM1 ganglioside multi-center acutespinal cord injury study. Spine 26(24 Suppl):S87–S98, 2001.
24. Geisler FH, Dorsey FC, Coleman WP: Correction: Recovery of motorfunction after spinal cord injury—A randomized, placebo-controlledtrial with GM-1 ganglioside. N Engl J Med 325:1659–1660, 1991.
25. Geisler FH, Dorsey FC, Coleman WP: Recovery of motor func-tion after spinal cord injury: A randomized, placebo-controlledtrial with GM-1 ganglioside. N Engl J Med 324:1829–1838, 1991.
26. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside in hu-man spinal cord injury. J Neurotrauma 9[Suppl 2]:S517–S530,1992.
27. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside forspinal cord injury. N Engl J Med 326:494, 1992.
28. George ER, Scholten DJ, Buechler CM, Jordan-Tibbs J, Mattice C,Albrecht RM: Failure of methylprednisolone to improve theoutcome of spinal cord injuries. Am Surg 61:659–664, 1995.
29. Gerhart KA, Johnson RL, Menconi J, Hoffman RE, LammertseDP: Utilization and effectiveness of methylprednisolone in apopulation-based sample of spinal cord injured persons. Para-plegia 33:316–321, 1995.
30. Gerndt SJ, Rodriguez JL, Pawlik JW, Taheri PA, Wahl WL,Micheals AJ, Papadopoulos SM: Consequences of high-dose ste-roid therapy for acute spinal cord injury. J Trauma 42:279–284,1997.
31. Hanigan WC, Anderson RJ: Commentary on NASCIS-2. J SpinalDisord 5:125–131, 1992.
32. Hurlbert RJ: Methylprednisolone for acute spinal cord injury: Aninappropriate standard of care. J Neurosurg 93[Suppl 1]:1–7, 2000.
33. Hurlbert RJ: The use of methylprednisolone. J Neurosurg93[Suppl 1]:340–341, 2000 (letter).
33a. Hurlbert RJ: The role of steroids in acute spinal cord injury: Anevidence-based analysis. Spine 26(24 Suppl):S39–S46, 2001.
34. Landi G, Ciccone A: GM-1 ganglioside for spinal cord injury.N Engl J Med 326:493, 1992 (letter).
35. Lyons MK, Partington MD, Meyer FB: A randomized, controlledtrial of methylprednisolone or naloxone in the treatment of acutespinal-cord injury. N Engl J Med 323:1207–1208, 1990.
36. Matsumoto T, Tamaki T, Kawakami M, Yoshida M, Ando M,Yamada H: Early complications of high-dose methylpred-nisolone sodium succinate treatment in the follow-up of acutecervical spinal cord injury. Spine 26:426–430, 2001.
37. Nesathurai S: Steroids and spinal cord injury: Revisiting theNASCIS 2 and NASCIS 3 trials. J Trauma 45:1088–1093, 1998.
38. Pointillart V, Petitjean ME, Wiart L, Vital JM, Lassie P, ThicoipeM, Dabadie P: Pharmacological therapy of spinal cord injuryduring the acute phase. Spinal Cord 38:71–76, 2000.
39. Poynton AR, O’Farrell DA, Shannon F, Murray P, McManus F,Walsh MG: An evaluation of the factors affecting neurological re-covery following spinal cord injury. Injury 28:545–548, 1997.
40. Rosner MJ: National acute spinal cord injury study of methyl-prednisolone or naloxone. Neurosurgery 28:628, 1991 (letter).
41. Rosner MJ: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–325, 1992.
42. Rosner MJ: Treatment of spinal cord injury. J Neurosurg 80:954–955, 1994.
43. Schönhöfer PS: GM-1 ganglioside for spinal cord injury. N EnglJ Med 326:493, 1992 (letter).
44. Shapiro SA: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–327, 1992.
45. Short DJ: Use of steroids for acute spinal cord injury must bereassessed. BMJ 321:1224, 2000 (letter).
46. Short DJ, El Masry WS, Jones PW: High dose methylpred-nisolone in the management of acute spinal cord injury: A system-atic review from a clinical perspective. Spinal Cord 38:273–286,2000.
47. Tator CH: Experimental and clinical studies of the pathophysi-ology and management of acute spinal cord injury. J SpinalCord Med 19:206–214, 1996.
48. Tator CH: Biology of neurological recovery and functional res-toration after spinal cord injury. Neurosurgery 42:696–708, 1998.
49. Wing PC, Nance P, Connell DG, Gagnon F: Risk of avascularnecrosis following short term megadose methylprednisolonetreatment. Spinal Cord 36:633–636, 1998.
50. Young W, Bracken MB: The Second National Acute Spinal CordInjury Study. J Neurotrauma 9[Suppl 1]:S397–S405, 1992.
51. Zeidman SM, Ling GS, Ducker TB, Ellenbogen RG: Clinicalapplications of pharmacologic therapies for spinal cord injury.J Spinal Disord 9:367–380, 1996.
S72 Guidelines for Management of Acute Cervical Spinal Injuries
CHAPTER 10
Deep Venous Thrombosis and Thromboembolism in Patientswith Cervical Spinal Cord Injuries
RECOMMENDATIONSSTANDARDS:• Prophylactic treatment of thromboembolism in patients with severe motor deficits due to spinal cord injury
is recommended.• The use of low-molecular-weight heparins, rotating beds, adjusted dose heparin, or a combination of
modalities is recommended as a prophylactic treatment strategy.• Low-dose heparin in combination with pneumatic compression stockings or electrical stimulation is rec-
ommended as a prophylactic treatment strategy.
GUIDELINES:• Low-dose heparin therapy alone is not recommended as a prophylactic treatment strategy.• Oral anticoagulation alone is not recommended as a prophylactic treatment strategy.
OPTIONS:• Duplex Doppler ultrasound, impedance plethysmography, and venography are recommended for use as
diagnostic tests for deep venous thrombosis in the spinal cord-injured patient population.• A 3-month duration of prophylactic treatment for deep venous thrombosis and pulmonary embolism is
recommended.• Vena cava filters are recommended for patients who do not respond to anticoagulation or who are not
candidates for anticoagulation therapy and/or mechanical devices.
RATIONALE
Deep venous thrombosis (DVT) and pulmonary embo-lism (PE) are problems frequently encountered inpatients who have sustained cervical spinal cord in-
juries (SCIs). Several means of prophylaxis and treatment areavailable, including anticoagulation, pneumatic compressiondevices, and vena cava filters. The purpose of this evidence-based medicine review is to evaluate the literature on themethods of prevention and identification of DVT and PEcomplications in patients after acute cervical SCI.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 through 2001was performed. The following medical subject headings wereused in combination with “spinal cord injury”: “deep venousthrombosis,” “pulmonary embolism,” and “thromboembo-lism.” The search was limited to human studies in the Englishlanguage. This resulted in 129 citations. Duplicate references,reviews, letters, and tangential reports were discarded.Thirty-seven articles dealing with the prophylaxis or treat-
ment of thromboembolic disease in adult SCI patients makeup the basis for this guideline and are summarized in Table10.1. Supporting references included four evidence-based re-views published by various organizations concerned withthromboembolism prophylaxis and treatment in different pa-tient populations. Finally, several series dealing with throm-boembolism in general trauma patients with results germaneto a discussion of SCI patients are included among the refer-ences as supporting documents.
SCIENTIFIC FOUNDATION
The incidence of thromboembolic complications in the un-treated SCI patient population is high. Depending on injuryseverity, patient age, and the methods used to diagnose athromboembolism, the incidence of thromboembolic eventsranges from 7 to 100% in reported series of patients receivingeither no prophylaxis or inadequate prophylaxis (3, 9, 11, 13,16, 22, 23, 26, 27, 29, 30, 33, 36). Substantial morbidity andmortality have been associated with the occurrence of DVTand PE events in the SCI patient population (6, 7).
S73Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE10
.1.
Sum
mar
yof
Rep
orts
onD
eep
Ven
ous
Thro
mbo
sis
and
Pulm
onar
yEm
bolis
min
Pati
ents
wit
hC
ervi
cal
Spin
alC
ord
Inju
ries
a
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Che
net
al.,
1999
(4a)
Larg
epo
pula
tion
ofA
SCI
patie
nts
(164
9)st
udie
dfr
omad
mis
sion
tore
habi
litat
ion
(mea
n19
d)to
disc
harg
e(m
ean,
50d)
.In
cide
nce
ofD
VT
�PE
decl
inin
gov
ertim
e,
but
rem
ains
6.1%
desp
itepr
ophy
laxi
s.
IIID
VT/
PEst
illpr
oble
ms
desp
itepr
ophy
laxi
s.(S
eeM
cKin
ley
etal
.(2
4)
for
follo
w-u
p).
McK
inle
yet
al.,
1999
(24)
Chr
onic
SCI
popu
latio
nst
udie
d.In
cide
nce
ofD
VT
high
est
duri
ng1s
tyr
(2.1
%)
but
then
drop
sof
fto
0.5–
1%/y
rth
erea
fter.
IIIR
isk
ofD
VT/
PEhi
ghes
tdu
ring
1st
yraf
ter
inju
ryan
dth
enri
skdr
ops
sign
ifica
ntly
.
Pow
ell
etal
.,19
99(3
0)In
cide
nce
ofD
VT
inSC
Ipo
pula
tion
(n�
189)
ontr
ansf
erto
reha
bilit
atio
n
(dia
gnos
isw
ithU
S)w
as4.
1%in
grou
pw
hore
ceiv
edpr
ophy
laxi
sve
rsus
16.4
%in
grou
pw
ithou
tpr
ophy
laxi
s.In
prop
hyla
xis
grou
p,D
VTs
only
occu
rred
inpa
tient
s
rece
ivin
ghe
pari
nal
one.
IIPr
ophy
laxi
sde
crea
ses
inci
denc
eof
DV
Tin
SCI
popu
latio
n.H
epar
in
alon
ew
asth
ele
ast
effe
ctiv
em
easu
re.
Rou
ssi
etal
.,19
99(3
2)6
of67
(9%
)of
SCI
patie
nts
deve
lope
dD
VT
desp
itepr
ophy
laxi
sw
ithLM
WH
.
D-d
imer
had
100%
nega
tive
pred
ictiv
eva
lue
com
pare
dw
ithdu
plex
Dop
pler
.
(How
ever
,sp
ecifi
city
only
34%
and
posi
tive
pred
ictiv
eva
lue
13%
).
Ifo
rdi
agno
stic
test
;III
othe
rwis
eIn
cide
nce
ofD
VT
desp
itepr
ophy
laxi
sw
ithLM
WH
still
9%.
D-d
imer
isse
nsiti
vebu
tno
tsp
ecifi
cte
stfo
rD
VT.
Win
emill
eret
al.,
1999
(40)
Ret
rosp
ectiv
est
udy
of42
8SC
Ipa
tient
s.TE
occu
rred
in19
.6%
.C
ompr
essi
on
stoc
king
san
dse
quen
tial
com
pres
sion
devi
ces
low
ered
risk
ofTE
.Ef
fect
sof
low
-
dose
hepa
rin
wer
ese
enin
first
14d
but
wer
eno
tsi
gnifi
cant
.TE
sal
loc
curr
edin
first
150
d.
IIISe
quen
tial
com
pres
sion
devi
ces
and
stoc
king
sre
duce
risk
ofTE
.
Low
-dos
ehe
pari
nm
aybe
effe
ctiv
ein
first
14d
afte
rin
jury
.
Tom
aio
etal
.,19
98(3
5a)
Enox
apar
in(L
MW
H)
vers
ushe
pari
nus
efo
rin
itial
DV
Ttr
eatm
ent
ingr
oup
of6
SCI
patie
nts.
IIIEn
oxap
arin
was
cost
-effe
ctiv
eal
tern
ativ
eto
intr
aven
ous
hepa
rin
for
initi
altr
eatm
ent
ofD
VT.
Har
ris
etal
.,19
96(1
8)R
etro
spec
tive
stud
yof
enox
apar
in(L
MW
H)
in10
5SC
Ipa
tient
s(o
ne-t
hird
inta
ct,
40
com
plet
e).
No
clin
ical
DV
T/PE
in10
5,no
US
evid
ence
in60
.
IIIEn
oxap
arin
issa
fean
def
fect
ive
for
DV
Tpr
ophy
laxi
sin
the
SCI
patie
nt.
Kha
nsar
inia
etal
.,19
95(2
0)R
etro
spec
tive
hist
oric
alco
hort
com
pari
son
ofpr
ophy
lact
icPG
Fin
324
gene
ral
trau
ma
patie
nts.
PGF
grou
pha
dfe
wer
PEs
than
cont
rol
grou
p.
IIIG
reen
field
filte
rsa
fean
def
fect
ive
for
PEpr
ophy
laxi
sin
gene
ral
trau
ma
popu
latio
n.
Gee
rts
etal
.,19
94(1
1)Pr
ospe
ctiv
eev
alua
tion
of71
6tr
aum
apa
tient
s(n
opr
ophy
laxi
s)w
ithV
OP
and
veno
grap
hy.
Inci
denc
eof
DV
Tin
SCI
popu
latio
n(n
�66
)w
as62
%.
IIID
VT
isve
ryco
mm
onin
SCI
patie
nts
ifno
prop
hyla
xis
used
.
Wils
onet
al.,
1994
(39)
Inse
rted
cava
lfil
ters
in15
SCI
patie
nts.
Non
eha
dD
VT
orPE
in1
yr.
Cla
ims
this
resu
ltsu
peri
orto
hist
oric
alco
ntro
ls(n
oev
iden
cepr
esen
ted
tosu
ppor
tth
iscl
aim
).
1-yr
pate
ncy
rate
was
81%
.
IIIIn
sert
ion
ofca
val
filte
rsap
pear
sto
besa
fein
SCI
patie
nts.
Gre
enet
al.,
1994
(12)
His
tori
cal
coho
rtco
mpa
riso
nof
LMW
Han
dst
anda
rdan
dad
just
eddo
sehe
pari
n
prop
hyla
xis.
Trau
ma
patie
nts
trea
ted
with
8-w
kco
urse
ofLM
WH
had
few
er
blee
ding
epis
odes
(P�
0.05
)an
dTE
com
plic
atio
ns(P
�0.
15)
than
thos
etr
eate
d
with
hepa
rin.
IIILM
WH
may
besa
fer
and
mor
eef
fect
ive
for
prop
hyla
xis
than
min
idos
eor
adju
sted
dose
hepa
rin.
Gun
duz
etal
.,19
93(1
6)31
SCI
patie
nts
onlo
w-d
ose
hepa
rin
ther
apy
unde
rwen
tve
nogr
aphy
.In
cide
nce
of
DV
Tw
as53
.3%
.
IIIIn
cide
nce
ofD
VT
high
inSC
Ipa
tient
sde
spite
low
-dos
ehe
pari
n
(ther
apy
star
ted
onre
habi
litat
ion
unit)
.
Bur
nset
al.,
1993
(3)
Pros
pect
ive
asse
ssm
ent
ofD
VT
inhi
gh-r
isk
trau
ma
patie
nts
with
US.
Foun
d
inci
denc
eof
21%
(12/
57)
desp
itelo
w-d
ose
hepa
rin
orpn
eum
atic
boot
sin
85%
.
IIID
VT
isco
mm
onde
spite
use
oflo
w-d
ose
hepa
rin
orpn
eum
atic
boot
s.
Lam
bet
al.,
1993
(23)
287
chro
nica
llyin
jure
dSC
Ipa
tient
sfo
llow
ed.
Ove
rall
inci
denc
eof
TEev
ents
was
10%
,m
ost
even
tsin
first
6m
o.
IIIPr
ophy
lact
icth
erap
yno
tne
cess
ary
beyo
nd6
mo
inSC
Ipo
pula
tion.
Kul
karn
iet
al.,
1992
(22)
100
SCI
patie
nts
pros
pect
ivel
ytr
eate
dw
ithlo
w-d
ose
hepa
rin.
26%
inci
denc
eof
clin
ical
lyde
tect
edD
VT
(9%
PE)
note
d.
IIID
VT
and
PEin
cide
nce
sign
ifica
ntde
spite
low
-dos
esu
bcut
aneo
us
hepa
rin.
Mer
liet
al.,
1992
(25)
Hep
arin
�pn
eum
atic
stoc
king
seq
ual
tohi
stor
ical
cont
rols
ofhe
pari
n�
stim
ulat
ion
and
bette
rth
anhi
stor
ical
cont
rols
ofhe
pari
nor
plac
ebo
inSC
Ipa
tient
s.
IILo
w-d
ose
hepa
rin
�pn
eum
atic
hose
safe
and
effe
ctiv
eas
DV
T
prop
hyla
xis
inSC
Ipa
tient
s.
War
ing
and
Kar
unas
,19
91(3
6)La
rge
data
base
(141
9)of
SCI
patie
nts.
Inci
denc
eof
DV
Tw
as14
.5%
,PE
4.6%
.
Seve
rity
ofin
jury
was
apr
edic
tor
ofD
VT
and
age
was
apr
edic
tor
ofPE
.N
o
men
tion
mad
eof
prop
hyla
ctic
mea
sure
s.
IIID
VT
and
PEar
esi
gnifi
cant
prob
lem
sin
SCI
popu
latio
n.A
gean
d
inju
ryse
veri
tyne
edto
bead
dres
sed
inst
udie
sco
mpa
ring
trea
tmen
t
mod
aliti
es.
S74 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE10
.1.
Con
tinu
ed
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Yel
nik
etal
.,19
91(4
1)Pr
ospe
ctiv
est
udy
of12
7SC
Ipa
tient
sw
ithph
lebo
grap
hy.
29/1
27ha
dD
VT
on
adm
issi
onto
reha
bilit
atio
nun
it.O
f87
patie
nts
with
initi
ally
nega
tive
stud
ies,
12
deve
lope
dD
VT
desp
itepr
ophy
laxi
sfo
rup
to80
d.
IIIIn
cide
nce
ofD
VT
inSC
Ipo
pula
tion
ishi
gh,
and
high
-ris
kpe
riod
exte
nds
toen
dof
3rd
mo.
Aut
hors
reco
mm
end
peri
odic
scre
enin
g
with
phle
bogr
aphy
.
Bal
shi
etal
.,19
89(1
)C
ase
seri
esof
13qu
adri
pleg
icpa
tient
sw
hoha
dve
naca
val
filte
rspl
aced
for
DV
T
orPE
.A
bnor
mal
ities
ofth
efil
ter
wer
ede
tect
edin
5/11
patie
nts
who
had
follo
w-u
p
x-ra
ys:
2pa
tient
sre
quir
edla
paro
tom
yto
rem
ove
filte
rs,
4ha
ddi
stal
mig
ratio
n,an
d
2ha
dna
rrow
ing
ofdi
amet
eras
soci
ated
with
cava
loc
clus
ion.
9of
thes
e11
patie
nts
wer
etr
eate
dw
ithth
e“q
uad
coug
h”te
chni
que.
IIIFi
lter
plac
emen
tm
aybe
asso
ciat
edw
ithsi
gnifi
cant
long
-ter
m
mor
bidi
tyin
the
SCI
popu
latio
n,pa
rtic
ular
lyth
ose
requ
irin
g
aggr
essi
vepu
lmon
ary
toile
t.
DeV
ivo
etal
.,19
89(7
)Ep
idem
iolo
gica
lst
udy
ofca
uses
ofde
ath
for
SCI
patie
nts.
Hig
hest
ratio
sof
actu
alto
expe
cted
caus
esof
deat
hw
ere
for
pneu
mon
ia,
PE,
and
sept
icem
ia.
The
risk
ratio
for
TEdr
oppe
dsi
gnifi
cant
lyaf
ter
the
1st
mo
post
inju
rybu
tre
mai
ned
elev
ated
at6
mo
post
inju
ry.
IIITE
isa
sign
ifica
ntpr
oble
mfo
rpa
tient
sw
hosu
rviv
eSC
I.Pe
riod
of
grea
test
risk
isin
first
few
mon
ths
afte
rin
jury
,bu
tri
skco
ntin
ues
even
afte
r6
mo.
Gre
enet
al.,
1988
(13)
Ran
dom
ized
cont
rolle
dtr
ial
oflo
w-d
ose
vers
usad
just
eddo
sehe
pari
nin
SCI
patie
nts.
Rat
eof
TElo
wer
inad
just
eddo
segr
oup
(7%
vers
us31
%)
(inte
ntto
trea
t,
P�
NS)
,bu
tal
soha
dhi
gher
rate
ofbl
eedi
ngco
mpl
icat
ions
(7/2
9).
IA
djus
ted
dose
hepa
rin
mor
eef
fect
ive
than
low
-dos
ehe
pari
n,
blee
ding
mor
eco
mm
onin
adju
sted
dose
grou
p.
Mer
liet
al.,
1988
(26)
Pros
pect
ive
rand
omiz
edtr
ial
ofpl
aceb
ove
rsus
min
idos
ehe
pari
nve
rsus
hepa
rin
�
elec
tric
alst
imul
atio
nin
grou
pof
48SC
Ipa
tient
s.H
epar
ingr
oup
�pl
aceb
ogr
oup
at50
%,
stim
ulat
edgr
oup
sign
ifica
ntly
few
erD
VT.
ILo
w-d
ose
hepa
rin
nobe
tter
than
plac
ebo,
hepa
rin
�el
ectr
ical
stim
ulat
ion
muc
hbe
tter
for
DV
Tpr
ophy
laxi
sin
SCI
patie
nts.
Wei
ngar
den
etal
.,19
88(3
8a)
Ret
rosp
ectiv
ere
view
of14
8SC
Ipa
tient
s.10
had
docu
men
ted
DV
Tor
PE.
Of
6
patie
nts
who
had
adeq
uate
reco
rds,
all
6ha
dfe
ver
asa
pres
entin
gsi
gn,
4ha
dno
othe
rcl
inic
alsi
gns
reco
rded
.A
llep
isod
esoc
curr
edin
first
12w
k.
IIIFe
ver
may
indi
cate
TEdi
seas
ein
SCI
patie
nts.
Bec
ker
etal
.,19
87(2
)R
ando
miz
edtr
ial
ofro
tatin
gve
rsus
nonr
otat
ing
beds
inth
eac
ute
setti
ngaf
ter
SCI
(10
d).
n�
15pa
tient
s.Pl
ethy
smog
raph
yan
dfib
rino
gen
leg
scan
sus
ed.
IR
otat
ing
beds
redu
ceth
ein
cide
nce
ofD
VT
duri
ngth
efir
st10
daf
ter
SCI.
Tato
r,19
87(3
3)17
%in
cide
nce
ofD
VT
inse
ries
of20
8SC
Ipa
tient
s.In
cide
nce
was
high
erin
oper
ated
patie
nts
(23%
)co
mpa
red
with
nono
pera
ted
(10%
).U
seof
prop
hyla
xis
is
not
men
tione
d.
IIIPa
tient
sre
quir
ing
surg
ery
may
have
high
erin
cide
nce
ofD
VT.
Chu
etal
.,19
85(5
)C
ompa
riso
nbe
twee
nD
oppl
erU
S,V
OP,
and
clin
ical
exam
inat
ion
inSC
Ipa
tient
s.
All
had
sens
itivi
tyan
dsp
ecifi
city
of10
0%in
smal
l(n
�21
)se
ries
.O
vera
ll
inci
denc
e,19
%(C
lass
IIIbe
caus
eno
gold
stan
dard
used
).
IIID
oppl
erU
S,V
OP,
and
clin
ical
exam
inat
ion
all
good
for
diag
nosi
sof
DV
T.
Myl
lyne
net
al.,
1985
(27)
Com
pare
din
cide
nce
ofD
VT
inim
mob
ilize
dsp
ine-
inju
red
patie
nts
with
and
with
out
para
lysi
s.Th
ose
with
para
lysi
sha
da
100%
DV
Tin
cide
nce
(fibr
inog
en
scan
)ve
rsus
0%fo
rpa
tient
sim
mob
ilize
daf
ter
spin
alfr
actu
rew
ithou
tpa
raly
sis.
IIIIn
cide
nce
ofD
VT
isve
ryhi
ghin
SCI
patie
nts
and
isno
tto
tally
depe
nden
ton
imm
obili
zatio
n.
ElM
asri
and
Silv
er,
1981
(8)
Ret
rosp
ectiv
ere
view
of10
2pa
tient
sw
ithSC
I.Th
ere
wer
e21
epis
odes
ofPE
in19
patie
nts.
No
patie
ntw
ithPE
was
adeq
uate
lyan
ticoa
gula
ted
atth
etim
eof
the
PE
(ora
lph
enin
dion
e).
Onl
y8/
19pa
tient
sha
dev
iden
ceof
DV
Tby
exam
inat
ion
or
VO
P.
IIIA
utho
rsre
com
men
dpr
olon
ged
trea
tmen
t(�
6m
o)in
patie
nts
with
obes
ityor
prio
rhi
stor
yof
DV
T.
Fris
bie
and
Sasa
hara
,19
81(9
)Sm
all,
pros
pect
ive
cont
rolle
dst
udy
oflo
w-d
ose
(500
0un
itstw
ice
daily
)he
pari
n
vers
usco
ntro
lgr
oup.
No
diffe
renc
ein
inci
denc
eof
DV
Tno
ted
(onl
y7%
inea
ch
grou
p).
Aut
hors
sugg
est
prot
ectiv
eef
fect
offr
eque
ntph
ysio
ther
apy.
IIN
odi
ffere
nce
betw
een
low
-dos
ehe
pari
nan
dco
ntro
lgr
oups
inSC
I
patie
nts
rece
ivin
gtw
ice
daily
phys
ioth
erap
y.
Perk
ash
etal
.,19
80(2
8a)
Trea
tmen
tof
8pa
tient
sw
ithTE
disc
usse
d.A
utho
rsus
edhe
pari
nan
dth
enw
arfa
rin
with
reas
onab
lere
sults
.
IIIA
ntic
oagu
latio
nis
effe
ctiv
etr
eatm
ent
for
SCI
patie
nts
with
TE.
Perk
ash
etal
.,19
78(2
9)In
cide
nce
ofTE
in48
SCI
patie
nts
was
18%
.C
linic
alex
amin
atio
nse
nsiti
vity
89%
,
spec
ifici
ty88
%,
nega
tive
pred
ictiv
eva
lue
97%
,po
sitiv
epr
edic
tive
valu
e62
%.
One
-thi
rdof
TEev
ents
occu
rred
�12
wk
afte
rin
jury
.
Ifo
rdi
agno
stic
test
;III
othe
rwis
eC
linic
alex
amin
atio
nap
pear
sto
bequ
itego
odfo
rde
tect
ion
ofD
VT
insu
bacu
tese
tting
.Pe
riod
ofri
skm
ayex
tend
beyo
nd12
wk.
Wat
son,
1978
(38)
Ret
rosp
ectiv
ehi
stor
ical
coho
rtst
udy
inve
stig
atin
glo
w-d
ose
hepa
rin
vers
usno
prop
hyla
xis.
IIIH
epar
ingr
oup
had
few
erTE
com
plic
atio
ns.
No
TEev
ents
afte
r3
mo
desp
itepr
ophy
laxi
sce
ssat
ion
at3
mo.
Deep Venous Thrombosis and Thromboembolism S75
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Prophylaxis
Prophylactic therapy has been shown to be effective for theprevention of DVT and PE. In a small randomized study,Becker et al. (2) demonstrated that the use of rotating bedsduring the first 10 days after SCI decreased the incidence ofDVT. Four of five control patients were diagnosed with DVT(by fibrinogen screening) compared with one of 10 treatedpatients. The use of low-dose heparin (5000 units given viasubcutaneous injection twice or three times daily) has beendescribed by several authors (4, 9, 16, 17, 22, 30, 38). Hachen(17) published the results of a retrospective historical compar-ison of low-dose heparin versus oral anticoagulation in agroup of 120 SCI patients. He found a lower incidence ofthromboembolic events in the low-dose heparin group com-pared with the oral anticoagulation group. In 1977, Casas et al.(4) reported the results of a prospective assessment of low-dose heparin in SCI patients. These authors administeredheparin for a mean period of 66 days in 18 SCI patients andnoted no thromboembolic events as detected by clinical ex-amination. Watson (38) reported a lower incidence of throm-boembolic events with the use of low-dose heparin than withno prophylaxis in a retrospective historical cohort study. Fris-bie and Sasahara (9), however, found that low-dose heparindid not affect the incidence of DVT in a prospective study of32 SCI patients compared with treatment with twice dailyphysical therapy alone. These authors thought that the lack ofeffect was due to the very low incidence of DVT in theircontrol group compared with other series because of theaggressive physical therapy paradigm used in their patients.Although they performed venous occlusion plethysmography(VOP) screening with confirmatory venography weekly, theincidence of DVT was only 7% in both groups. An identicalobserved frequency of DVT in both treatment groups cannotbe explained by anything other than that the treatments wereequivalent in this study. This incidence is substantially lowerthan that reported by two separate groups of investigators adecade later. In 1992, Kulkarni et al. (22) reported a muchhigher incidence of DVT (26%) and of PE (9%) in a group of100 SCI patients prospectively treated with low-dose heparin.In 1993, Gunduz et al. (16) reported a 53% incidence of DVTconfirmed by venography in 31 patients they managed withSCI treated with low-dose heparin. In a study published in1999, Powell et al. (30) noted that the incidence of DVT in 189SCI patients receiving prophylaxis was significantly lowerthan that identified in SCI patients who did not receive pro-phylaxis, 4.1% versus 16.4%. They found, in addition, thatDVT in the prophylaxis group occurred in patients who re-ceived low-dose heparin alone.
Several studies have demonstrated that better prophylactictherapies than low-dose heparin exist (13, 25, 26). Green et al.(13) published a randomized controlled study comparinglow-dose versus adjusted dose heparin (dose adjusted to 1.5times normal activated partial thromboplastin time) in SCIpatients. They found that patients treated with adjusted doseheparin had fewer thromboembolic events (7% versus 31%)during the course of their 10-week study, but the patients hada higher incidence of bleeding complications. Merli et al. (26),TA
BLE
10.1
.C
onti
nued
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Cas
aset
al.,
1978
(4)
Pros
pect
ive
asse
ssm
ent
oflo
w-d
ose
hepa
rin
in18
/21
patie
nts
with
SCI
(mea
n
dura
tion,
66d)
.N
opa
tient
trea
ted
had
sym
ptom
atic
DV
Tor
PE.
No
use
of
US/
plet
hysm
ogra
phy/
veno
grap
hy.
IIILo
w-d
ose
hepa
rin
may
beus
eful
for
prev
entio
nof
sym
ptom
atic
DV
T.
Todd
etal
.,19
76(3
4)U
sed
VO
P,fib
rino
gen
scan
,an
dve
nogr
aphy
tost
udy
20SC
Ipa
tient
sfo
r60
d.
Fibr
inog
ensc
anw
aspo
sitiv
ein
all
patie
nts
but
was
conf
irm
edby
anot
her
test
in
only
half
ofth
eca
ses.
IIID
VT
isco
mm
onin
SCI
popu
latio
n.
Hac
hen,
1974
(17)
Coh
ort
cont
rolle
dtr
ial
oflo
w-d
ose
hepa
rin
(500
0un
itsth
ree
times
daily
)ve
rsus
oral
war
fari
nin
SCI
patie
nts.
Hep
arin
grou
pha
dsi
gnifi
cant
lyfe
wer
TEev
ents
.
IILo
w-d
ose
subc
utan
eous
hepa
rin
bette
rth
anor
alw
arfa
rin
for
prop
hyla
xis
afte
rac
ute
SCI.
Nas
o,19
74(2
8)PE
occu
rred
in4/
26pa
tient
sw
ithac
ute
(�3
mo)
SCI
but
none
occu
rred
in17
patie
nts
with
chro
nic
(�3
mo)
SCI.
IIISC
Ipa
tient
spr
imar
ilyat
risk
duri
ngfir
st3
mo
afte
rin
jury
.
Wat
son,
1968
(37)
Inci
denc
eof
TEco
mpl
icat
ions
per
year
rang
esfr
om8
to40
%in
sam
eun
it(n
o
prop
hyla
xis)
.
IIITE
com
plic
atio
nsar
ea
sign
ifica
ntpr
oble
m,
and
ther
eis
vari
abili
ty
year
toye
arde
spite
iden
tical
trea
tmen
tst
rate
gies
.
aSC
I,sp
inal
cord
inju
ry;A
SCI,
acut
eSC
I;D
VT,
deep
veno
usth
rom
bosi
s;PE
,pul
mon
ary
embo
lism
;LM
WH
,low
-mol
ecul
ar-w
eigh
the
pari
ns;T
E,th
rom
boem
bolis
m;U
S,ul
tras
ound
;PG
F,pr
ophy
lact
icG
reen
field
filte
r;V
OP,
veno
usoc
clus
ion
plet
hysm
ogra
phy;
NS,
not
sign
ifica
nt.
S76 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
in 1988, reported their findings of the additive protectiveeffects of electrical stimulation in combination with low-doseheparin, heparin alone, and placebo in 48 SCI patients treatedfor 4 weeks. In this Class I prospective, randomized trial, theyfound that the heparin therapy alone group had an incidenceof DVT similar to that of the placebo group. The combinationof low-dose heparin and electrical stimulation significantlydecreased the incidence of DVT (1 of 15 patients comparedwith the other two treatment groups [8 of 16 low-dose heparinalone, and 8 of 17 placebo; P � 0.05]). In 1992, Merli et al. (25)reported that heparin in combination with pneumatic stock-ings was equal to the effectiveness of heparin plus electricalstimulation. The heparin in combination with electrical stim-ulation group and the placebo group for this comparison werea historical cohort. This constitutes Class III evidence becausethe comparison group is a historical one. Winemiller et al. (40)studied a large series of 428 SCI patients with a multivariateanalysis and found that the use of pneumatic compressiondevices for 6 weeks was associated with a significant decreasein thromboembolic events (odds ratio, 0.5; 95% confidenceinterval, 0.28–0.90). Low-dose heparin treatment seemed tohave a protective effect as well; however, the effect of heparinalone was not statistically significant.
Recently, low-molecular-weight heparins (LMWHs) havebeen studied as prophylactic therapy for thromboembolism inSCI patients. Green et al. (12) treated a series of SCI patientswith 8 weeks of LMWH (tinzaparin) and compared the resultswith those in a historical cohort of patients treated with low-dose or adjusted dose heparin. They found that the use ofLMWHs compared favorably with the use of either heparindosing regimen in terms of fewer thromboembolic events (16of 79 in heparin group versus 7 of 68 in LMWH group; P �0.15) and a significant decrease in bleeding complications (9 of79 in heparin group versus 1 of 68 in LMWH group; P � 0.04).More recently, Harris et al. (18) performed a retrospectivestudy of LMWH (enoxaparin) administration in a series of 105patients with spinal injuries. Forty of their 105 patients hadneurologically complete injuries. No patient treated with LM-WHs exhibited clinical or ultrasound evidence of DVT, and nopatient had a PE. Roussi et al. (32) reported a 9% incidence ofDVT in a study involving 69 SCI patients receiving LMWHs,testimony to the fact that no prophylactic therapy is 100%effective.
The use of inferior vena cava (IVC) filters as prophylacticdevices for thromboembolism has been advocated (19, 20, 31,39). Wilson et al. (39) placed caval filters in 15 SCI patientswho were concurrently treated with either low-dose heparinor pneumatic stockings. None had a PE during a 1-yearfollow-up period. The 1-year patency rate of the IVC was 81%.These authors reported that their results are superior to thosein a historical cohort of 111 patients treated without IVCfilters. Seven of the cohort patients had a PE; however, six ofthe seven were not receiving any prophylaxis at the time oftheir PE. The patient who was receiving DVT prophylaxis hada gunshot wound to the spine (39). Khansarinia et al. (20)described a historical cohort study of 108 general traumapatients treated with prophylactic IVC filter placement. Noneof these patients had a PE. They compared this group with a
historical cohort of 216 patients treated (apparently) witheither low-dose heparin or pneumatic compression devicesbefore the use of IVC filters. Thirteen of these 216 had a PE, ofwhich 9 were fatal. The overall mortality of the filter groupwas lower than that of the control group, but this differencewas not significant (16% versus 22%). Tola et al. (35) haveshown that percutaneous IVC filter placement in the ICUsetting is as safe and is less costly than IVC filter placement inthe operating room or the invasive radiology suite. Theseauthors suggest that IVC interruption is an effective means toprevent PE. Placement of filters is not without complications.Balshi et al. (1), Greenfield (15), and Kinney et al. (21) havedescribed distal migration, intraperitoneal erosion, and symp-tomatic IVC occlusion in patients with SCIs treated with IVCfilters. Balshi et al. (1) hypothesized that quadriplegic patientsare at higher risk for complications from IVC filter placementowing to loss of abdominal muscle tone, as well as the use of the“quad cough” maneuver. No study performed to date comparesthe use of prophylactic IVC filters to the use of modern methodsof pharmacological PE and DVT prophylaxis.
Duration of prophylaxis
Most thromboembolic events seem to occur within the first2 to 3 months after injury. Naso (28) described his experiencewith four patients with PEs in a group of 43 SCI patients. Allfour PE events occurred within 3 months injury. Perkash et al.(29) reported an 18% incidence of thromboembolism in aseries of 48 patients with acute SCIs and 2 patients withtransverse myelitis. Only one patient had a new onset PE 3months after injury; 2 other patients had recurrent PEs3 months after injury owing to existing DVT. Lamb et al. (23)described a series of 287 SCI patients. The overall risk ofthromboembolic events in their patient population was 10%.Most events occurred within the first 6 months after injury.Twenty-two of 28 events they documented occurred withinthe first 3 months of injury. El Masri and Silver (8) reported 21documented events of PE in a series of 102 spine-injuredpatients. Twenty of 21 events occurred within the first 3months after SCI. A pulmonary embolism occurred in a pa-tient with a history of PE whose therapeutic anticoagulationwas discontinued for gallbladder surgery. DeVivo et al. (7)found that patients had a 500-fold higher risk of dying fromPE in the first month after SCI than age- and sex-matchednoninjured patients. This risk decreased with time but re-mained approximately 20 times higher than for normativecontrols 6 months after injury. McKinley et al. (24) studiedchronic spine-injured patients in a rehabilitation center settingand found an incidence of DVT of 2.1% in the first year afterinjury. This incidence decreased to between 0.5 and 1% peryear thereafter. From these data, it is apparent that mostthromboembolic events (DVT and PE) occur within 3 monthsof acute spinal injury. Although late thromboembolic eventscan occur, the risk of these events must be balanced againstthe cost and risks of indefinite anticoagulation. Prolongedprophylactic anticoagulation therapy is not without risk andis associated with bleeding complications (12, 13). Most stud-ies addressing prophylactic treatment for DVT and PE have
Deep Venous Thrombosis and Thromboembolism S77
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
used treatment courses of 8 to 12 weeks with success. Forthese reasons, it is recommended that prophylactic treatmentbe continued for 8 to 12 weeks, in SCI patients without othermajor risk factors (previous thromboembolic events, obesity,advanced age), for DVT and PE. Prophylactic treatment maybe discontinued earlier in patients with useful motor functionin the lower extremities, as these patients seem to be at lessrisk for DVT and PE (6, 27).
Diagnosis
The diagnosis of DVT in various studies has been made onthe basis of clinical criteria, Doppler ultrasound examination,VOP, venography, fibrinogen scanning, or D-dimer measure-ment (2–4, 8–11, 13, 14, 16, 17, 22, 26–30, 32, 34, 36–38, 41).Although venography may be considered a “gold standard”examination for DVT, venography is not possible in all pa-tients, is invasive, and is expensive (11). Gunduz et al. (16)reported a 10% incidence of adverse effects from venographyincluding postvenographic phlebitis and allergic reactions.Doppler ultrasound examination and VOP are both less inva-sive, less expensive, and more broadly applicable (11, 30). Thesensitivity and specificity of these examinations when com-pared with venography has been generally reported to rangefrom 80 to 100% (5). Chu et al. (5) compared Doppler ultra-sound and VOP with the clinical examination and found allthree to agree 100% of the time in a small series of 21 patients.Perkash et al. (29) studied a series of 48 SCI patients with dailyphysical examinations and weekly VOP. They found that thesensitivity of the clinical examination compared with VOPwas 89%; the specificity was 88%, the negative predictivevalue was 97%, and the positive predictive value was 62%.Other authors have described the increased sensitivity of fi-brinogen scanning and the use of D-dimer measurements forthe diagnosis of DVT. Increased sensitivity is associated withdecreased specificity. For example, Roussi et al. (32) reported100% sensitivity and 100% negative predictive value withD-dimer determinations compared with Doppler ultrasoundand the clinical examination. However, the specificity ofD-dimer was only 34%, and the positive predictive value wasonly 13%. Similarly, Todd et al. (34) found that fibrinogenscanning was positive in all 20 patients studied in a prospec-tive fashion, but the diagnosis of DVT was confirmed byanother test in only half of the cases. Overall, no single test iscompletely applicable, accurate, and sensitive for the detec-tion of DVT in the SCI patient population. Furthermore, asubstantial number of patients who have PEs are found tohave negative lower extremity venograms (8, 11). The Con-sortium for Spinal Cord Medicine (6) has recommended theuse of ultrasound for the study of patients with suspectedDVT and venography for use when clinical suspicion is strongand the ultrasound examination is negative. On the basis ofthe medical evidence available, these recommendations seemto be sound.
SUMMARY
Thromboembolic disease is a common occurrence in pa-tients who have sustained a cervical SCI, and it is associated
with significant morbidity. Class I medical evidence existsdemonstrating the efficacy of several means of prophylaxisfor the prevention of thromboembolic events. Therefore, pa-tients with SCI should be treated with a regimen aimed atprophylaxis. Although low-dose heparin therapy has beenreported to be effective as prophylaxis for thromboembolismin several Class III studies, other Class I, II, and III medicalevidence indicates that better alternatives than low-dose hep-arin therapy exist. These alternatives include the use ofLMWH, adjusted dose heparin, or anticoagulation in conjunc-tion with pneumatic compression devices or electrical stimu-lation. Oral anticoagulation alone does not seem to be aseffective as these other measures used for prophylaxis.
The incidence of thromboembolic events seems to decreaseover time, and the prolonged use of anticoagulant therapy isassociated with a definite incidence of bleeding complica-tions. There are many reports of the beneficial effects of theprophylaxis therapy for 6 to 12 weeks after SCI. Very fewthromboembolic events occur beyond 3 months after injury.For these reasons, it is recommended that prophylactic ther-apy be discontinued after 3 months unless the patient is athigh risk (previous thromboembolic events, obesity, ad-vanced age). It is reasonable to discontinue therapy earlier inpatients with retained lower extremity motor function afterSCI, because the incidence of thromboembolic events in thesepatients is substantially lower than in those patients withmotor complete injuries.
Caval filters seem to be efficacious for the prevention of PEin SCI patients. The relative efficacy of caval filters versusprophylactic combination therapy with LMWH and pneu-matic compression stockings has not been studied. Cavalfilters are associated with long-term complications in SCIpatients, although these complications are relatively rare. Ca-val filters are recommended for SCI patients who have hadthromboembolic events despite anticoagulation and for SCIpatients with contraindications to anticoagulation and/or theuse of pneumatic compression devices.
There are several methods available for the diagnosis ofDVT. Venography is considered the “gold standard,” but it isinvasive, not applicable to all patients, and associated withintrinsic morbidity. Duplex Doppler ultrasound and VOPhave been reported to have sensitivities of approximately 90%and are not invasive. It is reasonable to use these noninvasivetests for the diagnosis of DVT and to reserve venography forthe rare situation when clinical suspicion is high and theresults of VOP and ultrasound testing are negative.
KEY ISSUES FOR FUTURE INVESTIGATION
Although thromboembolic events in the SCI patient areassociated with significant morbidity, no study has demon-strated improved outcomes in SCI patients as a result ofsurveillance testing for them. A prospective study evaluating6-month outcomes in patients treated with prophylaxis withor without surveillance ultrasound imaging would be a sub-stantial and potentially cost-saving contribution to theliterature.
S78 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Caval filters seem to be effective in preventing PEs, andmany institutions are using these devices as first-tier preven-tive therapy without trying other preventive measures. Cavalfilters have not been compared with LMWHs or combinationtherapy with anticoagulants and pneumatic compression de-vices for efficacy in the SCI patient population. As filters doseem to be associated with long-term morbidity in a fractionof SCI patients, a prospective study needs to be performed toestablish whether the potential increase in protection againstPE offsets the risks for long-term complications. A studycomparing the use of vena caval filters prophylactically ver-sus other modes of prevention with the use of filters placedonly after failure of alternative methods should be instituted,including cost-effectiveness outcomes of each mode of pre-vention currently used in SCI patients.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 615 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Balshi JD, Cantelmo NL, Menzoian JO: Complications of cavalinterruption by Greenfield filter in quadriplegics. J Vasc Surg9:558–562, 1989.
2. Becker DM, Gonzalez M, Gentili A, Eismont F, Green BA: Pre-vention of deep venous thrombosis in patients with acute spinalcord injuries: Use of rotating treatment tables. Neurosurgery20:675–677, 1987.
3. Burns GA, Cohn SM, Frumento RJ, Degutis LC, Hammers L:Prospective ultrasound evaluation of venous thrombosis inhigh-risk trauma patients. J Trauma 35:405–408, 1993.
4. Casas ER, Sanchez MP, Arias CR, Masip JP: Prophylaxis ofvenous thrombosis and pulmonary embolism in patients withacute traumatic spinal cord lesions. Paraplegia 15:209–214, 1977.
4a. Chen D, Apple DF, Hudson LM, Bode R: Medical complicationsduring acute rehabilitation following spinal cord injury: Currentexperience of the Model Systems. Arch Phys Med Rehabil 80:1397–1401, 1999.
5. Chu DA, Ahn JH, Ragnarsson KT, Helt J, Folcarelli P, Ramirez A:Deep venous thrombosis: Diagnosis in spinal cord injured pa-tients. Arch Phys Med Rehabil 66:365–368, 1985.
6. Consortium for Spinal Cord Medicine: Prevention of thromboem-bolism in spinal cord injury. Spinal Cord Med 20:259–283, 1997.
7. DeVivo MJ, Kartus PL, Stover SL, Rutt RD, Fine PR: Cause ofdeath for patients with spinal cord injuries. Arch Intern Med149:1761–1766, 1989.
8. El Masri WS, Silver JR: Prophylactic anticoagulant therapy inpatients with spinal cord injury. Paraplegia 19:334–342, 1981.
9. Frisbie JH, Sasahara AA: Low dose heparin prophylaxis for deepvenous thrombosis in acute spinal cord injury patients: A con-trolled study. Paraplegia 19:343–346, 1981.
10. Frisbie JH, Sharma GV: Pulmonary embolism manifesting asacute disturbances of behavior in patients with spinal cord in-jury. Paraplegia 32:570–572, 1994.
11. Geerts WH, Code KI, Jay RM, Chen E, Szalai JP: A prospectivestudy of venous thromboembolism after major trauma. N EnglJ Med 331:1601–1606, 1994.
12. Green D, Chen D, Chmiel JS, Olsen NK, Berkowitz M, Novick A,Alleva J, Steinberg D, Nussbaum S, Tolotta M: Prevention ofthromboembolism in spinal cord injury: Role of low molecularweight heparin. Arch Phys Med Rehabil 75:290–292, 1994.
13. Green D, Lee MY, Ito VY, Cohn T, Press J, Filbrandt PR,VandenBerg WC, Yarkony GM, Meyer PR Jr: Fixed- vs adjusted-dose heparin in the prophylaxis of thromboembolism in spinalcord injury. JAMA 260:1255–1258, 1988.
14. Green D, Lee MY, Lim AC, Chmiel JS, Vetter M, Pang T, Chen D,Fenton L, Yarkony GM, Meyer PR Jr: Prevention of thromboem-bolism after spinal cord injury using low-molecular-weight hep-arin. Ann Intern Med 113:571–574, 1990.
15. Greenfield LJ: Does cervical spinal cord injury induce a higherincidence of complications after prophylactic Greenfield filterusage? J Vasc Interv Radiol 8:719–720, 1997.
16. Gunduz S, Ogur E, Mohur H, Somuncu I, Acjksoz E, Ustunsoz B:Deep vein thrombosis in spinal cord injured patients. Paraplegia31:606–610, 1993.
17. Hachen HJ: Anticoagulant therapy in patients with spinal cordinjury. Paraplegia 12:176–187, 1974.
18. Harris S, Chen D, Green D: Enoxaparin for thromboembolismprophylaxis in spinal injury. Am J Phys Med Rehabil 75:326–327, 1996.
19. Jarrell BE, Posuniak E, Roberts J, Osterholm J, Cotler J,Ditunno J: A new method of management using the Kim-RayGreenfield filter for deep venous thrombosis and pulmonaryembolism in spinal cord injury. Surg Gynecol Obstet 157:316–320, 1983.
20. Khansarinia S, Dennis JW, Veldenz HC, Butcher JL, Hartland L:Prophylactic Greenfield filter placement in selected high-risktrauma patients. J Vasc Surg 22:231–235, 1995.
21. Kinney TB, Rose SC, Valji K, Oglevie SB, Roberts AC: Doescervical spinal cord injury induce a higher incidence of compli-cations after prophylactic Greenfield inferior vena cava filterusage? J Vasc Interv Radiol 7:907–915, 1996.
22. Kulkarni JR, Burt AA, Tromans AT, Constable PD: Prophylac-tic low dose heparin anticoagulant therapy in patients withspinal cord injuries: A retrospective study. Paraplegia 30:169–172, 1992.
23. Lamb GC, Tomski MA, Kaufman J, Maiman DJ: Is chronic spinalcord injury associated with increased risk of venous thrombo-embolism? J Am Paraplegia Soc 16:153–156, 1993.
24. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ: Long-term medical complications after traumatic spinal cord injury: Aregional model systems analysis. Arch Phys Med Rehabil 80:1402–1410, 1999.
25. Merli GJ, Crabbe S, Doyle L, Ditunno JF, Herbison GJ: Mechan-ical plus pharmacological prophylaxis for deep vein thrombosisin acute spinal cord injury. Paraplegia 30:558–562, 1992.
26. Merli GJ, Herbison GJ, Ditunno JF, Weitz HH, Henzes JH, ParkCH, Jaweed MM: Deep vein thrombosis: Prophylaxis in acutespinal cord injured patients. Arch Phys Med Rehabil 69:661–665, 1988.
27. Myllynen P, Kammonen M, Rokkanen P, Bostman O, Lalla M,Laasonen E: Deep venous thrombosis and pulmonary embolismin patients with acute spinal cord injury: A comparison withnonparalyzed patients immobilized due to spinal fractures.J Trauma 25:541–543, 1985.
28. Naso F: Pulmonary embolism in acute spinal cord injury. ArchPhys Med Rehabil 55:275–278, 1974.
28a. Perkash A: Experience with the management of deep vein throm-bosis in patients with spinal cord injury: Part II—A critical evalu-ation of the anticoagulant therapy. Paraplegia 18:2–14, 1980.
29. Perkash A, Prakash V, Perkash I: Experience with the manage-ment of thromboembolism in patients with spinal cord injury:Part I—Incidence, diagnosis and role of some risk factors. Para-plegia 16:322–331, 1978.
Deep Venous Thrombosis and Thromboembolism S79
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
30. Powell M, Kirshblum S, O’Connor KC: Duplex ultrasoundscreening for deep vein thrombosis in spinal cord injured pa-tients at rehabilitation admission. Arch Phys Med Rehabil 80:1044–1046, 1999.
31. Quirke TE, Ritota PC, Swan KG: Inferior vena caval filter use inU.S. trauma centers: A practitioner survey. J Trauma 43:333–337,1997.
32. Roussi J, Bentolila S, Boudaoud L, Casadevall N, Vallee C,Carlier R, Lortat-Jacob S, Dizien O, Bussel B: Contribution ofD-dimer determination in the exclusion of deep venous throm-bosis in spinal cord injury patients. Spinal Cord 37:548–552,1999.
33. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF: Com-parison of surgical and conservative management in 208 patientswith acute spinal cord injury. Can J Neurol Sci 14:60–69, 1987.
34. Todd JW, Frisbie JH, Rossier AB, Adams DF, Als AV, ArmeniaRJ, Sasahara AA, Tow DE: Deep venous thrombosis in acutespinal cord injury: A comparison of 125I fibrinogen leg scanning,impedance plethysmography and venography. Paraplegia 14:50–57, 1976.
35. Tola JC, Holtzman R, Lottenberg L: Bedside placement of infe-rior vena cava filters in the intensive care unit. Am Surg 65:833–837, 1999.
35a. Tomaio A, Kirshblum SC, O’Connor KC, Johnston M: Treatmentof acute deep vein thrombosis in spinal cord injured patientswith enoxaparin: A cost analysis. J Spinal Cord Med 21:205–210,1998.
36. Waring WP, Karunas RS: Acute spinal cord injuries and theincidence of clinically occurring thromboembolic disease. Para-plegia 29:8–16, 1991.
37. Watson N: Venous thrombosis and pulmonary embolism inspinal cord injury. Paraplegia 6:113–121, 1968.
38. Watson N: Anti-coagulant therapy in the prevention of venousthrombosis and pulmonary embolism in the spinal cord injury.Paraplegia 16:265–269, 1978.
38a. Weingarden SI, Weingarden DS, Belen J: Fever and thromboembolicdisease in acute spinal cord injury. Paraplegia 26:35–42, 1988.
39. Wilson JT, Rogers FB, Wald SL, Shackford SR, Ricci MA: Prophy-lactic vena cava filter insertion in patients with traumatic spinalcord injury: Preliminary results. Neurosurgery 35:234–239, 1994.
40. Winemiller MH, Stolp-Smith KA, Silverstein MD, Therneau TM:Prevention of venous thromboembolism in patients with spinalcord injury: Effects of sequential pneumatic compression andheparin. J Spinal Cord Med 22:182–191, 1999.
41. Yelnik A, Dizien O, Bussel B, Schouman-Claeys E, Frija G, Pan-nier S, Held JP: Systematic lower limb phlebography in acutespinal cord injury in 147 patients. Paraplegia 29:253–260, 1991.
Drawings by Leonardo da Vinci, revealing how thecervical musculature stabilizes the cervical spinal col-umn. Leonardo’s investigations were some of the ear-liest into the relationship of the spinal anatomy andits mechanics. Courtesy, Dr. Edwin Todd, Pasadena,California.
S80 Guidelines for Management of Acute Cervical Spinal Injuries
CHAPTER 11
Nutritional Support after Spinal Cord Injury
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Nutritional support of patients with spinal cord injuries is recommended. Energy expenditure is best
determined by indirect calorimetry in these patients because equation estimates of energy expenditure andsubsequent caloric need tend to be inaccurate.
RATIONALE
Hypermetabolism, an accelerated catabolic rate, andrampant nitrogen losses are consistent sequelae tomajor trauma, particularly acute traumatic brain in-
jury and acute spinal cord injury (ASCI) (7, 9–11, 13, 18, 19,22). A well-documented hypermetabolic, catabolic injury cas-cade is initiated immediately after central nervous systeminjury that results in depletion of whole-body energy stores,loss of lean muscle mass, reduced protein synthesis, and,ultimately, loss of gastrointestinal mucosal integrity and com-promise of immune competence (6, 9, 10, 12, 18, 19, 22).Severely brain-injured and spinal cord injury (SCI) patients,therefore, are at risk for prolonged nitrogen losses and ad-vanced malnutrition within 2 to 3 weeks after injury withresultant increased susceptibility for infection, impairedwound healing, and difficulty weaning from mechanical ven-tilation (5, 10, 13, 18, 19, 22). These factors, added to theinherent immobility, denervation, and muscle atrophy asso-ciated with SCI, provide the rationale for nutritional supportof SCI patients after trauma.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The following medical subject headings wereused in combination with “spinal cord injury”: “nutrition”and “nutritional support.” Approximately 105 citations wereacquired. Non-English language citations were deleted, andtitles and abstracts of the remaining publications were re-viewed. We focused on the specific issue of nutrition andhuman patients with ASCI and identified 18 articles. Relevantmanuscripts and reviews describing nutritional support ofhead-injured patients and several reports describing the nu-tritional status of chronic SCI patients are included among the23 citations in the reference list. These efforts identified oneClass II study and four Class III studies that describe metab-olism, nitrogen wasting, and the effect of feeding on nitrogenbalance and serum biochemistries in patients after ASCI.
These articles are summarized in Table 11.1. There were nostudies that examined the effects of nutritional support onoutcome after ASCI.
SCIENTIFIC FOUNDATION
Hypermetabolism, catabolism, and accelerated nitrogenlosses are well-recognized complications that follow trau-matic injury (10, 11, 19, 22). They have been identified andstudied in human patients who have sustained traumaticbrain injuries and SCI. A number of publications have de-scribed the increased energy requirements and nitrogen lossesof patients after acute head injury (7, 8, 10–12, 17, 18, 22, 23).Fewer studies have focused on hypermetabolism, catabolism,and nitrogen losses after ASCI (6, 13, 14, 20, 22). Althoughthere are metabolic similarities between isolated traumaticbrain injury and severe isolated SCI, it seems that there maybe important biological differences between the two centralnervous system injury types that have a bearing on supple-mental nutritional therapy (6, 13, 14, 18–20, 22).
Severe head injury is associated with a resting energy ex-penditure (REE) of approximately 140% of predicted normalbasal energy expenditure (BEE) (8, 10, 11, 17, 18, 22, 23).Indirect calorimetry is the most widely used reliable means todetermine individual energy requirements in hospitalized pa-tients after traumatic injury (10, 18, 19, 22). It requires the useof a portable metabolic cart and uses a technique that mea-sures respiratory gas exchange and the rate of oxygen use ina given patient. It provides an estimate of energy expenditureby the patient by determining the known caloric yield from 1L of oxygen based on differences in oxygen consumption andcarbon dioxide production. It is performed at the bedside inthe intensive care unit in severely injured patients. Metabolicexpenditure is expressed as a percentage of normal BEE at rest(predicted). Indirect calorimetry is typically performed oncedaily for the first several days after injury and periodicallythereafter (10, 18, 19, 22). The Harris-Benedict equation, withactivity and stress of injury variables, has been shown topredict energy expenditure after a traumatic brain injury with
S81Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
reasonable accuracy without indirect calorimetry (10, 14, 19,20, 22).
Nutritional support of head-injured patients is typicallybegun within days of admission and is guided by the meta-bolic information provided by indirect calorimetry and bypredicted energy expenditure values derived by equation.Hypermetabolism, accelerated catabolism, and excessive ni-trogen losses continue for at least 2 weeks after injury (8–11,18, 19, 22). The exact duration of this response to injury isunknown, may vary among similar patients, and can be af-fected by other traumatic injuries, pancreatitis, infection, orsepsis (2, 10, 18, 19, 22). Nutritional support in this setting isdesigned to provide nitrogen-rich, high-energy supplementalfuel to blunt excess catabolism and preserve energy stores,muscle mass, gastrointestinal integrity, and immune compe-tence (6, 10, 18, 19, 22). Nitrogen balance is difficult—oftenimpossible—to achieve, particularly within the first week ofinjury (7, 11, 13, 14, 20). Matching nutritional replacementwith caloric needs, therefore, has become the primary goal ofnutritional therapy.
The extent of neuronal connectivity and the neurogenicstimuli (muscle tone) to the musculoskeletal system seemsimportant to the level of metabolic expenditure after injury ofthe central nervous system (1, 3, 4, 13–16, 18–21, 22). Agitated,combative head-injured patients, for example, can have REElevels as high as 200% of expected BEE levels (10, 11, 18, 22).Conversely, pharmacological paralysis of head-injured pa-tients has been associated with reductions in resting energyexpenditure by 20 to 30% (10, 11, 18, 22). Patients who havesustained isolated ASCIs often have increased metabolic ex-penditure compared with normative energy expenditure lev-els (13, 14, 18–20, 22). However, because of the paralysis and
flaccidity associated with ASCI, measured REE values in thesepatients are considerably lower than those predicted by theHarris-Benedict equation based on age, sex, body surfacearea, activity, and injury severity (14, 19, 20, 23). Patients withthe greatest neurological deficits and the least muscle toneafter SCI (high-cervical level quadriplegic patients) havelower measured REE values than those found in patients withincomplete spinal injuries or lower spinal cord injuries (tho-racic level paraplegic patients) (13, 14, 19, 20, 22). Kaufman etal. (13), in 1985, described their experience with eight ASCIpatients managed at the University of Texas. They notedaccelerated nitrogen losses and ongoing negative nitrogenbalance greater than expected. Differences in initial andfollow-up nutritional assessments revealed deterioration innutritional status during the 2-week period of observation,partly caused by inadequate supply of protein and calories.Infective complications and prolonged respiratory supportwere common. The authors concluded that muscle atrophymight play an important role in the accelerated nitrogenlosses they identified in patients with paralysis due to com-plete SCI and that improved nutritional support might reducemedical complications after ASCI.
Young et al. (22) reported four quadriplegic ASCI patientsassessed with indirect calorimetry. They found that indirectcalorimetry provided more accurate REE values for their pa-tients than Harris-Benedict equation estimates, even if theequation estimates did not incorporate injury and activityfactors. The authors also noted marked daily nitrogen lossesand negative nitrogen balance. They concluded that equationestimates of REE values of SCI patients overestimate meta-bolic expenditure, and they emphasized the importance of
TABLE 11.1. Summary of Reports on Nutritional Support after Spinal Cord Injurya
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Cruse et al., 2000 (5) Comparison of nutritional,immune, endocrine status in 15ASCI patients versus 16 matchedcontrols.
II SCI patients have hormonal changes, poornutritional status, and decreased immunefunction compared with controls.
Rodriguez et al., 1997 (20) Prospective assessment andtreatment of 12 ASCI patients.
III REE less than predicted, marked“obligatory” nitrogen losses due toflaccidity and atrophy of denervatedmuscle after SCI.
Kearns et al., 1992 (14) Prospective assessment of 10 ASCIpatients over 4-wk period ofobservation.
III Exaggerated nitrogen and 3-methylhistidineexcretion, marked weight loss. Lower REEthan predicted after SCI.
Young et al., 1987 (22) 4 ASCI patients assessed viaindirect calorimetry.
III Indirect calorimetry best means todetermine energy expenditure after ASCI.
Kaufman et al., 1985 (13) Assessment of nutritional status of8 SCI patients over 2-wk period ofobservation.
III Deterioration in nutritional status despiteattempted treatment. Marked nitrogenlosses. Increased infectious and respiratorycomplications.
a ASCI, acute spinal cord injury; SCI, spinal cord injury; REE, resting energy expenditure.
S82 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
indirect calorimetry in predicting energy expenditure afterASCI.
Kearns et al. (14) prospectively assessed and provided nu-tritional support to 10 ASCI patients they managed and mon-itored for 4 weeks. Their 1992 report documents the use ofindirect calorimetry to determine REE and provide matchedcaloric supplementation. All patients had isolated SCIs with-out associated head injury or other organ system trauma.Initial measured REE values were 10% below predicted BEElevels. All patients experienced exaggerated nitrogen and3-methylhistidine losses, indicating excessive lean body massand muscle loss. A 10% decrease in body weight accompaniedthese losses despite caloric replacement matched to or exceed-ing measured REE values for each patient. The specifics ofnutrition administration (mix and route of delivery) were notpresented. The authors noted an increase in REE over time inpart attributable to reductions in body weight and in part dueto return of muscle tone. The authors concluded that acuteisolated SCI is associated with lower REE values comparedwith predicted values. ASCI patients have exaggerated nitro-gen and 3-methylhistidine losses owing to atrophy of dener-vated muscle. The authors attributed the reduced metabolicactivity seen in these patients to the flaccidity of denervatedmusculature after severe SCI and noted that as muscle lossand weight reductions progress, REE increases, particularly ifrecovery of motor function and/or return of muscle toneoccurs.
Rodriguez et al. (20) studied the metabolic response to SCIin 12 acute trauma patients. Assessment and nutritional sup-port were instituted immediately after injury and continuedfor 4 weeks after injury. Harris-Benedict estimations of energyexpenditure were compared with values obtained from indi-rect calorimetry in each patient. All patients had acceleratednitrogen losses and negative nitrogen balance. Eleven of the12 patients had negative nitrogen balance for the entire 4weeks of therapy despite matched caloric replacement. Thesingle patient in whom nitrogen balance was realized had anincomplete SCI. The Harris-Benedict equation with activityfactor of 1.2 and a stress/injury factor of 1.6 consistentlyoverestimated energy expenditure in these 12 patients andwould have resulted in excessive feeding. The authors con-cluded that large nitrogen losses after severe SCI are “oblig-atory” as a result of atrophy and wasting of denervatedmusculature below the level of injury. Patients with completetraumatic myelopathy had greater obligatory nitrogen lossesthan patients with incomplete SCI. They recommended thatindirect calorimetry be used as the energy expenditure assess-ment method after SCI, particularly in the early postinjuryperiod. If the Harris-Benedict equation is used in these pa-tients in this setting, they recommend that the activity factorshould be eliminated and the stress/injury factor of the equa-tion should be reduced.
Cruse et al. (5) examined the neurological, immune, endo-crine, and nutritional status of 15 male SCI patients and com-pared them with 16 healthy age-matched control subjects. Thetiming of assessment in relation to SCI for each patient was notspecified. Their report described decreased natural and adaptiveimmune responses in the SCI patient population beginning
within 2 weeks of injury that reached a nadir 3 months afterinjury. They noted increased adrenocorticotropic hormone andplasma cortisol levels, decreased zinc, albumin, and prealbuminserum levels, surface marker changes in both lymphocytes andgranulocytes, and decreased adhesion molecule-binding abilityafter SCI compared with findings in healthy control patients.They concluded that patients with severe ASCI have decreasedimmune function, impaired nutritional status, and a decreasednumber of adhesion molecules, all of which occurs within weeksafter acute injury. The authors note that these hormonal alter-ations, nutritional deficiencies, and changes in immune functionmay increase susceptibility to infection and may contribute todelayed wound healing.
The change in energy expenditure identified in patientsafter ASCI seems to persist long after the initial injury andrecovery phase (1, 3, 4, 15, 16, 19, 21). Several investigatorshave noted long-standing reductions in REE in SCI patients,reductions that correlate with the degree of neurological in-jury and the extent of lean body mass loss after paralysis (1, 3,4, 15, 16, 19, 21). Cox et al. (4) measured energy expenditure instable non-ASCI patients in the rehabilitation setting. Theyreported that quadriplegic patients required 22.7 kcal/kg/dcompared with 27.9 kcal/kg/d for paraplegic patients theystudied. Most investigators conclude that equation methodsto estimate energy expenditure in SCI patients are inaccurate,in both the acute and chronic settings (15, 16, 20–22).
There has been no report assessing the efficacy of the route offeeding (parenteral or enteral) for SCI patients in the acute set-ting. The literature on nutritional support for head injury pa-tients supports using the enteral route for nutritional supple-mentation if the gut is functional (8, 10, 11, 18–20, 22). Thisgeneral policy seems to have been followed by investigators ofnutritional support for ASCI patients (13, 14, 20, 22). The poten-tial benefits of enteral feeding over parenteral delivery includemaintenance of gut integrity and function, reduced expense,lower risk of infection, and avoidance of complications related tothe intravenous catheter (8, 10, 11, 18, 19, 22). Nasoduodenal ornasojejunal feeding tubes usually allow full caloric, high-nitrogen, high-volume feeding within days of injury. In patientswith bowel injury, mechanical bowel obstruction, or prolongedileus, it is recommended that parenteral nutrition be initiateduntil the bowel recovers and conversion to enteral nutrition canbe accomplished (8, 10, 18, 19, 22).
There has been no report assessing the mix or compositionof nutritional supplementation for SCI patients. The literatureon nutritional support for head injury patients suggests be-ginning with a high-nitrogen enteral or parenteral solutioncontaining at least 15% of calories as protein, no more than15% glucose/dextrose, a minimum of 4% of total energyneeds as essential fatty acids, and the addition of vitamins,essential elements, and trace minerals (10, 12, 17, 18, 19, 22,23). No study has been published that has examined the effectof nutritional support on neurological outcome after ASCI.
SUMMARY
Alterations in metabolism occur after ASCI, but the markedhypermetabolic response seen after acute traumatic brain in-
Nutritional Support after Spinal Cord Injury S83
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
jury seems to be blunted in SCI patients by the flaccidity ofdenervated musculature after spinal cord transection or in-jury. As a result, REE is lower than predicted after ASCI.Equation estimates of REE in these patients have proven to beinaccurate; therefore, indirect calorimetry is the recom-mended technique to assess energy expenditure in both theacute and chronic settings. Protein catabolism does occur afteracute, severe SCI, and marked losses in lean body mass due tomuscle atrophy result in huge nitrogen losses, prolongednegative nitrogen balance, and rapid weight loss. Nutritionalsupport of the SCI patient to meet caloric and nitrogen needs,not to achieve nitrogen balance, is safe and may reduce thedeleterious effects of the catabolic, nitrogen-wasting processthat occurs after acute SCI.
KEY ISSUES FOR FUTURE INVESTIGATION:
An assessment of the timing, route of administration, andcomposition of nutritional therapy on outcome, both neuro-logical and medical, should be performed. This could be bestaccomplished with a multicenter case-control study.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 615 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Barboriak JJ, Rooney CB, El Ghatit AZ, Spuda K, Anderson AJ:Nutrition in spinal cord injury patients. J Am Paraplegia Soc6:32–36, 1983.
2. Carey ME, Nance FC, Kirgis HD, Young HF, Megison LC Jr, KlineDG: Pancreatitis following spinal cord injury. J Neurosurg 47:917–922, 1997.
3. Clarke KS: Caloric costs of activity in paraplegic persons. ArchPhys Med Rehabil 47:427–435, 1966.
4. Cox SA, Weiss SM, Posuniak EA, Worthington P, Prioleau M,Heffley G: Energy expenditure after spinal cord injury: An eval-uation of stable rehabilitating patients. J Trauma 25:419–423, 1985.
5. Cruse JM, Lewis RE, Dilioglou S, Roe DL, Wallace WF, Chen RS:Review of immune function, healing of pressure ulcers, and nu-tritional status in patients with spinal cord injury. J Spinal CordMed 23:129–135, 2000.
6. Cruse JM, Lewis RE, Roe DL, Dilioglou S, Blaine MC, Wallace WF,Chen RS: Facilitation of immune function, healing of pressureulcers, and nutritional status in spinal cord injury patients. ExpMol Pathol 68:38–54, 2000.
7. Frankenfield DC, Smith JS, Cooney RN: Accelerated nitrogen lossafter traumatic injury is not attenuated by achievement of energybalance. JPEN J Parenter Enteral Nutr 21:324–329, 1997.
8. Grahm TW, Zadrozny DB, Harrington T: The benefits of earlyjejunal hyperalimentation in the head injured patient. Neuro-surgery 25:729–735, 1989.
9. Hadley MN: Hypermetabolism after CNS trauma: Arresting the“injury cascade.” Nutrition 5:143, 1989.
10. Hadley MN: Hypermetabolism following head trauma: Nutri-tional considerations, in Barrow DL (ed): Complications and Se-quelae of Head Injury (Neurosurgical Topics series). Park Ridge,AANS, 1992, pp 161–168.
11. Hadley MN, Grahm TW, Harrington T, Schiller WR, McDermottMK, Posillico DB: Nutritional support and neurotrauma: A criti-cal review of early nutrition in forty-five acute head injury pa-tients. Neurosurgery 19:367–373, 1986.
12. Hulsewe KW, van Acker BA, von Meyenfeldt MF, Soeters PB:Nutritional depletion and dietary manipulation: Effects on theimmune response. World J Surg 23:536–544, 1999.
13. Kaufman HH, Rowlands BJ, Stein DK, Kopaniky DR, GildenbergPL: General metabolism in patients with acute paraplegia andquadriplegia. Neurosurgery 16:309–313, 1985.
14. Kearns PJ, Thompson JD, Werner PC, Pipp TL, Wilmot CB: Nu-tritional and metabolic response to acute spinal cord injury. JPENJ Parenter Enteral Nutr 16:11–15, 1992.
15. Levine AM, Nash MS, Green BA, Shea JD, Aronica MJ: An exam-ination of dietary intakes and nutritional status of chronic healthyspinal cord injured individuals. Paraplegia 30:880–889, 1992.
16. Peiffer SC, Blust P, Leyson JF: Nutritional assessment of the spinalcord injured patient. J Am Diet Assoc 78:501–505, 1981.
17. Rapp RP, Young B, Twyman D, Bivins BA, Haack D, Tibbs PA,Bean JR: The favorable effect of early parenteral feeding on sur-vival in head injured patients. J Neurosurg 58:906–912, 1983.
18. Riley KO, May AK, Hadley MN: Neurological injury and nutri-tional support, in Batjer HH, Loftus CM (eds): Textbook of Neuro-logical Surgery: Principles and Practice. Philadelphia, LippincottWilliams & Wilkins (in press).
19. Rodriguez DJ, Benzel EC: Nutritional support, in Benzel EC (ed):Spine Surgery: Techniques, Complication Avoidance, and Management.New York, Churchill Livingstone, 1999, vol 2, pp 1321–1331.
20. Rodriguez DJ, Benzel EC, Clevenger FW: The metabolic responseto spinal cord injury. Spinal Cord 35:599–604, 1997.
21. Sedlock DA, Laventure SJ: Body composition and resting energyexpenditure in long term spinal cord injury. Paraplegia 28:448–454, 1990.
22. Young B, Ott L, Rapp RP, Norton J: The patient with criticalneurological disease. Crit Care Clin 3:217–233, 1987.
23. Young B, Ott L, Twyman D, Norton J, Rapp R, Tibbs P, Haack D,Brivins B, Dempsey R: The effect of nutritional support on out-come from severe head injury. J Neurosurg 67:668–676, 1987.
S84 Guidelines for Management of Acute Cervical Spinal Injuries
CHAPTER 12
Management of Pediatric Cervical Spine and SpinalCord Injuries
RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines:• In children who have experienced trauma and are alert, conversant, have no neurological deficit, no midline
cervical tenderness, and no painful distracting injury, and are not intoxicated, cervical spine x-rays are notnecessary to exclude cervical spine injury and are not recommended.
• In children who have experienced trauma and who are either not alert, nonconversant, or have neurological deficit,midline cervical tenderness, or painful distracting injury, or are intoxicated, it is recommended that anteroposteriorand lateral cervical spine x-rays be obtained.
Options:• In children younger than age 9 years who have experienced trauma, and who are nonconversant or have an
altered mental status, a neurological deficit, neck pain, or a painful distracting injury, are intoxicated, or haveunexplained hypotension, it is recommended that anteroposterior and lateral cervical spine x-rays be obtained.
• In children age 9 years or older who have experienced trauma, and who are nonconversant or have an altered mentalstatus, a neurological deficit, neck pain, or a painful distracting injury, are intoxicated, or have unexplained hypoten-sion, it is recommended that anteroposterior, lateral, and open-mouth cervical spine x-rays be obtained.
• Computed tomographic scanning with attention to the suspected level of neurological injury to excludeoccult fractures or to evaluate regions not seen adequately on plain x-rays is recommended.
• Flexion/extension cervical x-rays or fluoroscopy may be considered to exclude gross ligamentous instabilitywhen there remains a suspicion of cervical spine instability after static x-rays are obtained.
• Magnetic resonance imaging of the cervical spine may be considered to exclude cord or nerve rootcompression, evaluate ligamentous integrity, or provide information regarding neurological prognosis.
TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• Thoracic elevation or an occipital recess to prevent flexion of the head and neck when restrained supine on an
otherwise flat backboard may allow for better neutral alignment and immobilization of the cervical spine in childrenyounger than 8 years because of the relatively large head in these younger children and is recommended.
• Closed reduction and halo immobilization for injuries of the C2 synchondrosis between the body andodontoid is recommended in children younger than 7 years.
• Consideration of primary operative therapy is recommended for isolated ligamentous injuries of the cervicalspine with associated deformity.
RATIONALE
There are distinct, unique aspects of the management ofchildren with potential injuries of the cervical spinalcolumn and cervical spinal cord compared with adult
patients that warrant specific recommendations. The methodsof preadmission immobilization necessary to approximate
“neutral” cervical spinal alignment in a young child differfrom those methods commonly used for adults. The spinalinjury patterns among young children differ from those thatoccur in adults. The diagnostic studies and images necessaryto exclude a cervical spine injury in a child may be differentthan in the adult as well. The interpretation of pediatric x-raystudies must be made with knowledge of age-related devel-
S85Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
opment of the osseous and ligamentous anatomy. Methods ofreduction, stabilization, and subsequent treatment, surgicaland nonsurgical, must be customized to each child, takinginto account the child’s degree of physical maturation and hisor her specific injury. The purpose of this review is to addressthe unique aspects of children with real or potential cervicalspinal injuries and provide recommendations regarding theirmanagement.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “spinal cord inju-ries” and “child” yielded 1022 citations. These citations werereviewed in combination with “cervical vertebra,” “spinalinjuries,” and “child,” which yielded 152 citations. Non-English language citations were deleted. The remaining ab-stracts were reviewed for those that described children whohad sustained or were being evaluated for a cervical spinalcord or cervical spinal column injury. Articles describing theclinical aspects and management of children were used togenerate these guidelines. Case reports were excluded. Of the58 articles meeting selection criteria, none were Class I stud-ies. One Class II study addressed diagnostic imaging in chil-dren. All remaining articles were case series representingClass III evidence. Summaries of these 58 articles are providedin Tables 12.1 and 12.2. In addition, articles germane to thetopic but not meeting criteria for inclusion in the EvidentiaryTables, such as general review articles, are referenced in theScientific Foundation section and are included in thereferences.
SCIENTIFIC FOUNDATION
Preadmission immobilization
The primary goal of preadmission management of pediatricpatients with potential cervical spine or spinal cord injury(SCI) is to prevent further injury. Along with ensuring anadequate airway, ventilation, and perfusion, spinal immobi-lization likely plays an important role in preventing furtherinjury to the vertebral column and spinal cord. Immobiliza-tion of the child’s cervical spine in the neutral position isdesired. To achieve neutral alignment of the cervical spine inchildren younger than 8 years, allowances must be made forthe relatively large head compared with the torso, whichforces the neck into a position of flexion when the head andtorso are supine on a flat surface (39). Nypaver and Treloar(39) prospectively evaluated 40 children younger than 8 yearsseen in an emergency room for reasons other then head andneck trauma and assessed them with respect to neutral posi-tioning on a backboard. They found that all 40 children re-quired elevation of the torso to eliminate positional neckflexion and achieve neutral alignment as determined by twoindependent observers. The mean amount of elevation re-quired was 25 mm. Children younger than 4 years requiredmore elevation than those 4 years or older (P � 0.05). Becauseof these findings, it was recommended that, when immobiliz-
ing children younger than 8 years, either the torso be elevatedor an occipital recess be created to achieve a more neutralposition for immobilization of the cervical spine. In a separatereport, Treloar and Nypaver (60) similarly found that semi-rigid cervical collars placed on children younger than 8 yearsdid not prevent this positional forced flexion when placedsupine on standard, rigid spinal boards.
Herzenberg et al. (26) studied 10 children younger than 7years with cervical spine injuries who were positioned on abackboard. All had anterior angulations or translation at theinjured segment, which was reduced by allowing neck exten-sion into a more neutral position. They suggested that align-ment of the patient’s external auditory meatus with his/hershoulders would help to achieve neutral cervical spine posi-tioning. Curran et al. (8), however, found no correlation be-tween age and degree of cervical kyphosis identified in chil-dren transported on backboards. They did note, however, that30% of children had more than 10 degrees of kyphosis asdetermined by Cobb angle measurements between C2 and C6.No specific technique or device allowed superior neutral po-sitioning of the cervical spine in patients they studied. Noneof their patients were immobilized on boards with an occipitalrecess or thoracic padding. Huerta et al. (27) evaluated avariety of immobilization devices on children, infants, andchild-sized mannequins. They concluded that no collar pro-vided “acceptable immobilization” when used alone. Theyfound that the combination of a modified half-spine board,rigid cervical collar, and tape was the most effective means ofimmobilizing the cervical spine for transport in children.Schafermeyer et al. (51), however, cautioned that immobiliza-tion techniques that use taping across the torso to secure thechild to the spine board may have deleterious effects onrespiratory function. They studied 51 healthy children 6 to 15years of age by measuring forced vital capacity (FVC). FVCdecreased when the child moved from the upright to thesupine position. Taping across the torso to secure the volun-teer to the spine board caused further reductions in FVC of 41to 96% (mean, 80%), compared with the supine FVC withouttape. The authors cautioned that this restriction of FVC mightbe enough to create respiratory insufficiency in some traumapatients.
In summary, when spinal immobilization is indicated forchildren for transportation, the type of immobilization shouldtake into account the child’s age and physical maturity. Itshould allow for the relatively larger head with respect to thetorso in younger children. Although ideal spinal immobiliza-tion of pediatric patients who have sustained trauma seems tobe provided by a combination of a spinal board, rigid collar,and tape, these immobilization techniques may negativelyinfluence the child’s respiratory function.
Imaging
After immobilization and transport to an acute care facility,initial clinical evaluation and medical/hemodynamic supportand the need for and type of radiographic assessment must bedecided and undertaken. Several authors have evaluated theindications for radiographic assessment of children with a
S86 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
potential cervical spinal injury (4, 32). Laham et al. (32) inves-tigated the role of cervical spine x-ray evaluation of 268 chil-dren with apparent isolated head injuries. They retrospec-tively divided the children into high-risk (n � 133) and low-risk (n � 135) groups. High-risk children were those who
were incapable of verbal communication, either because ofage (younger than 2 yr) or head injury, and those who hadneck pain. They used the “three-view approach” of antero-posterior, lateral, and open-mouth x-rays. They discovered nocervical spine injuries in the low-risk group but 10 injuries
TABLE 12.1. Summary of Reports on Diagnosis of Pediatric Cervical Spinal Injuriesa
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Viccellio et al., 2001 (62) Prospective multicenter evaluation of cervicalspine x-rays obtained in 3065 childrenincurring trauma. Low-risk criteria of absenceof: neck tenderness, painful distracting injury,altered alertness, neurological deficit, orintoxication.
II No child fulfilling all 5 low-risk criteria had acervical spine injury. X-rays may not be necessary toclear the cervical spine in children fulfilling all 5criteria.
Ralston et al., 2001 (45) Blinded review of 129 children with bluntcervical trauma who had flexion and extensionx-rays.
III Flexion and extension views with normal cervicalspine x-rays or with only loss of cervical lordosis didnot unmask any new abnormalities.
Buhs et al., 2000 (6) Multi-institutional review of pediatric cervicalspine injuries and the x-rays needed to achievea diagnosis.
III Lateral cervical x-ray was diagnostic in 13/15children aged �9 yr. In no child �9 yr old was theopen-mouth view the diagnostic study. In only 1/36children �9 yr old was the open-mouth view thediagnostic study.
Dwek and Chung, 2000(11a)
Retrospective review of 247 children with ahistory of trauma who had flexion andextension cervical spine x-rays.
III All children (91%) with normal static cervical x-rayshad normal flexion/extension x-rays.
Swischuk et al., 2000 (59) Survey of pediatric radiologists regarding use ofopen-mouth view of the odontoid.
III �50% response. Approximately 40% of respondentsdid not use open-mouth views in children.
Scarrow et al., 1999 (50) Performed flexion/extension cervicalfluoroscopy with SSEP monitoring in 15comatose pediatric patients.
III None had radiographic abnormalities. 3 childrenhad changes in the SSEPs; 1 of these 3 children wasstudied with MRI, and findings were normal.
Shaw et al., 1999 (56) Retrospective review of the cervical x-rays in138 trauma patients �16 yr old.
III 22% incidence of pseudosubluxation of C2 on C3.Median age of pseudosubluxation group was 6.5 yrversus 9 yr for those without pseudosubluxation.
Berne et al., 1999 (3) 58 patients with severe blunt trauma underwenthelical CT of entire cervical spine.
III 20 had cervical spine injuries. Plain x-rays missed 8injuries. CT missed 2 injuries.
Keiper et al., 1998 (29) Retrospective review of evaluating 52 childrenby MRI with suspected cervical spine trauma orinstability without fracture.
III There were 16 abnormal studies. The most commonabnormality was posterior ligamentous injury. 4children underwent surgical stabilization. The MRIfindings caused the surgeon to extend the length ofstabilization in all 4 cases.
Davis et al., 1993 (9) Retrospective review of 15 children with spinalcord injury underwent MRI 12 h to 2 mo afterinjury. 7 with SCIWORA.
III MRI findings correlated with prognosis. Hemorrhagiccord contusions and cord “infarction” wereassociated with permanent deficits. No compressivelesions in SCIWORA cases. Normal MRI findingswere associated with no myelopathy.
Schleehauf et al., 1989 (51) 104 “high-risk” patients underwent CT asscreening tool for cervical spine injury.
III Sensitivity overall was 0.78. Sensitivity was 1.0 forunstable injuries not able to be seen by plain x-rays.2 upper cervical subluxations without fracture weremissed.
Kowalski et al., 1987 (31) 8 patients with occipitalatlantoaxial problemsand 6 normal subjects were studied with CT.
III CT findings looked similar for those with C1–C2rotary subluxation to normal subjects with theirheads maximally turned. CT with the head turned tothe contralateral side differentiated rotarysubluxation from normals and spasmodic torticollis.
Cattell and Filtzer, 1965 (7) Lateral upright cervical x-rays in neutral,flexion, and extension in 160 randomlyselected children aged 1–16 yr.
III C2–C3 subluxation was moderate to marked in 24%,predominantly in children �8 yr of age. Theatlantodens interval was �3 mm during flexion in20% of children �8 yr of age.
a SSEP, somatosensory evoked potential; CT, computed tomography; MRI, magnetic resonance imaging; SCIWORA, spinal cord injury withoutradiographic abnormality.
Management of Pediatric Cervical Spinal Injuries S87
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE12
.2.
Sum
mar
yof
Rep
orts
onTr
eatm
ent
ofPe
diat
ric
Cer
vica
lSp
inal
Inju
ries
a
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ce
Cla
ssC
oncl
usio
ns
Eler
aky
etal
.,20
00(1
2)R
etro
spec
tive
revi
ewof
102
child
ren
with
cerv
ical
spin
alin
juri
es.
III30
child
ren
(30%
)w
ere
trea
ted
surg
ical
ly.
Ode
ntet
al.,
1999
(40)
Rev
iew
of15
youn
gch
ildre
nw
ithod
onto
idin
juri
es.
III6
with
neur
olog
ical
defic
itsha
dce
rvic
otho
raci
cco
rdin
juri
es.
Exte
rnal
imm
obili
zatio
nw
asa
succ
essf
ulpr
imar
yth
erap
y.3
child
ren
who
unde
rwen
top
erat
ions
asth
eir
prim
ary
ther
apy
expe
rien
ced
com
plic
atio
ns.
Schw
arz,
1998
(52)
Are
view
of4
child
ren
pres
entin
g�
3m
oaf
ter
the
onse
tof
C1–
C2
rota
ry
subl
uxat
ion.
III2
child
ren
had
irre
duci
ble
subl
uxat
ions
.1
child
had
recu
rren
tsu
blux
atio
nin
aM
iner
vaca
st.
1
child
was
succ
essf
ully
trea
ted
with
clos
edre
duct
ion
and
8w
kin
aM
iner
vaca
st.
Suba
chet
al.,
1998
(58)
Are
view
of20
child
ren
with
C1–
C2
rota
rysu
blux
atio
n.III
4re
duce
dsp
onta
neou
sly.
15/1
6tr
eate
dw
ithtr
actio
nre
duce
din
am
ean
of4
d.6
child
ren
requ
ired
fusi
onbe
caus
eof
recu
rren
tsu
blux
atio
nor
irre
duci
ble
subl
uxat
ion.
No
child
expe
rien
ced
recu
rren
tsu
blux
atio
nif
redu
ced
with
in21
dof
sym
ptom
onse
t.
Finc
han
dB
arne
s,19
98(1
6)R
etro
spec
tive
revi
ewof
32ch
ildre
nw
ithm
ajor
cerv
ical
spin
ein
juri
es.
III8
child
ren
(25%
)w
ere
trea
ted
surg
ical
ly.
All
achi
eved
unio
nor
radi
olog
ical
stab
ility
.N
o
neur
olog
ical
dete
rior
atio
nfr
omsu
rger
yor
clos
edre
duct
ion.
Ope
rate
don
ligam
ento
usin
juri
es.
Rei
nges
etal
.,19
98(4
7)R
epor
tof
prim
ary
C1–
C2
fusi
onin
ayo
ung
child
with
anod
onto
id
inju
ryan
dlo
wer
cerv
ical
cord
inju
ry.
IIIN
one
urol
ogic
alim
prov
emen
t.Su
cces
sful
fusi
on.
Trel
oar
and
Nyp
aver
,19
97(6
0)M
easu
rem
ent
ofce
rvic
alsp
ine
flexi
onin
child
ren
with
sem
i-ri
gid
colla
rs
onsp
inal
boar
ds.
IIISe
mi-
rigi
dco
llars
did
not
prev
ent
the
cerv
ical
spin
efr
ombe
ing
forc
edin
tofle
xion
inch
ildre
n�
8
yrol
dw
hen
ona
spin
albo
ard.
Lui
etal
.,19
96(3
3)R
etro
spec
tive
revi
ewof
C1–
C2
inju
ries
in22
child
ren.
12ch
ildre
nha
d
odon
toid
inju
ries
.9
child
ren
had
ligam
ento
usin
juri
es(A
AD
s)on
ly.
IIIFl
exio
n/ex
tens
ion
x-ra
ysw
ere
need
edto
diag
nose
4od
onto
idin
juri
esan
d6
AA
Ds.
9/12
odon
toid
inju
ries
redu
ced
easi
ly.
5/7
odon
toid
inju
ries
wer
etr
eate
dsu
cces
sful
lyw
ithha
lo.
2od
onto
id
inju
ries
wer
eop
erat
edon
imm
edia
tely
.2
odon
toid
inju
ries
faile
dex
tern
alim
mob
iliza
tion.
5A
AD
s
wer
ein
itial
lytr
eate
dw
ithsu
rgic
alfu
sion
.2
AA
Ds
initi
ally
trea
ted
with
halo
requ
ired
surg
ical
stab
iliza
tion.
Giv
ens
etal
.,19
96(2
0)R
evie
wof
34ch
ildre
nw
ithce
rvic
alsp
ine
inju
ries
over
a3-
yrpe
riod
.III
18in
juri
esoc
curr
edbe
low
C3.
The
leve
lof
inju
rydi
dno
tco
rrel
ate
with
age.
You
ngag
eis
not
asso
ciat
edw
ithex
clus
ivel
yup
per
cerv
ical
spin
ein
juri
es.
Turg
utet
al.,
1996
(61)
Ret
rosp
ectiv
ere
view
of82
child
ren
with
spin
alco
rdor
colu
mn
inju
ries
.III
14ch
ildre
n(1
7%)
wer
etr
eate
dsu
rgic
ally
.
Dor
man
set
al.,
1995
(11)
Are
view
of37
child
ren
with
halo
ring
san
dve
sts
aged
3–16
yr.
Arb
itrar
ilydi
vide
din
toth
ose
�10
yrol
d,an
dol
der.
IIIO
vera
ll68
%co
mpl
icat
ion
rate
.Pi
n-si
tein
fect
ion
was
the
mos
tco
mm
onco
mpl
icat
ion.
Puru
lent
infe
ctio
nsoc
curr
edm
ore
freq
uent
lyin
the
olde
rgr
oup.
Bot
hlo
osen
ing
and
infe
ctio
noc
curr
ed
mor
efr
eque
ntly
inth
ean
teri
orpi
nsi
tes.
Men
ticog
lou
etal
.,19
95(3
7)R
etro
spec
tive
case
seri
esof
15ne
onat
esw
ithbi
rth-
rela
ted
high
cerv
ical
cord
inju
ries
.
IIIA
ll15
wer
ece
phal
icpr
esen
tatio
nsin
whi
chfo
rcep
san
dat
tem
pted
rota
tion
wer
eus
ed.
All
but
1
wer
eap
neic
atbi
rth.
Cur
ran
etal
.,19
95(8
)Pr
ospe
ctiv
est
udy
of11
8ch
ildre
nw
hoar
rive
dim
mob
ilize
dto
asi
ngle
emer
genc
yro
om.
The
cerv
ical
spin
eal
ignm
ent
was
mea
sure
dan
d
com
pare
dw
ithag
ean
dty
peof
imm
obili
zatio
n.
IIIN
oco
rrel
atio
nw
ithde
gree
ofky
phos
isor
lord
osis
was
foun
dw
ithag
e.30
%ha
da
kyph
osis
of
�10
degr
ees.
No
sing
leim
mob
iliza
tion
tech
niqu
ew
assu
peri
or.
Schw
arz
etal
.,19
94(5
3)R
evie
wof
10ch
ildre
nw
ithve
rteb
ral
frac
ture
san
dky
phot
ican
gula
tion.
IIITh
eky
phot
ican
gula
tion
rem
aine
dun
chan
ged
orw
orse
ned
whe
nex
tern
alim
mob
iliza
tion
alon
e
(n�
7)or
dors
alfu
sion
(n�
1)w
asus
ed.
Onl
yth
ose
unde
rgoi
nga
vent
ral
fusi
on(n
�2)
had
a
stab
lere
duct
ion
ofth
eky
phot
icde
form
ity.
Nyp
aver
and
Trel
oar,
1994
(39)
40ch
ildre
nw
ere
plac
edon
spin
ebo
ards
,an
dob
serv
ers
judg
edw
heth
er
the
cerv
ical
spin
ew
asin
the
“neu
tral
”po
sitio
n.
IIIC
hild
ren
�8
yrof
age
requ
ired
tors
oel
evat
ion
toac
hiev
ene
utra
lal
ignm
ent.
Chi
ldre
n�
4yr
ofag
e
requ
ired
the
grea
test
amou
ntof
elev
atio
n.
Laha
met
al.,
1994
(32)
Div
ided
head
-inj
ured
child
ren
into
high
-(�
2yr
ofag
e,
nonc
omm
unic
ativ
e,or
with
neck
pain
)an
dlo
w-r
isk
grou
psfo
rce
rvic
al
spin
ein
jury
.
IIIN
oce
rvic
alsp
ine
inju
ries
dete
cted
inth
elo
w-r
isk
grou
p.10
inju
ries
(7.5
%)
wer
ede
tect
edin
the
high
-ris
kgr
oup.
Fotte
ret
al.,
1994
(17)
Rep
ort
ofbi
rth-
rela
ted
spin
alco
rdin
juri
esim
aged
with
ultr
asou
ndan
d
MR
I.
IIIA
neon
ate
with
com
plet
ein
jury
had
norm
alpl
ain
x-ra
ysw
ithsp
inal
ultr
asou
ndsh
owin
g
inho
mog
eneo
usec
hoge
nici
tyan
ddi
srup
ted
cord
surf
ace.
Ane
onat
ew
ithan
inco
mpl
ete
inju
ryha
d
inta
ctco
rdsu
rfac
ew
ithin
crea
sed
cord
echo
geni
city
.M
RI
corr
obor
ated
thes
efin
ding
s.
Mar
kset
al.,
1993
(36)
Rev
iew
of8
child
ren,
ages
3m
oto
12yr
,im
mob
ilize
din
aha
loja
cket
for
6w
kto
12m
o(m
ean,
2m
o).
IIITh
eon
lyco
mpl
icat
ion
was
aja
cket
chan
gere
quir
edfo
ra
fore
ign
body
(coi
n).
Onl
y3
ofth
ese
child
ren
had
cerv
ical
inst
abili
ty.
S88 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE12
.2.
Con
tinu
ed
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ce
Cla
ssC
oncl
usio
ns
Shac
ked
etal
.,19
93(5
4)R
etro
spec
tive
revi
ewof
6ch
ildre
n(3
–14
yrol
d)w
ithce
rvic
alsp
ine
inju
ries
trea
ted
via
anan
teri
orap
proa
ch.
IIIA
utog
raft
with
out
inst
rum
enta
tion
afte
rco
rpec
tom
yw
asus
ed.
They
wer
est
abili
zed
post
oper
ativ
ely
with
hard
colla
ror
Min
erva
cast
.A
llw
ithso
lidfu
sion
s,go
odal
ignm
ent,
and
norm
alce
rvic
al
grow
th.
Follo
w-u
p,3–
8yr
.
Gro
gaar
det
al.,
1993
(23)
Atla
ntoa
xial
rota
rysu
blux
atio
nde
scri
bed
in9
child
ren.
8w
ere
diag
nose
dw
ithin
5d,
1w
asdi
agno
sed
afte
r8
wk.
III8
child
ren
wer
etr
eate
dsu
cces
sful
lyw
ith“m
ild”
trac
tion
and
then
aco
llar
for
4–6
wk.
1ch
ild
pres
entin
gla
tere
quir
ed1
wk
oftr
actio
nfo
rre
duct
ion.
Ther
ew
ere
2re
disl
ocat
ions
.A
llev
entu
ally
heal
edin
alig
nmen
tw
ithou
tsu
rger
y.
Man
daba
chet
al.,
1993
(35)
13ch
ildre
nw
ithax
isin
juri
esw
ere
revi
ewed
.10
wer
etr
eate
dpr
imar
ily
with
clos
edre
duct
ion
and
halo
imm
obili
zatio
n.
III8
ofth
e10
trea
ted
prim
arily
with
clos
edre
duct
ion
and
halo
imm
obili
zatio
nfu
sed.
2re
quir
ed
surg
ical
stab
iliza
tion
and
fusi
on.
Mac
Kin
non
etal
.,19
93(3
4)R
etro
spec
tive
case
seri
esof
22ne
onat
esw
ithbi
rth-
rela
ted
spin
alco
rd
inju
ries
.Th
eyex
clud
edne
onat
esw
ithSC
IWO
RA
.
IIIA
ll14
with
high
cerv
ical
inju
ries
had
ceph
alic
pres
enta
tions
with
atte
mpt
edfo
rcep
sro
tatio
n.A
ll6
with
cerv
icot
hora
cic
inju
ries
had
bree
chpr
esen
tatio
ns.
Bot
hne
onat
esw
ithth
orac
olum
bar
inju
ries
wer
epr
emat
ure.
Ros
sitc
han
dO
akes
,19
92(4
8)R
etro
spec
tive
revi
ewof
5ne
onat
esw
ithpe
rina
tal
spin
alco
rdin
jury
.N
o
flexi
on/e
xten
sion
view
sre
port
ed.
III4
ofth
e5
had
noab
norm
ality
onst
atic
spin
alx-
rays
.R
espi
rato
ryin
suffi
cien
cyan
dhy
poto
nia
wer
e
com
mon
sign
s.M
yelo
gram
sw
ere
unre
veal
ing.
All
3w
ithhi
ghce
rvic
alin
juri
esdi
edby
age
3yr
.
Ose
nbac
han
dM
enez
es,
1992
(41)
Ret
rosp
ectiv
ere
view
of17
9ch
ildre
nw
ithsp
inal
inju
ries
.III
59(3
3%)
unde
rwen
tsu
rgic
altr
eatm
ent
for
irre
duci
ble
unst
able
inju
ries
.83
%of
thos
etr
eate
d
surg
ical
lyw
ere
�9
yrof
age.
No
child
with
com
plet
eor
seve
repa
rtia
lm
yelo
path
yre
gain
edus
eful
func
tion.
Rat
hbon
eet
al.,
1992
(46)
Ret
rosp
ectiv
ere
view
of12
child
ren
with
pres
umed
spin
alco
rd
conc
ussi
ondu
ring
athl
etic
sw
ere
inve
stig
ated
for
the
pres
ence
ofce
rvic
al
sten
osis
.
III3
had
aTo
rgra
tio�
0.8,
and
4ha
da
cana
lan
tero
post
erio
rdi
amet
er�
13.4
mm
.M
RI
was
not
used
toev
alua
tefo
rst
enos
is.
Ham
ilton
and
Myl
es,
1992
(25)
Ret
rosp
ectiv
ere
view
ofal
lpe
diat
ric
spin
alin
juri
esdu
ring
14-y
rpe
riod
.
73ch
ildre
nha
dce
rvic
alin
juri
es.
IIISu
rger
yw
aspe
rfor
med
in26
%of
child
ren.
13%
ofch
ildre
nw
ithfr
actu
rean
dno
subl
uxat
ion,
50%
with
subl
uxat
ion
alon
e,an
d57
%w
ithfr
actu
rean
dsu
blux
atio
nw
ere
trea
ted
surg
ical
ly.
Of
39
child
ren
with
com
plet
em
yelo
path
y,4
impr
oved
one
ortw
oFr
anke
lgr
ades
.
Scha
ferm
eyer
etal
.,19
91(5
5)FV
Cw
asst
udie
din
heal
thy
child
ren
whe
nup
righ
t,su
pine
,an
dsu
pine
tape
dto
asp
inal
boar
d.
IIITa
ping
the
child
toth
esp
inal
boar
dca
used
FVC
tode
crea
se41
–96%
(mea
n,80
%)
ofsu
pine
FVC
.
Boh
net
al.,
1990
(4)
16/1
9ch
ildre
npr
esen
ting
with
abse
ntvi
tal
sign
sor
seve
rehy
pote
nsio
n
unex
plai
ned
bybl
ood
loss
unde
rwen
tpo
stm
orte
mex
amin
atio
n.
III13
/16
had
cord
lace
ratio
nor
tran
sect
ion.
2of
thes
ech
ildre
nha
da
norm
alce
rvic
alx-
ray.
Gas
kill
and
Mar
lin,
1990
(18)
6ch
ildre
nag
ed2–
4yr
wer
epl
aced
inM
iner
vaja
cket
sfo
rce
rvic
alsp
ine
inst
abili
ty.
III1
child
had
skin
brea
kdow
nof
the
chin
.Ea
ting
and
othe
rda
ilyac
tiviti
esw
ere
not
impa
ired
.2
wer
epl
aced
inM
iner
vaja
cket
saf
ter
com
plic
atio
nsof
halo
ring
and
vest
imm
obili
zatio
n.
Phill
ips
and
Hen
sing
er,
1989
(44)
Are
view
of23
child
ren
with
C1–
C2
rota
rysu
blux
atio
n.III
16ch
ildre
nse
enw
ithin
1m
oof
onse
tha
dei
ther
spon
tane
ous
redu
ctio
nor
redu
ced
with
trac
tion.
Of
the
7ch
ildre
npr
esen
ting
with
sym
ptom
sfo
r�
1m
o,1
subl
uxat
ion
was
irre
duci
ble,
and
4ha
d
recu
rren
tsu
blux
atio
ns.
Kaw
abe
etal
.,19
89(2
8)R
evie
wof
the
radi
olog
yof
17ch
ildre
nw
ithC
1–C
2ro
tary
subl
uxat
ion.
IIIC
lass
ified
acco
rdin
gto
Fiel
ding
and
Haw
kins
as10
Type
I,5
Type
II,2
Type
III,
and
noTy
peIV
.
Ben
zel
etal
.,19
89(2
)A
com
pari
son
ofce
rvic
alm
otio
nof
inju
red
patie
nts
(onl
y1
child
)
imm
obili
zed
inha
loan
dM
iner
vaja
cket
s.
IIITh
eM
iner
vaja
cket
allo
wed
less
mot
ion
than
the
halo
jack
etat
ever
yle
vel
exce
ptC
1–C
2.
Bau
met
al.,
1989
(1)
Are
view
com
pari
ngth
eha
loco
mpl
icat
ions
in13
child
ren
and
80
adul
ts.
III39
%co
mpl
icat
ion
rate
inch
ildre
nve
rsus
8%in
adul
ts.
The
child
ren
had
4pi
n-si
tein
fect
ions
and
1in
ner
tabl
ecr
ania
lpi
npe
netr
atio
n.
Mub
arak
etal
.,19
89(3
8)R
evie
wof
3ch
ildre
n�
2yr
old
who
wer
epl
aced
inha
lori
ngs
for
2–31
⁄2m
o.
III10
pins
tight
ened
“fin
ger-
tight
”in
a7-
mo-
old
child
,an
d2
in/lb
ina
16-
and
24-m
o-ol
d.2
of3
deve
lope
dm
inor
pin-
site
infe
ctio
nsne
cess
itatin
gpi
nre
mov
al.
Her
zenb
erg
etal
.,19
89(2
6)R
epor
ted
10ch
ildre
n�
7yr
ofag
ew
ithce
rvic
alsp
ine
inju
ries
posi
tione
don
afla
tba
ckbo
ard.
IIITh
ein
juri
esw
ere
ante
rior
lyan
gula
ted
ortr
ansl
ated
whe
non
afla
tba
ckbo
ard
beca
use
the
head
was
forc
edin
tofle
xion
.El
evat
ing
the
tors
oal
low
edfo
rm
ore
neut
ral
alig
nmen
tan
dre
duct
ion
of
the
inju
red
segm
ent.
Evan
san
dB
ethe
m,
1989
(14)
Rev
iew
of24
cons
ecut
ive
cerv
ical
spin
ein
juri
esin
child
ren
�18
yrol
d.III
Hal
fof
the
child
ren
had
inju
ries
atC
3or
abov
e.1
child
was
trea
ted
with
lam
inec
tom
yan
d2
with
fusi
on.
Frac
ture
she
aled
in21
/22
with
nono
pera
tive
ther
apy.
Bir
ney
and
Han
ley,
1988
(3a)
Ret
rosp
ectiv
ere
view
of61
child
ren
with
cerv
ical
spin
ein
juri
es.
23of
thes
ein
juri
esw
ere
C1–
C2
rota
rysu
blux
atio
n.
IIIR
otar
ysu
blux
atio
nun
asso
ciat
edw
ithne
urol
ogic
alde
ficit.
The
defo
rmity
reso
lved
with
halte
r
trac
tion
(n�
10)
orce
rvic
albr
acin
g.1
child
had
are
curr
ence
.A
child
with
tran
sver
selig
amen
t
disr
uptio
nw
astr
eate
dsu
cces
sful
lyw
itha
soft
colla
ron
ly.
Management of Pediatric Cervical Spinal Injuries S89
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE12
.2.
Con
tinu
ed
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ce
Cla
ssC
oncl
usio
ns
Had
ley
etal
.,19
88(2
4)R
etro
spec
tive
revi
ewof
122
child
ren
with
spin
alin
juri
es.
Ther
ew
ere
97
cerv
ical
inju
ries
.
IIIO
nly
12ce
rvic
alin
juri
esw
ere
trea
ted
surg
ical
ly.
Hue
rta
etal
.,19
87(2
7)Ev
alua
tion
ofth
eim
mob
iliza
tion
ofco
mm
erci
ally
avai
labl
ein
fant
and
pedi
atri
cce
rvic
alco
llars
.
IIIN
oco
llar
used
alon
epr
ovid
edac
cept
able
imm
obili
zatio
n.Th
eus
eof
am
odifi
edha
lf-sp
ine
boar
d,
rigi
dco
llar,
and
tape
prov
ided
the
best
imm
obili
zatio
n.
Penn
ecot
etal
.,19
84(4
3)R
evie
wof
16ch
ildre
nw
ithlig
amen
tous
inju
ries
ofth
ece
rvic
alsp
ine.
5
with
C1–
C2
inju
ries
.
IIIO
fth
e11
child
ren
with
inju
ries
belo
wC
2,8
unde
rwen
tsu
rgic
alst
abili
zatio
n.Th
eyre
com
men
ded
a3-
mo
tria
lof
exte
rnal
imm
obili
zatio
nin
child
ren
with
ligam
ento
usin
juri
esbu
tno
neur
olog
ical
defic
itor
disl
ocat
ion.
El-K
hour
yet
al.,
1984
(13)
Rev
iew
of3
child
ren
with
C1–
C2
rota
rysu
blux
atio
n.III
All
3w
ere
trea
ted
succ
essf
ully
with
trac
tion
orm
anua
lre
duct
ion
with
in24
hof
pres
enta
tion.
One
child
had
recu
rren
tsu
blux
atio
nth
ene
xtda
yan
dw
astr
eate
dsu
cces
sful
lyw
ithm
anua
lre
duct
ion.
Exte
rnal
orth
oses
wer
eus
edfo
r10
wk,
3m
o,an
d4
mo,
resp
ectiv
ely.
Koo
pet
al.,
1984
(30)
Ret
rosp
ectiv
ere
view
of13
child
ren
with
cerv
ical
inst
abili
tytr
eate
dw
ith
post
erio
rar
thro
desi
san
dha
loim
mob
iliza
tion.
Onl
y3
had
trau
mat
ic
lesi
ons.
III1
faile
dfu
sion
whe
nba
nk-b
one
was
used
.O
ther
ssu
cces
sful
lyfu
sed
with
auto
geno
usili
accr
est
or
rib.
Inte
rnal
wir
ing
used
in2
child
ren.
Ave
rage
halo
imm
obili
zatio
nw
as15
0d.
Sher
ket
al.,
1978
(57)
Rep
ort
of11
child
ren
with
odon
toid
inju
ries
,an
dre
view
of24
from
the
liter
atur
e.
IIIM
ajor
ityof
inju
red
odon
toid
sar
ean
gled
ante
rior
ly.
All
but
1ch
ildw
astr
eate
dsu
cces
sful
lyw
ith
exte
rnal
imm
obili
zatio
n.
Fiel
ding
and
Haw
kins
,19
77(1
5)Th
era
diog
raph
icfin
ding
sof
17pa
tient
sw
ithat
lant
oaxi
alro
tary
fixat
ion
are
desc
ribe
dan
dcl
assi
fied
into
4ty
pes.
III4
clas
ses
ofC
1–C
2ro
tary
subl
uxat
ion
wer
ede
scri
bed,
Type
sI–
IV.
Type
I:od
onto
idac
tsas
pivo
tw
ithco
mpe
tent
tran
sver
selig
amen
t.
Type
II:on
ela
tera
lar
ticul
arpr
oces
sac
tsas
pivo
tw
ithup
to5
mm
ofan
teri
ordi
spla
cem
ent.
Type
III:
both
C1
infe
rior
face
tsar
esu
blux
edan
teri
orly
with
�5
mm
ofan
teri
ordi
spla
cem
ent,
whi
chsu
gges
tsan
inco
mpe
tent
tran
sver
selig
amen
t.
Type
IV:
post
erio
rdi
spla
cem
ent
with
abse
ntor
inco
mpe
tent
odon
toid
.
Gau
finan
dG
oodm
an,
1975
(19)
Are
view
of3
child
ren
�20
mo
old
with
cerv
ical
spin
ein
juri
es.
2of
thes
ech
ildre
nw
ere
trea
ted
with
trac
tion
deliv
ered
via
22-g
auge
wir
e
plac
edth
roug
hbi
late
ral
pari
etal
burr
hole
s.
IIISu
cces
sful
trac
tion
appl
ied
toth
e10
-wk-
old
and
16-m
o-ol
dch
ild.
�9
lbus
edin
the
10-w
k-ol
d
infa
nt.
No
com
plic
atio
nsw
ere
enco
unte
red
with
the
trac
tion
inpl
ace
for
14an
d41
d,
resp
ectiv
ely.
aA
AD
,at
lant
oaxi
aldi
sloc
atio
n;M
RI,
mag
neti
cre
sona
nce
imag
ing;
SCIW
OR
A,
spin
alco
rdin
jury
wit
hout
radi
ogra
phic
abno
rmal
ity;
FVC
,fo
rced
vita
lca
paci
ty.
S90 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
(7.5%) in the high-risk group. The authors concluded thatcervical spine x-rays are not necessary in children with iso-lated head injuries who can communicate and have no neckpain or neurological deficit. Bohn et al. (4) emphasized thatunexplained hypotension or absent vital signs in pediatrictrauma patients are likely to result from a severe cervical cordinjury. Therefore, they advocate suspicion for a cervical SCI inchildren with either multisystem trauma or an isolated headinjury with hypotension or cardiopulmonary arrest.
Viccellio et al. (62) evaluated the cervical spines in childrenyounger than 18 years using the National EmergencyX-radiography Utilization Study (NEXUS) decision instru-ment in a Class II prospective multicenter study. They usedfive low-risk criteria. These criteria were: 1) the absence ofmidline cervical tenderness, 2) evidence of intoxication, 3)altered level of alertness, 4) focal neurological deficit, and 5) apainful distracting injury. X-rays were obtained at the discre-tion of the treating physician. When x-rays were obtained, aminimum of three views were obtained. Only those patientsfor whom x-rays were obtained were included in the study. Ifall five criteria were met, the child was considered to be at lowrisk. If any one of the five criteria was present, the child wasconsidered to be at high risk. Of 3065 children evaluated, 603fulfilled the low-risk criteria. None of these 603 children de-fined as low-risk had a documented cervical spine injury byradiographic evaluation. Thirty injuries (0.98%) were docu-mented in children not fulfilling the low-risk criteria. Theauthors concluded that applying the NEXUS criteria to chil-dren would reduce cervical spine x-ray use by 20% and wouldnot result in missed injuries. They cautioned that they hadstudied relatively small numbers of young children youngerthan 2 years (n � 88). Statistically, this created large confi-dence intervals for the sensitivity of their instrument whenapplied to younger children. From this Class II study, they“cautiously” endorsed the application of NEXUS criteria inchildren, particularly those from birth to age 9 years. Theirconclusions are consistent with the Class III evidence previ-ously described by Laham et al. (32) on this topic.
The need for and usefulness of open-mouth odontoid viewsin pediatric trauma patients has been questioned (6, 59). Swis-chuk et al. (59) surveyed 984 pediatric radiologists to deter-mine how many injuries were missed on lateral cervical spinex-rays but detected on an open-mouth view. There were 432responses. One hundred sixty-one respondents did not rou-tinely use open-mouth views. Of the 271 radiologists whoobtained open-mouth views in young children, 191 (70%)would not persist beyond a single attempt. Seventy-one radi-ologists (26%) would make up to five attempts to obtain anadequate image. Twenty-eight (7%) of the 432 respondentsreported missing a total of 46 fractures on the lateral view thatwere detected on the open-mouth view. The types of injurieswere not classified (i.e., odontoid versus C1 injury). The au-thors calculated a missed fracture rate of 0.007 per year perradiologist in their study. They concluded that the open-mouth view x-ray might not be needed routinely in childrenyounger than 5 years. Buhs et al. (6) also investigated theusefulness of open mouth views in children. They performeda multi-institutional retrospective review of a large metropol-
itan population of patients younger than 16 years who wereassessed for cervical spine trauma over a 10-year period.Fifty-one children with cervical spinal injuries were identi-fied. The lateral cervical spine x-ray confirmed the diagnosisin 13 of 15 children younger than 9 years. In none of the 15younger patients did the open-mouth view provide the diag-nosis. In only 1 of 36 patients in the group aged 9 to 16 yearswas the open-mouth view the diagnostic study (a Type IIIodontoid injury). The authors concluded that the open-mouthview x-ray is not necessary for clearing the cervical spine inchildren younger than 9 years.
Lui et al. (33), in their review of 22 children with C1–C2injuries, commented that flexion/extension x-rays were re-quired to “identify the instability” of traumatic injuries to thedens in 4 of 12 children with odontoid fractures, and in 6 of 9children with purely ligamentous injuries resulting in atlan-toaxial dislocation. The authors did not state whether anabnormality on the static x-ray led to the dynamic studies, orwhether the initial static studies were normal. Because theydid not describe flexion/extension x-rays as part of theirroutine for the assessment of children with potential cervicalspine injuries, it is likely that some imaging or clinical findingprompted the decision to obtain dynamic films in thesechildren.
The experience of Ruge et al. (49) highlighted the propen-sity for upper cervical injuries in children younger than 9years. They reported no injuries below C3. Evans and Bethem(14) described 24 children with cervical spine injuries. In halfof the patients, the injury was at C3 or higher. Givens et al.(20), however, described the occurrence of important injuriesoccurring at all levels of the cervical spine in young children.They described 34 children with cervical spine injuries. Therewas no correlation of level of injury with age. Two of thechildren they managed had injuries at C7–T1. Hence, it wouldbe dangerous to assume that lower cervical spine injuries donot occur in young children and irresponsible to discount theneed for adequate imaging of the lower cervical spine andcervical-thoracic junction in these young patients.
Scarrow et al. (50) attempted to define a protocol to evaluatethe cervical spine in obtunded children after trauma. Theyused somatosensory evoked responses during flexion/extension fluoroscopy. Of the 15 children evaluated with thisprotocol, none showed pathological motion during flexion/extension fluoroscopy. Three children were thought to have achange in the evoked responses during flexion/extension.Only one of the three children with an abnormal evokedresponse underwent magnetic resonance imaging (MRI) thatwas normal. Their investigation failed to demonstrate anyusefulness for evoked responses, flexion/extension fluoros-copy, or MRI of the cervical spine in the evaluation of thecervical spine in children with altered mental status aftertrauma. Larger numbers of children investigated in this man-ner might define a role for one or more of these diagnosticmaneuvers, but as yet there is no evidence to support theiruse.
Ralston et al. (45) retrospectively analyzed the cervicalspine x-rays of 129 children who had flexion/extension x-raysperformed after an initial static x-ray. They found that if the
Management of Pediatric Cervical Spinal Injuries S91
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
static x-ray was normal or depicted only loss of lordosis, theflexion/extension views would reveal no abnormality. Theauthors concluded that the value of the dynamic x-rays wasconfirmation of cervical spinal stability when there was aquestionable finding on the static lateral x-ray. Likewise,Dwek and Chung (11a) retrospectively reviewed 247 childrenwith a history of trauma in whom flexion-extension cervicalx-rays were obtained. Flexion-extension studies were normalin all children who had normal neutral x-rays. The authorsconcluded that the use of flexion-extension x-rays after ob-taining normal static x-rays is of “questionable use.”
The interpretation of cervical spine x-rays must account forthe age and anatomic maturation of the patient. Commonnormal findings on cervical spine x-rays obtained on youngchildren are pseudosubluxation of C2 on C3, overriding of theanterior atlas in relation to the odontoid on extension, exag-gerated atlantodens intervals, and the radiolucent synchon-drosis between the odontoid and C2 body. These normalfindings can be mistaken for acute traumatic injuries in chil-dren after trauma. Cattell and Filtzer (7) obtained lateralcervical x-rays in neutral, flexion, and extension in 160 ran-domly selected children who had no history of trauma orhead and neck problems. The subjects’ ages ranged from 1 to16 years, with 10 children for each year of age. They found a24% incidence of moderate to marked C2 on C3 subluxation inchildren aged 1 through 7 years. Thirty-two (46%) of 70 chil-dren younger than 8 years had 3 mm or more of anterior-posterior motion of C2 on C3 on flexion/extension x-rays.Fourteen percent of all children had radiographic pseudosub-luxation of C3 on C4. Twenty percent of children aged 1through 7 years had an atlantodens interval of 3 mm or more.Overriding of the anterior arch of the atlas on the odontoidwas present in 20% of children younger than 8 years. Thesynchondrosis between the odontoid and axis body was notedas a lucency in all children imaged up to age 4 years. Thesynchondrosis remained visible in half the children up to age11 years. The authors also described an absence of the normalcervical lordosis in 14% of subjects, most commonly in the age8- to 16-year groups. Shaw et al. (56), in a retrospective reviewof cervical spine x-rays in 138 children younger than 16 yearswho were evaluated after trauma, found a 22% incidence ofradiographic pseudosubluxation of C2 on C3. The only factorthat correlated with the presence of pseudosubluxation intheir study was patient age. The pseudosubluxation grouphad a median age of 6.5 years versus 9 years in the groupwithout this finding. It was identified, however, in children asold as 14 years. Intubation status, injury severity score, andsex had no correlation with pseudosubluxation of C2 on C3.To differentiate between physiological and traumatic subluxa-tions, they recommend a method that involves drawing a linethrough the posterior arches of C1 and C3. In the circum-stance of pseudosubluxation of C2 on C3, the C1–C3 lineshould pass through, touch, or lie up to 1 mm anterior to theanterior cortex of the posterior arch of C2. If the anteriorcortex of the posterior arch of C2 is 2 mm or more behind theline, then a true dislocation (rather than pseudosubluxation)should be assumed.
Keiper et al. (29) reviewed their experience of using MRI inthe evaluation of children with clinical evidence of cervicalspine trauma who had no evidence of fracture by plain x-raysor computed tomography (CT), but who had persistent ordelayed symptoms or instability. There were 16 abnormalMRI examinations in 52 children. Posterior soft tissue andligamentous changes were described as the most commonabnormalities. Only one child had a bulging disc. Four ofthese 52 children underwent surgical treatment. In each of thefour surgical cases, the MRI findings led the surgeon to sta-bilize more levels than otherwise would have been under-taken without the MRI information. Davis et al. (9) describedthe use of MRI in evaluating pediatric SCI and found that itdid not reveal any lesion that would warrant surgical decom-pression. These authors did note, however, that MRI findingswere correlated with neurological outcome. Evidence of he-matomyelia was associated with permanent neurological def-icit. Although little information is available on this subject, itseems that preoperative MRI studies of children with unstablecervical spinal injuries, who require surgical stabilization,may affect the specifics of the surgical management.
There are no studies that have systematically reviewed therole of CT in the evaluation of the cervical spines of pediatricpatients after trauma. In children younger than 10 years withcervical spinal injuries, most patients will have ligamentousinjuries without fracture (10, 12, 24, 25, 41). In older childrenwith cervical spinal injuries, the incidence of a fracture ismuch greater than ligamentous injury without fracture, 80%versus 20%, respectively (14, 62). Therefore, normal osseousanatomy as depicted on an axial computed tomographic im-age should not be used alone to exclude injury to the pediatriccervical spine. Schleehauf et al. (52), in reporting a series ofpediatric and adult trauma patients in 1989, concluded thatCT should not be relied on to exclude ligamentous injuries.They reported two false-negative computed tomographicstudies in patients with C1–C2 ligamentous injuries in theirstudy of the merits of CT to evaluate the cervical spine inhigh-risk trauma patients. The authors favored CT for theevaluation of regions that could not be viewed adequatelywith plain x-rays (e.g., C7–T1) and for the investigation of theosseous integrity of specific vertebrae suspicious for fractureon plain x-rays.
In a series consisting almost entirely of adults, the role ofhelical CT in the evaluation of the cervical spine in “high-risk”patients after severe, blunt, multisystem trauma was prospec-tively studied (3). The plain spine x-rays and computed to-mographic images were reviewed by a radiologist blinded tothe patients and their history. The investigators found 20cervical spine injuries (12 stable, 8 unstable) in 58 patients(34%). Eight of these injuries (five stable, three unstable) werenot detected on plain x-rays. The authors concluded thathelical cervical spinal CT should be used to assess the cervicalspine in high-risk trauma patients. In young children inwhom the entire cervical spine is often easily and accuratelyvisualized on plain x-ray studies, the need for cervical spinalhelical CT is likely not as great. In older high-risk childrenwho have spinal biomechanics and injury patterns more con-
S92 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
sistent with those of adult trauma patients, helical CT of thecervical spine may be fruitful.
In summary, to clear a child’s cervical spine, Class II andClass III evidence supports obtaining lateral and anteropos-terior cervical spine x-rays in children who have experiencedtrauma and cannot communicate because of age or headinjury, have a neurological deficit, have neck pain, have apainful distracting injury, or are intoxicated. In children whoare alert, have no neurological deficit, no midline cervicaltenderness, and no painful distracting injury, and are notintoxicated, cervical spine x-rays are not necessary to excludecervical spine injury (32, 62). Unexplained hypotensionshould raise the suspicion of SCI. Open-mouth views of theodontoid do not seem to be useful in children younger than 9years. Open-mouth views should be attempted in childrenaged 9 years and older. Flexion/extension studies (fluorosco-py or x-rays) are likely to be unrevealing in children withstatic x-rays proven to be normal. Dynamic studies could beconsidered, however, when the static x-rays or the child’sclinical findings suggest but do not definitively demonstratecervical spinal instability. CT of the cervical spine should beused judiciously to define bony anatomy at specific levels butis not recommended as a means to clear the entire cervicalspine in children. MRI may provide important informationabout ligamentous injury that may influence surgical manage-ment, and it may provide prognostic information regardingneurological outcome.
Injury management
Injury patterns that have a strong predilection for or areunique to children merit discussion because of the specializedmanagement paradigms used to treat them. SCIs withoutradiographic abnormality and atlanto-occipital dislocation in-juries have been addressed elsewhere (see Chapters 13 and14). SCIs secondary to birth-related trauma and epiphysiolysisof the axis are injuries unique to children. Common but notunique to children are C1–C2 rotary subluxation injuries.These entities will be discussed below in light of the availableliterature. It should be noted that there is no informationprovided in the literature describing the medical managementof pediatric patients with SCI. The issue of corticosteroidadministration after acute pediatric SCI, for example, has notbeen addressed. Although prospective, randomized clinicaltrials such as the Second and Third National Acute SpinalCord Injury Studies have evaluated pharmacological therapyafter acute SCI, children younger than 13 years were excludedfrom the study (5).
Neonatal SCI
Birth injuries of the spinal cord occur in approximately 1per 60,000 births (63). The most common level of injury isupper cervical and then cervicothoracic (34). MacKinnon et al.(34) described 22 neonates with birth-related SCIs. The diag-nosis was defined by the following criteria: clinical findings ofacute cord injury for at least 1 day and evidence of spinal cordor spinal column injury by imaging or electrophysiologicalstudies. Fourteen neonates had upper cervical injuries, six had
cervicothoracic injuries, and two had thoracolumbar injuries.All upper cervical cord injuries were associated with cephalicpresentation and the use of forceps for rotational maneuvers.Cervicothoracic injuries were associated with the breech pre-sentation. All infants had signs of “spinal shock,” defined asflaccidity, no spontaneous motion, and no deep tendon re-flexes. Of the nine infants with upper cervical injuries surviv-ing longer than 3 months, seven were alive at last follow-up.Six of these seven are dependent on mechanical ventilation.The two neonates with upper cervical injuries who hadbreathing movements on Day 1 of life were the only twothought to have satisfactory outcomes. All survivors withupper cervical cord injuries whose first respiratory effort wasbeyond the first 24 hours of life have remained ventilator-dependent. Only two children of six who sustained cervico-thoracic SCI lived and remained paraplegic. One requiresmechanical ventilation. Hypoxic and ischemic encephalopa-thy was noted in 9 of 14 newborns with upper cervical cordinjuries, and in 1 of 6 with a cervicothoracic cord injury. Theauthors did not describe any treatment provided for the un-derlying spinal column or cord injury, or whether survivorsexperienced progression of any spinal deformities.
Menticoglou et al. (37), drawing partly from the same pa-tient data as MacKinnon et al. (34), reported 15 neonates withbirth-related upper cervical SCIs. All were associated withcephalic deliveries requiring rotational maneuvers with for-ceps. All but one child was apneic at birth with quadriplegia.There is no description of postinjury spinal column or spinalcord management, medical or surgical, in their report. Ros-sitch and Oakes (48) described five neonates with birth-related SCI. They reported that incorrect diagnoses weremade in four. They consisted of Werdnig-Hoffmann syn-drome, occult myelodysplasia, and birth asphyxia. Only oneneonate had an abnormal plain x-ray (atlanto-occipital dislo-cation). They provided no description of the management ofthe spinal cord or column injuries in these five neonates.Fotter et al. (17) reported the use of bedside ultrasound todiagnose neonatal SCI. They found excellent correlation withMRI studies with respect to the extent of cord injury in theirtwo cases. Pang and Hanley (42) provided the only descrip-tion of an external immobilization device for neonates. Theydescribed a thermoplastic molded device that is contoured tothe occiput, neck, and thorax. Velcro straps cross the foreheadand torso, securing the infant and immobilizing the spinalcolumn.
In summary, cervical instability after birth-related SCI is notaddressed in the literature. The extremely high mortality rateassociated with birth-related SCI may have generated thera-peutic nihilism for this entity, hence the lack of aggressivemanagement. The literature suggests that the presentation ofapnea with flaccid quadriplegia after cephalic presentationwith forceps manipulation is the hallmark of upper cervicalSCI. Absence of respiratory effort within the first 24 hours oflife is associated with dependence on long-term mechanicalventilation. It seems reasonable to treat these neonates withspinal immobilization for a presumed cervical spinal injury.The method and length of immobilization is at presentarbitrary.
Management of Pediatric Cervical Spinal Injuries S93
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Odontoid epiphysiolysis
The neurocentral synchondrosis of C2, which may not fusecompletely until age 7 years, represents a vulnerable site ofinjury in young children (22). The lateral cervical spine x-rayis the diagnostic imaging modality of choice to depict thisinjury. It will often reveal the odontoid process to be angu-lated anteriorly, and rarely posteriorly (57). Although injuriesto the neurocentral or subdental synchondrosis may be seenin children up to age 7 years, it most commonly occurs inpreschool-aged children (35). Mandabach et al. (35) described13 children with odontoid injuries ranging in age from 9months to 7 years. The authors reported that 8 of 10 childrenwho were initially managed with halo immobilization aloneachieved stable fusion. The average time to fusion was 13weeks, with a range of 10 to 18 weeks. Because the injuryoccurs through the epiphysis, it has a high likelihood ofhealing if closed reduction and immobilization are used.Mandabach et al., in their review, cited several other reportsdescribing the successful treatment of young children withodontoid injuries who were managed with a variety of exter-nal immobilization devices. Sherk et al. (57) reported 11 chil-dren with odontoid injuries and reviewed an additional 24from the literature. Only 1 of these 35 children required sur-gical fusion. Although the literature describes the use of Min-erva jackets, soft collars, hard collars, and the halo vest asmeans of external immobilization to achieve successful fusionin young children with odontoid injuries, the halo is the mostwidely used immobilization device in the contemporary lit-erature for these injuries (35, 40, 57).
To obtain injury reduction in these children, Mandabach etal. (35) advocate the application of the halo device underketamine anesthesia and then realignment of the dens utiliz-ing C-arm fluoroscopy. Other reports describe using tractionto obtain alignment before immobilizing the child in an ex-ternal orthosis (22). Compared with halo application and im-mediate reduction and immobilization, traction requires aperiod of bed rest and is associated with the potential risk ofoverdistraction (35).
The literature is scant regarding the operative treatment ofC2 epiphysiolysis. Most reports describe using operative in-ternal fixation and fusion only if external immobilization hasfailed to maintain reduction or achieve stability. Reinges et al.(47) noted that only three “young” children have been re-ported in the literature who have had odontoid injuries pri-marily treated with surgical stabilization. This underscoresthe near universal application of external immobilization asthe primary means of treating odontoid injuries in youngchildren. Odent et al. (40) reported that of the 15 youngchildren with odontoid injuries they managed, three that weretreated with surgical stabilization and fusion experiencedcomplications. The other 12 children with similar injuries,managed without operation, did well. Wang et al. (64) de-scribed using anterior odontoid screw fixation as the primarytreatment option in a 3-year-old child with C2 epiphysiolysis.A hard cervical collar was used postoperatively. Halo immo-bilization was not used either preoperatively or postopera-tively. They successfully used anterior odontoid screw fixa-
tion as the primary treatment in two older children (ages 10and 14 years) and then hard collar immobilization. It is likelythat these two children had true Type II odontoid fracturesand not C2 epiphysiolysis. Likewise, Godard et al. (21) per-formed anterior odontoid screw fixation in a 2-year-old childwith a severe head injury. They used skeletal traction to alignthe fracture preoperatively. The rationale for proceeding tooperative stabilization without an attempt at treatment withexternal immobilization was to avoid the halo orthosis and toallow for more aggressive physiotherapy in this severely in-jured child. They believe that anterior odontoid screw fixationis advantageous because no motion segments are fused, nor-mal motion is preserved, and the need for halo immobiliza-tion is obviated.
For management of injuries of the C2 neurocentral syn-chondrosis, the literature supports the use of closed reductionand external immobilization for approximately 10 weeks. Thisstrategy is associated with an 80% fusion success rate. Al-though primary surgical stabilization of this injury has beenreported, the experience in the literature is limited. Surgicalstabilization seems to play a role when external immobiliza-tion is unable to maintain alignment of the odontoid atop theC2 body. Although both anterior and posterior surgical ap-proaches have been successfully used in this setting, there aremore reports describing posterior C1–C2 techniques than re-ports describing anterior operative techniques.
Atlantoaxial rotary subluxation
Fixed rotary subluxation of the atlantoaxial complex is notunique to children, but it is more common during childhood.It can present after minor trauma, in association with anupper respiratory infection, or without an identifiable incitingevent. The head is rotated to one side with the head tilted tothe other side, causing the so-called cock-robin appearance.The child is unable to turn his or her head past the midline.Attempts to move the neck are often painful. The neurologicalstatus is almost always normal (13, 31, 44, 58).
It can be difficult to differentiate atlantoaxial rotary sublux-ation from other causes of head rotation on clinical groundsalone. Several reports describe the radiographic characteriza-tion and diagnosis of this entity. Fielding and Hawkins (15)described 17 children and adults with atlantoaxial rotary sub-luxation and classified their dislocations into four types basedon radiographic features. Type I was the most common type,identified in 8 of the 17 patients. It was described as unilateralanterior rotation of the atlas pivoting around the dens with acompetent transverse ligament. Type II was identified in fivepatients. It was described as unilateral anterior subluxation ofthe atlas with the pivot being the contralateral C1–C2 facet.The atlantodens interval is increased to no more than 5 mm.Type III is described as anterior subluxation of both C1 facetswith an incompetent transverse ligament. Type IV is posteriordisplacement of C1 relative to C2 with an absent or hypoplas-tic odontoid process.
Kawabe et al. (28) reviewed the x-rays of a series of 17children with C1–C2 rotary subluxation and classified themaccording to Fielding and Hawkins (15). There were 10 Type
S94 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
I, five Type II, two Type III, and no Type IV subluxations intheir experience. CT has been used to help define the C1–C2complex in cases of suspected rotary subluxation. Kowalski etal. (31) demonstrated the superiority of dynamic computedtomographic studies compared with information obtainedwith static computed tomographic studies. They comparedthe computed tomographic scans of 8 patients with C1–C2pathology to computed tomographic studies of six normalsubjects. The computed tomographic scans obtained with nor-mal subjects maximally rotating their heads could not bedifferentiated from the computed tomographic scans of thosewith known C1–C2 rotary subluxation. When CT was per-formed with the head rotated as far as possible to the con-tralateral side, computed tomographic studies of normal sub-jects could be easily differentiated from those performed onpatients with rotary subluxation.
Type I and Type II subluxation account for most rotaryatlantoaxial subluxations in reports describing these injuries.Grogaard et al. (23) and Subach et al. (58) published retro-spective reviews on the success of conservative therapies inchildren presenting early after C1–C2 rotary subluxation.Grogaard et al. (23) described eight children who presentedwithin 5 days of subluxation, and one child who presented 8weeks after injury. All were successfully treated with closedreduction and immobilization. The child presenting late re-quired 1 week of skeletal traction to achieve reduction andwas ultimately treated with halo immobilization for 10 weeks.The children who presented early had their injuries reducedwith manual manipulation. They were treated in a hard collarfor 4 to 6 weeks. Two patients had recurrent subluxation. Bothwere reduced and treated successfully without surgical inter-vention. Subach et al. (58) reported 20 children with C1–C2rotary subluxation, in whom four injuries reduced spontane-ously. Injury reduction was accomplished in 15 of 16 patientstreated with traction for a mean duration of 4 days. Six chil-dren required fusion because of recurrent subluxation (n � 5)or irreducible subluxation (n � 1). No child experienced re-current subluxation if reduced within 21 days of symptomonset.
El-Khoury et al. (13) reported three children who presentedwithin 24 hours of traumatic rotary subluxation. All threewere successfully treated with traction or manual reductionwithin 24 hours of presentation. One child experienced recur-rent subluxation the next day that was successfully reducedmanually. External orthoses were used from 10 weeks to 4months. Phillips and Hensinger (44) reviewed 23 childrenwith C1–C2 rotary subluxation. Sixteen children were seenwithin 1 month of subluxation onset and either experiencedspontaneous reduction or were reduced with traction. Ofseven children presenting with a duration of symptoms ofmore than 1 month, one subluxation was irreducible, and fourrecurred after initial reduction. Schwarz (53) described fourchildren who presented more than 3 months after the onset ofC1–C2 rotary subluxation. Two children had irreducible sub-luxations. One child had recurrent subluxation despite the useof a Minerva cast. Only one child had successful treatmentwith closed reduction and a Minerva cast immobilization for8 weeks. These experiences highlight the ease and success of
nonsurgical management for these injuries when the sublux-ation is treated early rather than late. If the subluxation iseasily reducible and treated early, 4 weeks in a rigid collarseems to be sufficient for healing. Because C1–C2 rotary sub-luxation can reduce spontaneously in the first week, tractionor manipulation can be reserved for those subluxations thatdo not reduce spontaneously in the first few days. The use ofmore restrictive external immobilization devices (e.g., halovest, Minerva cast) for longer periods of treatment up to 4months has been described in those children presenting late,or those who have recurrent subluxations (44). Operativetreatment for C1–C2 rotary subluxations has been reserved forrecurrent subluxations or those that cannot be reduced byclosed means. Subach et al. (58) operated on 6 of the 20children they reported with rotary subluxation using theseindications. The authors used a posterior approach and ac-complished atlantoaxial fusion. There were no complications,and all fusions were successful.
In summary, the diagnosis of atlantoaxial rotary subluxa-tion is suggested when findings of a “cock-robin” appearanceare present: the patient is unable to turn the head past themidline to the contralateral side and experiences spasm of theipsilateral (the side to which the chin is turned) sternocleido-mastoid muscle (44). Plain cervical spine x-rays may revealthe lateral mass of C1 rotated anterior to the odontoid on alateral view. The anteroposterior x-ray may demonstrate ro-tation of the spinous processes toward the ipsilateral side in acompensatory motion to restore alignment. If the diagnosis ofC1–C2 rotary subluxation is not certain after clinical exami-nation and plain radiographic study, a dynamic computedtomographic study should be considered. It seems that thelonger a C1–C2 rotary subluxation is present before attemptedtreatment, the less likely reduction can be accomplished. Ifreduction is accomplished in these older injuries, it is lesslikely to be maintained. Therefore, rotary subluxations that donot reduce spontaneously should undergo attempted reduc-tion with manipulation or traction. The subsequent period ofimmobilization should be proportional to the length of timethat the subluxation was present before treatment. Surgicalarthrodesis can be considered for those with irreducible sub-luxations, recurrent subluxations, or subluxations present formore than 3 weeks duration.
Other injuries
Lui et al. (33) described nine children with ligamentousinjuries resulting in atlantoaxial dislocation. Unlike childrenwith traumatic injuries to the dens who can be managed withclosed reduction and immobilization, these children with at-lantoaxial dislocation required surgical stabilization and fu-sion. The authors attempted to treat two children with haloimmobilization for 3 months duration; both attempts failed toachieve stability. All nine children with atlantoaxial instabilityrequired operative stabilization and fusion. Finally, Rathboneet al. (46) described a series of 12 children who sustained a“spinal cord concussion” while participating in athleticevents. They found that four of these children had plain spinex-rays consistent with cervical spinal stenosis. The authors
Management of Pediatric Cervical Spinal Injuries S95
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
raised the concern that children with congenital cervical ste-nosis may be more susceptible to SCI in contact sports.
Therapeutic cervical spine immobilization
Once an injury to the pediatric cervical spine has been diag-nosed, some form of external immobilization is usually neces-sary to allow for application of traction either to restore align-ment or to immobilize the spine to allow for healing of the injury.This section will discuss the literature available concerningmethods of skeletal traction in children and various externalorthoses used to immobilize the pediatric cervical spine.
Traction for the purpose of restoring alignment or reducingneural compression in children is rarely addressed in the litera-ture. Unique concerns of cervical traction in children exist be-cause of the relatively thinner cranium with a higher likelihoodof inner cranial table penetration, lighter body weight that pro-vides less counterforce to traction, more elastic ligaments, andless well-developed musculature, increasing the potential foroverdistraction. The placement of bilateral pairs of parietal burrholes and passing 22-gauge wire through them to provide apoint of fixation for traction has been described for infants withcervical spinal injuries. Gaufin and Goodman (19) reported aseries of three infants with cervical injuries, two of whom hadinjuries reduced in this fashion. Up to 9 pounds was used in a10-week-old infant and a 16-month-old boy. They experiencedno complications with 14 and 41 days of traction, respectively.Other techniques of cervical traction application in children arenot described in the literature.
Mubarak et al. (38) described halo application in infants forthe purpose of immobilization, but not halo-ring traction.They described three infants aged 7 months, 16 months, and24 months. Ten pins were used in each child. The pins in theyoungest child were inserted to finger tightness only, whereasthe older children had 2 inches per pound of torque applied(38). The children were maintained in the halo devices for 2 to3.5 months. Only the youngest child had a minor complica-tion of frontal pin-site infection necessitating removal of twoanterior pins.
Marks et al. (36) described eight children aged 3 months to12 years who were immobilized in halo vests for 6 weeks to 12months with a mean duration of 2 months. Only three of thesechildren had cervical spinal instability. Five had thoracic spi-nal disorders. The only complication they reported was theneed to remove and replace the vest when a foreign bodybecame lodged under the vest. Dormans et al. (11) reported on37 children aged 3 to 12 years whom they managed in haloimmobilization devices. They had a 68% complication rate.Pin-site infections were most common. They arbitrarily di-vided their patient population into those younger than 10years and those 10 years or older. Purulent pin-site infectionsoccurred more commonly in the older group. Loosening ofpins occurred more commonly in the younger group. Bothloosening and infection occurred more often at the anteriorpin sites. They also reported one incident of dural penetrationand one transient supraorbital nerve injury. Baum et al. (1)compared halo use complications in children and adults. Thecomplication rates in their series were 8% for adults and 39%
for children. The complications reported for the children wereone cranial penetration and four pin-site infections. Althoughthe halo device seems to provide adequate immobilization ofthe cervical spine in children, the rate of minor complicationswith halo use is higher in children than in adults.
Gaskill and Marlin (18) described six children aged 2 yearsto 4 years who had cervical spinal instability managed with athermoplastic Minerva orthosis as an alternative to a haloimmobilization device. Two of the children they describedhad halo devices removed because of complications beforebeing placed in Minerva orthoses. The authors described noproblems with eating or with activities of daily living in thesechildren. Only one child had a minor complication from Min-erva use, a site of skin breakdown. The authors concludedthat immobilization with a thermoplastic Minerva orthosisoffered a reliable and satisfactory alternative to halo immobi-lization in young children.
Benzel et al. (2) analyzed cervical motion during spinal immo-bilization in adults serially treated with halo and Minerva de-vices. They found that the Minerva offered superior immobili-zation at all intersegmental levels of the cervical spine, with theexception of C1–C2. Although this study was carried out inadults with cervical spine instability, it underscores the useful-ness of the Minerva as a cervical immobilization device. Becausea large proportion of pediatric cervical spine injuries occur be-tween the occiput and C2, the Minerva device may not be idealfor many pediatric cervical spine injuries.
In summary, the physical properties of young skin, cranialthickness, and small body size likely contribute to the highercomplication rate among children who require traction orlong-term cervical spinal immobilization, as compared withadults. The literature includes descriptions of options avail-able for reduction and immobilization of cervical spine inju-ries in children, but it does not provide evidence for a singlebest method.
Surgical treatment
There are no reports in the literature that address the topicof early versus late surgical decompression after acute pedi-atric cervical SCI. Pediatric spinal injuries account for only 5%of all vertebral column injuries. Recent reports that describethe management of pediatric spinal injuries have been offeredby Turgut et al. (61), Finch and Barnes (16), and Elaraky et al.(12). These authors managed pediatric spinal injuries opera-tively in 17, 25, and 30% of patients, respectively. The reportby Elaraky et al. (12) in 2000 suggests that operative treatmentof pediatric cervical spine injuries is being used more fre-quently than in the past. Specific details of the operativemanagement, including timing of intervention, the approach(anterior versus posterior), and the method of internal fixationas an adjunct to fusion, are scarce in the literature. Finch andBarnes (16) used primary operative stabilization in most chil-dren they managed with ligamentous injuries of the cervicalspine. They stated that although external immobilization mayhave resulted in ligamentous healing, they elected to inter-nally fixate and fuse such injuries. They based their approachon two cases of ligamentous injuries of the cervical spine that
S96 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
they managed with external immobilization, which failed toheal and later required operative fusion. Shacked et al. (55)described six children aged 3 years to 14 years who hadcervical spine injuries that they treated surgically via an an-terior approach. They reported successful fusion with goodalignment and normal cervical spine growth in follow-up forall six children. The procedure varied (i.e., total or partialcorpectomy versus discectomy only) depending on the pa-thology. All underwent autograft fusion without instrumen-tation. The authors described severe hyperflexion injury withfracture and avulsion of the vertebral body, fracture-dislocation with disruption of the posterior elements and disc,and major anatomic deformity of the cervical spine with cordcompression as indications for an anterior approach.
Pennecot et al. (43) described 16 children with ligamentousinjuries of the cervical spine. They managed minor ligamentousinjuries (atlantodens interval of 5–7 mm, or interspinous widen-ing without dislocation or neurological deficit) with reductionand immobilization. Of 11 children with injuries below C2, 8required operative treatment with fusion via a posterior ap-proach. They used interspinous wiring techniques in younger(preschool-aged) children and posterior plates and screws inolder children as adjuncts to fusion. All had successful fusion atlast follow-up. All children were immobilized in a plaster or halocast postoperatively. Similarly, Koop et al. (30) described 13children with acute cervical spine injuries who required poste-rior arthrodesis and halo immobilization. They reported success-ful fusion in 12 patients. The single failure was associated withallograft fusion substrate. All other children had autologousgrafts. Internal fixation with wire was used in only two children.Halo immobilization was used for an average of 150 days. Theyreduced the length of postoperative halo immobilization to 100days in their most recent cases. They commented that carefultechnique allowed successful posterior fusion in children withminimal complications. Schwarz et al. (54) described 10 childrenwith traumatic cervical kyphosis. Two children who underwentanterior reconstruction with fusion had successful deformityreduction. All others managed with either external immobiliza-tion with or without traction (n � 7) or posterior fusion (n � 1)had either progression of the posttraumatic deformity or a stableunreduced kyphotic angulation.
In summary, pediatric spinal injuries are relatively infre-quent. Most pediatric spinal injuries are managed withoutoperations. Selection criteria for operative intervention in chil-dren with cervical spine injuries are difficult to glean from theliterature. Anatomic reduction of deformity, stabilization ofunstable injuries and decompression of the spinal cord, andisolated ligamentous injuries associated with deformity areindications for surgical treatment cited by various authors (16,30, 33, 43, 54, 55, 64). These reports provide Class III evidence.
SUMMARY
The available medical evidence does not allow the generationof diagnostic or treatment standards for managing pediatricpatients with cervical spine or cervical SCI. Only diagnosticguidelines and options and treatment options are supported bythis evidence. The literature suggests that obtaining neutral cer-
vical spine alignment in a child may be difficult when standardbackboards are used. The determination that a child does nothave a cervical spine injury on clinical grounds alone is sup-ported by Class II and Class III evidence. When the child is alertand communicative and is without neurological deficit, necktenderness, painful distracting injury, or intoxication, cervicalx-rays are not necessary to exclude cervical spinal injury. Whencervical spine x-rays are used to verify or rule out a cervicalspinal injury in children younger than 9 years, only lateral andanteroposterior cervical spine views need be obtained. The tra-ditional three-view x-ray assessment may increase the sensitivityof plain spine x-rays in children aged 9 years and older. Mostpediatric cervical spine injuries can be effectively treated withoutoperating. The most effective immobilization seems to be accom-plished with either halo devices or Minerva jackets. Halo immo-bilization is associated with acceptable but considerable minormorbidity in children, typically pin-site infection and pin loos-ening. The only specific pediatric cervical spine injury for whichmedical evidence supports a particular treatment paradigm is anodontoid injury in children younger than 7 years. These childrenare effectively treated with closed reduction and immobilization.Primarily ligamentous injuries of the cervical spine in childrenmay heal with external immobilization alone but are associatedwith a relatively high rate of progressive deformity when treatedwithout operating. Pharmacological therapy and intensive careunit management schemes for children with SCI have not beendescribed in the literature.
KEY ISSUES FOR FUTURE INVESTIGATION
Prospective epidemiological data may be the best source ofinformation that could lead to methods of prevention byidentifying the more common mechanisms of spinal injury inchildren. Future studies involving pediatric cervical spineinjury patients should be multi-institutional because of theinfrequency of these injuries treated at any single institution.Further defining the indications and methods for cervicalspine clearance in young children (younger than 9 yr) withprospective gathering of data would be a valuable addition tothe literature. The role of flexion/extension x-rays is poorlydefined in the literature, and a prospective evaluation of theirsensitivity and specificity for spinal column injury in specificclinical scenarios would be a valuable addition to the litera-ture. The incidence and clinical significance of complicationsof cervical spine injuries in children, such as syringomyeliaand vertebral artery injury, are unknown and could be stud-ied by prospectively gathering data in a multi-institutionalsetting.
More common injuries, such as odontoid injuries, could bestudied prospectively in a randomized fashion (e.g., closedreduction and immobilization versus anterior screw fixation),although it would be difficult from technical and feasibilitystandpoints. Prospectively collected data could also providethe basis for case-control or other comparative studies togenerate Class II evidence.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
Management of Pediatric Cervical Spinal Injuries S97
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
REFERENCES
1. Baum JA, Hanley EN Jr, Pullekines J: Comparison of halo com-plications in adults and children. Spine 14:251–252, 1989.
2. Benzel EC, Hadden TA, Saulsbery CM: A comparison of theMinerva and halo jackets for stabilization of the cervical spine.J Neurosurg 70:411–414, 1989.
3. Berne JD, Velmahos GC, El-Tawil Q, Demetriades D, Asensio JA,Murray JA, Cornwell EE, Belzberg H, Berne TV: Value of com-plete cervical helical computed tomographic scanning in identi-fying cervical spine injury in the unevaluable blunt traumapatient with multiple injuries: A prospective study. J Trauma47:896–902, 1999.
3a. Birney TJ, Hanley EN: Traumatic cervical spine injuries in child-hood and adolescence. Spine 14:1277–1282, 1989.
4. Bohn D, Armstrong D, Becker L, Humphreys R: Cervical spineinjuries in children. J Trauma 30:463–466, 1990.
5. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized trial of methylpred-nisolone or naloxone in the treatment of acute spinal cord injury:Results of the Second National Acute Spinal Cord Injury Study(NASCIS II). N Engl J Med 322:1405–1411, 1990.
6. Buhs C, Cullen M, Klein M, Farmer D: The pediatric traumaC-spine: Is the “odontoid” view necessary? J Pediatr Surg 35:994–997, 2000.
7. Cattell HS, Filtzer DL: Pseudosubluxation and other normalvariations in the cervical spine in children: A study of onehundred and sixty children. J Bone Joint Surg Am 47A:1295–1309, 1965.
8. Curran C, Dietrich AM, Bowman MJ, Ginn-Pease ME, King DR,Kosnik E: Pediatric cervical-spine immobilization: Achievingneutral position? J Trauma 39:729–732, 1995.
9. Davis PC, Reisner A, Hudgins PA, Davis WE, O’Brien MS:Spinal injuries in children: Role of MR. AJNR Am J Neuroradiol14:607–617, 1993.
10. Dickman CA, Rekate HL, Sonntag VKH, Zabramski JM: Pediat-ric spinal trauma: Vertebral column and spinal cord injuries inchildren. Pediatr Neurosci 15:237–256, 1989.
11. Dormans JP, Criscitiello AA, Drummond DS, Davidson RS:Complications in children managed with immobilization in ahalo vest. J Bone Joint Surg Am 77A:1370–1373, 1995.
11a. Dwek JR, Chung CB: Radiography of cervical spine injury inchildren: Are flexion-extension radiographs useful for acutetrauma? AJR Am J Roentgenol 174:1617–1619, 2000.
12. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VKH:Pediatric cervical spine injuries: Report of 102 cases and reviewof the literature. J Neurosurg 92[Suppl 1]:12–17, 2000.
13. El-Khoury GY, Clark CR, Gravett AW: Acute traumatic rotatoryatlanto-axial dislocation in children: A report of three cases.J Bone Joint Surg Am 66A:774–777, 1984.
14. Evans DL, Bethem D: Cervical spine injuries in children.J Pediatr Orthop 9:563–568, 1989.
15. Fielding JW, Hawkins RJ: Atlanto-axial rotary fixation: Fixedrotatory subluxation of the atlanto-axial joint. J Bone Joint SurgAm 59A:37–44, 1977.
16. Finch GD, Barnes MJ: Major cervical spine injuries in childrenand adolescents. J Pediatr Orthop 18:811–814, 1998.
17. Fotter R, Sorantin E, Schneider U, Ranner G, Fast C, Schober P:Ultrasound diagnosis of birth-related spinal cord trauma: Neo-natal diagnosis and follow-up and correlation with MRI. PediatrRadiol 24:241–244, 1994.
18. Gaskill SJ, Marlin AE: Custom fitted thermoplastic Minervajackets in the treatment of cervical spine instability in preschoolage children. Pediatr Neurosurg 16:35–39, 1990.
19. Gaufin LM, Goodman SJ: Cervical spine injuries in infants.J Neurosurg 42:179–185, 1975.
20. Givens TG, Polley KA, Smith GF, Hardin WD Jr: Pediatric cer-vical spine injury: A three-year experience. J Trauma 41:310–314,1996.
21. Godard J, Hadji M, Raul JS: Odontoid fractures in the child withneurological injury: Direct anterior osteosynthesis with a cortico-spongious screw and literature review. Childs Nerv Syst 13:105–107, 1997.
22. Griffiths SC: Fracture of the odontoid process in children.J Pediatr Surg 7:680–683, 1972.
23. Grogaard B, Dullerud R, Magnaes B: Acute torticollis in childrendue to atlanto-axial rotary fixation. Arch Orthop Trauma Surg112:185–188, 1993.
24. Hadley MN, Zabramski JM, Browner CM, Rekate H, SonntagVKH: Pediatric spinal trauma: Review of 122 cases of spinal cordand vertebral column injuries. J Neurosurg 68:18–24, 1988.
25. Hamilton MG, Myles ST: Pediatric spinal injury: Review of 174hospital admissions. J Neurosurg 77:700–704, 1992.
26. Herzenberg JE, Hensinger RN, Dedrick DK, Phillips WA: Emer-gency transport and positioning of young children who have aninjury of the cervical spine: The standard backboard may bedangerous. J Bone Joint Surg Am 71A:15–22, 1989.
27. Huerta C, Griffith R, Joyce SM: Cervical spine stabilization inpediatric patients: Evaluation of current techniques. Ann EmergMed 16:1121–1126, 1987.
28. Kawabe N, Hirotani H, Tanaka O: Pathomechanism of atlanto-axial rotary fixation in children. J Pediatr Orthop 9:569–574,1989.
29. Keiper MD, Zimmerman RA, Bilaniuk LT: MRI in the assess-ment of the supportive soft tissues of the cervical spine in acutetrauma in children. Neuroradiology 40:359–363, 1998.
30. Koop SE, Winter RB, Lonstein JE: The surgical treatment ofinstability of the upper part of the cervical spine in children andadolescents. J Bone Joint Surg Am 66A:403–411, 1984.
31. Kowalski HM, Cohen WA, Cooper P, Wisoff JH: Pitfalls in theCT diagnosis of atlantoaxial rotary subluxation. AJR Am JRoentgenol 149:595–600, 1987.
32. Laham JL, Cotcamp DH, Gibbons PA, Kahana MD, Crone KR:Isolated head injuries versus multiple trauma in pediatric pa-tients: Do the same indications for cervical spine evaluationapply? Pediatr Neurosurg 21:221–226, 1994.
33. Lui TN, Lee ST, Wong CW, Yeh YS, Tzaan WC, Chen TY, HungSY: C1–C2 fracture-dislocations in children and adolescents.J Trauma 40:408–411, 1996.
34. MacKinnon JA, Perlman M, Kirpalani H, Rehan V, Sauve R,Kovacs L: Spinal cord injury at birth: Diagnostic and prognosticdata in twenty-two patients. J Pediatr 122:431–437, 1993.
35. Mandabach M, Ruge JR, Hahn YS, McLone DG: Pediatric axisfractures: Early halo immobilization, management and outcome.Pediatr Neurosurg 19:225–232, 1993.
36. Marks DS, Roberts P, Wilton PJ, Burns LA, Thompson AG: Ahalo jacket for stabilisation of the paediatric cervical spine. ArchOrthop Trauma Surg 112:134–135, 1993.
37. Menticoglou SM, Perlman M, Manning FA: High cervical spinalcord injury in neonates delivered with forceps: Report of 15cases. Obstet Gynecol 86:589–594, 1995.
38. Mubarak SJ, Camp JF, Vuletich W, Wenger DR, Garfin SR: Haloapplication in the infant. J Pediatr Orthop 9:612–614, 1989.
S98 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
39. Nypaver M, Treloar D: Neutral cervical spine positioning inchildren. Ann Emerg Med 23:208–211, 1994.
40. Odent T, Langlais J, Glorion C, Kassis B, Bataille J, Pouliquen JC:Fractures of the odontoid process: A report of 15 cases in chil-dren younger than six years. J Pediatr Orthop 19:51–54, 1999.
41. Osenbach RK, Menezes AH: Pediatric spinal cord and vertebralcolumn injury. Neurosurgery 30:385–390, 1992.
42. Pang D, Hanley EN: Special problems of spinal stabilization inchildren, in Cooper PR (ed): Management of Posttraumatic SpinalInstability (Neurosurgical Topics No. 3). Park Ridge, AANS, 1990,pp 181–206.
43. Pennecot GF, Leonard P, Peyrot Des Gachons S, Hardy JR,Pouliquen JC: Traumatic ligamentous instability of the cervicalspine in children. J Pediatr Orthop 4:339–345, 1984.
44. Phillips WA, Hensinger RN: The management of rotatoryatlanto-axial subluxation in children. J Bone Joint Surg Am71A:664–668, 1989.
45. Ralston ME, Chung K, Barnes PD, Emans JB, Schutzman SA:Role of flexion-extension radiographs in blunt pediatric cervicalspine injury. Acad Emerg Med 8:237–245, 2001.
46. Rathbone D, Johnson G, Letts M: Spinal cord concussion inpediatric athletes. J Pediatr Orthop 12:616–620, 1992.
47. Reinges MH, Mayfrank L, Rohde V, Spetzger U, Gilsbach JM:Surgically treated traumatic synchondrotic disruption of theodontoid process in a 15-month-old girl. Childs Nerv Syst 14:85–87, 1998.
48. Rossitch E Jr, Oakes WJ: Perinatal spinal cord injury. PediatrNeurosurg 18:149–152, 1992.
49. Ruge JR, Sinson GP, McLone DG, Cerullo LJ: Pediatric spinalinjury: The very young. J Neurosurg 68:25–30, 1988.
50. Scarrow AM, Levy EI, Resnick DK, Adelson PD, Sclabassi RJ: Cervicalspine evaluation in obtunded or comatose pediatric trauma patients: Apilot study. Pediatr Neurosurg 30:169–175, 1999.
51. Schafermeyer RW, Ribbeck BM, Gaskins J, Thomason S, HarlanM, Attkisson A: Respiratory effects of spinal immobilization inchildren. Ann Emerg Med 20:1017–1019, 1991.
52. Schleehauf K, Ross SE, Civil ID, Schwab CW: Computed tomog-raphy in the initial evaluation of the cervical spine. Ann EmergMed 18:815–817, 1989.
53. Schwarz N: The fate of missed atlanto-axial rotatory subluxationin children. Arch Orthop Trauma Surg 117:288–289, 1998.
54. Schwarz N, Genelin F, Schwarz AF: Post-traumatic cervical ky-phosis in children cannot be prevented by non-operative meth-ods. Injury 25:173–175, 1994.
55. Shacked I, Ram Z, Hadani M: The anterior cervical approach fortraumatic injuries to the cervical spine in children. Clin Orthop292:144–150, 1993.
56. Shaw M, Burnett H, Wilson A, Chan O: Pseudosubluxation of C2on C3 in polytraumatized children: Prevalence and significance.Clin Radiol 54:377–380, 1999.
57. Sherk HH, Nicholson JT, Chung SM: Fractures of the odontoid processin young children. J Bone Joint Surg Am 60A:921–924, 1978.
58. Subach BR, McLaughlin MR, Albright AL, Pollack IF: Currentmanagement of pediatric atlantoaxial rotatory subluxation.Spine 23:2174–2179, 1998.
59. Swischuk LE, John SD, Hendrick EP: Is the open-mouth odon-toid view necessary in children under 5 years? Pediatr Radiol30:186–189, 2000.
60. Treloar DJ, Nypaver M: Angulation of the pediatric cervicalspine with and without cervical collar. Pediatr Emerg Care13:5–8, 1997.
61. Turgut M, Akpinar G, Akalan N, Ozcan OE: Spinal injuries inthe pediatric age group: A review of 82 cases of spinal cord andvertebral column injuries. Eur Spine J 5:148–152, 1996.
62. Viccellio P, Simon H, Pressman BD, Shah MN, Mower WR,Hoffman JR: A prospective multicenter study of cervical spineinjury in children. Pediatrics 108:E20, 2001.
63. Vogel LC: Unique management needs of pediatric spinal cordinjury patients: Etiology and pathophysiology. J Spinal CordMed 20:10–13, 1997.
64. Wang J, Vokshoor A, Kim S, Elton S, Kosnik E, Bartkowski H:Pediatric atlantoaxial instability: Management with screw fixa-tion. Pediatr Neurosurg 30:70–78, 1999.
Plate from Gautier D, Duverney M: Essai D’anatomie, en Tab-leaux Imprimés. . . . Paris, 1745. Courtesy, Dr. Irwin J. Pincus,Los Angeles, California.
Management of Pediatric Cervical Spinal Injuries S99
CHAPTER 13
Spinal Cord Injury without Radiographic Abnormality
RECOMMENDATIONSDIAGNOSIS:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options:• Plain spinal x-rays of the region of injury and computed tomographic scanning with attention to the
suspected level of neurological injury to exclude occult fractures are recommended.• Magnetic resonance imaging of the region of suspected neurological injury may provide useful diagnostic
information.• Plain x-rays of the entire spinal column may be considered.• Neither spinal angiography nor myelography is recommended in the evaluation of patients with spinal cord
injury without radiographic abnormality.
TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• External immobilization is recommended until spinal stability is confirmed by flexion/extension x-rays.• External immobilization of the spinal segment of injury for up to 12 weeks may be considered.• Avoidance of “high-risk” activities for up to 6 months after spinal cord injury without radiographic
abnormality may be considered.
PROGNOSIS:Standards: There is insufficient evidence to support prognostic standards.Guidelines: There is insufficient evidence to support prognostic guidelines.Options: Magnetic resonance imaging of the region of neurological injury may provide useful prognostic
information about neurological outcome after spinal cord injury without radiographic abnormality.
RATIONALE
Diagnosis
Pang and Wilberger defined the term SCIWORA (spinalcord injury without radiographic abnormality) in 1982 asobjective signs of myelopathy as a result of trauma with no
evidence of fracture or ligamentous instability on plain spinex-rays and tomography (12). In their 1982 article, the authorscautioned that if the early warning signs of transient symptomscould be recognized and promptly acted upon before the onsetof neurological signs, the tragic fate of some of these childrenmight be duly averted (12). Hamilton and Myles (8), Osenbachand Menezes (9), and Pang and Wilberger (12) documented thedelayed onset of SCIWORA in children as late as 4 days afterinjury. Therefore, a concern is whether a child with a normalneurological examination, but with a history of transient neuro-logical symptoms or persisting subjective neurological symp-toms referable to traumatic myelopathy, should be assigned the
diagnosis of SCIWORA and managed accordingly, despite theabsence of “objective signs of myelopathy.”
Pang and Pollack (11) recommended obtaining a computedtomographic scan focused at the neurological level of injury toexclude an occult fracture in a child with a neurological deficitreferable to the spinal cord, but without abnormalities onplain x-rays of the spine. In addition, dynamic flexion/extension x-rays or fluoroscopy has been advocated to ex-clude pathological intersegmental motion consistent with lig-amentous injury without fracture. If paraspinous musclespasm, pain, or uncooperation prevents dynamic studies, theyrecommended external immobilization until the child can flexand extend the spine for dynamic x-ray assessment. The find-ing of fracture, subluxation, or abnormal intersegmental mo-tion at the level of neurological injury excludes SCIWORA asa diagnosis. In the initial report by Pang and Wilberger (12), 1of 24 children exhibited pathological motion on initial dy-namic x-rays. By their own definition of SCIWORA, this one
S100 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
child would not be diagnosed with SCIWORA because theinitial flexion and extension x-rays were abnormal. Althoughconcern exists for the development of pathological interseg-mental motion in children with SCIWORA after normalflexion/extension studies, there has not been documentationof such instability ever developing.
Magnetic resonance imaging (MRI) findings in childrenwith SCIWORA have spanned the spectrum from normal tocomplete cord disruption, along with evidence of ligamentousand disc injury in some (3, 6). Possible roles for MRI ofchildren with SCIWORA include 1) excluding compressivelesions of the cord or roots or ligamentous disruption thatmight warrant surgical intervention, 2) guiding treatmentregarding length of external immobilization, and/or 3) deter-mining when to allow patients to return to full activity.
Treatment
Because there exists no subluxation or malalignment inSCIWORA, the mainstay of treatment has been immobiliza-tion and avoidance of activity that may either lead to exacer-bation of the present injury or increase the potential for re-current injury. Medical management issues, such as bloodpressure support and pharmacological therapy, are of concernto this population as well and have been addressed in otherguidelines. (Of note, the often-cited prospective studies ofpharmacological therapy in the treatment of acute spinal cordinjury did not include children younger than 13 yr [1].)
Pang and Pollack (11) have recommended 12 weeks of exter-nal immobilization to allow adequate time for healing of thepresumed ligamentous strain/injury and to prevent exacerba-tion of the myelopathy. It is unclear, however, what role immo-bilization plays in this population once dynamic x-rays havedisplayed no instability. The length of and even the need forimmobilization remain debatable, given the current literature. Ifthe incidence of delayed pathological intersegmental motion inchildren with SCIWORA, who have been proven to have normaldynamic x-rays, approaches zero, then the role of spinal immo-bilization for SCIWORA patients needs to be considered in lightof the available literature. If physiological motion (normal) of thespinal column can potentiate SCIWORA in these patients whenthere is no malalignment, subluxation, or lesion causing cordcompression, then immobilization is warranted.
Prognosis
SCIWORA has been shown to be associated with a highincidence of complete neurological injuries, particularly inchildren younger than 9 years. Hadley et al. (7) reported fourcomplete injuries in six children younger than 10 years withSCIWORA. The regions of complete injury tend to be cervicaland upper thoracic. Pang and Wilberger (12) found the pre-senting neurological examination to relate strongly to out-come. There are some data to suggest that MRI abnormalities(or lack of abnormalities) of the cord may be more predictive
of outcome than presenting neurological status (3, 6). Becauseno child has been documented to develop spinal instabilityafter the diagnosis of SCIWORA and the patient has, bydefinition, normal flexion/extension x-rays, there has beenlittle impetus to define predictors of instability. On the otherhand, children have been documented to have recurrentSCIWORA (13), and predictors of a “high-risk” subgroup ofchildren with SCIWORA for recurrent injury may exist.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 was un-dertaken. The medical subject headings “pediatric” and “SCI-WORA” combined with “spinal cord injury” yielded approxi-mately 145 citations. Non-English language citations weredeleted. The articles were reviewed for those that identifiedchildren who had incurred a SCIWORA. Those articles thatdescribed the clinical aspects and management of children withSCIWORA were used to generate these guidelines. Case reportswere excluded from review. Of the 15 articles meeting the selec-tion criteria, none were Class I or Class II studies. All were caseseries representing Class III data. The articles are summarized inTable 13.1. In addition, articles germane to the topic but notmeeting criteria for inclusion in the Evidentiary Tables, such asgeneral review articles, are referenced in the Scientific Founda-tion section and included in the references.
SCIENTIFIC FOUNDATION
One concern is whether the child with a normal neurologicalexamination and either a history of transient neurological deficit(e.g., paraparesis or quadriparesis) or persisting subjectivesymptoms (e.g., numbness or dysesthesias) would be a candi-date for the diagnosis of SCIWORA. Pang and Wilberger (12)described 13 of their 24 children to have a “latent” period from30 minutes to 4 days (mean, 1.2 d) before the onset of objectivesensorimotor deficits. All 13 of these children had transient sub-jective complaints at the time of their initial trauma that clearedwithin 1 hour before their neurological decline. Those who de-veloped mild neurological deficits often improved to normal,whereas those who developed severe neurological deficits wereoften left with permanent neurological dysfunction. Hamiltonand Myles (8), Osenbach and Menezes (9), and Pang and Pollack(11) also reported a 22, 23, and 27% incidence, respectively, ofdelayed onset of myelopathy within their series of children withSCIWORA. Dickman et al. (4), Eleraky et al. (5), and Hadley et al.(7) described no child having a latent period of neurologicalnormalcy after injury. The observations of delayed deteriorationby different investigators, however, raise the concern that anychild presenting with a history of transient neurological deficit orsymptoms after an appropriate mechanism of injury may beconsidered for the diagnosis of SCIWORA, despite the absenceof objective evidence of myelopathy on the initial neurologicalexamination.
Spinal Cord Injury without Radiographic Abnormality
S101Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TABLE 13.1. Summary of Reports on Spinal Cord Injury without Radiographic Abnormalitya
Series (Ref. No.) Description of StudyEvidence
ClassResults
Eleraky et al., 2000 (5) Retrospective review of 102 children withcervical spinal injuries. Young (0–9 yr) childrencompared with older children. MRI performedin 12/18 children with SCIWORA.
III SCIWORA in 18%. MRI findings did not altermanagement (external immobilization).
Turgut et al., 1996 (18) Retrospective review of 11/82 children withspinal injuries with SCIWORA.
III SCIWORA represented 13% of spinal injuries inchildren.
Grabb and Pang, 1994 (6) Retrospective review of 7 children withSCIWORA underwent MRI. Neurological statusat presentation and follow-up was correlated toMRI findings.
III No compressive lesions found. Prognosis correlated withMRI findings. Hematomyelia involving �50% of corddiameter was associated with permanent severe deficits.Lesser degrees of hematomyelia and edema only wereassociated with incomplete recovery, and normal MRIpredicted full recovery.
Davis et al., 1993 (3) Retrospective review of 15 children with spinalcord injury underwent MRI 12 h to 2 mo afterinjury. 7 children with SCIWORA.
III MRI correlated with prognosis. Hemorrhagic cordcontusions and cord “infarction” were associated withpermanent deficits. No compressive lesions in SCIWORAcases. Normal MRI was associated with no myelopathy.
Hamilton and Myles, 1992 (8) Retrospective review of 174 pediatric spinalinjuries during 14-yr period.
III SCIWORA represented 13% of spinal injuries. Ofchildren aged 0–9 yr with spinal injuries, 42% hadSCIWORA, whereas of children aged 10–14 yr, only14% had SCIWORA.
Osenbach and Menezes, 1992 (10) Retrospective review of 34/179 children withspinal injuries with SCIWORA.
III SCIWORA represented 19% of spinal injuries inchildren. Younger children (�9 yr) had higher incidenceof SCIWORA.
Rathbone et al., 1992 (15) Retrospective review of 12 children withpresumed spinal cord concussion duringathletics was investigated for the presence ofcervical stenosis.
III 3 had a Torg ratio �0.8, and 4 had a canalanteroposterior diameter �13.4 mm. MRI was not usedto evaluate for stenosis.
Rossitch and Oakes, 1992 (16) Retrospective review of 5 neonates withperinatal spinal cord injury. 4 of the 5 had noabnormality on static spinal x-rays. No flexion/extension views reported. Myelograms wereunrevealing.
III Children with perinatal spinal cord injury often havenormal x-rays. The neonates are often initiallymisdiagnosed. Respiratory insufficiency and hypotoniaare common signs.
Dickman et al., 1991 (4) Retrospective review of 26 children withSCIWORA over 19-yr period. Clinical andepidemiological features were analyzed.
III SCIWORA represents 16% of spinal injuries in children.Motor vehicle accident was most common mechanism. 7children had MRI. 5 were normal studies, 2 showed cordsignal abnormalities. Younger children tended to havemore severe injuries.
Osenbach and Menezes, 1989 (9) Retrospective review of 31 children withSCIWORA.
III 26 cervical and 5 thoracic injuries. Complete cord injuryin 12. Delayed onset of deficits in 7. No surgical lesionsfound by MRI or CT-myelography. Spinal angiogramsdone in 4 thoracic cases were normal. No delayedinstability at follow-up.
Pang and Pollack, 1989 (11) Retrospective review of 55 children withSCIWORA (43 cervical, 12 thoracic). Clinicalprofiles reported to illustrate syndrome.
III 22 “severe” injuries; 33 “mild” injuries. Age �8 yrassociated with more severe injuries. 8 cases of recurrentinjury from 3 d to 10 wk after initial injury. No recurrentinjuries with 12 wk of Guilford brace.
Hadley et al., 1988 (7) Retrospective review of 122 children withspinal injuries. Young (0–9 yr) compared witholder children.
III 17% with SCIWORA. Higher incidence of SCIWORA in0–9 yr olds versus 10–16 yr olds. 5 studied with MRI, noabnormalities detected.
Pollack et al., 1988 (13) Retrospective review of 8 children withrecurrent SCIWORA compared with 12children treated with longer immobilization.
III Recurrent SCIWORA occurred from 3 to 10 wk afterinitial injury. Recurrent injuries were more severe. Norecurrent injuries with 12 wk of Guilford brace.
Ruge et al., 1988 (17) Retrospective review comparing 0–3 yr olds to4–12 yr olds with spinal injury.
III n � 47, 21% with SCIWORA.
Pang and Wilberger, 1982 (12) Retrospective review of 24 children withSCIWORA.
III 1 child with instability on flexion/extension at 1 wk.
a SCIWORA, spinal cord injury without radiographic abnormality; MRI, magnetic resonance imaging; CT, computed tomography.
S102 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Pang and Wilberger (12) had 1 child of 24 demonstrate whatwas considered to be pathological intersegmental motion onflexion/extension x-rays 1 week after injury, after resolutionof neck pain and paraspinous muscle spasm. By definition,this child would not be considered to have had SCIWORAbecause the initial flexion/extension x-rays were abnormal.This child was treated successfully with external immobiliza-tion alone for 8 weeks. No child with SCIWORA has beendocumented in the literature to have had normal dynamicx-rays and subsequently develop intersegmental instability.
In 1994, a series of seven children with SCIWORA weredemonstrated to have ligamentous, disc, and intramedullaryabnormalities identified by MRI (6). Soft tissue findings con-sisted of anterior longitudinal ligament disruption in associ-ation with a hyperextension injury, posterior longitudinalligament disruption, and a noncompressive C2–C3 disc her-niation associated with lateral flexion, and one case of C6–C7disc abnormality associated with hyperflexion. Intramedul-lary findings included cord transection and rostral stumphemorrhage, severe hematomyelia, a minor intramedullaryhemorrhage, and edema without hemorrhage. Davis et al. (3)described seven children with SCIWORA who were exam-ined with MRI. They described no abnormalities of muscles,ligaments, or discs but did correlate the presence of intramed-ullary hemorrhage or cord “infarction” with permanent neu-rological deficit. The lack of intramedullary findings corre-lated with a normal neurological outcome. Dickman et al. (4)commented on seven children with SCIWORA who wereexamined with MRI. Five of the seven studies revealed noabnormality, and two studies documented intramedullarysignal changes. Osenbach and Menezes (9) commented intheir series of childhood SCIWORA that MRI and computedtomography-myelography performed on their patients didnot demonstrate a single compressive lesion. In addition, theyperformed spinal arteriograms in four of five children withthoracic SCIWORA and found no angiographic abnormalities.Rossitch and Oakes (16) performed myelograms on neonateswith SCIWORA from birth injury and found no abnormalitiesthat changed their management. Hadley et al. (7), before 1988,obtained MRI studies on five children with SCIWORA andidentified no abnormalities. These results need to be viewedin the context of the technology available at the time of study.
There has been no report of any situation in which the careof a child with SCIWORA has been altered by the results ofMRI and/or myelography imaging studies. No child withMRI-documented ligamentous injury and SCIWORA has de-veloped spinal instability, early or delayed. To date, there hasbeen no correlation between the ligamentous findings on MRIin SCIWORA patients and subsequent spinal instability. Theappearance of the spinal cord on MRI does provide prognos-tic information regarding ultimate neurological outcome.
Hadley et al. (7) noted a 16% incidence of multiple noncon-tiguous injuries of the spine or spinal cord in children withany type of spinal column or spinal cord injury. Ruge et al.(17) had a similar incidence (17%) of multiple levels of spinalinjury in children. Although neither of these studies dealtwith an isolated population of children with SCIWORA, theydo provide consistent observations that one in six children
with spinal trauma will have multiple levels of injury. Pangand Wilberger (12) reported 1 of 24 children with a second-level injury (L2 Chance fracture) who had a T6 neural injury(SCIWORA), but they did not obtain complete spine x-rays onevery child. Because of these observations, one should con-sider x-rays of the entire spinal column when any traumaticspinal injury, SCIWORA or otherwise, is identified in a child.
In the initial series of children with SCIWORA reported byPang and Wilberger (12), treatment routinely consisted of 4weeks of external immobilization with a “cervical collar” forcervical injuries. In cases of thoracic injury, if subsequent plainx-rays showed no abnormality after 1 week of bed rest, the childwas mobilized without a brace. In a later report, in 1989, Pangand Pollack (11) recommended 12 weeks of external immobili-zation for SCIWORA patients to allow for healing of the pre-sumed ligamentous strain/injury and to prevent exacerbation ofthe myelopathy. They also advocated external immobilizationfor this timeframe to prevent recurrent injury during the healingphase. They reported seven children who sustained recurrentSCIWORA of greater severity with lesser degrees of force whenexternal immobilization was removed before 12 weeks or theywere allowed to participate in activities against physician in-structions within 6 months of the initial injury. For these reasons,the authors recommend 12 weeks of external immobilizationand 12 additional weeks of activity restriction after SCIWORA.
Dickman et al. (4), Eleraky et al. (5), and Hadley et al. (7)reported no neurological deterioration in any patient withSCIWORA after admission or discharge. None of these re-ports described the length of time children with SCIWORAwere immobilized. It has not been routine among treatingphysicians to prescribe 12 weeks of immobilization for chil-dren with SCIWORA (2). Although a single report by Pollacket al. (13) describes recurrent SCIWORA within 12 weeks ofthe original injury, this has not been validated by other ob-servations. Because MRI evaluation was not available forthose with recurrent injury, it is not known whether certainMRI characteristics (e.g., ligamentous disruption) could pre-dict an “at risk” group for recurrent SCIWORA.
Although Pang and Wilberger (12) reported that, in theirseries, neurological outcome correlated with the presentingneurological status, others have shown that the MRI appear-ance of the spinal cord is predictive of neurological outcomein children with SCIWORA (3, 6). Absence of signal changewithin the cord is associated with an excellent outcome. Sig-nal change consistent with edema or microhemorrhages, butnot frank hematomyelia, is associated with significant im-provement of neurological function over time. The presenceof frank hematomyelia or cord disruption is associated with asevere, permanent neurological injury (3, 6). The correlationof neurological outcome with spinal cord MRI findings inSCIWORA remains consistent with the findings in muchlarger numbers of patients with spinal cord injury (non-SCIWORA) who have been studied with MRI (14).
SUMMARY
Children presenting with a history of transient neurologicalsigns or symptoms referable to traumatic myelopathy, despite
Spinal Cord Injury without Radiographic Abnormality S103
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
having normal x-rays and the absence of objective evidence ofmyelopathy, may develop SCIWORA in a delayed fashion.No child with SCIWORA has developed pathological inter-segmental motion with instability after demonstrating normalflexion/extension x-rays. MRI has not identified any lesion ina child with SCIWORA for which the management schemewould be changed by the results of the MRI examination.Similarly, no child with MRI-documented ligamentous injuryand SCIWORA has developed evidence of spinal instability.Hard collar immobilization for patients with cervical levelSCIWORA for 12 weeks and avoidance of activities that en-courage flexion/extension of the neck for an additional 12weeks has not been associated with recurrent injury. Thespinal cord findings on MRI provide prognostic informationregarding long-term neurological outcome in patients withSCIWORA. Myelography and angiography have no definedrole in the evaluation of children with SCIWORA.
KEY ISSUES FOR FUTURE INVESTIGATION
The treatment end points of spinal immobilization and activ-ity restriction for patients with SCIWORA have been arbitrarilychosen. MRI may be helpful to guide the length of time a childis immobilized and has activities restricted. The absence of evi-dence of ligamentous injury on MRI may indicate that there is noneed for external immobilization or activity restriction. It hasbeen observed that recurrent SCIWORA can occur despite a lackof evidence of spinal instability. An investigation that obtainedMRI on all children with SCIWORA and followed their clinicalstatus longitudinally might highlight the usefulness of MRI inthe management of children who go on to develop recurrentSCIWORA. The literature provides little guidance as to the like-lihood for subsequent catastrophic injury in children presentingwith SCIWORA of any severity who are found to have a preex-isting spinal or neurological abnormality, such as congenitalcervical stenosis or a Chiari malformation (15). Longitudinalclinical follow-up of SCIWORA patients of this type may pro-vide information to appropriately counsel these children. Thereare no data to elucidate the role of age in the success or failure ofvarious treatments for this condition. A longitudinal study of apatient population of reasonable size could be undertaken.
Serious attempts to address these topics cannot be forthcom-ing from a single institution or investigator because of the rela-tively small numbers of children who sustain SCIWORA annu-ally (10, 18). A multi-institutional, protocol-directed study ofSCIWORA patients may provide answers to some of the ques-tions that accompany this unique spinal cord injury subtype.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35924-3295.
REFERENCES
1. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon J,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC,Wilberger JE, Winn HR: A randomized trial of methylpred-nisolone or naloxone in the treatment of acute spinal cord injury:Results of the Second National Acute Spinal Cord Injury Study(NASCIS II). N Engl J Med 322:1405–1411, 1990.
2. Bruce D: Spinal cord injury without radiographic abnormality inchildren. Pediatr Neurosci 15:175, 1989 (comment).
3. Davis PC, Reisner A, Hudgins PA, Davis WE, O’Brien MS: Spinalinjuries in children: Role of MR. AJNR Am J Neuroradiol 14:607–617, 1993.
4. Dickman CA, Zabramski JM, Hadley MN, Rekate HL, SonntagVKH: Pediatric spinal cord injury without radiographic abnor-malities. J Spinal Disord 4:296–305, 1991.
5. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VKH:Pediatric cervical spine injuries: Report of 102 cases and review ofthe literature. J Neurosurg 92[Suppl 1]:12–17, 2000.
6. Grabb PA, Pang D: Magnetic resonance imaging in the evaluationof spinal cord injury without radiographic abnormality in chil-dren. Neurosurgery 35:406–414, 1994.
7. Hadley MN, Zabramski JM, Browner CM, Rekate H, SonntagVKH: Pediatric spinal trauma. J Neurosurg 68:18–24, 1988.
8. Hamilton MG, Myles ST: Pediatric spinal injury: Review of 174hospital admissions. J Neurosurg 77:700–704, 1992.
9. Osenbach RK, Menezes AH: Spinal cord injury without radiographicabnormality in children. Pediatr Neurosci 15:168–175, 1989.
10. Osenbach RK, Menezes AH: Pediatric spinal cord and vertebralcolumn injury. Neurosurgery 30:385–390, 1992.
11. Pang D, Pollack IF: Spinal cord injury without radiographic ab-normality in children: The SCIWORA syndrome. J Trauma 29:654–664, 1989.
12. Pang D, Wilberger JE: Spinal cord injury without radiographicabnormalities in children. J Neurosurg 57:114–129, 1982.
13. Pollack IF, Pang D, Sclabassi R: Recurrent spinal cord injurywithout radiographic abnormalities in children. J Neurosurg 69:177–182, 1988.
14. Ramon S, Dominguez R, Ramirez L, Paraira M, Olona M, CastelloT, Garcia-Fernandez L: Clinical and magnetic resonance imagingcorrelation in acute spinal cord injury. Spinal Cord 35:664–673,1997.
15. Rathbone D, Johnson G, Letts M: Spinal cord concussion in pedi-atric athletes. J Pediatr Orthop 12:616–620, 1992.
16. Rossitch E Jr, Oakes WJ: Perinatal spinal cord injury: Clinical,radiographic, and pathological features. Pediatr Neurosurg 18:149–152, 1992.
17. Ruge JR, Sinson GP, McLone DG, Cerullo LJ: Pediatric spinalinjury: The very young. J Neurosurg 68:25–30, 1988.
18. Turgut M, Akpinar G, Akalan N, Ozcan OE: Spinal injuries in thepediatric age group: A review of 82 cases of spinal cord andvertebral column injuries. Eur Spine J 5:148–152, 1996.
S104 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
CHAPTER 14
Diagnosis and Management of Traumatic Atlanto-occipitalDislocation Injuries
RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options:• A lateral cervical x-ray is recommended for the diagnosis of atlanto-occipital dislocation. If a radiological
method for measurement is used, the basion-axial interval-basion-dental interval method is recommended.• The presence of upper cervical prevertebral soft tissue swelling on an otherwise nondiagnostic plain x-ray
should prompt additional imaging.• If there is clinical suspicion of atlanto-occipital dislocation, and plain x-rays are nondiagnostic, computed
tomography or magnetic resonance imaging is recommended, particularly for the diagnosis of non-Type IIdislocations.
TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment with internal fixation and arthrodesis using one of a variety of methods is recommended.
Traction may be used in the management of patients with atlanto-occipital dislocation, but it is associatedwith a 10% risk of neurological deterioration.
RATIONALE
Although traumatic atlanto-occipital dislocation (AOD)is perceived to be an uncommon injury frequentlyresulting in death, improvements in emergency man-
agement of the patient in the field, rapid transport, and betterrecognition have resulted in more survivors of AOD in thepast 2 decades. Infrequent observation of patients with AODand missed diagnoses may impair outcomes of patients withthis unusual injury (44). An assimilation of the reported ex-periences of clinicians evaluating and managing AOD mayfacilitate development of diagnostic and treatment options forthis traumatic disorder. Specific questions that were evalu-ated include the sensitivity of plain x-rays, computed tomog-raphy (CT), and magnetic resonance imaging (MRI) in thediagnosis of AOD, as well as the safety and efficacy of varioustreatment modalities for AOD, including no treatment, trac-tion, external immobilization, and internal fixation withfusion.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “atlanto-occipital
joint” and “dislocation” yielded 690 and 86,205 citations, re-spectively. A subset of 233 citations contained both headings.The reference lists of the articles were reviewed to identifyadditional case reports. Because fewer than 100 cases of sur-vivors of AOD were identified, even single case reports wereconsidered, provided that basic inclusion criteria were met.The articles were reviewed using the following criteria forinclusion in diagnosis: human survivors, type of traumaticatlanto-occipital dislocation, and plain radiographic findings.The articles were also reviewed using the following criteriafor inclusion in treatment: human survivors, type of traumaticAOD, management, and outcome. The observations from thereports were combined because the usual methods for analy-sis were precluded by the infrequent observation of this in-jury. The type of dislocation was classified according toTraynelis et al. (51) into Type I (anterior), Type II (longitudi-nal), and Type III (posterior) dislocations. Lateral, rotational,and multidirectional dislocations that could not be classifiedinto one of these types were considered separately and arenoted as “Other Type.” The duration of follow-up rangedfrom several weeks to 4 years. Of the articles meeting thediagnostic selection criteria, 48 articles with 79 patients pro-vided data on 29 Type I, 32 Type II, 4 Type III, and 14 othertypes of AOD. Two of these articles (10, 44) included one
S105Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
patient each from two previously published individual casereports (41, 42). Of the articles meeting the treatment selectioncriteria, 43 articles with 62 patients provided data on 24 TypeI, 23 Type II, 3 Type III, and 12 other types of AOD. Two ofthese articles (10, 44) included one patient each from twopreviously published individual case reports (41, 42). All ar-ticles contained Class III medical evidence consisting of eithersingle case studies or small case series with no report contain-ing more than six patients. The information provided by thesereports was compiled and scrutinized, and it constitutes thebasis for this guideline. Summaries of these reports are pro-vided in Tables 14. 1 and 14.2.
SCIENTIFIC FOUNDATION
Diagnosis
A variety of radiographic measurements (17) have beenproposed for the diagnosis of AOD on a lateral cervical x-ray(Fig. 14.1). A displacement of more than 10 mm between thebasion and dens is considered abnormal by Wholey et al. (53).A ratio of the basion-posterior atlas arch distance to theopisthion-anterior atlas arch distance of more than 1 is con-sidered abnormal by Powers et al. (43). A distance of morethan 13 mm between the posterior mandible and anterior atlasor 20 mm between the posterior mandible and dens is con-sidered abnormal by Dublin et al. (12). Failure of a line fromthe basion to the axis spinolaminar junction to intersect C2 ora line from the opisthion to the posterior inferior corner of thebody of the axis to intersect C1 are considered abnormal byLee et al. (32). Finally, a displacement of more than �12 mmor more than �4 mm between the basion and posterior C2line, or a displacement of more than 12 mm from the basion todens (2 mm more than the Wholey recommendation) is con-sidered abnormal by Harris et al. (24, 25). A comparativestudy by Lee et al. (32) found a 50% sensitivity of the Wholeymethod, 33% sensitivity of the Powers ratio, and a 25% sen-sitivity of the Dublin method. The authors applied their X-linemethod with a 75% sensitivity (32). Although neither thePowers ratio nor X-line method could be applied in nearlyhalf their patients, a comparative study by Harris et al. (25)found a 60% sensitivity of the Powers ratio, a 20% sensitivityof the Lee method, and 100% sensitivity of the basion-axialinterval-basion-dental interval (BAI-BDI) method among pa-tients in whom the required landmarks could be identified.Przybylski et al. (44) reported failure to diagnose AOD in twoof five patients with the Powers ratio, one of five patients withthe X-line method, and two of five patients with the BAI-BDImethod. No radiographic method reviewed has complete sen-sitivity. The BAI-BDI method proposed by Harris et al. (whichincorporates the basion-dens distance described by Wholey) isat present the most reliable means to diagnose AOD on alateral cervical spine x-ray.
Many of the case reports and case series in the literature donot describe the method(s) used for diagnosis of AOD. Be-cause the most sensitive method was proposed by Harris et al.(25) in 1994, this method was probably not used for many ofthe evaluations. Although, retrospectively, a diagnosis waspossible on the first lateral x-ray in 60 of 79 patients (sensi-
tivity, 0.76), the diagnosis was actually made in only 45 of the79 patients (sensitivity, 0.57) on the first lateral x-ray. Of the 15patients whose diagnosis could have been made on the firstlateral x-ray, 3 were not stratified by type, whereas 11 of theremaining 12 were not Type II dislocations (1, 7, 9, 13, 27, 29,41, 44, 46, 49, 52, 54). A second lateral x-ray (nine cases),tomography (one case), fluoroscopy (two cases), CT (twocases), and MRI (five cases) were required for diagnosis in 19of 79 patients (1, 3, 4, 8, 10, 19–21, 26, 28, 36, 40, 42–44). Thesensitivities stratified by type of dislocation are: Type I, 0.83(24 of 29 patients); Type II, 0.72 (23 of 32 patients); Type III,0.75 (3 of 4 patients), other type, 0.71 (10 of 14 patients).Because these data were obtained from case reports and smallcase series, comparison with the accuracy of plain x-rays inpatients without AOD could not be performed. As a result,specificity, predictive values, and likelihood ratios cannot bediscerned from the available literature.
Of the 15 patients in whom the diagnosis was missed on theinitial plain x-rays, the initial neurological condition of 3patients was unknown (1). Of the remaining 12 patients, 4were neurologically normal (one Type I, one Type III, twoother type) (13, 29). Two of those four patients originallyreported as normal developed a monoparesis (one Type I, oneother type) (7, 49). Neither recovered completely. Eight of theremaining 12 patients had neurological abnormalities fromthe outset, five of whom worsened. Four of the five transientlyworsened, including one Type I injury patient with quadri-paresis and Cranial Nerve IX, X, and XII palsies (9) who wasonly spastic at last follow-up. One patient with a Type I injurydeveloped a hemiparesis but recovered (27). One Type I in-jury patient developed quadriparesis and was hemiparetic atfollow-up (46). One lateral dislocation patient with parapare-sis and torticollis had recovered at last follow-up (52). Onepatient (Type I) with initial monoparesis experienced perma-nent worsening and was quadriplegic at follow-up (54).
Although plain x-rays do not reliably identify AOD, theindex of suspicion may be increased with the identification ofprevertebral soft tissue swelling. Although plain x-rays wereobtained in all cases considered, the presence or absence ofsoft tissue swelling was described in only half (1, 3, 4, 6, 9–11,15, 19, 21, 22, 26, 27, 31, 33–36, 38, 39, 42, 45, 48, 52, 54). Thesensitivity of soft tissue swelling is 0.90 (37 of 41 cases). Acutecraniocervical CT was performed in 40 of 79 patients withAOD (1–4, 6–11, 20, 22, 26–30, 33, 34, 38, 41, 42, 44, 45, 48, 52,55). However, for 15 of 40 patients, the authors did not reportwhether AOD was diagnosed by CT. The diagnosis of AODwas made by CT in 21 of 25 patients (sensitivity, 0.84). Al-though no other computed tomographic findings were re-ported in 11 of 40 patients, 24 of the remaining 29 patientswith AOD studied with CT had hemorrhages (19 subarach-noid hemorrhages, 1 subdural hemorrhage, 4 contusions).Five patients had no computed tomographic evidence of as-sociated hemorrhage. Nine of 15 patients in whom the diag-nosis of AOD was missed on the first plain x-ray had subse-quent acute computed tomographic scanning; 8 hadsubarachnoid or other associated hemorrhage (1, 9, 8, 44).Craniocervical MRI was performed in 18 of 79 patients withAOD (6–8, 10, 15, 21, 23, 26, 27, 30, 34, 37, 40, 42, 49, 55). The
S106 Guidelines for Management of Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TABLE 14.1. Summary of Reports on Imaging Diagnosis of Atlanto-occipital Dislocation Injuriesa
Series (Ref. No.) AOD Type Diagnosis Made by X-ray Findings CT Findings MRI Findings
Grabb et al., 1999 (21) I Plain x-ray STS, Powers � Unreported Partial tear tectorial membrane
II Plain x-ray STS, Powers � None performed Tear posterior AOL
II MRI STS, Powers � None performed Partial tear tectorial membrane
Naso et al., 1997 (37) I/II Plain x-ray No mention STS Unreported Delayed study
Sponseller and Cass, 1997 (49) I Plain x-ray (missed) No mention STS None performed None performed
II Plain x-ray No mention STS Unreported BS contusion
Przybylski et al., 1996 (44) I MRI Powers/BDI/X-line � SAH, � Dx BS contusion, � Dx
II Plain x-ray (missed) Powers/BDI/X-line � SAH, � Dx BS contusion, � Dx
Pang and Wilberger, 1980 (41) II 2nd plain x-ray Powers/BDI �/X-line � SAH, � Dx None performed
I/lateral Plain x-ray (missed) Powers/BDI/X-line � Normal, head only None performed
I/lateral Plain x-ray (missed) Powers/BDI/X-line � SAH, � Dx None performed
Yamaguchi et al., 1996 (55) I Plain x-ray No mention STS SAH, � tomography BS contusion, � Dx
Guigui et al., 1995 (22) I Plain x-ray STS � Dx None performed
Ahuja et al., 1994 (1) I Fluoroscopy STS, Powers � SAH, unknown None performed
II 5 plain x-ray (3
missed)
STS, Powers � None performed None performed
II STS, Powers � SAH, � Dx None performed
II STS, Powers � SAH, unknown None performed
I/II STS, Powers � None performed None performed
I/II STS, Powers � SAH, � Dx None performed
Donahue et al., 1994 (11) I Plain x-ray STS None performed None performed
II Plain x-ray STS, 5 mm distraction None performed None performed
II Plain x-ray STS None performed None performed
II Plain x-ray 6 mm distraction Intracerebral bleed None performed
Palmer and Turney, 1994 (40) II CT No mention STS Unreported Cord contusion, � Dx
Dickman et al., 1993 (10) II Plain x-ray 15 mm distraction None performed None performed
Papadopoulos et al., 1991 (42) Rotatory CT STS � Dx None performed
Rotatory MRI STS No blood, � Dx Epidural, � Dx
II/rotatory 2nd plain x-ray STS � Dx Epidural, � Dx
Harmanli and Koyfman, 1993 (23) II Plain x-ray No mention STS None performed � Dx
Hosono et al., 1993 (27) I Plain x-ray (missed) STS Edema, head only Delayed study
Matava et al., 1993 (34) II Plain x-ray STS Delayed study None performed
II Plain x-ray No mention STS None, � Dx None performed
II Plain x-ray No mention STS SAH, � Dx BS contusion
Nischal et al., 1993 (38) II Plain x-ray STS BS contusion, � Dx None performed
II Plain x-ray STS � Dx None performed
Bundshuh et al., 1992 (6) I Plain x-ray STS SAH, � Dx SAH, � Dx
I Plain x-ray STS, Powers/X-line � SAH � Dx
Farley et al., 1992 (15) I Plain x-ray STS, Power � None performed Cord contusion
Belzberg and Tranmer, 1991 (3) II 2nd plain x-ray STS SAH, � Dx None performed
Hladky et al., 1991 (26) II MRI No mention STS Contusion, head only � Dx
II MRI No STS Normal, head only � Dx
Lee et al., 1991 (33) II Plain x-ray STS SAH, � Dx None performed
I/rotatory Plain x-ray STS � Dx None performed
Maves et al., 1991 (35) II Plain x-ray No mention STS None performed None performed
II Plain x-ray No mention STS None performed None performed
III Plain x-ray No mention STS None performed None performed
Montane et al., 1991 (36) I Plain x-ray STS None performed None performed
II 2nd plain x-ray STS None performed None performed
II 2nd plain x-ray No STS None performed None performed
Dibenedetto and Lee, 1990 (9) I Plain x-ray (missed) STS ICH, � Dx None performed
Jones et al., 1990 (30) I Plain x-ray No mention STS � Dx Premedullary edema
Colnet et al., 1989 (8) Lateral rotatory Tomography Late study SAH, � Dx Delayed study
Jevtich, 1989 (29) Lateral Plain x-ray (missed) No mention STS Delayed study None performed
Hummel and Plave, 1988 (28) I 2nd plain x-ray No mention STS Subdural, head only None performed
Zampella et al., 1988 (56) II Plain x-ray No mention STS SAH, head only Delayed study
Georgopoulos et al., 1987 (20) I Cineradiography No mention STS Delayed study None performed
Traumatic Atlanto-occipital Dislocation Injuries S107
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
MRI findings were not reported for 4 of the 18 patients stud-ied. The diagnosis of AOD could be made in 12 of 14 casesstudied with MRI (sensitivity, 0.86).
In summary, physicians often miss the diagnosis of AOD onplain x-rays (sensitivity, 0.57), particularly in the circumstanceof nonlongitudinal dislocations (non-Type II). Although im-proved interpretation may increase the sensitivity of plainx-rays to 0.76, additional imaging of the craniovertebral junc-tion with CT or MRI is recommended in patients suspected ofhaving AOD, given their superior sensitivity over plainx-rays. Other methods, such as fluoroscopy, tomography, andmyelography, have also been used to confirm the diagnosis ofAOD. Neurological abnormalities, including lower cranialnerve paresis (particularly Cranial Nerves VI, X, and XII),monoparesis, hemiparesis, quadriparesis, respiratory dys-function including apnea, and complete high cervical cordmotor deficits in the setting of normal plain spinal x-raysshould prompt additional imaging with CT or MRI. The pres-ence of prevertebral soft tissue swelling on plain x-rays andsubarachnoid hemorrhage at the craniovertebral junction oncomputed tomographic scans should prompt consideration ofthe diagnosis of AOD (5, 44).
Treatment
Ten patients in the literature did not receive initial treat-ment for AOD, nine of whom were not correctly diagnosed
until neurological worsening occurred (7, 8, 10, 20, 46, 48, 49,52, 54). Five of the nine patients had Type I injuries, and fourof the nine had other injury types. Four of nine had persistentdeficits at last follow-up that were worse in comparison withtheir examinations on presentation (7, 10, 49, 54). Two of thesepatients were normal initially. At last follow-up, one had aCranial Nerve X deficit with spasticity (Type I) (49), and onehad monoparesis (7). The other two patients had mild initialdeficits. One patient with an initial Cranial Nerve VI palsyhad hemiparesis at last follow-up (10), and another withinitial monoparesis was quadriplegic at follow-up (54). Fivepatients who worsened initially without treatment eventuallyimproved from their initial neurological condition. Finally,one quadriplegic patient with Type II AOD (56) who was nottreated improved to quadriparesis at last follow-up. In sum-mary, failure to treat AOD resulted in worsening of all pa-tients with incomplete injuries. Nearly half of these patientsfailed to improve to their initial examination baselineconditions.
Of 21 patients with AOD initially treated with traction, 2worsened transiently and developed worsening quadriparesisand Cranial Nerve VI deficits. Both patients had resolution oftheir Cranial Nerve VI deficits but not of their quadriparesiswith discontinuation of traction. One patient had a Type IIinjury (40), and one patient had a rotational dislocation (10).Four patients were initially normal and remained normal at
TABLE 14.1. Continued
Series (Ref. No.) AOD Type Diagnosis Made by X-ray Findings CT Findings MRI Findings
Bools and Rose, 1986 (4) I Plain x-ray STS SAH, � Dx None performed
III 2nd plain x-ray No mention STS None performed None performed
Collalto et al., 1986 (7) I/lateral Plain x-ray (missed) No STS SAH, head only Delayed study
Putnam et al., 1986 (45) I Plain x-ray STS, Powers � SAH, � Dx None performed
Ramsay et al., 1986 (46) I Plain x-ray (missed) No mention STS None performed None performed
Roy-Camille et al., 1986 (48) I Late plain x-ray No mention STS Delayed study None performed
I Plain x-ray STS None performed None performed
Zigler et al., 1986 (57) I Plain x-ray No mention STS None performed None performed
Watridge et al., 1985 (52) Lateral Plain x-ray (missed) No STS Delayed study None performed
Banna et al., 1983 (2) Rotatory Plain x-ray No mention STS � Dx None performed
Kaufman et al., 1982 (31) II Plain x-ray STS None performed None performed
II Plain x-ray STS None performed None performed
Woodring et al., 1981 (54) I Plain x-ray No mention STS None performed None performed
I Plain x-ray (missed) STS None performed None performed
Powers et al., 1979 (43) I Plain x-ray Late study None performed None performed
II 2nd plain x-ray No mention STS None performed None performed
Rockswold and Seljeskog, 1979 (47) II Plain x-ray No mention STS None performed None performed
Eismont and Bohlman, 1978 (13) III Plain x-ray (missed) No mention STS None performed None performed
Fruin and Pirotte, 1977 (18) I Plain x-ray No mention STS None performed None performed
Page et al., 1973 (39) I Plain x-ray STS None performed None performed
Evarts, 1970 (14) I Plain x-ray No mention STS None performed None performed
Gabrielsen and Maxwell, 1966 (19) I 2nd plain x-ray STS None performed None performed
Farthing, 1948 (16) III Plain x-ray No mention STS None performed None performed
a One patient was eliminated because the plain x-ray interpretation was not reported: Ferrara and Bartfield (17) (1 patient). Three articles wereeliminated because the type of dislocation was not reported: Georgopoulos et al. (20) (2 of 3 patients); Hladky et al. (26) (1 of 3 patients); Nasoet al. (37) (1 of 2 patients). One article (5 patients) was eliminated because individual patient data was not reported: Bulas et al. (5) (5 of 5patients). AOD, atlanto-occipital dislocation; MRI, magnetic resonance imaging; CT, computed tomography; STS, soft tissue swelling; BDI,basion-dental interval; SAH, subarachnoid hemorrhage; Dx, diagnosis. �, done or positive; �, not done or negative; BS, brainstem; ICH,intracerebral hemorrhage; AOL, atlanto-occipital ligament.
S108 Guidelines for Management of Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TABLE 14.2. Summary of Reports on Treatment of Atlanto-occipital Dislocation Injuriesa
Series (Ref. No.) Type Initial Examination Treatment Outcome
Naso et al., 1997 (37) Mixed I/II Quadriplegia Supportive Death 5 wk
Sponseller and Cass, 1997 (49) I Normal None (neuro worse), traction, fusion � brace Spastic, CN 10
II Normal Brace failed (6 wk), fusion Normal
Przybylski et al., 1996 (44) I Quadriplegia Collar � fusion Quadriplegia
Pang and Wilberger, 1980 (41) II Quadriplegia Halo failed (22 wk), fusion Quadriplegia
II Normal Fusion � collar CN 10
Mixed I/lateral Hemiplegia Collar � fusion Monoparesis
Mixed I/lateral Quadriparesis, CN 6/7/12 Fusion � collar CN 12
Yamaguchi et al., 1996 (55) I Quadriplegia, CN 10, 11, 12 Brace failed (10 wk), fusion Quadriplegia, CN 10, 11, 12
Guigui et al., 1995 (22) I Normal Fusion � brace Normal
Donahue et al., 1994 (11) I Hemiparesis Halo distracted (temporarily neuro worse),
fusion
Hyperreflexic
II CN 6 Halo � fusion Normal
II Quadriplegia, CN 7/10 Collar/traction � fusion Quadriparesis, CN 7/10
II Quadriparesis, CN 3/7 Fusion Quadriparesis
Palmer and Turney, 1994 (40) II Quadriparesis, CN 6 Traction (neuro worse), brace � fusion Quadriparesis
Dickman et al., 1993 (10) II Quadriplegia, CN 9/10 Brace Unchanged (sepsis, death at 3 mo)
Papadopoulos et al., 1991 (42) Rotatory Quadriparesis, CN 6 Traction (neuro worse), fusion � halo Quadreparesis
Rotatory CN 6 None (neuro worse), fusion � halo Hemiparesis
Mixed II/rotatory Hemiparesis, CN 3/6 Halo � fusion Normal
Harmanli and Koyfman, 1993 (23) II Hemiparesis, CN 12 Fusion � brace Normal
Hosono et al., 1993 (27) I Hemiparesis Brace (neuro worse), fusion � brace Normal
Matava et al., 1993 (34) II Hemiplegia, CN 6/12 Fusion � brace Spastic, CN 6
II Hemiparesis, CN 6 Fusion � brace Normal
II CN 6/9/10 Fusion � brace Spastic
Nischal et al., 1993 (38) II Quadriparesis, CN 3, 6, 9, 10 Brace � fusion Hemiparesis, CN 3, 6, 9, 10
II Quadriplegia, CN 9, 10 Brace � fusion Hemiparesis
Bundshuh et al., 1992 (6) I Quadriparesis, CN 6, 9, 10, 12 Traction � fusion CN 6, 12
Farley et al., 1992 (15) I Quadriplegia, CN 10 Traction � brace Quadriplegia
Belzberg and Tranmer, 1991 (3) II Quadriparesis, CN 6, 9, 10 Traction � brace � fusion Monoparesis, CN 6
Lee et al., 1991 (33) II Normal Traction � fusion Normal
Mixed I/rotatory CN 6 Brace � fusion CN 6
Montane et al., 1991 (36) I Hemiparesis Fusion � brace Spastic
II Quadriparesis Traction, fusion � brace Normal
II Quadriplegia Fusion � brace Quadriplegia
Dibenedetto and Lee, 1990 (9) I Quadriparesis, CN 9, 10, 12 Collar (neuro worse, 6 wk), fusion � brace Spastic
Colnet et al., 1989 (8) Mixed lateral/rotatory Hemiplegia, CN 6, 9, 10 None (neuro worse), traction � shunt �
decompression
Hemiparesis
Jevtich, 1989 (29) Lateral Normal Traction � brace Normal
Hummel and Plaue, 1988 (28) I Hemiparesis Fusion � brace Normal
Zampella et al., 1988 (56) II Quadriplegia, CN 5–12 None Quadriplegia, CN 6
Georgopoulos et al., 1987 (20) I Normal None (neuro worse), fusion � brace Normal
Bools and Rose, 1986 (4) III Normal Traction, fusion � brace Normal
Collalto et al., 1986 (7) Mixed I/lateral Normal None (neuro worse), fusion � brace Monoparesis
Putnam et al., 1986 (45) I Quadriplegia, CN 6 Brace Death (sepsis 8 mo)
Ramsay et al., 1986 (46) I Quadriparesis None (neuro worse), traction � brace Hemiplegia
Roy-Camille et al., 1986 (48) I CN 6, 11 None, brace failed (3 mo), traction � fusion CN 6
I Quadriplegia, CN 6, 9–12 Traction � fusion Quadriplegia
Zigler et al., 1986 (57) I Quadriplegia, CN 11 Traction � brace � fusion Quadriplegia
Watridge et al., 1985 (52) Lateral Paraparesis None (neuro worse), traction � fusion �
decompression � brace
Normal
Banna et al., 1983 (2) Rotatory Normal Traction (2 wk) Normal
Kaufman et al., 1982 (31) II Quadriplegia Brace � fusion Quadriparesis, CN 9, 10
II Monoparesis Brace Normal
Traumatic Atlanto-occipital Dislocation Injuries S109
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
follow-up (2, 4, 16, 33). The remaining 15 patients had im-proved neurological function compared with their initial find-ings at last follow-up (3, 6, 14, 15, 18, 19, 36, 39, 43, 47, 48, 54,57). Ten had Type I injuries, five had Type II injuries, two hadType III injuries, and two had other dislocations. In total, oneof six patients with Type II injuries and one of three patientswith other translational injuries had transient worsening withthe use of craniocervical traction. In summary, traction forAOD has been reported to cause occasional neurologicalworsening. In both circumstances, the worsening did notpersist after discontinuation of traction. Because the fre-quency of neurological worsening with traction for AOD isapproximately 10%, 10 times higher than for subaxial injuries,the use of traction should be considered with caution inpatients with AOD.
Of 19 patients initially treated with external immobilizationexcluding traction, 8 patients were immobilized in anticipa-tion of internal fixation and fusion and none worsened duringthe presurgical interval (one Type I, four Type II, three othertype) (10, 11, 31, 33, 38, 44). Of the remaining 11 patientstreated with external immobilization alone excluding traction,4 worsened transiently (three Type I, one Type II) (9, 11, 13,27). All subsequently underwent craniocervical fixation andfusion. Two of these patients were normal at follow-up (oneinitially normal, one initially hemiparetic), and two werespastic (one initially quadriparetic and one hemiparetic). Ofthe remaining seven patients managed with external immo-bilization alone who did not worsen while in external immo-bilization, two patients managed in collars and one patienttreated in a halo were unstable after 6 to 22 weeks of immo-bilization (one Type I, two Type II). Two were quadriplegic,and one was normal. All three underwent internal fixationand fusion without change in their initial neurological condi-tion at last follow-up. Only four patients with AOD weresuccessfully treated with external immobilization alone (oneType I, two Type II, one other dislocation). Of the 21 patientsinitially treated with traction, 6 were subsequently managedwith external immobilization and none developed neurolog-
ical worsening. Two of the six (both Type I) remained unsta-ble after 3 to 5 months of bracing and were subsequentlytreated with craniocervical fixation and fusion. Five of thosesix patients had improvement in their neurological conditionat follow-up. The sixth patient remained normal.
In summary, 5 of 13 patients with AOD who did not worsenneurologically while treated with external immobilization(with or without traction) failed to achieve bony union withstability without internal fixation and fusion. In addition, sixpatients transiently worsened with external immobilization(with or without initial traction). Factors affecting fusion orpersistent nonunion (e.g., degree and type of displacement,patient age, and association with occipital condyle fractures)could not be identified. Because 11 (28%) of 40 patients man-aged with external immobilization either deteriorated neuro-logically or failed to achieve craniocervical stability withoutsurgical internal fixation and fusion, treatment of AOD withexternal immobilization alone should be considered withcaution.
Finally, 19 patients in the literature were treated withplanned early craniocervical fusion with internal fixation.Only one patient worsened neurologically after surgery. Thispatient with a Type II injury was normal initially and devel-oped a Cranial Nerve X deficit that persisted at follow-up (44).All but 3 of the remaining 18 patients improved neurologi-cally at follow-up. Four had Type I, 10 had Type II, and 4 hadother types of dislocation. None of the patients treated withcraniocervical fusion and internal fixation had late instabilityrequiring reoperation or further treatment.
SUMMARY
AOD is an uncommon traumatic injury that is difficult todiagnose and is frequently missed on initial lateral cervicalx-rays. Patients who survive often have neurological impair-ment, including lower cranial neuropathies, unilateral or bi-lateral weakness, or quadriplegia. But nearly 20% of patients
TABLE 14.2. Continued
Series (Ref. No.) Type Initial Examination Treatment Outcome
Woodring et al., 1981 (54) I Hemiparesis, CN 6 Traction CN 6
I Monoparesis None (neuro worse), traction � fusion Quadriplegia
Powers et al., 1979 (43) I Hemiparesis, CN 6 Traction � brace Hemiparesis
II Hemiparesis, CN 7 Traction � brace Normal
Rockswold and Seljeskog, 1979 (47) II Hemiparesis, CN 6 Traction, brace � fusion Ambulates
Eismont and Bohlman, 1978 (13) III Normal Collar (neuro worse), fusion � brace Normal
Fruin and Pirotte, 1977 (18) I Hemiparesis, CN 6, 9–12 Traction � fusion CN 6, 11
Page et al., 1973 (39) I Quadriplegia, CN 10, 12 Traction, brace failed (5 mo), fusion Quadriparesis, CN 10
Evarts, 1970 (14) I Hemiparesis, CN 6, 9, 10, 12 Traction, brace � fusion CN 6
Gabrielsen and Maxwell, 1966 (19) I Hyperreflexic, CN 6 Traction, brace failed (3 mo), fusion Numb scalp
Farthing, 1948 (16) III Normal Traction � brace Normal
a Three articles were eliminated because the type of dislocation was not reported: Georgopoulos et al. (20) (2 of 3 patients); Bulas et al. (5)(5 of 5 patients); Naso et al. (37) (1 of 2 patients). Two articles (8 patients) were eliminated because the initial examination was not reported:Grabb et al. (21) (3 patients); Ahuja et al. (1) (5 patients). Two articles (6 patients) were eliminated because the treatment was not reported: Maveset al. (35) (3 patients); Hladky et al. (26) (3 patients). Two articles were eliminated because the outcome was not reported: Jones et al. (30) (1patient); Bools and Rose (4) (1 of 2 patients). CN, cranial nerve; neuro, neurological examination.
S110 Guidelines for Management of Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
with acute traumatic AOD will have a normal neurologicalexamination on presentation. The lack of localizing featuresmay impede diagnosis in the patient with a normal cervicalx-ray. A high index of suspicion must be maintained to diag-nose AOD. Prevertebral soft tissue swelling on a lateral cer-vical x-ray or craniocervical subarachnoid hemorrhage onaxial CT has been associated with AOD and may promptconsideration of the diagnosis. Additional imaging, includingCT and MRI, may be required to confirm the diagnosis ofAOD if plain x-rays are inadequate. All patients with AODshould be treated. Without treatment, nearly all patients de-veloped neurological worsening, and some did not recover.Although treatment with traction and external immobiliza-tion has been used successfully in some patients, transient orpermanent neurological worsening and late instability havebeen reported more often in association with these treatmentsthan with surgical treatment. Consequently, craniocervicalfusion with internal fixation is recommended for the treat-ment of patients with acute traumatic AOD.
KEY ISSUES FOR FUTURE INVESTIGATION
Although the use of external immobilization for AOD wasoften associated with late instability, several patients achievedstability without operative management. CT with three-dimensional reconstruction for more precise measurement of
the magnitude of displacement and MRI for differentiation ofpartial and complete ligament tears from stretch injuries maybe useful in identifying a subgroup of patients in whomstability might be achieved with external immobilizationalone. Because AOD remains relatively infrequent, coopera-tive prospective collection of plain radiographic, computedtomographic, and MRI data in patients with AOD is recom-mended to determine whether a subgroup of patients withAOD can be treated with external immobilization alone withfewer occurrences of late instability.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, Division of Neuro-logical Surgery, 516 Medical Education Building, 1813 6th AvenueSouth, Birmingham, AL 35294-3295.
REFERENCES
1. Ahuja A, Glasauer FE, Alker GJ Jr, Klein DM: Radiology insurvivors of traumatic atlanto-occipital dislocation. Surg Neurol41:112–118, 1994.
2. Banna M, Stevenson GW, Tumiel A: Unilateral atlanto-occipitaldislocation complicating an anomaly of the atlas. J Bone JointSurg Am 65A:685–687, 1983.
3. Belzberg AJ, Tranmer BI: Stabilization of traumatic atlanto-occipital dislocation. J Neurosurg 75:478–482, 1991.
FIGURE 14.1. Midsagittal diagrams of thecraniocervical junction show the variousmethods for identifying AOD on a lateralcervical x-ray. A, the Wholey measure; B,the Powers ratio; C, the Dublin measure; D,the X-line method; E, the BAI-BDI method.
Traumatic Atlanto-occipital Dislocation Injuries S111
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
4. Bools JC, Rose BS: Traumatic atlantooccipital dislocation: Twocases with survival. AJNR Am J Neuroradiol 7:901–904, 1986.
5. Bulas DI, Fitz CR, Johnson DL: Traumatic atlanto-occipital dislo-cation in children. Radiology 188:155–158, 1993.
6. Bundschuh CV, Alley JB, Ross M, Porter IS, Gudeman SK: Mag-netic resonance imaging of suspected atlanto-occipital disloca-tion: Two case reports. Spine 17:245–248, 1992.
7. Collalto PM, DeMuth WW, Schwentker EP, Boal DK: Traumaticatlanto-occipital dislocation: Case report. J Bone Joint Surg Am68A:1106–1109, 1986.
8. Colnet G, Chabannes J, Commun C, Rigal MC, Alassaf M:Atlanto-occipital luxation and syringomyelia: 2 rare complica-tions of cervical injury—Diagnosis and therapeutic effects: Apro-pos of a case [in French]. Neurochirurgie 35:58–63, 1989.
9. Dibenedetto T, Lee CK: Traumatic atlanto-occipital instability: Acase report with follow-up and a new diagnostic technique. Spine595–597, 1990.
10. Dickman CA, Papadopoulos SM, Sonntag VKH, Spetzler RF,Rekate HL, Drabier J: Traumatic occipitoatlantal dislocations.J Spinal Disord 6:300–313, 1993.
11. Donahue DJ, Muhlbauer MS, Kaufman RA, Warner WC, SanfordRA: Childhood survival of atlantooccipital dislocation:Underdiagnosis, recognition, treatment and review of the litera-ture. Pediatr Neurosurg 21:105–111, 1994.
12. Dublin AB, Marks WM, Weinstock D, Newton TH: Traumaticdislocation of the atlanto-occipital articulation (AOA) with short-term survival: With a radiographic method of measuring theAOA. J Neurosurg 52:541–546, 1980.
13. Eismont FJ, Bohlman HH: Posterior atlanto-occipital dislocationwith fractures of the atlas and odontoid process: Report of a casewith survival. J Bone Joint Surg Am 60A:397–399, 1978.
14. Evarts CM: Traumatic occipito-atlantal dislocation. J Bone JointSurg Am 52A:1653–1660, 1970.
15. Farley FA, Graziano GP, Hensinger RN: Traumatic atlanto-occipital dislocation in a child. Spine 17:1539–1541, 1992.
16. Farthing JW: Atlantocranial dislocation with survival: A casereport. N C Med J 9:34–36, 1948.
17. Ferrera PC, Bartfield JM: Traumatic atlanto-occipital dislocation:A potentially survivable injury. Am J Emerg Med 14:291–296,1996.
18. Fruin AH, Pirotte TP: Traumatic atlantooccipital dislocation: Casereport. J Neurosurg 46:663–665, 1977.
19. Gabrielsen TO, Maxwell JA: Traumatic atlanto-occipital disloca-tion: With case report of a patient who survived. Am JRoentgenol Radium Ther Nucl Med 97:624–629, 1966.
20. Georgopoulos G, Pizzutillo PD, Lee MS: Occipito-atlantal insta-bility in children: A report of five cases and review of the litera-ture. J Bone Joint Surg Am 69A:429–436, 1987.
21. Grabb BC, Frye TA, Hedlund GL, Vaid YN, Grabb PA, Royal SA:MRI diagnosis of suspected atlanto-occipital dissociation in child-hood. Pediatr Radiol 29:275–281, 1999.
22. Guigui P, Milaire M, Morvan G, Lassale B, Deburge A: Traumaticatlantooccipital dislocation with survival: Case report and reviewof the literature. Eur Spine J 4:242–247, 1995.
23. Harmanli O, Koyfman Y: Traumatic atlanto-occipital dislocationwith survival: A case report and review of the literature. SurgNeurol 39:324–330, 1993.
24. Harris JH, Carson GC, Wagner LK: Radiologic diagnosis of trau-matic occipitovertebral dissociation: Part 1—Normal oc-cipitovertebral relationships on lateral radiographs of supine sub-jects. Am J Radiol 162:881–886, 1994.
25. Harris JH Jr, Carson GC, Wagner LK, Kerr N: Radiologic diagno-sis of traumatic occipitovertebral dissociation: Part 2—Compari-son of three methods of detecting occipitovertebral relationshipson lateral radiographs of supine subjects. Am J Radiol 162:887–892, 1994.
26. Hladky JP, Lejeune JP, Leclercq F, Dhellemmes P, Christiaens JL: Trau-matic atlanto-occipital dislocation [in French]. Neurochirurgie 37:312–317, 1991.
27. Hosono N, Yonenobu K, Kawagoe K, Hirayama N, Ono K: Trau-matic anterior atlanto-occipital dislocation: A case report withsurvival. Spine 18:786–790, 1993.
28. Hummel A, Plaue R: Diagnosis and treatment of atlanto-occipitalruptures [in German]. Unfallchirurgie 14:311–319, 1988.
29. Jevtich V: Traumatic lateral atlanto-occipital dislocation withspontaneous bony fusion: A case report. Spine 14:123–124, 1989.
30. Jones DN, Knox AM, Sage MR: Traumatic avulsion fracture of theoccipital condyles and clivus with associated unilateral atlantooc-cipital distraction. AJNR Am J Neuroradiol 11:1181–1183, 1990.
31. Kaufman RA, Dunbar JS, Botsford JA, McLaurin RL: Traumaticlongitudinal atlanto-occipital distraction injuries in children.AJNR Am J Neuroradiol 3:415–419, 1982.
32. Lee C, Woodring JH, Goldstein SJ, Daniel TL, Young AB, TibbsPA: Evaluation of traumatic atlantooccipital dislocations. AJNRAm J Neuroradiol 8:19–26, 1987.
33. Lee C, Woodring JH, Walsh JW: Carotid and vertebral injury insurvivors of atlanto-occipital dislocation: Case reports and liter-ature review. J Trauma 31:401–407, 1991.
34. Matava MJ, Whitesides TE Jr, Davis PC: Traumatic atlanto-occipital dislocation with survival: Serial computerized tomogra-phy as an aid to diagnosis and reduction—A report of three cases.Spine 18:1897–1903, 1993.
35. Maves CK, Souza A, Prenger EC, Kirks DR: Traumatic atlanto-occipital disruption in children. Pediatr Radiol 21:504–507, 1991.
36. Montane I, Eismont FJ, Green BA: Traumatic occipitoatlantal dis-location. Spine 16:112–116, 1991.
37. Naso WB, Cure J, Cuddy BG: Retropharyngealpseudomeningocele after atlanto-occipital dislocation: Report oftwo cases. Neurosurgery 40:1288–1291, 1997.
38. Nischal K, Chumas P, Sparrow O: Prolonged survival afteratlanto-occipital dislocation: Two case reports and review. Br JNeurosurg 7:677–682, 1993.
39. Page CP, Story JL, Wissinger JP, Branch CL: Traumatic atlantooc-cipital dislocation: Case report. J Neurosurg 39:394–397, 1973.
40. Palmer MT, Turney SZ: Tracheal rupture and atlanto-occipitaldislocation: Case report. J Trauma 37:314–317, 1994.
41. Pang D, Wilberger JE Jr: Traumatic atlanto-occipital dislocationwith survival: Case report and review. Neurosurgery 7:503–508,1980.
42. Papadopoulos SM, Dickman CA, Sonntag VKH, Rekate HL,Spetzler RF: Traumatic atlantooccipital dislocation with survival.Neurosurgery 28:574–579, 1991.
43. Powers B, Miller MD, Kramer RS, Martinez S, Gehweiler JA Jr:Traumatic anterior atlanto-occipital dislocation. Neurosurgery4:12–17, 1979.
44. Przybylski GJ, Clyde BL, Fitz CR: Craniocervical junction sub-arachnoid hemorrhage associated with atlanto-occipital disloca-tion. Spine 21:1761–1768, 1996.
45. Putnam WE, Stratton FT, Rohr RJ, Stitzell W, Roat G: Traumaticatlanto-occipital dislocations: Value of the Powers ratio in diag-nosis. J Am Osteopath Assoc 86:798–804, 1986.
46. Ramsay AH, Waxman BP, O’Brien JF: A case of traumatic atlanto-occipital dislocation with survival. Injury 17:412–413, 1986.
S112 Guidelines for Management of Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
47. Rockswold GL, Seljeskog EL: Traumatic atlantocranial dislocationwith survival. Minn Med 62:151–152, 1979.
48. Roy-Camille R, Benazet JP, Saillant G, Henry P, Mamoudy P,Leonard P: Traumatic atlanto-occipital luxation: Value of newradiologic signs (apropos of 2 cases) [in French]. Rev Chir OrthopReparatrice Appar Mot 72:303–309, 1986.
49. Sponseller PD, Cass JR: Atlanto-occipital fusion for dislocation inchildren with neurologic preservation: A case report. Spine 22:344–347, 1997.
50. Deleted in proof.51. Traynelis VC, Marano GD, Dunker RO, Kaufman HH: Traumatic
atlanto-occipital dislocation: Case report. J Neurosurg 65:863–870, 1986.52. Watridge CB, Orrison WW, Arnold H, Woods GA: Lateral atlanto-
occipital dislocation: Case report. Neurosurgery 17:345–347, 1985.
53. Wholey MH, Bruwer AJ, Baker HL: The lateral roentgenogram ofthe neck (with comments on the atlanto-odontoid-basion relation-ship). Radiology 71:350–356, 1958.
54. Woodring JH, Selke AC Jr, Duff DE: Traumatic atlantooccipitaldislocation with survival. AJR Am J Radiol 137:21–24, 1981.
55. Yamaguchi N, Ikeda K, Ishise J, Yamashita J: Traumatic atlanto-occipital dislocation with long-term survival. Neurol Med Chir(Tokyo) 36:36–39, 1996.
56. Zampella EJ, Duvall ER, Langford KH: Computed tomographyand magnetic resonance imaging in traumatic locked-in syn-drome. Neurosurgery 22:591–593, 1988.
57. Zigler JE, Waters RL, Nelson RW, Capen DA, Perry J: Occipito-cervico-thoracic spine fusion in a patient with occipito-cervicaldislocation and survival. Spine 11:645–646, 1986.
Drawings by Leonardo da Vinci. Courtesy, Dr. Edwin Todd, Pasadena, California.
Traumatic Atlanto-occipital Dislocation Injuries S113
CHAPTER 15
Occipital Condyle Fractures
RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: Computed tomographic imaging is recommended for establishing the diagnosis of occipital
condyle fractures. Clinical suspicion should be raised by the presence of one or more of the followingcriteria: blunt trauma patients sustaining high-energy craniocervical injuries, altered consciousness, occip-ital pain or tenderness, impaired cervical motion, lower cranial nerve paresis, or retropharyngeal soft tissueswelling.
Options: Magnetic resonance imaging is recommended to assess the integrity of the craniocervical ligaments.
TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Treatment with external cervical immobilization is recommended.
RATIONALE
Although the traumatic occipital condyle fracture (OCF)was first described by Bell (2) in 1817, more frequentobservation of this injury has only been reported dur-
ing the past 2 decades. Improvements in computed tomo-graphic imaging technology and use of computed tomo-graphic imaging of the head-injured patient that includes thecraniovertebral junction have resulted in more frequent rec-ognition of this injury. However, the overall infrequent occur-rence of OCF and missed diagnoses in patients with OCF mayresult in late neurological deficits in these patients. An anal-ysis of the reported cases of OCF may facilitate developmentof diagnostic and treatment recommendations for this disor-der and is undertaken in this chapter. Specific questions thatwere addressed include the accuracy of plain x-rays andcomputed tomography (CT) in the diagnosis of OCF and thesafety and efficacy of various treatment modalities, includingno treatment, traction, external immobilization, decompres-sion, and internal fixation with fusion.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “occipital bone”and “fracture” (spinal, cranial, or fracture alone) yielded 1,830and 33,537 citations, respectively. A subset of 218 citationscontained both headings. The references of the articles werereviewed to identify additional case reports. The articleswere reviewed using the following criteria for inclusion indiagnosis: human survivors, type of fracture, and tomo-graphic or plain radiographic findings. The articles were also
reviewed using the following criteria for inclusion in treat-ment: human survivors, type of fracture, management, andoutcome. Because fewer than 100 cases of survivors wereidentified, even single case reports were considered, providedthat basic inclusion criteria were met. The observations fromthe reports were combined because the usual methods foranalysis were precluded by the infrequent occurrence of thisinjury. Forty-seven articles met the selection criteria, provid-ing data on a total of 91 patients. All but two articles containedClass III data of either single case studies or small case series;none contained more than 15 patients. The two exceptionswere prospective studies to evaluate the use of clinical criteriain blunt trauma patients to prompt computed tomographicimaging of the cranial base (5, 26). The duration of follow-upin all articles ranged from several weeks to 5 years. The dataprovided by these reports was compiled and constitute thebasis for this guideline. Summaries of the 43 articles mostgermane to this topic are provided in Table 15.1.
SCIENTIFIC FOUNDATION
Diagnosis
Plain x-rays of the cervical spine were obtained in nearly all91 patients culled from the literature review. Normal imagingwas reported in 42 patients. Eight patients had prevertebralsoft tissue swelling, only four of whom did not have associ-ated cervical fractures (5, 17, 19, 28, 30, 41). Ten patients withcervical fractures or displacements were described withoutmention of the presence or absence of soft tissue swelling (3,4, 16, 25, 31, 35, 40, 41). Three patients had multiple cervicalfractures (5, 31). Associated fractures included four atlas, two
S114 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Type II odontoid, three axis, two C5 fractures, one each of C3,C6, and C7 fractures, and one unspecified cervical fracture.One patient had atlantoaxial widening, and one had C5–C6subluxation (16, 41). Only two patients were identified withOCF on plain x-rays of the cranium or cervical spine (21, 44).The results of plain x-rays were not reported in 28 patients,and plain x-rays were not obtained in one patient with an oldhealed fracture identified by CT (12). The calculated sensitiv-ity of plain x-rays from these reports in the diagnosis of OCFis 3.2% (2 of 62 patients). Because the data were obtained fromcase reports and small case series of patients known to haveOCF, comparison with the findings of plain x-rays in patientswithout OCF could not be performed. As a result, specificity,positive predictive value, and negative predictive value couldnot be determined.
The type of fracture was classified according to Andersonand Montesano (1) into Type I (comminuted from impact),Type II (extension of a linear basilar cranial fracture) (20), andType III (avulsion of a fragment) fractures (Fig. 15.1). The 91patients in this review population provided data on 85 uni-lateral fractures (12 Type I, 24 Type II, and 49 Type III), 4bilateral fractures (one Type I, two Type III, and one mixedType I and Type III), and 2 old fractures.
All but one patient (44) underwent tomographic imaging (6polytomography alone, 83 computed tomographic imagingalone, and 1 both). One OCF was missed with polytomogra-phy and subsequently identified on CT (33). Two patients hadOCF diagnosed from retrospective review of computed tomo-graphic images that were initially interpreted as normal (12).The diagnosis of OCF could be made in every patient withOCF. Bloom et al. (5) performed a prospective study during a1-year period to identify the frequency of OCF in patientsmeeting certain clinical criteria. Fifty-five consecutive patientswith high-energy blunt craniocervical trauma underwentthin-section craniocervical junction computed tomographicimaging. Supplemental criteria included reduced GlasgowComa Score at admission, occipitocervical tenderness, re-duced craniocervical motion, lower cranial nerve abnormal-ity, and retropharyngeal soft tissue swelling. Nine (16.4%) of55 patients were identified with OCF. Other reports haveestimated a 1 to 3% frequency of OCF in patients sustainingblunt craniocervical trauma (24, 31). Similarly, Link et al. (26)reported the results of craniocervical CT on 202 patients witha Glasgow Coma Score between 3 and 6. OCF was identifiedin 9 (4.4%) of 202 patients.
Loss of consciousness was observed in 36 of 44 patients.Among 64 patients who had a sufficiently detailed neurolog-ical examination reported, 25 were normal, 24 had acute ordelayed cranial nerve deficits alone, 6 had cranial nerve def-icits with limb weakness, 6 had mild to severe limb weaknesswithout cranial nerve deficits, 1 had a delayed onset of ver-tigo, 1 had hyperreflexia, and 1 had diplopia. Only 4 patientswere found who did not have occipitocervical pain in theabsence of significantly impaired consciousness (28, 32, 41).
One patient was intoxicated, one had severe extremity pain,and two had severe facial trauma.
Only 11 patients were investigated with magnetic reso-nance imaging (MRI) (3, 12, 13, 15, 19, 21–23, 41, 42). Earlycraniocervical MRI was performed in eight patients, and lateMRI studies were obtained in three patients. Cervicomedul-lary hemorrhages were seen in three patients, two had normalimaging, one had a retrodental hemorrhage, one had a torntectorial membrane, and one had demonstration of the frac-ture. Displaced fracture fragments were observed in all threepatients with delayed MRI. Although early MRI has beeninfrequently reported after OCF, Tuli et al. (41) proposed anew classification scheme using MRI to differentiate stablefrom unstable OCF. However, the case example they gavedemonstrated concurrent atlantoaxial instability thatprompted occipitocervical fusion (rather than atlanto-occipital instability and OCF fracture).
In summary, the diagnosis of OCF is rarely made on plainx-rays. Imaging of the craniovertebral junction with CT orother tomographic methods is recommended in patients sus-pected of having OCF. Blunt trauma patients sustaining high-energy craniocervical injuries may be more likely to sustainOCF. Consequently, cranial imaging should include evalua-tion of the craniocervical junction. Other clinical criteria, in-cluding altered consciousness, occipital pain or tenderness,impaired craniocervical motion, lower cranial nerve paresis,or retropharyngeal soft tissue swelling, should prompt com-puted tomographic imaging of the craniocervical junction.
Treatment
Twenty-three patients (2 Type I, 14 Type II, 5 Type III, 2unknown type) did not receive treatment (6, 9, 12, 13, 18, 31,33, 34, 41, 42, 44, 45). Nine of these patients (one Type I, fourType II, four Type III) developed cranial nerve deficits withindays to weeks after injury (6, 9, 12, 13, 31, 33, 34, 42, 45). Onehypoglossal nerve palsy resolved, two hypoglossal nerve def-icits improved, three other cranial nerve deficits persisted(two hypoglossal, one glossopharyngeal, and one vagal), andthree outcomes were not reported. Six additional patientsdeveloped delayed deficits or symptoms. Two initially un-treated patients (one Type II, one Type III) developed multi-ple lower cranial nerve deficits that improved after 6 weeks ofcervical immobilization (23). Another initially untreated pa-tient (Type III) developed vertigo after 3 months that resolvedafter 8 weeks of collar immobilization (7). One patient (TypeIII) developed nystagmus and a lateral rectus palsy afterprecautionary collar immobilization was discontinued. Thedeficit resolved after cervical immobilization was resumed(14). One patient (Type III) developed double vision duringcervical traction that resolved with surgical decompression(45). Finally, one patient (Type III) developed delayed vagal,spinal accessory, and hypoglossal nerve palsies during cervi-cal immobilization in a cervical collar (8). The Cranial Nerve
Occipital Condyle Fractures
S115Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE15
.1.
Sum
mar
yof
Rep
orts
onO
ccip
ital
Con
dyle
Frac
ture
sa
Seri
es(R
ef.
No.
)A
ge(y
r)/S
exTy
peLo
catio
nPa
inPl
ain
X-r
ayC
TM
RI
Exam
inat
ion
Trea
tmen
tO
utco
me
Legr
oset
al.,
2000
(23)
71/F
III�
Unr
epU
nrep
L,�
Epid
ural
Del
CN
6,7,
106
wk
colla
r18
mo
CN
10
44/M
II�
Unr
epU
nrep
R,
�N
orm
alD
elC
N6,
9–12
6w
kco
llar
3.5
mo
CN
10
Ide
etal
.,19
98(1
9)25
/MIII
��
STS,
C1
FxR
,�
Tect
oria
lm
embr
ane
tear
Nor
mal
10w
kco
llar
10w
kno
rmal
Dem
isch
etal
.,19
98(1
3)45
/FII
Unr
epU
nrep
Unr
epR
,�
FxD
elC
N12
Non
e1
yrim
pC
N12
Blo
omet
al.,
1997
(5)
21/M
IIIU
nrep
Unr
epST
S,C
6,7
FxR
,�
Unr
epN
orm
al�
8w
kco
llar
Nor
mal
36/F
IIIU
nrep
Unr
epU
nrep
L,�
Unr
epN
orm
al�
8w
kco
llar
Pain
15/F
I/IU
nrep
Unr
epU
nrep
B,
�U
nrep
Qpa
resi
s�
8w
kco
llar
Imp
Qpa
resi
s
45/F
III/I
Unr
epU
nrep
Unr
epB
,�
Unr
epC
N12
�8
wk
colla
rPa
in,
CN
12
22/F
IIU
nrep
Unr
epU
nrep
R,
�U
nrep
Nor
mal
�8
wk
colla
rU
nrep
21/M
IU
nrep
Unr
epST
S,C
1,2,
5Fx
R,
�U
nrep
Nor
mal
�8
wk
colla
rU
nrep
41/M
IU
nrep
Unr
epU
nrep
R,
�U
nrep
Nor
mal
�8
wk
colla
rN
orm
al
6/F
IIU
nrep
Unr
epU
nrep
L,�
Unr
epN
orm
al�
8w
kco
llar
Nor
mal
25/F
IU
nrep
Unr
epST
S,C
2Fx
L,�
Unr
epN
orm
al�
8w
kco
llar
Unr
ep
20/M
IU
nrep
Unr
epC
FxR
,�
Unr
epPp
legi
a�
8w
kco
llar
Unr
ep
Tuli
etal
.,19
97(4
1)64
/FIII
Unr
ep�
STS
R,
�N
one
Nor
mal
12w
kco
llar
3m
ono
rmal
69/F
IIIU
nrep
�A
Aw
ide
L,�
FxM
pare
sis,
CN
OC
fusi
onIm
p
27/M
Old
Unr
ep�
Nor
mal
L,�
Non
eN
orm
alN
one
3yr
norm
al
Cot
talo
rda
etal
.,19
96(1
0)15
/FI
Unr
ep�
Nor
mal
R,
�N
one
Nor
mal
7w
kM
iner
vaTr
colla
r4
mo
norm
al
Lam
and
Stra
tford
,19
96(2
2)20
/FIII
Unr
epU
nrep
Nor
mal
R,
�C
ontu
sion
Hpa
,C
N12
3m
oha
lo5
yrim
pC
N12
Urc
ulo
etal
.,19
96(4
2)62
/MIII
Unr
epU
nrep
Nor
mal
R,
�Fx
Del
CN
9,10
Non
e6
mo
sam
e
Nob
lean
dSm
oker
,19
96(3
1)33
/MI
Unr
epU
nrep
Unr
ep?,
�N
one
Del
CN
12N
one
Unr
ep
26/M
IU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
15N
one
Unr
ep
16/M
IIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
13N
one
Unr
ep
32/M
IIU
nrep
Unr
epC
2Fx
?,�
Non
eC
N7,
12N
one
Unr
ep
53/F
IIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
8N
one
Unr
ep
47/F
IIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
15N
one
Unr
ep
37/M
IIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
8N
one
Unr
ep
11/M
IIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
13N
one
Unr
ep
33/M
IIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
15N
one
Unr
ep
23/M
IIU
nrep
Unr
epU
nrep
?,�
Non
eU
nrep
Unr
epU
nrep
39/M
IIIU
nrep
Unr
epII
Od
Fx?,
�N
one
CN
7H
alo
Unr
ep
88/M
IIIU
nrep
Unr
epC
1,II
Od
Fx?,
�N
one
GC
S15
Hal
oU
nrep
29/M
IIIU
nrep
Unr
epU
nrep
?,�
Non
eU
nrep
Unr
epU
nrep
14/F
IIIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
11C
olla
rU
nrep
17/F
IIIU
nrep
Unr
epU
nrep
?,�
Non
eG
CS
7N
one
Unr
ep
Cas
tling
and
Hic
ks,
1995
(9)
21/M
II�
�N
orm
alR
,�
Non
eD
elC
N12
Non
e2
yrno
rmal
Emer
yet
al.,
1995
(15)
26/M
IIIU
nrep
�N
orm
alL,
�Fr
actu
reH
yper
refle
xic
Col
lar
4m
ono
rmal
Pale
yan
dW
ood,
1995
(34)
21/M
IIIU
nrep
�N
orm
alL,
�N
orm
alD
elC
N12
Non
e6
mo
imp
CN
12
Stro
oban
tset
al.,
1994
(39)
27/M
III�
�N
orm
alR
,�
Non
eN
orm
al10
wk
colla
r21
mo
norm
al
12/F
III�
�C
1Fx
L,�
Non
eN
orm
al4
wk
Min
erva
Nor
mal
Was
serb
erg
and
Bar
tlett,
1995
(45)
39/M
III�
Unr
epN
orm
alL,
�N
one
Del
CN
12N
one
12
24/M
III�
�N
orm
alL,
�N
one
Del
dipl
opia
Dec
ompr
essi
onN
orm
al
16/M
III�
Unr
epN
orm
alR
,�
Non
eB
rain
inju
ryC
olla
r3
mo
CN
12
34/M
IIIU
nrep
Unr
epN
orm
alR
,�
Non
eU
nrep
Hal
oU
nrep
S116 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE15
.1.
Con
tinu
edSe
ries
(Ref
.N
o.)
Age
(yr)
/Sex
Type
Loca
tion
Pain
Plai
nX
-ray
CT
MR
IEx
amin
atio
nTr
eatm
ent
Out
com
e
You
nget
al.,
1994
(47)
26/F
III�
Unr
epN
orm
alL,
�N
one
Hpa
,C
N9–
1212
wk
halo
14m
oim
pC
N9–
12
20/M
III�
Unr
epN
orm
alR
,�
Non
eG
CS
7C
olla
r1
yrH
pa
Man
nan
dC
ohen
,19
94(2
7)23
/MIII
��
Nor
mal
R,
�N
one
Nor
mal
6w
kco
llar
Nor
mal
Ols
son
and
Kun
z,19
94(3
2)43
/MIII
Unr
ep�
Nor
mal
L,�
Non
eN
orm
alC
olla
rN
orm
al
Shar
ma
etal
.,19
94(3
7)35
/MII
Unr
epU
nrep
Nor
mal
L,�
Non
eC
N9,
10D
ecom
pres
sion
3m
oim
pC
N9,
10
Mas
saro
and
Lano
tte,
1993
(29)
21/M
IIIU
nrep
Unr
epN
orm
alL,
�N
one
Hse
nsor
y,C
N12
8w
kM
iner
va2
yrC
N12
Rai
laet
al.,
1993
(35)
25/M
III�
�N
orm
alL,
�N
one
Nor
mal
6w
kco
llar
Nor
mal
67/M
III�
�C
1ab
norm
alL,
�N
one
Nor
mal
Col
lar
Nor
mal
Bet
tini
etal
.,19
93(3
)39
/FI
Unr
ep�
C3
FxL,
�N
one
Nor
mal
Unr
epU
nrep
24/M
II�
Unr
epN
orm
alR
,�
Non
eC
oma
Unr
epU
nrep
21/F
III�
Unr
epU
nrep
?,�
Con
tusi
onC
oma
Unr
epU
nrep
21/M
III/II
IU
nrep
�N
orm
alB
,�
Non
eN
orm
alU
nrep
Unr
ep
Leve
ntha
let
al.,
1992
(25)
42/F
II�
Unr
epN
orm
alL,
�N
one
CN
6,7
3m
oco
llar
Unr
ep
19/F
III�
�N
orm
alL,
�N
one
Nor
mal
Col
lar
Unk
now
n
43/M
IIIU
nrep
�C
5Fx
R,
�N
one
Nor
mal
3m
oco
llar
Nor
mal
17/F
II�
Unr
epL1
FxR
,�
Non
eG
CS
103
mo
colla
rN
orm
al
36/M
I�
GC
S8
T1Fx
R,
�N
one
GC
S8
3m
oha
loN
orm
al
17/M
I�
GC
S4
Nor
mal
R,
�N
one
GC
S4
3m
oco
llar
Nor
mal
Mod
yan
dM
orri
s,19
92(3
0)21
/MIII
�U
nrep
STS
L,�
Non
eU
nrep
6w
kco
llar
18m
ono
sym
ptom
s
Boz
boga
etal
.,19
92(7
)34
/FIII
��
Nor
mal
L,�
Non
eL
Hpa
,di
plop
iaLa
tede
com
pres
sion
4yr
norm
al
37/M
III�
Unr
epU
nrep
L,�
Non
eD
elve
rtig
oD
el8
wk
colla
r3
yrno
rmal
Bri
dgm
anan
dM
cNab
,19
92(8
)32
/MIII
��
Nor
mal
L,�
Non
eD
elC
N10
–12
Col
lar
1yr
imp
CN
10–1
2
Wan
iet
al.,
1991
(44)
67/M
II�
Unr
ep�
Con
dfx
L,no
neN
one
CN
9–12
Non
eC
N10
,12
Wes
sels
,19
90(4
6)26
/MIII
��
Unr
epR
,�
Non
eC
N7–
12C
olla
r6
wk
imp
CN
7–12
7m
o/M
II�
Unr
epU
nrep
L,�
Non
eC
N5,
7–12
Col
lar
4m
oC
N7–
12
27/M
II�
Unr
epU
nrep
R,
�N
one
CN
7–12
Col
lar
6w
kim
p
Mar
iani
,19
90(2
8)30
/MIII
��
STS
R,
�N
one
Nor
mal
8w
kco
llar
Nor
mal
Jone
set
al.,
1990
(21)
43/M
III/II
I�
Unr
ep�
Con
dFx
B,
�C
ontu
sion
Qpl
egia
OC
Fx4
wk
Qpl
egia
Des
aiet
al.,
1990
(14)
33/M
III�
�N
orm
alL,
�N
one
CN
6C
olla
r4
mo
norm
al
Val
aska
tzis
and
Ham
mer
,19
90(4
3)19
/MIII
��
Nor
mal
R,
�N
one
Nor
mal
6w
kco
llar
Nor
mal
Orb
ayet
al.,
1989
(33)
37/M
IIIU
nrep
�N
orm
alL,
�(to
mo
�)
Non
eD
elC
N12
Non
e15
mo
CN
12
Savo
lain
eet
al.,
1989
(36)
71/F
III�
�N
orm
alR
,�
Non
eH
pleg
ia,
CN
6Tr
,ha
loLM
pare
sis
And
erso
nan
dM
onte
sano
,19
88(1
)3/
MI
�U
nrep
Nor
mal
R,
�N
one
Unc
onSo
ft24
mo
norm
al
18/F
III�
Unr
epN
orm
al?,
�N
one
Unr
epM
iner
va36
mo
22/M
III�
Unr
epN
orm
alR
,to
mo
�N
one
Unc
onH
alo
12m
ono
rmal
23/M
III�
Unr
epN
orm
alL,
�N
one
Unc
onC
olla
rD
eath
25/M
III�
Unr
epN
orm
al?,
tom
o�
Non
eU
nrep
Min
erva
17m
o
37/M
II�
Unr
epN
orm
alL,
�N
one
Unc
onC
olla
r12
mo
norm
al
Cur
riet
al.,
1988
(11)
16/F
III�
Unr
epN
orm
alR
,�
Non
eD
ecer
ebra
teC
olla
r6
mo
unre
p
Has
him
oto
etal
.,19
88(1
8)71
/MII
�U
nrep
Nor
mal
L,�
Non
eC
N9–
12N
one
6m
oC
N9–
12
Dee
bet
al.,
1988
(12)
25/F
IIU
nrep
Unr
epN
orm
alD
elL,
�N
one
CN
12N
one
Unr
ep
66/F
Old
Unr
ep�
Non
eD
elL,
�Fx
Nor
mal
Non
eU
nrep
Spen
cer
etal
.,19
84(3
8)19
/MI
�G
CS
8N
orm
alL,
�N
one
GC
S8
Col
lar,
halo
BC
N9,
10
Gol
dste
inet
al.,
1982
(16)
24/F
IIIU
nrep
�C
5,6
slx
L,to
mo
�N
one
Nor
mal
2m
oco
llar
Nor
mal
Har
ding
-Sm
ithet
al.,
1981
(17)
18/M
III�
Unr
epST
SR
,to
mo
�N
one
Unc
onC
olla
r16
mo
norm
al
Bol
ende
ret
al.,
1978
(6)
23/M
IIIU
nrep
Unr
epN
orm
alR
,to
mo
�N
one
CN
9–12
Non
eU
nrep
22/M
IIU
nrep
Unr
epN
orm
alR
,to
mo
�N
one
Del
CN
6,9,
10N
one
Unr
ep
aC
T,co
mpu
ted
tom
ogra
phy;
MR
I,m
agne
tic
reso
nanc
eim
agin
g;U
nrep
,unr
epor
ted;
Del
,del
ayed
;�,d
one
orpo
siti
ve;�
,not
done
orne
gati
ve;F
x,fr
actu
re;I
mp,
impr
ovem
ent;
STS,
soft
tiss
uesw
ellin
g;C
,cer
vica
l;A
A,a
tlan
toax
ial;
Q,q
uadr
i-;P
,par
a-;M
,mon
o-;O
C,o
ccip
ital
cond
yle;
CN
,cra
nial
nerv
e;G
CS,
Gla
sgow
Com
aSc
ale;
Od,
odon
toid
;L,l
eft;
R,
righ
t;B
,bi
late
ral;
tom
o,to
mog
raph
y;Tr
,tr
acti
on;
Unc
on,
unco
nfir
med
;H
pa,
hem
ipar
esis
;C
ond,
cond
ylar
;sl
x,su
blux
atio
n.
Occipital Condyle Fractures S117
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
X and XI palsies improved. However, the hypoglossal palsypersisted at 1 year.
Forty-four patients were treated with cervical collar immo-bilization (8 Type I, 8 Type II, 28 Type III) (1, 4, 5, 7, 8, 10, 11,14–17, 19, 23, 25, 27, 28, 30–32, 35, 39, 41, 43, 45, 46). Thirteenpatients were treated with halo/Minerva immobilization (2Type I, 11 Type III) (1, 22, 25, 29, 31, 36, 38, 40, 45, 47).Treatment was unreported in six patients (3, 31). Five patients(one Type II, four Type III) underwent surgery. Two patientswith Type III OCF were treated with occipitocervical fusion(one with concurrent atlanto-occipital dislocation and onewith atlantoaxial instability) (21, 41). One patient (Type III)with delayed diplopia had symptom resolution after removalof the fracture fragment (45), whereas one patient (Type II)with lower cranial nerve deficits (37) and one (Type III) withdiplopia and hemiparesis (7) remained unchanged severaldays after surgery. The latter patient subsequently recoverednormal function.
In summary, 12 of 15 patients who developed delayedsymptoms or deficits were not initially treated. Only 3 of these12 patients were subsequently treated with cervical immobi-lization. All three improved. In comparison, only three of sixpatients demonstrated improvement in deficits without treat-ment. Only one patient (Type III) developed a deficit duringtreatment that persisted (hypoglossal nerve palsy) despitecollar use. Only three patients underwent surgery for decom-pression of the brainstem, one of whom had immediate andlasting improvement in symptoms postoperatively. Because12 of 23 patients developed delayed deficits without treat-ment and another developed a deficit after premature discon-tinuation of treatment, the literature suggests that patientswith Type III OCF should be treated with external immobili-zation. Treatment of patients with OCF Types I and II mayinclude external immobilization.
SUMMARY
OCF is an uncommon injury requiring computed tomo-graphic imaging for diagnosis. Patients sustaining high-energy blunt craniocervical trauma, particularly in the settingof loss of consciousness, impaired consciousness, occipitocer-vical pain or motion impairment, and lower cranial nervedeficits, should undergo computed tomographic imaging ofthe craniocervical junction. Untreated patients with OCF often
develop lower cranial nerve deficits that usually recover orimprove with external immobilization. Identification of TypeIII OCF should prompt external immobilization. Additionaltreatment may be dictated by the presence of associated cer-vical fractures or instability.
KEY ISSUES FOR FUTURE INVESTIGATION
Although Type III OCF is considered by many authors tobe unstable, not all patients, treated or not, developed neuro-logical deficits. Computed tomographic imaging with three-dimensional reconstruction for more precise measurement ofthe magnitude of fracture displacement and MRI for differ-entiation of partial and complete ligamentous injuries may beuseful in identifying subgroups of patients who do not re-quire treatment or, conversely, require more rigid halo immo-bilization, rather than collar immobilization. Because OCFinjuries remain relatively infrequent, cooperative retrospec-tive collection of plain x-ray, computed tomographic, andMRI data in patients with OCF is recommended.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Anderson PA, Montesano PX: Morphology and treatment of oc-cipital condyle fractures. Spine 13:731–736, 1988.
2. Bell C: Surgical observations. Middlesex Hosp J 4:469, 1817.3. Bettini N, Malaguti MC, Sintini M, Monti C: Fractures of the
occipital condyles: Report of four cases and review of the litera-ture. Skeletal Radiol 2:187–190, 1993.
4. Bloom AI, Neeman Z, Floman Y, Gomori J, Bar-Ziv J: Occipitalcondyle fracture and ligament injury: Imaging by CT. PediatrRadiol 26:786–790, 1996.
5. Bloom AI, Neeman Z, Slasky BS, Floman Y, Milgrom M, RivkindA, Bar-Ziv J: Fracture of the occipital condyles and associatedcraniocervical ligament injury: Incidence, CT imaging and impli-cations. Clin Radiol 52:198–202, 1997.
6. Bolender N, Cromwell LD, Wendling L: Fracture of the occipitalcondyle. AJR Am J Radiol 131:729–731, 1978.
7. Bozboga M, Unal F, Hepgul K, Izgi N, Turantan MI, Turker K:Fracture of the occipital condyle: Case report. Spine 17:1119–1121,1992.
FIGURE 15.1. Classification of occipital condyle fractures according to Anderson and Montesano (1). A, Type I fractures mayoccur with axial loading. B, Type II fractures are extensions of a basilar cranial fracture. C, Type III fractures may result froman avulsion of the condyle during rotation, lateral bending, or a combination of mechanisms.
S118 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
8. Bridgman SA, McNab W: Traumatic occipital condyle fracture,multiple cranial nerve palsies, and torticollis: A case report andreview of the literature. Surg Neurol 38:152–156, 1992.
9. Castling B, Hicks K: Traumatic isolated unilateral hypoglossalnerve palsy: Case report and review of the literature. Br J OralMaxillofac Surg 33:171–173, 1995.
10. Cottalorda J, Allard D, Dutour N: Fracture of the occipital con-dyle. J Pediatr Orthop 5:161–163, 1996.
11. Curri D, Cervellini P, Zanusso M, Benedetti A: Isolated fracture ofoccipital condyle: Case report. J Neurosurg Sci 32:157–159, 1988.
12. Deeb ZL, Rothfus WE, Goldberg AL, Daffner RH: Occult occipitalcondyle fractures presenting as tumors. J Comput Tomogr 12:261–263, 1988.
13. Demisch S, Lindner A, Beck R, Zierz S: The forgotten condyle:Delayed hypoglossal nerve palsy caused by fracture of the occip-ital condyle. Clin Neurol Neurosurg 100:44–45, 1998.
14. Desai SS, Coumas JM, Danylevich A, Hayes E, Dunn EJ: Fractureof the occipital condyle: Case report and review of the literature.J Trauma 30:240–241, 1990.
15. Emery E, Saillant G, Ismail M, Fohanno D, Roy-Camille R: Frac-ture of the occipital condyle: Case report and review of theliterature. Eur Spine J 4:191–193, 1995.
16. Goldstein SJ, Woodring JH, Young AB: Occipital condyle fractureassociated with cervical spine injury. Surg Neurol 17:350–352,1982.
17. Harding-Smith J, MacIntosh PK, Sherbon KJ: Fracture of theoccipital condyle: A case report and review of the literature.J Bone Joint Surg Am 63A:1170–1171, 1981.
18. Hashimoto T, Watanabe O, Takase M, Koniyama J, Kobota M:Collet-Sicard syndrome after minor head trauma. Neurosurgery23:367–370, 1988.
19. Ide C, Nisolle JF, Misson N, Trigaux JP, Gustin T, De Coene B,Gilliard C: Unusual occipitoatlantal fracture dissociation with noneurological impairment: Case report. J Neurosurg 88:773–776,1998.
20. Jacoby CG: Fracture of the occipital condyle. AJR Am JRoentgenol 132:500, 1979 (letter).
21. Jones DN, Knox AM, Sage MR: Traumatic avulsion fracture of theoccipital condyles and clivus with associated unilateral atlantooc-cipital distraction. AJNR Am J Neuroradiol 11:1181–1183, 1990.
22. Lam CH, Stratford J: Bilateral hypoglossal nerve injury withoccipital condylar fracture. Can J Neurol Sci 23:145–148, 1996.
23. Legros B, Fournier P, Chiaroni P, Ritz O, Fusciardi J: Basal frac-ture of the skull and lower (IX, X, XI, XII) cranial nerves palsy:Four case reports including two fractures of the occipital con-dyle—A literature review. J Trauma 48:342–348, 2000.
24. Leone A, Cerase A, Colosimo C, Lauro L, Puca A, Marano P:Occipital condylar fractures: A review. Radiology 216:635–644,2000.
25. Leventhal MR, Boydston WR, Sebes JI, Pinstein ML, Watridge CB,Lowrey R: The diagnosis and treatment of fractures of the occip-ital condyle. Orthopedics 15:944–947, 1992.
26. Link TM, Schuierer G, Hufendiek A, Horch C, Peters PE: Sub-stantial head trauma: Value of routine CT examination of thecervicocranium. Radiology 196:741–745, 1995.
27. Mann FA, Cohen W: Occipital condyle fracture: Significance inthe assessment of occipitoatlantal stability. AJR Am J Radiol163:193–194, 1994.
28. Mariani PJ: Occipital condyle fracture presenting as aretropharyngeal hematoma. Ann Emerg Med 19:1447–1449, 1990.
29. Massaro F, Lanotte M: Fracture of the occipital condyle. Injury24:419–420, 1993.
30. Mody BS, Morris EW: Fracture of the occipital condyle: Casereport and review of the world literature. Injury 23:350–352, 1992.
31. Noble ER, Smoker WR: The forgotten condyle: The appearance,morphology, and classification of occipital condyle fractures.AJNR Am J Neuroradiol 17:507–513, 1996.
32. Olsson R, Kunz R: Fracture of the occipital condyle as an inciden-tal finding during CT-evaluation of a maxillary fracture. ActaRadiol 35:90–91, 1994.
33. Orbay T, Aykol S, Seckin Z, Ergun R: Late hypoglossal nervepalsy following fracture of the occipital condyle. Surg Neurol31:402–404, 1989.
34. Paley MD, Wood GA: Traumatic bilateral hypoglossal nervepalsy. Br J Oral Maxillofac Surg 33:239–241, 1995.
35. Raila FA, Aitken AT, Vickers GN: Computed tomography andthree-dimensional reconstruction in the evaluation of occipitalcondyle fracture. Skeletal Radiol 22:269–271, 1993.
36. Savolaine ER, Ebraheim NA, Jackson WT, Rusin JJ: Three-dimensional computed tomography in evaluation of occipitalcondyle fracture. J Orthop Trauma 3:71–75, 1989.
37. Sharma BS, Mahajan RK, Bhatia S, Khosla VK: Collet-Sicard syn-drome after closed head injury. Clin Neurol Neurosurg 96:197–198, 1994.
38. Spencer JA, Yeakley JW, Kaufman HH: Fracture of the occipitalcondyle. Neurosurgery 15:101–103, 1984.
39. Stroobants J, Fidlers L, Storms JL, Klaes R, Dua G, Van Hoye M:High cervical pain and impairment of skull mobility as the onlysymptoms of an occipital condyle fracture: Case report.J Neurosurg 81:137–138, 1994.
40. Stroobants J, Seynaeve P, Fidlers L, Klaes R, Brabants K, VanHoye M: Occipital condyle fracture must be considered in thepediatric population: Case report. J Trauma 36:440–441, 1994.
41. Tuli S, Tator CH, Fehlings MG, Mackay M: Occipital condylefractures. Neurosurgery 41:368–377, 1997.
42. Urculo E, Arrazola M, Arrazola M Jr, Riu I, Moyua A: Delayedglossopharyngeal and vagus nerve paralysis following occipitalcondyle fracture: Case report. J Neurosurg 84:522–525, 1996.
43. Valaskatzis EP, Hammer AJ: Fracture of the occipital condyle: Acase report. S Afr Med J 77:47–48, 1990.
44. Wani MA, Tandon PN, Banerji AK, Bhatia R: Collet-Sicard syn-drome resulting from closed head injury: Case report. J Trauma31:1437–1439, 1991.
45. Wasserberg J, Bartlett RJV: Occipital condyle fractures diagnosedby high-definition CT and coronal reconstructions. Neuroradiol-ogy 37:370–373, 1995.
46. Wessels LS: Fracture of the occipital condyle: A report of 3 cases.S Afr J Surg 28:155–156, 1990.
47. Young WF, Rosenwasser RH, Getch C, Jallo J: Diagnosis andmanagement of occipital condyle fractures. Neurosurgery 34:257–261, 1994.
Occipital Condyle Fractures S119
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
CHAPTER 16
Isolated Fractures of the Atlas in Adults
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Treatment options in the management of isolated fractures of the atlas are based on the specific
atlas fracture type. It is recommended that isolated fractures of the atlas with an intact transverse atlantalligament be treated with cervical immobilization alone. It is recommended that isolated fractures of theatlas with disruption of the transverse atlantal ligament be treated with either cervical immobilization aloneor surgical fixation and fusion.
RATIONALE
The atlas vertebra is subject to a variety of acute fractureinjuries and may be associated with other cervical frac-tures and ligamentous traumatic injuries (4, 8, 25, 26,
31). Although the treatment of atlas fractures in combinationwith other cervical fracture injuries is most commonly linkedto the treatment of the associated injury (8), isolated fracturesof the atlas occur with sufficient frequency to warrant review.
The medical literature addressing the management of frac-tures of the atlas was examined using evidence-based medi-cine techniques to determine the optimal treatment for iso-lated atlas fractures, including isolated anterior or posteriorarch fractures, anterior and posterior arch fractures (burstfractures), lateral mass fractures, comminuted fractures, andtransverse process fractures.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “vertebral fracture”combined with “atlas” and “human” yielded 360 references.The abstracts were reviewed, and articles addressing clinicalmanagement and follow-up of atlas fractures were selectedfor inclusion. The relative infrequency of these fractures, thesmall number of collected case series, and the numerous casereports with pertinent information required rather broad in-clusion and exclusion criteria. Several papers addressing rel-evant background information such as biomechanics and ra-diology were included. The bibliographies of the selectedarticles were also reviewed to provide additional referencesand to assess completeness of the literature review. Theseefforts resulted in 32 articles describing acute traumatic atlasfractures. Ten Class III articles (eight case series and two casereports) documenting treatment of patients with atlas frac-tures are summarized in Table 16.1. The remaining 22 articlesare included in the reference list and contribute to the scien-
tific foundation. Treatment options are summarized in Table16.2.
SCIENTIFIC FOUNDATION
Atlas fractures account for approximately 1 to 2% of allfractures of the human spinal column and roughly 2 to 13% ofall acute cervical spine fractures (8, 21, 30). The first knownfracture of the atlas was demonstrated at autopsy by Cooperin 1822 and has been the subject of a series of historicalpublications (28). In 1920, Jefferson (15) reviewed 42 previ-ously described cases of atlas fracture and added 4 new cases.Although his article documents a variety of atlas fracturepatterns, it is best known for the characterization of the “Jef-ferson fracture,” a burst fracture injury of the atlas ring (10). In1945, Hinchey and Bickel (13a) added 112 cases of atlas frac-ture to the literature. Sherk and Nicholson summarized anadditional 73 cases in 1970 (30).
Spence et al. (31), in 1970, reported their findings of a studyof the mechanism of atlas fracture and potential rupture of thetransverse atlantal ligament. Using 10 cadaveric specimens,the authors studied the application of force required to frac-ture C1 and to rupture the transverse ligament (range, 38–104kg; mean, 58 kg). The sum of the excursion of the C1 lateralmasses over the C2 lateral masses after traumatic injuryranged from 4.8 to 7.6 mm (mean, 6.3 mm). The authorsconcluded that if the sum of lateral mass displacement (LMD)of C1 over C2 on the anteroposterior radiographic image ismore than 6.9 mm, then the transverse atlantal ligament is“probably torn.” In a follow-up clinical and biomechanicalstudy, Fielding et al. (5) confirmed these findings. These twostudies, completed before the era of magnetic resonance im-aging (MRI), are the basis for the widely quoted “rules ofSpence” (i.e., �6.9 mm LMD � transverse atlantal ligamentdisruption) offered to assist in the management of patientswith isolated atlas fractures. Subsequently, Heller et al. (12)reported their observations on 35 open-mouth odontoid filmsusing calibration markings to assess radiographic magnifica-
S120 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
tion. They found an 18% magnification factor on open-mouthodontoid-view x-rays. Applying this information to the eval-uation of atlas burst fractures by means of the rules of Spencesuggests that the sum of the LMD measurements indicatingatlantal transverse ligament disruption should be increasedfrom 6.9 to 8.1 mm. This study pointed out the difficulty inusing plain radiographic measurements to assess the integrityof the transverse atlantal ligament after acute traumatic atlasfracture.
Hadley et al. (8) reported a series of 57 patients with atlasfractures, representing 6.6% of their series of cervical fracturesin 1988. The authors managed 32 patients with isolated atlasfractures, none of whom sustained neurological injury. Theauthors based their treatment recommendations on the degreeof LMD in each patient. Twelve patients had nondisplacedatlas fractures. Of these, 10 were treated with a rigid collar,one with a soft collar, and one with a suboccipital-mandibularimmobilizer (SOMI). The duration of treatment for these pa-tients was 8 to 12 weeks. Fifteen patients had an LMD of lessthan 7 mm. Eight were treated with a rigid collar, three witha SOMI, and four with a halo immobilization brace for 10 to 12weeks. The remaining five patients had an LMD of more than7 mm and were managed with a halo orthosis. These fivepatients were treated for 12 to 16 weeks. Of the 32 patients, 29were available for long-term follow-up (median, 40 mo).Three complained of neck pain. All were successfully treated.No patient required subsequent surgical fixation. The authorsconcluded that isolated fractures of the atlas are effectivelymanaged with external immobilization alone for 12 weeks(median duration). Atlas fractures with an LMD of more than6.9 mm required more rigid immobilization (halo orthosis)than those with an LMD of less than 6.9 mm (cervical collar).Levine and Edwards (21) described their experience with 34patients with isolated atlas fractures in 1991. They followed asimilar treatment algorithm with similar success.
Fowler et al. (6) reported a series of 48 consecutive atlasfracture patients, representing 5.5% of all cervical fractures intheir experience. In their series, 33% of their patients had otherassociated cervical spine fractures. Atlas fractures were di-vided into burst (n � 30), posterior arch (n � 17), and anteriorarch fractures (n � 1). None of the patients with an isolatedatlas fracture presented with neurological deficit. These au-thors suggested treatment with closed traction reduction ofisolated atlas fractures if the LMD is more than 7.0 mm, andthen immobilization in a rigid collar. No patient in this seriesunderwent surgical fixation. All were effectively treated withthis management scheme at last follow-up, although the du-ration of treatment was not specified. In 1991, Kesterson et al.(16) reported their series of 17 cases of atlas fractures. Thirteenwere isolated atlas fractures and were considered stable. Allwere successfully managed with rigid cervical immobilization(nine collar, one SOMI, one halo, two Minerva). Again, theduration of treatment was not specified. Several other authorshave described the successful treatment of isolated atlas frac-
tures with rigid cervical immobilization, using similar man-agement principles (9, 13, 20, 29, 32).
Landells and Van Peteghem (18) described a series of 35patients with atlas fractures, representing 4.7% of their insti-tutional experience with acute cervical fracture injuries. Theauthors categorized atlas fractures into three types. Type Ifractures involved a single arch and occurred in 16 of their 35patients. Type II fractures were burst fractures and repre-sented 13 of the 35 isolated fractures they treated. Type IIIfractures were atlas lateral mass fractures identified in 6 of the35 patients. The authors used the original rules of Spence toassist with the identification of stability and noted four pa-tients with an LMD of more than 6.9 mm. Regardless of thefracture type or stability, all fractures except one were initiallytreated with external immobilization for an unreported lengthof time. The one exception was a patient with a Type I fracturewho underwent early surgery with C1–C2 wiring and fusion.The reason for the exception is not made clear in the text. Only1 of 34 patients treated with external immobilization requiredsurgery for late instability. The authors observed no relation-ship between successful treatment and the amount of initialLMD. They recommended that atlas fractures be initiallytreated with rigid external immobilization. They noted thatlate instability can occur and recommended clinical follow-upof these patients.
Clinically observed atlas fracture patterns can be repro-duced in cadaveric experimental models of axial loading (11).In a series of biomechanical studies, Oda et al. (24, 25) andPanjabi et al. (26) reproduced these atlas fracture patternswith axial loading and found that the burst fracture wasassociated with postinjury hypermobility at C1–C2. Theseauthors described a 42% increase in flexion/extension motion,a 24% increase in lateral bending, and a 5% increase in axialrotation. They found that in all instances of transverse atlantalligament disruption, the atlantodens interval was more than 3mm. The authors concluded that the atlantodens interval wasthe most reliable predictor of transverse ligament disruptionin adults after acute C1 fracture.
McGuire and Harkey (22), in 1995, described two cases ofunstable atlas burst fractures treated with surgical fixation andfusion. The fractures were thought to be unstable based on apredental space more than 5 mm and an LMD more than 9 mm,respectively. Both were treated with posterior C1–C2 transartic-ular screw fixation and fusion with good results. The authorsreported that transarticular screw fixation obviated the need forhalo immobilization postoperatively. Several other authors havereported successful surgical fixation and fusion for atlas frac-tures when they are associated with disruption of the transverseligament with resultant instability (17, 18, 27). These few patientswere treated with posterior C1–C2 wiring and fusion proceduresand were managed in rigid orthoses (halo, Guilford brace) for 12to 16 weeks postoperatively.
More recently, it has been proposed that MRI is a more sen-sitive indicator of transverse atlantal ligament disruption than
Isolated Fractures of the Atlas in Adults
S121Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE16
.1.
Sum
mar
yof
Rep
orts
onTr
eatm
ent
ofFr
actu
res
ofth
eA
tlas
inA
dult
sa
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Lee
etal
.,19
98(2
0)R
etro
spec
tive
revi
ewin
clud
ing
12ca
ses
ofis
olat
edat
las
frac
ture
.III
All
trea
ted
succ
essf
ully
with
exte
rnal
imm
obili
zatio
n.Tr
eatm
ent
algo
rith
mpr
opos
ed:
Stab
le,
trea
tw
ithco
llar
12w
k.U
nsta
ble,
prop
oses
surg
ical
fixat
ion.
(Inst
abili
tyde
fined
asla
tera
lm
ass
disp
lace
men
t�
7m
mor
MR
Iev
iden
ceof
tran
sver
selig
amen
tdi
srup
tion)
.
McG
uire
and
Har
key,
1995
(22)
2ca
ses
ofun
stab
leat
las
burs
tfr
actu
retr
eate
dw
ithpo
ster
ior
tran
sart
icul
arsc
rew
fixat
ion
and
fusi
on.
IIIC
onsi
dere
dun
stab
leba
sed
onpr
eden
tal
spac
e�
5m
man
dLM
D�
9m
m.
Bot
htr
eate
dsu
cces
sful
ly.
Ace
rvic
alco
llar
was
used
post
oper
ativ
ely.
Levi
nean
dEd
war
ds,
1991
(21)
Ret
rosp
ectiv
ere
view
of34
patie
nts
with
atla
sfr
actu
res.
Med
ian
follo
w-u
p,4.
5yr
.III
Ifla
tera
lm
ass
disp
lace
men
t�
7m
m,
patie
nts
trea
ted
with
colla
r,an
dif
�7
mm
,pa
tient
str
eate
dw
ithei
ther
halo
alon
eor
redu
ced
intr
actio
nan
dm
aint
aine
dun
tilhe
aled
(6w
kin
trac
tion
and
6w
kin
halo
).N
opa
tient
str
eate
dsu
rgic
ally
.
Kes
ters
onet
al.,
1991
(16)
Ret
rosp
ectiv
ere
view
incl
udin
g13
patie
nts
with
isol
ated
atla
sbu
rst
(Jeffe
rson
)fr
actu
re.
IIIA
llsu
cces
sful
lytr
eate
dw
ithim
mob
iliza
tion.
Fow
ler
etal
.,19
90(6
)R
etro
spec
tive
revi
ewof
48co
nsec
utiv
eat
las
frac
ture
sdi
vide
din
tobu
rst
(30)
,po
ster
ior
arch
(17)
,an
dan
teri
orar
chfr
actu
res
(1).
IIIA
utho
rssu
gges
tre
duct
ion
bytr
actio
nif
LMD
�7.
0,fo
llow
edby
colla
r.N
opa
tient
sun
derw
ent
surg
ical
fixat
ion.
Had
ley
etal
.,19
88(8
)R
etro
spec
tive
revi
ewin
clud
ing
32is
olat
edfr
actu
res
ofth
eat
las.
Med
ian
follo
w-u
p,40
mo
on29
/32
frac
ture
s.
IIITh
efo
llow
ing
trea
tmen
tpa
ttern
sw
ere
reco
gniz
ed:
LMD
�7
mm
(5pa
tient
s):
trea
ted
with
halo
.LM
D0–
7m
m(1
5pa
tient
s):
4tr
eate
din
halo
,11
trea
ted
inSO
MI.
LMD
none
(12
patie
nts)
:tr
eate
din
rigi
dco
llar.
Non
eof
thes
eis
olat
edC
1fr
actu
res
sust
aine
dne
urol
ogic
alin
jury
orre
quir
edsu
rger
y.3
com
plai
ned
ofne
ckpa
in;
othe
rwis
e,al
lw
ere
succ
essf
ully
trea
ted.
Aut
hors
’re
com
men
datio
n:is
olat
edC
1fr
actu
res
can
bem
anag
edw
ithou
tea
rly
surg
ical
fixat
ion.
Ifth
eLM
D�
6.9
mm
,th
enha
loim
mob
iliza
tion
isin
dica
ted.
Land
ells
and
Van
Pete
ghem
,19
88(1
8)R
etro
spec
tive
revi
ewof
35pa
tient
sw
ithfr
actu
reof
the
atla
s.III
The
auth
ors
outli
nea
clas
sific
atio
nsc
hem
e:Ty
peI:
sing
lear
ch(1
6),
mos
tpr
eval
ent
and
mos
tof
ten
asso
ciat
edw
ithot
her
frac
ture
s.Ty
peII:
burs
tfr
actu
re(1
3),
mos
tof
ten
inis
olat
ion,
only
1/13
with
defic
it.Ty
peIII
:la
tera
lm
ass
frac
ture
(6).
Trea
tmen
tno
tst
anda
rdbu
t34
patie
nts
wer
etr
eate
dw
ithri
gid
exte
rnal
imm
obili
zatio
n.O
nly
1pa
tient
trea
ted
with
earl
ysu
rger
y(T
ype
Ifr
actu
retr
eate
dw
ithC
1–C
2fu
sion
).O
nepa
tient
requ
ired
surg
ery
info
llow
-up.
56%
ofpa
tient
sre
port
edsi
gnifi
cant
sym
ptom
sat
1yr
(nec
kpa
in,
scal
pdy
sest
hesi
as).
Aut
hors
argu
efo
rco
nser
vativ
em
easu
res
with
trac
tion
and
imm
obili
zatio
nw
ithca
refu
lfo
llow
-up.
Sega
let
al.,
1987
(28)
Ret
rosp
ectiv
ere
view
incl
udin
g8
isol
ated
atla
sfr
actu
res.
Med
ian
follo
w-u
p,46
mo.
III2/
4pa
tient
sw
itha
com
min
uted
frac
ture
,de
scri
bed
asa
unila
tera
lav
ulsi
onof
the
tran
sver
selig
amen
tat
tach
men
tan
dad
jace
ntar
chfr
actu
re,
deve
lope
da
nonu
nion
and
rem
aine
dsy
mpt
omat
icat
follo
w-u
p.Th
eau
thor
sre
com
men
dth
atth
ese
patie
nts
beco
nsid
ered
for
the
“mos
tef
fect
ive
imm
obili
zatio
n.”
Non
eof
the
patie
nts
unde
rwen
tsu
rgic
alfix
atio
n.
Kor
nber
g,19
86(1
7)R
epor
tof
asi
ngle
case
ofun
stab
leat
las
burs
tfr
actu
re.
IIIA
utho
rfe
els
that
fusi
onis
appr
opri
ate
for
unst
able
burs
tfr
actu
res
ofth
eat
las
(LM
D�
6.9
mm
)an
dde
scri
bes
aca
seof
post
erio
rar
chdi
srup
tion
inw
hich
they
wer
est
illab
leto
perf
orm
C1–
C2
post
erio
rfu
sion
beca
use
one
arch
rem
aine
dco
nnec
ted
tola
tera
lm
ass.
Schl
icke
and
Cal
laha
n,19
81(2
7)R
epor
tof
asi
ngle
case
ofun
stab
leat
las
burs
tfr
actu
re.
IIIC
ase
ofun
stab
lebu
rst
frac
ture
ofth
eat
las
with
LMD
of12
mm
.Th
eau
thor
spr
opos
ea
trea
tmen
tal
gori
thm
ofco
nsid
erin
gsu
rger
yfo
rdi
srup
tion
ofth
etr
ansv
erse
ligam
ent
(LM
D�
6.9
mm
).
aM
RI,
mag
neti
cre
sona
nce
imag
ing;
LMD
,la
tera
lm
ass
disp
lace
men
t;SO
MI,
subo
ccip
ital
-man
dibu
lar
imm
obili
zer.
S122 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
the rules of Spence (3, 4, 7). Dickman et al. (4) described twotypes of isolated transverse atlantal ligament injuries they iden-tified on MRI. In Type I, the substance of the ligament is injuredwithout associated fracture of the atlas. Type II involves anavulsion fracture of the atlas at the insertion of the transverseatlantal ligament. The authors concluded that patients with TypeI injuries should be treated with early surgical fixation because ofthe inherent instability at C1–C2 after ligamentous disruption.They favor rigid external immobilization for patients with TypeII ligament fracture injures. Applying MRI to their series of 39patients with atlas and/or axis fractures, the authors reportedthat the use of standard cervical x-rays and the rules of Spencewould have failed to identify 60% of the fractures they foundwith associated disruption of the transverse atlantal ligament (asdetermined by MRI) (3).
Lee et al. (20) described 16 patients with atlas fractures.These included six isolated anterior or posterior arch fractures(Landell’s Type I), four burst fractures (Landell’s Type II), andsix lateral mass fractures (Landell’s Type III). Twelve of the 16fracture injuries were isolated atlas fractures and were judgedto be stable as determined by integrity of the transverseligament either by MRI or by LMD criteria. All 12 weresuccessfully treated with rigid collar immobilization for 10 to12 weeks. The authors recommended a treatment algorithm ofcervical immobilization for stable atlas fractures and surgicalfixation and fusion for unstable atlas fractures and unstableC1–C2 combination fracture injuries. Their series, however,did not include any patient with an unstable isolated atlasfracture, nor any patient with an isolated atlas fracture whorequired surgical management.
Unusual isolated atlas fractures have been described in theliterature, often as radiographic curiosities (1, 11, 14, 19, 23).None of the cases for which clinical information was providedneeded surgical treatment. Fractures of the transverse processof the atlas have been described, including one of the casesdescribed by Jefferson in 1920 (2, 15). Although injury to thevertebral artery has been associated with fractures throughthe C1 transverse foramen, the bony C1 injury has not re-quired surgical fixation and has been treated with immobili-zation alone.
SUMMARY
There are no Class I or Class II studies that address themanagement of patients with isolated atlas fractures. All ofthe articles reviewed described case series or case reports thatprovide Class III evidence supporting several treatment strat-
egies for patients with acute C1 fracture injuries. Isolatedanterior or posterior atlas arch fractures and fractures of theatlas lateral mass have been effectively treated with externalcervical immobilization devices. Rigid collars, SOMI braces,and halo ring-vest orthoses have all been used for a durationof treatment of 8 to 12 weeks with good results. No study hasprovided evidence for using one of these devices over theother. Combined anterior and posterior arch fractures of theatlas (burst fractures) with an intact transverse atlantal liga-ment (implying C1–C2 stability) have been effectively man-aged with the use of a rigid collar, a SOMI brace, or a haloorthosis for a duration of 10 to 12 weeks. Combined anteriorand posterior arch fractures of the atlas (burst fractures) withevidence of transverse atlantal ligament disruption have beeneffectively treated with either rigid immobilization alone(halo orthosis) for a period of 12 weeks, or surgical stabiliza-tion and fusion. The type of C1–C2 internal fixation andfusion procedure performed may influence the need for andduration of postoperative immobilization.
Criteria proposed to determine transverse atlantal ligamentinjury with associated C1–C2 instability include the sum ofthe displacement of the lateral masses of C1 on C2 of morethan 8.1 mm on plain x-rays (rules of Spence corrected formagnification), a predental space of more than 3.0 mm inadults, and MRI evidence of ligamentous disruption oravulsion.
KEY ISSUES FOR FUTURE INVESTIGATION
The ability to identify the atlas fracture types at greatestrisk of nonunion and subsequent instability is a key issue indetermining appropriate management. Prospective data col-lection generating case-control studies at multiple institutionswould be feasible and useful in examining this issue. Therelative infrequency of isolated atlas fractures would make arandomized study less likely to be implemented. A uniformand clinically useful definition of instability in associationwith isolated atlas fractures would be of benefit. The sub-group of patients with isolated atlas fractures with transverseligament disruption that can be managed either by externalimmobilization alone or surgical fixation and fusion should beexamined in terms of long-term success, economic benefit,patient satisfaction, and return to preinjury activities. Therelative paucity of patients with atlas fractures treated with
TABLE 16.2. Treatment Options for Atlas Fractures
Atlas Fracture Type Treatment Options
Anterior or posterior arch fractures CollarAnterior and posterior arch (burst):
Stable (transverse atlantal ligament intact) Collar, haloUnstable (transverse atlantal ligament disrupted) Halo, C1–C2 stabilization and fusion
Lateral mass fractures:Comminuted fracture Collar, haloTransverse process fractures Collar
Isolated Fractures of the Atlas in Adults S123
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
surgical stabilization and fusion described in the literaturelimits the ability to address these issues at present.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tional Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Barker EG Jr, Krumpelman J, Long JM: Isolated fracture of themedial portion of the lateral mass of the atlas: A previouslyundescribed entity. AJR Am J Roentgenol 126:1053–1058, 1976.
2. Contostavlos DL: Massive subarachnoid hemorrhage due to lac-eration of the vertebral artery associated with fracture of thetransverse process of the atlas. J Forensic Sci 16:40–56, 1971.
3. Dickman CA, Sonntag VKH: Injuries involving the transverseatlantal ligament: Classification and treatment guidelines basedupon experience with 39 injuries. Neurosurgery 40:886–887,1997 (letter).
4. Dickman CA, Greene KA, Sonntag VKH: Injuries involving thetransverse atlantal ligament: Classification and treatment guide-lines based upon experience with 39 injuries. Neurosurgery38:44–50, 1996.
5. Fielding JW, Cochran GV, Lawsing JF III, Hohl M: Tears of thetransverse ligament of the atlas: A clinical and biomechanicalstudy. J Bone Joint Surg Am 56A:1683–1691, 1974.
6. Fowler JL, Sandhu A, Fraser RD: A review of fractures of theatlas vertebra. J Spinal Disord 3:19–24, 1990.
7. Greene KA, Dickman CA, Marciano FF, Drabier J, Drayer BP,Sonntag VKH: Transverse atlantal ligament disruption associ-ated with odontoid fractures. Spine 19:2307–2314, 1994.
8. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acutetraumatic atlas fractures: Management and long term outcome.Neurosurgery 23:31–35, 1988.
9. Han SY, Witten DM, Mussleman JP: Jefferson fracture of theatlas: Report of six cases. J Neurosurg 44:368–371, 1976.
10. Hays MB, Alker GJ Jr: Fractures of the atlas vertebra: The two-part burst fracture of Jefferson. Spine 13:601–603, 1988.
11. Hays MB, Bernhang AM: Fractures of the atlas vertebra: Athree-part fracture not previously classified. Spine 17:240–242,1992.
12. Heller JG, Viroslav S, Hudson T: Jefferson fractures: The role ofmagnification artifact in assessing transverse ligament integrity.J Spinal Disord 6:392–396, 1993.
13. Highland TR, Salciccioli GG: Is immobilization adequate treat-ment of unstable burst fractures of the atlas? A case report withlong-term follow-up evaluation. Clin Orthop 201:196–200, 1985.
13a. Hinchey JJ, Bickel WH: Fracture of the atlas: Review and pre-sentation of data on eight cases. Ann Surg 121:826–830, 1945.
14. Jakim I, Sweet MB, Wisniewski T, Gantz ED: Isolated avulsionfracture of the anterior tubercle of the atlas. Arch OrthopTrauma Surg 108:377–379, 1989.
15. Jefferson G: Fractures of the atlas vertebra: Report of four cases and areview of those previously reported. Br J Surg 7:407–422, 1920.
16. Kesterson L, Benzel EC, Orrison W, Coleman J: Evaluation andtreatment of atlas burst fractures (Jefferson fractures).J Neurosurg 75:213–220, 1991.
17. Kornberg M: Atypical unstable burst fracture of the atlas:Treated by primary atlantoaxial fusion. Orthop Rev 15:727–729,1986.
18. Landells CD, Van Peteghem PK: Fractures of the atlas: Classifi-cation, treatment and morbidity. Spine 13:450–452, 1988.
19. Lee C, Woodring JH: Unstable Jefferson variant atlas fractures:An unrecognized cervical injury. AJNR Am J Neuroradiol 12:1105–1110, 1991.
20. Lee TT, Green BA, Petrin DR: Treatment of stable burst fractureof the atlas (Jefferson fracture) with rigid cervical collar. Spine23:1963–1967, 1998.
21. Levine AM, Edwards CC: Fractures of the atlas. J Bone JointSurg Am 73A:680–691, 1991.
22. McGuire RA Jr, Harkey HL: Primary treatment of unstable Jef-ferson’s fractures. J Spinal Disord 8:233–236, 1995.
23. Merianos P, Tsekouras G, Koskinas A: An unusual fracture ofthe atlas. Injury 22:489–490, 1991.
24. Oda T, Panjabi MM, Crisco JJ III, Oxland TR: Multidirectionalinstabilities of experimental burst fractures of the atlas. Spine17:1285–1290, 1992.
25. Oda T, Panjabi MM, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part II—Relevance to clin-ical diagnosis and treatment. Spine 16[Suppl 10]:S466–S673,1991.
26. Panjabi MM, Oda T, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part I—Biomechanical anal-ysis of their mechanisms and fracture patterns. Spine 16[Suppl10]:S460–S465, 1991.
27. Schlicke LH, Callahan RA: A rational approach to burst fracturesof the atlas. Clin Orthop 154:18–21, 1981.
28. Segal LS, Grimm JO, Stauffer ES: Non-union of fractures of theatlas. J Bone Joint Surg Am 69A:1423–1434, 1987.
29. Seljeskog EL: Non-operative management of acute upper cervi-cal injuries. Acta Neurochir (Wien) 41:87–100, 1978.
30. Sherk HH, Nicholson JT: Fractures of the atlas. J Bone Joint SurgAm 52A:1017–1024, 1970.
31. Spence KF Jr, Decker S, Sell KW: Bursting atlantal fracture asso-ciated with rupture of the transverse ligament. J Bone Joint SurgAm 52A:543–549, 1970.
32. Zimmerman E, Grant J, Vise WM, Yashon D, Hunt WE: Treat-ment of Jefferson fracture with a halo apparatus: Report of twocases. J Neurosurg 44:372–375, 1976.
S124 Guidelines for Management of Acute Cervical Spinal Injuries
CHAPTER 17
Isolated Fractures of the Axis in Adults
RECOMMENDATIONSFRACTURES OF THE ODONTOID:Standards: There is insufficient evidence to support treatment standards.Guidelines: Type II odontoid fractures in patients 50 years and older should be considered for surgical
stabilization and fusion.Options: Type I, Type II, and Type III fractures may be managed initially with external cervical immobilization.
Type II and Type III odontoid fractures should be considered for surgical fixation in cases of densdisplacement of 5 mm or more, comminution of the odontoid fracture (Type IIA), and/or inability toachieve or maintain fracture alignment with external immobilization.
TRAUMATIC SPONDYLOLISTHESIS OF THE AXIS (HANGMAN’S FRACTURE):Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: Traumatic spondylolisthesis of the axis may be managed initially with external immobilization in
most cases. Surgical stabilization should be considered in cases of severe angulation of C2 on C3 (FrancisGrade II and IV, Effendi Type II), disruption of the C2–C3 disc space (Francis Grade V, Effendi Type III),or inability to establish or maintain alignment with external immobilization.
FRACTURES OF THE AXIS BODY (MISCELLANEOUS FRACTURES):Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options: External immobilization is recommended for treatment of isolated fractures of the axis body.
RATIONALE
Fractures of the axis represent unique cervical vertebralinjuries owing to the unique anatomy and biomechanicsof the C2 vertebra and the stresses applied to the dy-
namic atlantoaxial complex during trauma. Fractures of theaxis may be associated with other cervical fractures or liga-mentous injuries. Isolated fractures of the axis are commonand warrant independent consideration. Fractures of the axishave been divided into three general subtypes: fractures of theodontoid process, traumatic spondylolisthesis of the axis(hangman’s fractures), and miscellaneous non-odontoid non-hangman’s fractures of the C2 vertebra. Each of these fracturesubtypes has been further subdivided on the basis of theanatomic features and the functional significance of the indi-vidual fracture injury. The purpose of this review is to iden-tify evidence-based management strategies for each injurysubtype of traumatic fractures of the second cervical vertebra.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “spinal cord injury”
in combination with “axis,” “vertebrae,” “fracture,” and “hu-man” yielded 711 articles. Those manuscripts focusing on theclinical management of acute traumatic axis fractures wereselected for review. The bibliographies of these papers werescanned for additional references to confirm completeness ofthe literature review. Relevant papers addressing the mecha-nism of injury or the biomechanics and radiology of the C2vertebra were included. The articles were reviewed and classi-fied using established methodology. Thirty-eight articles forodontoid fracture, 17 for traumatic spondylolisthesis, and 8 formiscellaneous axis fractures provided the basis for the scientificfoundation of this guideline. Data from articles describing axisfractures and/or their management were categorized and areprovided in Tables 17.1 to 17.4). Fifteen additional articles arereferenced in the reference list as supporting information.
SCIENTIFIC FOUNDATION
Odontoid fractures
Overview
The most common traumatic axis injury is fracture throughthe odontoid process, either through the tip of the dens (Type
S125Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
I), through its base (Type II), or involving the odontoid butextending into the C2 body (Type III) (1, 6, 40). The anatomyand biomechanics of the C1–C2 complex provide for weight-bearing support for the head on the spine and for the mostmotion of any intervertebral unit in the cervical spine. Motionat C1–C2 is primarily rotational, accounting for one-half of theaxial rotation of the head on the neck (76). Translationalmotion of C1 on C2 is restricted by the transverse atlantalligament that approximates and secures the odontoid processto the anterior arch of the ring of C1. With a fracture of theodontoid process, restriction of translational movement of C1on C2 may be lost. Anterolisthesis or retrolisthesis of theC1-odontoid complex may occur relative to the body of C2. Ifsubstantial subluxation of C1 on C2 occurs, spinal cord injurymay result. The atlantoaxial complex is one of the most com-mon sites of dislocation in fatal cervical spinal injuries (21).
Earlier publications using evidence-based methodology forevaluating the literature on odontoid fracture managementhave focused on fusion as the primary outcome criterion witha minimum follow-up of 18 months (44, 72). Articles on odon-toid fractures containing this information were included inthis survey. Although it has been argued that the radiographicdetermination of fusion may be difficult and subject to ob-server variability, it seems to be the most appropriate out-come measure and is described in most of the clinical articlesaddressing odontoid fractures. It is recognized that outcomemeasures that incorporate patient satisfaction, quality of lifemeasures, and function would perhaps be superior, but thisinformation is sparse and less objective than the fusion criteriadescribed in the literature.
Classification of odontoid fractures
In 1974, Anderson and D’Alonzo (1) classified fractures ofthe odontoid into three types. This categorization has metwith general acceptance and remains in use with minor mod-ification. On the basis of a series of 49 patients managed from1954 through 1972 with an average follow-up of 22 months,
the authors defined three odontoid fracture types. Type Ifractures are oblique fractures through the upper portion ofthe odontoid process. Type II fractures cross the base of theodontoid process at the junction with the axis body. Type IIIfractures are fractures through the odontoid that extend intothe C2 body. The authors considered the Type III odontoidfracture to be more accurately described as a fracture of thebody of the axis. Using this scheme, the authors identified andtreated 2 Type I fractures, 32 Type II fractures, and 15 Type IIIfractures.
In 1988, Hadley et al. (38) added another fracture subtype tothis classification scheme. They described the Type IIA odon-toid fracture as a comminuted fracture involving the base ofthe dens with associated free fracture fragments. The inci-dence of a Type IIA fracture was estimated as 5% of all TypeII fractures (3 of 62 Type II fractures in their series) and wasassociated with severe instability and inability to obtain andmaintain fracture reduction and realignment. The authorsproposed that Type IIA odontoid fractures be managed withearly posterior surgical fixation and fusion of C1–C2.
Treatment
A variety of treatment strategies have been proposed forodontoid fractures based on the fracture type, the degree ofinitial dens displacement, the extent of angulation of the denswith respect to the body of C2, and the age of the patient.These include nonoperative and operative methods (1, 15, 16,30, 34, 38–40, 44, 50, 58, 72). Patients with odontoid fractureinjuries have been treated with external immobilization usinga variety of orthoses with varying results (1, 15, 16, 30, 34,38–40, 44, 50, 58, 72). Surgical options include posterior cer-vical fusion with or without transarticular screw fixation oranterior odontoid screw fixation techniques.
No treatment
In 1985, the Cervical Spine Research Society (16) publisheda multicenter review addressing the management of odontoidfractures. This report includes 18 patients with Type II odon-toid fractures and 3 patients with Type III odontoid fractureswho received no treatment. None of these cases achievedsubsequent bony fusion. The authors concluded that nontreat-ment of odontoid fractures should be eliminated as a man-agement choice.
Traction
Reviews by Traynelis (72) and Julien et al. (44) includeevidentiary tables describing seven articles containing ClassIII medical evidence addressing the treatment of odontoidfractures with traction and subsequent immobilization in acervical collar (1, 15, 16, 30, 34, 50, 58). All patients with TypeI odontoid fractures achieved radiographic fusion (3 of 3patients). Of patients with Type III fractures, 87% (55 of63 patients) achieved fusion. The failure rate for patients withType II fractures treated in this fashion was 43% (42 of 97patients); 57% achieved bony union. It seems that traction andthen cervical collar immobilization may be considered a man-agement option for patients with odontoid fractures, particu-
TABLE 17.1. Initial Management of Isolated Axis Fracture inthe Adult
Fracture Type Treatment Options
Odontoid fractureType I Collar, haloType II Consider for early
surgery or halo, collarType IIA Consider for early
surgery or haloType III Collar, halo, surgical
fusion
Traumatic spondylolisthesis of theaxis (hangman’s fracture)
Effendi Type I, Francis Type I, II Halo, collarEffendi Type II, III, Francis Type III,IV, V
Halo, consider surgicalstabilization
Miscellaneous axis fractures Collar or halo
S126 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE17
.2.
Sum
mar
yof
Rep
orts
onO
dont
oid
Frac
ture
s
Seri
es(R
ef.
No.
)St
udy
Des
ign
Evid
ence
Cla
ssC
omm
ents
And
erss
onet
al.,
2000
(2)
Ret
rosp
ectiv
eno
nran
dom
ized
repo
rtof
29pa
tient
sag
e�
65w
ithod
onto
id
frac
ture
sm
anag
edw
ithpo
ster
ior
fusi
on,
ante
rior
odon
toid
fixat
ion,
or
imm
obili
zatio
n.
IIIPo
ster
ior
fusi
onre
sulte
din
7/7
fuse
d(1
00%
),an
teri
orod
onto
idsc
rew
resu
lted
in3/
11fu
sed
(27%
),
and
halo
imm
obili
zatio
nre
sulte
din
3/10
fuse
d(3
0%).
The
auth
ors
argu
efo
rpo
ster
ior
fusi
onin
the
elde
rly
patie
ntw
ithan
odon
toid
frac
ture
.
Apf
elba
umet
al.,
2000
(3)
Ret
rosp
ectiv
ere
view
of2-
inst
itutio
nex
peri
ence
with
ante
rior
odon
toid
scre
w
fixat
ion.
147
odon
toid
frac
ture
s(T
ype
II[n
�13
8]an
dTy
peIII
[n�
9])
divi
ded
into
rece
nt(�
6m
o,12
9pa
tient
s)an
dre
mot
e(�
6m
ofr
omin
jury
,18
patie
nts)
grou
ps.
IIITh
efu
sion
rate
was
sign
ifica
ntly
high
erin
the
rece
ntgr
oup
com
pari
ngfu
sion
rate
sof
88%
vers
us25
%
(P�
0.05
)w
itha
mea
nfo
llow
-up
of18
mo.
Apo
sitiv
eco
rrel
atio
nw
asse
enbe
twee
nfu
sion
and
frac
ture
sor
ient
edin
the
hori
zont
alor
post
erio
rob
lique
plan
es.
No
effe
ctof
age,
sex,
num
ber
ofsc
rew
s
plac
ed,
ordi
spla
cem
ent
was
dem
onst
rate
d.
Dai
etal
.,20
00(2
0)R
evie
wof
57ca
ses
offa
iled
man
agem
ent
for
odon
toid
frac
ture
.III
50tr
eate
dw
ithoc
cipi
toce
rvic
alfu
sion
and
7w
ithat
lant
oaxi
alfu
sion
.2
case
sof
nonu
nion
afte
r
atla
ntoa
xial
fusi
onal
one.
38ac
hiev
edan
exce
llent
resu
lt.
Lenn
arso
net
al.,
2000
(46)
33pa
tient
sw
ithis
olat
edTy
peII
odon
toid
frac
ture
str
eate
dw
ithha
love
st
imm
obili
zatio
n.C
ases
defin
edas
nonu
nion
sin
halo
and
cont
rols
defin
edas
unio
ns.
IIC
ase-
cont
rol
Patie
nts
age
�50
yrha
da
risk
for
failu
re21
times
high
erth
anag
e�
50yr
.N
osi
gnifi
cant
diffe
renc
ein
med
ical
cond
ition
s,se
x,am
ount
offr
actu
redi
spla
cem
ent,
dire
ctio
nof
frac
ture
disp
lace
men
t,le
ngth
of
hosp
ital
stay
,or
leng
thof
follo
w-u
pbe
twee
ngr
oups
.
Julie
net
al.,
2000
(44)
Evid
ence
-bas
edre
view
ofm
anag
emen
tof
odon
toid
frac
ture
s.III
Aut
hors
conc
lude
that
ther
eis
insu
ffici
ent
evid
ence
for
stan
dard
sor
guid
elin
es.
Type
Ian
dIII
odon
toid
frac
ture
sca
nbe
man
aged
with
exte
rnal
imm
obili
zatio
n(1
00%
and
84%
fusi
on,
resp
ectiv
ely)
.A
nter
ior
fixat
ion
for
Type
IIIod
onto
idfr
actu
res
appe
ars
toim
prov
eth
efu
sion
rate
tone
arly
100%
.Ty
peII
frac
ture
sca
nbe
man
aged
with
exte
rnal
imm
obili
zatio
nw
ithan
expe
cted
fusi
onra
teof
appr
oxim
atel
y
65%
.Su
rgic
alin
stru
men
tatio
nan
dfu
sion
appe
arto
impr
ove
the
fusi
onra
tean
din
clud
epo
ster
ior
cerv
ical
fusi
on(7
4%)
oran
teri
orod
onto
idsc
rew
fixat
ion
(90%
)w
ithac
cept
able
mor
bidi
ty.
Cam
pane
lliet
al.,
1999
(14)
7pa
tient
sw
ithdi
spla
ced
Type
IIod
onto
idfr
actu
res
unde
rwen
tpo
ster
ior
tran
sart
icul
arsc
rew
fixat
ion.
III6/
7(8
6%)
achi
eved
rigi
dim
mob
iliza
tion.
1ve
rteb
ral
inju
ry.
The
auth
ors
conc
lude
that
this
isa
reas
onab
leop
tion.
Mul
ler
etal
.,19
99(5
6)R
etro
spec
tive
revi
ewof
23pa
tient
s�
70yr
ofag
ew
ithod
onto
idfr
actu
res.
IIIC
ompl
icat
ion
rate
sign
ifica
ntly
incr
ease
din
the
elde
rly
grou
p(5
2%ve
rsus
33%
)pr
imar
ilyow
ing
to
nonu
nion
afte
rno
nope
rativ
etr
eatm
ent.
The
auth
ors
sugg
est
that
the
elde
rly
patie
ntis
athi
ghri
skfo
r
mor
bidi
tyan
dm
orta
lity
and
sugg
est
earl
yha
lofix
atio
nor
prim
ary
stab
iliza
tion.
Mor
andi
etal
.,19
99(5
5)17
case
sof
odon
toid
scre
wfix
atio
n.III
Fusi
onin
16/1
7(9
4%).
The
auth
ors
sugg
est
patie
ntse
lect
ion
for
ante
rior
fixat
ion
beba
sed
onth
e
orie
ntat
ion
ofth
efr
actu
relin
e.
Suba
chet
al.,
1999
(69)
26pa
tient
s(m
ean
age,
35yr
)w
ithTy
peII
frac
ture
str
eate
dw
ithan
teri
orod
onto
id
scre
wfix
atio
n(s
ingl
esc
rew
)pl
usco
llar
(med
ian,
7.2
wk)
.
III25
/26
fusi
on(9
6%).
1pa
tient
requ
ired
post
erio
rfu
sion
for
inad
equa
tere
duct
ion.
Seyb
old
and
Bay
ley,
1998
(67)
Ret
rosp
ectiv
ere
view
of37
Type
IIan
d20
Type
IIIod
onto
idfr
actu
res
divi
ded
into
age
grou
ps.
Age
�60
and
�60
yr.
IIIA
ge�
60:
Type
II,7/
12fu
sed
(58%
).A
ge�
60:
Type
II,8/
11fu
sed
(73%
).Ty
peIII
:al
ltr
eate
dw
ith
halo
,95
%fu
sion
rega
rdle
ssof
type
.Fu
sion
rate
sdi
dno
tdi
ffer
sign
ifica
ntly
betw
een
the
2gr
oups
.Th
e
auth
ors
note
da
decr
ease
dto
lera
nce
inth
eel
derl
ypa
tient
for
halo
imm
obili
zatio
nan
d,as
are
sult,
favo
rsu
rger
yin
sele
cted
case
s.
Jenk
ins
etal
.,19
98(4
3)C
ompa
riso
nof
1ve
rsus
2sc
rew
sin
nons
elec
ted
patie
nts
with
Type
IIod
onto
id
frac
ture
sdi
vide
din
to2
grou
ps:
20(1
scre
w)
vers
us22
(2sc
rew
s).
Follo
w-u
p,9
mo.
IIITh
edi
ffere
nce
infu
sion
rate
,81
%(1
scre
w)
vers
us85
%(2
scre
ws)
,w
asno
tsi
gnifi
cant
.
Ber
lem
ann
and
Schw
arze
nbac
h,19
97(7
)
19pa
tient
s�
65yr
with
Type
IIod
onto
idfr
actu
res
trea
ted
with
ante
rior
odon
toid
scre
wfix
atio
nw
itha
follo
w-u
pof
4.5
yr.
IIIB
ony
fusi
onin
16/1
9(8
4%).
15/1
9(7
9%)
wer
eas
ympt
omat
ic.
The
auth
ors
favo
ran
teri
orfix
atio
nin
this
age
grou
p.
Tray
nelis
,19
97(7
2)Ev
iden
ce-b
ased
revi
ewof
Type
IIod
onto
idfr
actu
res.
IIIFi
rst
evid
ence
-bas
edre
port
onod
onto
idfr
actu
rem
anag
emen
t.In
dica
tes
that
insu
ffici
ent
data
wer
e
avai
labl
efo
rst
anda
rds
orgu
idel
ines
and
that
4tr
eatm
ent
optio
nsfo
rTy
peII
odon
toid
frac
ture
sw
ere
avai
labl
e,in
clud
ing:
trac
tion
follo
wed
byim
mob
iliza
tion,
imm
obili
zatio
nw
ithha
loor
Min
erva
,
post
erio
rce
rvic
alfu
sion
,or
ante
rior
scre
wfix
atio
n.Th
eau
thor
note
sth
atth
ehi
gher
fusi
onra
te
repo
rted
with
ante
rior
scre
wfix
atio
nm
ight
beof
fset
byits
high
erco
mpl
icat
ion
rate
and
lear
ning
curv
e.
Isolated Fractures of the Axis in Adults S127
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE17
.2.
Con
tinu
ed
Seri
es(R
ef.
No.
)St
udy
Des
ign
Evid
ence
Cla
ssC
omm
ents
Gre
ene
etal
.,19
97(3
6)R
etro
spec
tive
revi
ewof
340
case
sof
axis
frac
ture
s,in
clud
ing
199
odon
toid
frac
ture
s.
IIITy
peI:
2pa
tient
s,2/
2he
aled
with
halo
imm
obili
zatio
n(1
2w
k).
Type
II:12
0pa
tient
s,20
trea
ted
with
earl
ysu
rger
y.8
had
Type
IIw
ith�
6m
m,
4Ty
peIIA
with
inst
abili
tyde
spite
exte
rnal
imm
obili
zatio
n(1
Type
IIAtr
eate
dsu
cces
sful
lyw
itha
halo
),7
patie
nts
unde
rwen
tsu
rgic
alfu
sion
toav
oid
halo
imm
obili
zatio
n.95
wer
etr
eate
dw
ithex
tern
alim
mob
iliza
tion
(med
ian,
13w
k).
88w
ere
avai
labl
efo
r
follo
w-u
p.So
lidfu
sion
faile
din
25(2
8.4%
).7
wer
esu
cces
sful
lytr
eate
dw
ithad
ditio
nal
imm
obili
zatio
n,an
d18
wer
esu
cces
sful
lytr
eate
dw
ithpo
ster
ior
fusi
on(la
tesu
rger
y).
Sign
ifica
ntfa
ctor
was
dens
disp
lace
men
t�
6m
m(�
233
.74,
P�
0.00
01),
givi
ngan
86%
failu
rera
tein
the
halo
trea
tmen
tgr
oup;
5di
ed.
Type
III:
77pa
tient
s,69
man
aged
nono
pera
tivel
yw
ithex
tern
al
imm
obili
zatio
n.68
fuse
d(m
edia
n,12
wk)
.Th
eon
eth
atfa
iled
also
had
aC
1po
ster
ior
arch
frac
ture
and
requ
ired
post
erio
rfu
sion
.6
patie
nts
wer
etr
eate
dw
ithea
rly
surg
ery:
5be
caus
eth
eha
lofa
iled
to
mai
ntai
nal
ignm
ent
and
1be
caus
eof
aco
mbi
ned
C2–
C3
subl
uxat
ion.
2ha
dco
ncom
itant
late
ral
mas
s
frac
ture
sof
the
atla
sw
ithav
ulsi
onof
the
ligam
ento
usin
sert
ion
onth
etu
berc
le;
2di
ed.
Con
clus
ions
:
The
high
est
nonu
nion
rate
was
obse
rved
inTy
peII
odon
toid
disp
lace
d�
6m
m.
Surg
ery
was
reco
mm
ende
dfo
r1)
acut
efr
actu
rein
stab
ility
desp
iteex
tern
alim
mob
iliza
tion,
2)tr
ansv
erse
ligam
ent
disr
uptio
n,an
d3)
Type
IIod
onto
idfr
actu
rew
ith�
6m
mdi
spla
cem
ent.
Polin
etal
.,19
96(6
0)R
etro
spec
tive
revi
ewof
36Ty
peII
frac
ture
str
eate
dw
ithPh
ilade
lphi
aco
llar
(16)
or
halo
vest
imm
obili
zatio
n(2
0).
IIITy
peII:
54%
fuse
dw
ithco
llar.
74%
fuse
dw
ithha
lo.
Chi
baet
al.,
1996
(15)
104
patie
nts
with
odon
toid
frac
ture
s:Ty
peI,
2pa
tient
s.Ty
peII,
62pa
tient
s.Ty
pe
III,
32pa
tient
s.2
grou
ps:
Fres
hgr
oup,
72pa
tient
sw
hose
frac
ture
sw
ere
iden
tifie
d
with
in3
wk
oftr
aum
atic
even
t.O
ldgr
oup,
32pa
tient
sw
hoha
dan
exte
nded
peri
odbe
fore
defin
itive
trea
tmen
t:1
Type
I,21
Type
II,an
d8
Type
III.
IIITy
peI:
2pa
tient
s,co
llar
2/2,
both
fuse
d(1
00%
).Ty
peII:
62pa
tient
s,im
mob
iliza
tion
10/6
2,su
rger
y
52/6
2.In
fres
hfr
actu
regr
oup
trea
ted
with
surg
ery,
31/3
2fu
sed
(97%
).In
dela
yed
frac
ture
grou
p
trea
ted
with
surg
ery,
13/1
9fu
sed
(68%
).Ty
peIII
:32
patie
nts,
surg
ery
15/3
2(4
7%)
fuse
d,
imm
obili
zatio
n17
/32
(53%
)fu
sed,
10/1
5(6
6%)
trea
ted
with
surg
ery
fuse
d.11
/17
(65%
)tr
eate
dw
ith
imm
obili
zatio
nfu
sed.
Ever
ypa
tient
trea
ted
with
aha
lofu
sed,
5/5
(100
%).
Surg
ical
proc
edur
es:
66
patie
nts:
Post
erio
rce
rvic
alfu
sion
:10
patie
nts.
Type
II:9/
9su
cces
sful
fusi
ons
(100
%).
Type
III:
1/1
succ
essf
ulfu
sion
s(1
00%
).A
nter
ior
scre
wfix
atio
n:46
patie
nts.
36Ty
peII,
10Ty
peIII
,42
/45
patie
nts
achi
eved
bony
unio
n(9
3.3%
).Tr
anso
ral
fusi
on:
9pa
tient
s,6/
8Ty
peII.
Succ
essf
ulfu
sion
s(7
5%),
1/1
Type
IIIsu
cces
sful
fusi
on(1
00%
).Th
eau
thor
sof
this
larg
ese
ries
conc
lude
:Ty
peI
frac
ture
sca
n
gene
rally
bem
anag
edno
nope
rativ
ely.
Ant
erio
rsc
rew
fixat
ion
reco
mm
ende
dfo
rm
ost
Type
IIan
d
unst
able
Type
IIIfr
actu
res.
Con
trai
ndic
atio
nsin
clud
eol
des
tabl
ishe
dno
nuni
ons,
irre
duci
ble
frac
ture
s,
caud
aldi
spla
cem
ent,
seve
reos
teop
oros
is.
Type
IIIfr
actu
res
can
betr
eate
dw
ithha
loim
mob
iliza
tion
or
ante
rior
scre
wfix
atio
n.Es
tabl
ishe
dno
nuni
ons
and
irre
duci
ble
frac
ture
ssh
ould
betr
eate
dw
ith
post
erio
rfu
sion
.Tr
anso
ral
fusi
onre
serv
edfo
rra
reca
ses
ofan
teri
orco
rdco
mpr
essi
on.
Bed
nar
etal
.,19
95(5
)Pr
ospe
ctiv
ere
port
ofea
rly
surg
ical
stab
iliza
tion
in11
geri
atri
cpa
tient
sw
ith
odon
toid
frac
ture
s.
IIITh
eau
thor
ssu
gges
tth
atm
orta
lity
can
bere
duce
dby
surg
ical
inte
rven
tion
and
avoi
ding
the
use
of
halo
imm
obili
zatio
n.
Dic
kman
and
Sonn
tag,
1995
(22)
14pa
tient
sw
ithei
ther
acut
eor
suba
cute
Type
IIfr
actu
res
trea
ted
with
ante
rior
odon
toid
scre
wfix
atio
n.R
adio
grap
hic
crite
ria
for
fusi
on:
post
oper
ativ
ex-
rays
and
com
pute
dto
mog
raph
icsc
ans.
IIITy
peII:
14/1
4su
cces
sful
fusi
ons
(100
%).
Dic
kman
etal
.,19
95(2
3)D
escr
ibes
salv
age
proc
edur
esfo
rfa
iled
atla
ntoa
xial
nonu
nion
s.III
Rep
ort
incl
udes
2ca
ses
inw
hich
ante
rior
atla
ntoa
xial
tran
sart
icul
arsc
rew
sw
ere
used
and
8ca
ses
of
post
erio
rtr
ansa
rtic
ular
scre
ws.
Coy
neet
al.,
1995
(18)
15pa
tient
str
eate
dw
ithpo
ster
ior
wir
efu
sion
and
imm
obili
zed
post
oper
ativ
ely
in
eith
erPh
ilade
lphi
aco
llar
orha
lo.
Min
imum
follo
w-u
p,2
yr;
mea
n,4.
7yr
.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:ab
senc
eof
C1–
C2
mov
emen
ton
late
ral
flexi
on/
exte
nsio
nx-
rays
and
evid
ence
ofco
ntin
uity
oftr
abec
ular
bone
form
atio
nbe
twee
n
C1
and
C2
acro
ssth
egr
aft.
IIITy
peII:
13/1
4su
cces
sful
fusi
ons
(93%
).Ty
peIII
:2/
2su
cces
sful
fusi
ons
(100
%).
S128 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE17
.2.
Con
tinu
ed
Seri
es(R
ef.
No.
)St
udy
Des
ign
Evid
ence
Cla
ssC
omm
ents
Han
igan
etal
.,19
93(4
1)19
patie
nts
�80
yrof
age
with
odon
toid
frac
ture
s(1
6Ty
peII,
3Ty
peIII
).III
5pa
tient
sw
ithdi
spla
cem
ent
�5
mm
requ
ired
post
erio
rsu
rgic
alfix
atio
nw
ithgo
odre
sults
.Th
ere
was
27%
mor
talit
yin
the
cons
erva
tive
trea
tmen
tgr
oup,
with
prol
onge
dim
mob
iliza
tion
note
das
one
ofth
e
cont
ribu
ting
fact
ors.
Wad
dell
and
Rea
rdon
,
1983
(74)
24pa
tient
sw
ithod
onto
idfr
actu
re:
20Ty
peII
and
4Ty
peIII
frac
ture
s.16
/20
Type
IIfr
actu
res
wer
etr
eate
dw
ithC
1–C
2ar
thro
desi
s(G
allie
proc
edur
e).
All
Type
III
frac
ture
sw
ere
trea
ted
nono
pera
tivel
y.
IIITy
peII:
15/1
6su
cces
sful
fusi
ons
(94%
);1
patie
ntw
aslo
stto
follo
w-u
p.Ty
peIII
:3/
4su
cces
sful
fusi
ons
(75%
),1/
4no
nuni
on(2
5%).
Rya
nan
dTa
ylor
,19
93
(63)
30pa
tient
s�
60yr
ofag
ew
ithTy
peII
frac
ture
s.III
The
fusi
onra
tein
the
patie
nts
age
�60
trea
ted
with
imm
obili
zatio
nw
ason
ly7/
29(2
3%).
Des
pite
the
low
fusi
onra
tefo
rth
isag
egr
oup,
the
auth
ors
favo
rha
loim
mob
iliza
tion
over
surg
ical
fixat
ion.
Buc
holz
,19
81(1
2)26
patie
nts;
0Ty
peI,
17Ty
peII,
9Ty
peIII
.Pa
tient
sw
ere
imm
obili
zed
inha
lofo
r
am
inim
umof
3m
oan
d,if
nom
ovem
ent
onfle
xion
/ext
ensi
onx-
rays
,pl
aced
ina
Phila
delp
hia
colla
rfo
ran
addi
tiona
l4
wk.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:no
mov
emen
tor
subl
uxat
ion
atth
efr
actu
resi
teon
flexi
on/e
xten
sion
x-ra
ys.
IIITy
peII:
15/1
7su
cces
sful
fusi
ons
(88%
).2/
17no
nuni
ons
(12%
).Ty
peIII
:9/
9su
cces
sful
fusi
ons
(100
%).
3de
aths
:2
patie
nts
had
Type
IIfr
actu
res
whi
lebe
ing
trea
ted
inha
lo,
and
1pa
tient
with
Type
III
frac
ture
.
Had
ley
etal
.,19
88(3
8)R
etro
spec
tive
stud
yin
clud
ing
62pa
tient
sw
ithTy
peII
odon
toid
frac
ture
s,in
clud
ing
3w
ithco
mm
inut
ion
atth
eba
se.
IIITh
esu
bgro
upof
Type
IIod
onto
idfr
actu
rew
ithco
mm
inut
ion
atth
eba
sew
asde
fined
asth
eTy
peIIA
odon
toid
frac
ture
.Th
ecl
inic
alsi
gnifi
canc
eof
this
obse
rvat
ion
was
that
the
frac
ture
fuse
dpo
orly
with
imm
obili
zatio
nan
dw
asco
nsid
ered
for
earl
ysu
rger
y.
Gov
ende
ran
d
Gro
otbo
om,
1988
(34)
Rev
iew
of41
patie
nts
with
odon
toid
frac
ture
s:26
Type
II,15
Type
III.
1m
oin
trac
tion
(2–4
kg),
then
ari
gid
colla
rfo
r6–
8w
k,an
das
sess
edat
3m
o.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:bo
nyco
ntin
uity
acro
ssfr
actu
resi
tean
dno
mov
emen
ton
flexi
on/e
xten
sion
tom
ogra
ms.
IIITy
peII:
19/2
6su
cces
sful
fusi
ons
(73%
).2/
26fib
rous
unio
ns(8
%).
5/26
nonu
nion
s(1
9%).
Type
III:
15/1
5su
cces
sful
fusi
ons
(100
%).
No
mor
talit
y.7
halo
pin-
site
infe
ctio
ns.
3pa
tient
sha
dsk
in
exco
riat
ion
over
chin
seco
ndar
yto
halte
rtr
actio
n.
Fujii
etal
.,19
88(3
0)R
etro
spec
tive
revi
ewof
52pa
tient
sw
ithod
onto
idfr
actu
res,
incl
udin
gda
taon
24
trea
ted
with
imm
obili
zatio
n,10
trea
ted
with
ante
rior
scre
wfix
atio
n,an
d7
trea
ted
with
post
erio
rfu
sion
.R
adio
grap
hic
crite
ria
for
fusi
on:
ante
ropo
ster
ior
and
late
ral
tom
ogra
ms.
IIIIm
mob
iliza
tion:
Type
I:1/
1su
cces
sful
fusi
on(1
00%
).Ty
peII:
3/7
succ
essf
ulfu
sion
s(4
3%).
Type
III:
10/1
4su
cces
sful
fusi
ons
(72%
).Po
ster
ior
fusi
on:
Type
II:7/
7su
cces
sful
fusi
ons
(100
%).
Ant
erio
rsc
rew
fixat
ion:
Type
II:6/
8su
cces
sful
fusi
ons
(75%
).Ty
peIII
:2/
2su
cces
sful
fusi
ons
(100
%).
Lind
etal
.,19
87(4
8)R
evie
wof
14pa
tient
sw
ithod
onto
idfr
actu
res
man
aged
with
halo
imm
obili
zatio
n
and
eval
uate
dat
12w
kw
ithfle
xion
/ext
ensi
onx-
rays
.In
clud
ed9
Type
IIan
d5
Type
IIIfr
actu
res
with
a2-
yrfo
llow
-up.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:fle
xion
-
exte
nsio
nx-
rays
.
III10
/11
succ
essf
ulfu
sion
s(9
1%)
com
bine
dTy
peII
and
Type
IIIfr
actu
res.
Aut
hors
supp
ort
the
use
of
halo
imm
obili
zatio
nas
the
initi
altr
eatm
ent
for
Type
IIan
dIII
odon
toid
frac
ture
s.
Dun
nan
dSe
ljesk
og,
1986
(24)
Ret
rosp
ectiv
ere
port
of80
patie
nts
with
odon
toid
frac
ture
sin
clud
ing
data
on74
patie
nts
trea
ted
prim
arily
with
rigi
dbr
acin
gfo
r3–
6m
ofo
llow
edby
addi
tiona
l
colla
rsu
ppor
tfo
r6
wk
and
41pa
tient
sun
derg
oing
post
erio
rce
rvic
alfu
sion
.
Min
imum
follo
w-u
ppe
riod
was
6m
o;80
%of
the
patie
nts
had
follo
w-u
plo
nger
than
8m
o.R
adio
grap
hic
crite
ria
for
fusi
on:
late
ral
flexi
on-e
xten
sion
x-ra
ysat
3–4
mo.
IIIR
igid
imm
obili
zatio
n:Ty
peII:
40/5
9su
cces
sful
fusi
ons
(68%
).19
/59
nonu
nion
s(3
2%).
Type
III:
15/1
5
succ
essf
ulfu
sion
s(1
00%
).Po
ster
ior
fusi
on:
40/4
1su
cces
sful
fusi
ons
(98%
)fo
rco
mbi
ned
Type
IIan
d
Type
IIIfr
actu
res.
Cla
rkan
dW
hite
,19
85
(16)
Mul
ticen
ter
revi
ewin
clud
ing
144
patie
nts
man
aged
by27
diffe
rent
surg
eons
.
Fusi
onra
tes
repo
rted
base
don
frac
ture
type
and
trea
tmen
t.R
adio
grap
hic
crite
ria
for
fusi
on:
evid
ence
oftr
abec
ulat
ion
acro
ssth
efr
actu
resi
tean
dab
senc
eof
mov
emen
t
onla
tera
lfle
xion
-ext
ensi
onx-
rays
.
IIIN
otr
eatm
ent:
Type
II:0/
18su
cces
sful
fusi
ons
(0%
).Ty
peIII
:0/
3su
cces
sful
fusi
ons
(0%
).C
olla
r:Ty
pe
II:0/
3su
cces
sful
fusi
ons
(0%
).Typ
eIII
:5/
10su
cces
sful
fusi
ons
(50%
).Tr
actio
n:Ty
peII:
2/3
succ
essf
ul
fusi
ons
(66%
).Ty
peIII
:7/
8su
cces
sful
fusi
ons
(88%
).H
alo:
Type
II:25
/38
succ
essf
ulfu
sion
s(6
6%).
Type
III:
13/1
6su
cces
sful
fusi
ons
(81%
).A
nter
ior
fusi
on:
Type
II:7/
8su
cces
sful
fusi
ons
(88%
).Ty
peIII
:
2/2
succ
essf
ulfu
sion
s(1
00%
).Po
ster
ior
fusi
on:
Type
II:25
/26
succ
essf
ulfu
sion
s(9
6%).
Type
III:
4/4
succ
essf
ulfu
sion
s(1
00%
).
Isolated Fractures of the Axis in Adults S129
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE17
.2.
Con
tinu
ed
Seri
es(R
ef.
No.
)St
udy
Des
ign
Evid
ence
Cla
ssC
omm
ents
Pepi
net
al.,
1985
(58)
Ret
rosp
ectiv
ere
view
of41
patie
nts
with
odon
toid
frac
ture
sin
clud
ing
26tr
eate
d
cons
erva
tivel
yw
ithto
ngs,
four
-pos
ter
brac
e,co
llars
,an
d/or
halo
vest
s(0
Type
I,13
Type
II,13
Type
III).
12pa
tient
sun
derw
ent
post
erio
rce
rvic
alfu
sion
(1Ty
peI,
4
Type
II,an
d7
Type
III).
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:un
ion
onpl
ain
x-ra
ysan
d
tom
ogra
ms
asw
ell
asla
tera
lfle
xion
-ext
ensi
onvi
ews.
Non
unio
nw
asde
fined
as
mov
emen
tof
the
dens
frag
men
ton
late
ral
flexi
on-e
xten
sion
x-ra
ys.
IIIH
alo/
trac
tion:
Type
II:6/
13su
cces
sful
fusi
ons
(46%
).Ty
peIII
:11
/13
succ
essf
ulfu
sion
s(8
5%).
Post
erio
r
cerv
ical
fusi
on:
Type
I:1/
1su
cces
sful
fusi
ons
(100
%).
Type
II:4/
4su
cces
sful
fusi
ons
(100
%).
Type
III:
7/7
succ
essf
ulfu
sion
s(1
00%
).Th
eau
thor
sno
ted
that
the
halo
vest
sw
ere
poor
lyto
lera
ted
inpa
tient
s
age
�75
yr.
Wan
get
al.,
1984
(75)
Ret
rosp
ectiv
ere
view
of25
patie
nts
with
odon
toid
frac
ture
str
eate
dw
itha
vari
ety
of
cerv
ical
imm
obili
zatio
nte
chni
ques
.
IIITy
peI:
1/1
fuse
dw
ithha
lo(1
00%
).Ty
peII:
4/7
fuse
dw
ithco
llar
(57%
).4/
5fu
sed
inha
lo(8
0%).
Type
III:
2/2
fuse
dw
ithco
llar
(100
%).
10/1
2fu
sed
with
halo
(83%
).
Boh
ler,
1982
(9)
15pa
tient
s.W
ithod
onto
idfr
actu
res,
both
acut
ean
dch
roni
ctr
eate
dw
ithan
teri
or
scre
wfix
atio
nfo
llow
edby
ape
riod
ofce
rvic
alfix
atio
nin
apl
astic
colla
rfo
ra
peri
odof
4–16
wk.
Frac
ture
dist
ribu
tion:
0Ty
peI,
8Ty
peII,
and
7Ty
peIII
.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:no
tgi
ven.
IIITy
peII:
8/8
succ
essf
ulfu
sion
s(1
00%
).Ty
peIII
:7/
7su
cces
sful
fusi
ons
(100
%).
Mai
man
and
Lars
on,
1982
(49)
Ret
rosp
ectiv
ere
view
of49
case
sof
odon
toid
frac
ture
,in
clud
ing
34Ty
peII
frac
ture
str
eate
dw
ithea
rly
post
erio
rw
ire/
graf
tst
abili
zatio
n.Po
st-o
pera
tive
imm
obili
zatio
nw
itha
Min
erva
for
anav
erag
eof
5w
k.2
Type
IIIfr
actu
res
wer
e
incl
uded
.R
adio
grap
hic
crite
ria
for
nonu
nion
:to
mog
raph
icev
iden
ceof
avas
cula
r
necr
osis
,gr
oss
inst
abili
tyw
itha
dem
onst
rabl
ega
pat
the
frac
ture
line,
and
no
evid
ence
ofhe
alin
g.Fu
sion
resu
ltsev
alua
ted
6m
opo
stsu
rger
y.
IIITh
eau
thor
sob
serv
eda
100%
fusi
onra
teat
the
post
erio
rsu
rgic
alsi
te,
but
only
a35
%fu
sion
rate
acro
ssth
efr
actu
resi
te.
Rya
nan
dTa
ylor
,19
82
(62)
Ret
rosp
ectiv
ere
view
of23
patie
nts
with
odon
toid
frac
ture
sov
era
10-y
rpe
riod
,
incl
udin
g1
Type
I,16
Type
II,an
d6
Type
III.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:no
mov
emen
ton
late
ral
flexi
on-e
xten
sion
x-ra
ys.
IIIH
alo/
Min
erva
/Sub
occi
pita
l-m
andi
bula
rim
mob
ilize
r:Ty
peI:
1/1
succ
essf
ulfu
sion
.Ty
peII:
9/15
succ
essf
ulfu
sion
s(6
0%).
Type
III:
6/6
succ
essf
ulfu
sion
s(1
00%
).
Ekon
get
al.,
1981
(26)
Ret
rosp
ectiv
ere
view
of22
case
sof
odon
toid
frac
ture
trea
ted
with
halo
imm
obili
zatio
nfo
r3
mo.
Type
I,0
patie
nts.
Type
II,16
patie
nts.
Type
III,
6
patie
nts.
Incl
udes
outc
ome
on17
patie
nts
with
anav
erag
efo
llow
-up
of30
mo.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:la
tera
lfle
xion
-ext
ensi
onx-
rays
.
IIITy
peII:
6/12
succ
essf
ulfu
sion
s(5
0%).
Type
III:
4/5
succ
essf
ulfu
sion
s(8
0%).
Mar
aran
dTa
y,19
76(5
0)R
evie
wof
26ca
ses
ofod
onto
idfr
actu
rein
clud
ing
24Ty
peII
and
2Ty
peIII
trea
ted
with
cerv
ical
trac
tion
for
�10
wk.
Rad
iogr
aphi
ccr
iteri
afo
rfu
sion
:fib
rous
unio
nat
frac
ture
site
.
IIITy
peII:
9/24
succ
essf
ulfu
sion
s(3
7.5%
).Ty
peIII
:2/
2su
cces
sful
fusi
ons
(100
%).
And
erso
nan
dD
’Alo
nzo,
1974
(1)
Ret
rosp
ectiv
ere
view
of49
patie
nts
with
odon
toid
frac
ture
scl
assi
fied
into
Type
I,II,
and
IIIba
sed
onfr
actu
re.
IIIN
on-o
pera
tive
trea
tmen
t:37
patie
nts.
Type
I:co
llar/
brac
e,2/
2su
cces
sful
fusi
ons
(100
%).
Type
II:ha
lo,
14/2
2su
cces
sful
fusi
ons
(64%
).8/
22no
nuni
ons
(36%
).Ty
peIII
:ha
lo,
12/1
3su
cces
sful
fusi
ons
(92%
).
1/13
nonu
nion
(8%
).O
pera
tive
trea
tmen
t:12
patie
nts.
Type
II:8/
10su
cces
sful
fusi
ons
(80%
).Ty
peIII
:
2/2
(100
%).
S130 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE17
.3.
Sum
mar
yof
Rep
orts
onTr
aum
atic
Spon
dylo
listh
esis
ofth
eA
xis
Seri
es(R
ef.
No.
)St
udy
Des
ign
Evid
ence
Cla
ssC
omm
ents
Bar
ros
etal
.,19
99(4
)C
ase
repo
rtof
surg
ical
fixat
ion
inha
ngm
an’s
frac
ture
.III
Surg
ical
trea
tmen
tfo
rha
ngm
an’s
frac
ture
isan
optio
n.
Ver
hegg
enan
dJa
nsen
,19
98(7
3)R
etro
spec
tive
stud
yof
16pa
tient
str
eate
dw
ithea
rly
post
erio
rsc
rew
fixat
ion
ofth
e
neur
alar
chaf
ter
hang
man
’sfr
actu
re.
IIITh
eau
thor
ssu
gges
tth
atth
isis
the
optim
alth
erap
yfo
rEd
war
dsan
dLe
vine
(Effe
ndi)
Type
IIan
dIII
frac
ture
s,de
scri
bing
exce
llent
resu
ltsin
thei
rse
ries
.
Gre
ene
etal
.,19
97(3
6)34
0ca
ses
ofax
isfr
actu
res,
incl
udin
g74
patie
nts
with
trau
mat
icsp
ondy
lolis
thes
isof
the
axis
.Fo
llow
-up
avai
labl
eon
72pa
tient
s.
IIIM
ost
com
mon
:Ef
fend
iTy
peI
(72%
),Fr
anci
sG
rade
I(6
5%).
65tr
eate
dsu
cces
sful
lyw
ith
imm
obili
zatio
n(1
2w
k).
7re
quir
edea
rly
surg
ery
(pos
teri
orfu
sion
)ow
ing
topo
oral
ignm
ent
inth
eha
lo
(Effe
ndi
II,6
patie
nts;
Effe
ndi
III,
1pa
tient
;Fr
anci
sG
rade
I,1
patie
nt;
II,1
patie
nt;
III,
2pa
tient
s;IV
,3
patie
nts)
.33
%of
all
Effe
ndi
Type
sII
and
IIIan
d36
%of
all
Fran
cis
Type
sIII
,IV
,V
patie
nts
requ
ired
surg
ery.
Stro
ngco
rrel
atio
nob
serv
edbe
twee
nEf
fend
iI
and
Fran
cis
Ian
dEf
fend
iIII
and
Fran
cis
IV.
Con
clus
ions
:Im
mob
iliza
tion
isge
nera
llysu
ffici
ent
trea
tmen
t.Su
rger
ym
aybe
cons
ider
edfo
rse
vere
Fran
cis
orEf
fend
ity
peha
ngm
an’s
frac
ture
s.
Cor
icet
al.,
1996
(17)
Ret
rosp
ectiv
ere
view
ofha
ngm
an’s
frac
ture
incl
udin
g39
nond
ispl
aced
(�6
mm
C2
onC
3)tr
eate
dw
ithno
nrig
idim
mob
iliza
tion
(Phi
lade
lphi
aco
llar
for
anav
erag
eof
12
wk)
and
10di
spla
ced
(�6
mm
)tr
eate
dw
ithha
lo(3
),co
llar
(6),
orsu
rger
y(1
).
IIIN
ondi
spla
ced
grou
p:39
/39
fuse
dus
ing
colla
ral
one.
Dis
plac
edgr
oup:
also
fuse
dre
gard
less
of
trea
tmen
t.C
1–C
3fu
sion
requ
ired
in1
patie
ntfo
rfa
ilure
ofcl
osed
redu
ctio
n.
Star
ran
dEi
smon
t,19
93(6
8)R
evie
wof
19ca
ses
ofax
isfr
actu
rein
clud
ing
6ca
ses
ofa
patte
rnoc
curr
ing
thro
ugh
the
post
erio
ras
pect
ofth
eve
rteb
ral
body
cont
inui
tyof
the
post
erio
rco
rtex
with
subl
uxat
ion
resu
lting
inna
rrow
ing
ofth
esp
inal
cana
l.
IIIH
angm
an’s
frac
ture
vari
atio
noc
curr
edin
6/19
patie
nts,
incl
udin
g2
with
spin
alco
rdin
jury
from
the
asso
ciat
edsu
blux
atio
n.
Tan
and
Bal
acha
ndra
n,19
92(7
0)R
etro
spec
tive
stud
yof
33pa
tient
sw
ithha
ngm
an’s
frac
ture
.C
lass
ified
byEf
fend
i:21
Type
I,11
Type
II,an
d1
Type
III.
III20
/26
had
none
urol
ogic
alde
ficit
onad
mis
sion
.28
/33
with
com
plet
ere
cove
ryaf
ter
1yr
.
Torr
eman
,19
90(7
1)Lo
ng-t
erm
stud
yof
23pa
tient
sw
ithha
ngm
an’s
frac
ture
str
eate
dw
ithim
mob
iliza
tion.
Ave
rage
follo
w-u
p,9.
6yr
.
III10
0%lo
ng-t
erm
fusi
onra
tew
ithce
rvic
alim
mob
iliza
tion.
Gov
ende
ran
dC
harl
es,
1987
(33)
Pros
pect
ive
stud
yof
39pa
tient
s.III
All
patie
nts
succ
essf
ully
man
aged
with
trac
tion
and
imm
obili
zatio
n.
Gra
dyet
al.,
1986
(35)
Ret
rosp
ectiv
ere
view
of27
patie
nts
incl
udin
g16
man
aged
with
halo
,8
with
aco
llar,
and
3w
ithbe
dres
t.
IIIA
llac
hiev
edfu
sion
with
nore
sidu
alsy
mpt
oms.
The
auth
ors
reco
mm
end
the
use
ofa
Phila
delp
hia
colla
ral
one
infr
actu
res
with
min
imal
disp
lace
men
t.
Levi
nean
dEd
war
ds,
1985
(47)
Ret
rosp
ectiv
eca
sese
ries
of52
patie
nts
with
trau
mat
icsp
ondy
lolis
thes
isof
the
axis
clas
sifie
dus
ing
the
Effe
ndi
crite
ria.
IIIIs
olat
edTy
peI,
II,an
dIIa
wer
eal
lm
anag
edno
nope
rativ
ely.
3of
5Ty
peIII
patie
nts
unde
rwen
tsu
rgic
al
stab
iliza
tion
for
failu
reto
obta
inor
mai
ntai
nre
duct
ion
ina
halo
.Th
eau
thor
sid
entif
yth
eTy
peIIa
subg
roup
ofth
eEf
fend
iTy
peII
patie
nts
who
dist
ract
sign
ifica
ntly
with
the
appl
icat
ion
oftr
actio
nan
d
note
the
mec
hani
smof
inju
ryfo
rth
isgr
oup
islik
ely
flexi
on-d
istr
actio
n.3/
3Ty
peIIa
patie
nts
wer
e
trea
ted
with
gent
leex
tens
ion
and
com
pres
sion
unde
rflu
oros
copi
cgu
idan
cefo
llow
edw
ithha
lo
imm
obili
zatio
n.
Bor
neet
al.,
1984
(10)
Ret
rosp
ectiv
ere
view
of18
case
sof
“ped
icle
”fr
actu
reof
the
axis
trea
ted
with
dire
ct
inte
rnal
fixat
ion.
IIIA
ggre
ssiv
esu
rgic
alap
proa
chfo
rfix
atio
nof
pedi
cle-
isth
mus
frac
ture
sof
the
axis
with
100%
fusi
onra
te.
Mes
tdag
het
al.,
1984
(52)
Com
bine
dcl
inic
alan
dan
atom
icst
udy
desc
ribi
ng41
frac
ture
sof
the
post
erio
rne
ural
arch
ofth
eax
is.
11ca
ses
trea
ted
with
ante
rior
C2–
C3
inte
rbod
yfu
sion
.30
trea
ted
with
trac
tion
and
imm
obili
zatio
n.Fo
llow
-up
avai
labl
eon
30pa
tient
s.
IIIC
adav
eric
stud
yde
mon
stra
ted
that
frac
ture
sw
ithdi
spla
cem
ent
of�
5m
mw
ere
stab
le.
Cer
vica
l
mob
ility
was
mai
ntai
ned
bette
rin
the
cons
erva
tive
man
agem
ent
grou
p.Th
eau
thor
sre
com
men
d
cons
erva
tive
mea
sure
sex
cept
inca
ses
ofm
arke
din
stab
ility
orno
nuni
on.
Fran
cis
etal
.,19
81(2
9)C
lass
ifica
tion
pape
rba
sed
on12
3ca
ses
offr
actu
res
ofth
epo
ster
ior
arch
ofth
eax
is.
Gra
deba
sed
ondi
spla
cem
ent
and
angu
latio
n.
IIIG
rade
I(1
5%of
tota
lse
ries
):0%
nonu
nion
with
imm
obili
zatio
n.G
rade
II(7
%):
33%
nonu
nion
.G
rade
III(3
7%):
0%no
nuni
on.
Gra
deIV
(34%
):2%
nonu
nion
.G
rade
V(6
%):
28%
nonu
nion
.
Buc
holz
,19
81(1
2)A
utop
syst
udy
of17
0ca
ses
oftr
aum
atic
deat
h.III
38ha
dce
rvic
alsp
ine
frac
ture
san
d8/
38ha
dtr
aum
atic
spon
dylo
listh
esis
ofth
eax
is.
Pepi
nan
dH
awki
ns,
1981
(57)
Def
ined
anea
rly
clas
sific
atio
nsc
hem
efo
rha
ngm
an’s
frac
ture
base
don
42ca
ses.
IIITy
peI:
ano
ndis
plac
edfr
actu
reof
the
post
erio
rel
emen
ts.
Type
II:di
spla
ced
frac
ture
invo
lvin
gpo
ster
ior
and
ante
rior
stru
ctur
es.
All
42pa
tient
sw
ere
succ
essf
ully
trea
ted
nons
urgi
cally
.
Effe
ndi
etal
.,19
81(2
5)C
lass
ifica
tion
pape
rba
sed
on13
1ca
ses
ofpa
tient
sw
ithfr
actu
res
ofth
eri
ngof
the
axis
.Fr
actu
res
divi
ded
into
3gr
oups
base
don
mec
hani
smof
inju
ry,
disp
lace
men
t,
and
stab
ility
.
IIITy
peI
(65%
ofto
tal
grou
p):
Isol
ated
hair
line
frac
ture
sof
the
ring
ofth
eax
isw
ithm
inim
aldi
spla
cem
ent
ofth
ebo
dyof
C2
caus
edby
axia
llo
adin
gan
dhy
pere
xten
sion
.Ty
peII
(28%
):D
ispl
acem
ent
ofan
teri
or
frag
men
tw
ithdi
srup
tion
ofth
edi
scsp
ace
belo
wth
eax
isca
used
byhy
pere
xten
sion
and
rebo
und
flexi
on.
Type
III(7
%).
Dis
plac
emen
tof
ante
rior
frag
men
tw
ithC
2–C
3fa
cet
disl
ocat
ion
caus
edby
prim
ary
flexi
onan
dre
boun
dex
tens
ion.
Alth
ough
5pa
tient
sun
derw
ent
surg
ery,
the
auth
ors
conc
lude
that
the
vast
maj
ority
ofth
ese
patie
nts
are
best
man
aged
with
cerv
ical
imm
obili
zatio
n.
Bra
shea
ret
al.,
1975
(11)
29pa
tient
sw
ithha
ngm
an’s
frac
ture
sfo
llow
edfo
ran
aver
age
of6
yr.
IIIN
oca
seof
neur
olog
ical
defic
it.23
/23
(100
%)
trea
ted
with
imm
obili
zatio
nac
hiev
edfu
sion
.Su
ppor
ts
nono
pera
tive
man
agem
ent.
Isolated Fractures of the Axis in Adults S131
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
larly those with Type I and Type III fractures. The low fusionsuccess rate reported for Type II odontoid fractures managedwith traction and collar immobilization (57%) implies thatperhaps collar immobilization is not the ideal strategy forType II fracture patients.
Cervical collar
Several authors have proposed treatment of odontoid frac-tures with cervical collars. Polin et al. (60) in 1996 describe aseries of 36 Type II fractures treated with either a Philadelphiacollar or halo vest immobilization. The fusion rate was lowerin the patients treated with collars (53%) compared with 74%for patients managed in halos. An earlier report from the sameinstitution described a similar rate of fusion (57%) in a studyincluding seven Type II fractures treated with a collar alone(75). The infrequent Type I odontoid fracture seems to have anacceptable rate of fusion with rigid cervical collar immobili-zation, approaching 100% in one study (1, 15, 16). Type IIIodontoid fractures have been treated with cervical collars aswell, but they have a less favorable union rate, with fusionrates ranging from 50 to 65% in small series (16, 75).
Halo immobilization
In a series of publications resulting in the largest institu-tional series of axis fractures published to date, 340 cases ofaxis fractures were reviewed, including 199 odontoid frac-tures (2 Type I, 116 Type II, 4 Type IIA, 77 Type III (36, 39, 40).Excellent results were obtained with rigid external immobili-zation in the Type I and Type III fracture patients (2 of 2 and68 of 69 patients with successful fusion, respectively). Of theType II patients, 95 were treated with external immobilizationfor a median of 13 weeks. The authors reported a 28% failurerate. Seven failures were successfully treated with additionalexternal immobilization, and 18 patients underwent subse-quent posterior C1–C2 fusion. The authors found that a dis-placement of the dens of 6 mm or more was associated with ahigh nonunion rate, irrespective of patient age, direction ofdisplacement, or neurological deficit (86% failure rate; �2
33.74; P � 0.001). The degree of dens displacement and anegative correlation with fusion was noted by at least fourother investigators (16, 24, 26, 48). The amount of odontoiddisplacement observed ranged from 2 to 6 mm in thesestudies.
TABLE 17.4. Summary of Reports on Miscellaneous Axis Fractures
Series (Ref. No.) Study DesignEvidence
ClassComments
Greene et al., 1997 (36) 340 cases of axis fractures, including 67non-odontoid, non-hangman’s fractures(miscellaneous), most involving the bodyof lateral masses.
III 60/61 (98%) were successfully treated withexternal mobilization in all but 1 patient (1.6%nonfusion rate). 4 patients died, and 1underwent early surgery for 5-mm luxation ofC2 on C3.
Fujimura et al., 1996 (31) 31 cases of axis body fracturescategorized into 4 types based onradiographic imaging.
III 4 types: Avulsion: 9/9 fused withimmobilization. Transverse: 2/2 healed withimmobilization. Burst: 2/3 treated with C2–C3fusion. Sagittal fractures: 15/17 healed withimmobilization. 8 sagittally oriented fracturepatients still had pain despite a bony union.
Benzel et al., 1994 (6) Retrospective report of 15 patientsdescribed with fractures of the axis body.
III The authors propose classification into: Type 1:coronal (n � 12). Type 2: sagittal (n � 3). Type3: oblique and equivalent to the Type IIIodontoid fracture.
Korres et al., 1994 (45) 14 cases of avulsion fracture of theanteroinferior portion of the axissecondary to extension-type injuries.Mean follow-up, 8.5 yr.
III 3% of the cervical spine trauma cases over a12-yr period. All patients treated successfullywith cervical immobilization.
Bohay et al., 1992 (8) Describes 3 cases of vertical fractures ofthe axis.
III Notes that this is an unusual variant fracture ofthe axis body. All treated with immobilization.
Craig and Hodgson, 1991(19)
Describes 9 cases of superior facetfracture of the axis vertebra.
III 5 treated with reduction and immobilization. 3required open reduction and posterior fusion.
Burke and Harris, 1989(13)
Review of 165 patients with axisfractures. 31 miscellaneous bodyfractures identified and classified onmechanism of injury.
III Identified 31 patients with axis body fractures.21/38 (68%) were extension teardrop; 10/31(32%) were hyperextension.
Jakim and Sweet, 1988(42)
Case report of a transverse fracture of theaxis and literature review. Aclassification scheme is proposed.
III 3 types of axis body fractures were described:the Type III odontoid fracture of Anderson andD’Alonzo, the transverse body fracture, and theavulsion fracture.
S132 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Julien et al. (44) reviewed nine articles that dealt withtreatment of odontoid fractures (total, 269 patients) usinghalo/Minerva fixation for 8 to 12 weeks (12, 15, 16, 24, 26, 30,48, 58, 62). All patients with Type I odontoid fractures werefound to have successful fusion (3 of 3 patients) (15, 30, 62).One hundred sixty-eight patients with Type II odontoid frac-tures were treated with halo immobilization; 110 (65%) hadsuccessful fusion. There was a 30% nonunion rate (50 of 168patients). Eight patients were described as having a malunion.Of patients with Type III odontoid fractures, 84% (67 of 80patients) achieved a solid fusion. There was an 8% failure rate(6 of 80 patients), and 7 cases were described as malunions.The authors of these series generally concluded that rigidexternal immobilization can be considered a viable treatmentoption for Type I, Type II, and Type III odontoid fractures.Rigid external immobilization seems to be most successful forpatients with Type I, Type III, and nondisplaced Type IIodontoid fractures, but it should be considered with cautionin elderly patients.
Posterior cervical fixation
Posterior cervical fixation and fusion has been successfullyused in the treatment of acute traumatic odontoid fractures.Although no criteria defining the indications for surgical fix-ation have been established, a number of retrospective caseseries suggest treatment options (15, 16, 18, 24, 30, 49, 58, 74).These papers describe a total of 147 patients who underwentposterior cervical fixation and fusion for Type II odontoidfractures and 29 patients treated similarly for Type III frac-tures. One patient with a Type I fracture was treated success-fully with posterior fusion. The overall fusion rates for Type IIand Type III fractures managed with surgical fixation andfusion were 87 and 100%, respectively, in these series. Thereport of Maiman and Larson (49) described a fusion rate ofonly 35% across the fracture line but a fusion rate of 100% atthe posterior operative site.
The aforementioned series typically describe an instru-mented (wire or cable) posterior C1–C2 arthrodesis and thencervical immobilization in a rigid orthosis. More recently,transarticular screw fixation and fusion of C1–C2 has beenused for traumatic odontoid fractures, particularly in cases offailed fusion after initial management (14, 43). The reportedsurgical morbidity and mortality is 2 to 4% and includesfailure of fracture reduction, vertebral artery injury, and thenew onset of neurological deficit. Loss of motion at the atlan-toaxial joint after posterior C1–C2 fusion results from of dor-sal C1–C2 arthrodesis. Despite this, several authors favorposterior C1–C2 fusion rather than anterior odontoid screwfixation as the ideal treatment of unstable odontoid fractures(2, 14, 56).
Anterior cervical fixation
Anterior single and double screw fixation of odontoid frac-tures has been accomplished with success. The technical chal-lenges associated with this procedure have limited wide-spread application. If successful, this technique has thepotential to maintain rotational motion at the atlantoaxial
joint. It has been suggested that this is an appropriate strategywhen the odontoid fracture line is either horizontal or obliqueand posterior and that it is contraindicated in situations wherethe fracture line is oblique and anterior (2, 3, 20, 55). In casesof transverse atlantal ligament disruption, anterior screw fix-ation can result in an unsatisfactory outcome despite union ofthe odontoid fracture owing to persistent transverse atlantalligament incompetence. Julien et al. (44) summarized a seriesof articles that describe retrospective experiences with ante-rior screw fixation for odontoid fractures (9, 15, 22, 30, 43). Thecombined fusion rate of Type II fractures treated in thesereports is 89% (112 of 126 patients). Patients with Type IIIodontoid fractures achieved radiographic fusion in 20 of 20patients (100%). In a recent series reported by Subach et al.(69), 26 patients with Type II fractures (mean age, 35 yr)underwent anterior odontoid fixation with a single screw andthen immobilization in a cervical collar (median, 7.2 wk).Twenty-five (96%) of 26 patients achieved successful fusion.The one failure was attributed to inadequate fracture reduc-tion. That patient required subsequent posterior C1–C2 fu-sion. Jenkins et al. (43), in 1998, described a retrospectivenonrandomized series of 42 patients undergoing anteriorscrew fixation for Type II odontoid fractures. The authorscompared single-screw with two-screw techniques. The fu-sion rate in their experience was similar for single-screwfixation (81%) compared with two-screw fixation (85%). Useof lag screws to achieve anterior odontoid fixation is recom-mended. Complications of the procedure include retropha-ryngeal wall injury, screw fracture, infection, and screw mis-placement with injury to surrounding vascular and neuralstructures (9, 22, 30). Attempts at anterior odontoid fixationusing a transoral approach was associated with multiple sig-nificant complications (15).
Apfelbaum et al. (3) compared anterior screw fixation forrecent and remote odontoid fractures at two institutions. Onehundred forty-seven patients with Type II (n � 138) and TypeIII (n � 9) odontoid fractures underwent anterior screw fixa-tion either within 6 months of injury (129 patients) or morethan 18 months after injury (18 patients). The fusion rateswere 88% in the �6-month group versus 25% in the remotefracture injury group (P � 0.05), with a mean follow-up of 18months. A positive correlation was identified between fusionand fractures oriented in the horizontal or posterior obliqueplanes. No effect of age, sex, number of screws placed, ordegree of dens displacement was identified. Their experiencesuggests that anterior odontoid screw fixation for odontoidfractures is most effective when performed early after injury,particularly within 6 months of fracture.
Odontoid fracture management in the elderly patient
One of the controversial issues in the management of odon-toid fractures is the influence of age on treatment selection. Anumber of studies have examined the circumstance of acuteodontoid fracture in the older patient. Three case series argueagainst surgical fixation in the elderly patient (36, 62, 67).Seven other case series favor surgical fixation in this agegroup. There is also one case-control study by Lennarson et al.
Isolated Fractures of the Axis in Adults S133
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
(46) providing Class II medical evidence for surgical treat-ment of elderly patients. Ryan and Taylor (63) described 30patients 60 years and older with Type II odontoid fractures.The fusion success rate in patients older than 60 years treatedwith external immobilization was only 23%. The authorsthought that the high fracture nonunion rate was secondary toinadequate immobilization and delays in diagnosis in mostcases. If these issues were eliminated, no significant differencein outcome between surgical and nonsurgical managementwould have been demonstrated. They concluded that surgicalfixation and fusion for elderly patients with odontoid frac-tures should be reserved for unusual circumstances. Greene etal. (36) reported the largest series (120 patients) of retrospec-tively reviewed cases of traumatic odontoid Type II axis frac-tures. Patients with dens displacement of 6 mm or more intheir experience had a nonunion rate of 86%, compared witha nonunion rate of 18% for patients with displacement of lessthan 6 mm. The authors reported no significant relationshipbetween fracture nonunion and age using �2 analysis. It mightbe argued that statistical tests of association would be moreappropriate in this circumstance, and age might have beenshown to be a factor had it been used.
Andersson et al. (2) described 29 patients 65 years and olderwith odontoid fractures managed by surgical and nonsurgicalmeans. In their series, six (86%) of seven patients achievedsuccessful fusion after posterior cervical C1–C2 arthrodesis.Worse results were observed in patients treated with anteriorodontoid screw fixation (20% fusion rate) and in patientsmanaged with external immobilization alone (20% fusionrate). These authors favored posterior cervical fusion overother management options in elderly patients with Type IIodontoid fractures. Pepin et al. (58) reported their experiencewith 41 acute odontoid fractures (1 Type I, 19 Type II, 21 TypeIII). The authors found that halo immobilization was poorlytolerated in patients 75 years and older. They suggested thatearly C1–C2 fixation and fusion was appropriate in thisgroup. Hanigan et al. (41) described 19 patients 80 years andolder with odontoid fractures (16 Type II, 3 Type III). Fivepatients with displacement of more than 5 mm were treatedwith posterior cervical fixation and fusion with good results.Three of the five had stable nonunions. The authors reporteda mortality rate of 26% in patients managed conservativelywith prolonged immobilization rather than surgical fixationand fusion. On the other hand, they noted that no patienttreated with external immobilization alone developed clini-cally significant instability. Pitzen et al. (59) described theirexperience with surgical therapy in seven patients 70 yearsand older with odontoid fractures. Two patients died of re-lated medical comorbidity. Five patients did well and weremobilized early. The authors concluded that early surgicalfixation in this age group is the preferred management strat-egy. This view is shared by several other investigators, includ-ing Seybold and Bayley (67), Campanelli et al. (14), and Mul-ler et al. (56). Bednar et al. (5) reported a prospectiveassessment of elderly patients with odontoid fractures man-aged with early operative stabilization and fusion. Elevenpatients were included in their study. The authors found a91% fusion rate (10 of 11 patients). One patient died of unre-
lated causes. The authors argued in favor of early surgicalintervention for elderly patients with odontoid fractures. In1997, Berleman and Schwarzenbach (7) offered a retrospectivereview of their experience with 19 patients 65 years and olderwith Type II odontoid fractures treated with anterior odon-toid screw fixation. Radiographic fusion with nearly 5-yearfollow-up was obtained in 16 (85%) of 19 patients. The authorsconcluded that anterior odontoid screw fixation is a successfultherapy for elderly patients with Type II odontoid fractures.
In the only case-control Class II evidence study publishedon this topic, Lennarson et al. (46) examined 33 patients withisolated Type II odontoid fractures treated with halo vestimmobilization. The authors found that age older than 50years was a significant factor for failure of fusion in a haloimmobilization device. Patients 50 years and older had a riskfor nonunion 21 times higher than that found for patientsyounger than 50 years. No significant effect on outcome wasfound attributable to other medical conditions, sex of thepatient, degree of fracture displacement, direction of fracturedisplacement, length of hospital stay, or length of follow-up.
Traumatic spondylolisthesis of the axis(hangman’s fracture)
Overview
Traumatic fractures of the posterior elements of the axis,often related to hyperextension injuries from motor vehicleaccidents, diving, and falls, are reminiscent of the injury in-duced to the axis by judicial hangings (65, 78). A distinctionhas been made between the two fracture types because themechanisms of injury are different. The mechanism of injuryassociated with judicial hanging is one of distraction andhyperextension. The more common hangman’s fracture injuryinduced by motor vehicular trauma is typically a result ofhyperextension, compression, and possible rebound flexion.The incidence of head injury is high with the latter hangman’sfracture injury type.
Wood-Jones (78) described the cervical fracture-dislocationinjury induced by hanging in 1913. Garber (32) used the term“traumatic spondylolisthesis” of the axis in 1964. He de-scribed eight patients with “pedicular” fractures of the axisafter motor vehicle accidents. The term “hangman’s fracture”has been attributed to Schneider (65), who described a seriesof eight patients and noted the similarity between the fractureof the posterior elements of the axis to the pattern of frac-ture injury induced by judicial hanging. Williams (77) docu-mented four cases of hangman’s fracture injury in 1975, not-ing that three occurred associated with motor vehicleaccidents and the fourth with a fall. A number of authors havesuggested that a more appropriate term for this axis injurytype may be “traumatic spondylolisthesis of the axis” becauseof the differences in the mechanism of injury between hangingand the deceleration injuries of falls and motor vehicle acci-dents (28, 29). Most traumatic spondylolisthesis fractures ofC2 caused by motor vehicle accidents seem to result fromhyperextension and compression, rather than the hyperexten-sion and distraction associated with hangings. These differ-ences in the mechanism of injury, along with the wide range
S134 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
of neurological deficits identified with these injuries,prompted a series of investigators to attempt to better char-acterize and classify traumatic spondylolisthesis injuries ofthe axis.
Classification of hangman’s fractures
In 1981, Pepin and Hawkins (57) published a two-typeclassification scheme for hangman’s fractures. Type I wasdescribed as a nondisplaced fracture of the posterior elementsalone. Type II was a displaced fracture involving the posteriorelements and the body of C2. The authors successfully treated42 patients without surgery using their scheme, which in-volved reduction (Type II injuries) and immobilization. Theynoted a low incidence of associated spinal cord injury, but afrequent association with head injury. Although simple andeffective, Pepin and Hawkins’ scheme has not gained popularacceptance and is not widely used. In the same year, Franciset al. (29) published a collaborative experience in treating 123patients with traumatic spondylolisthesis of the axis. Injurieswere divided into five grades based on displacement andangulation of C2 on C3. Grade I was defined as displacementof less than 3.5 mm and angulation of less than 11 degrees.Grade V was defined as complete C2–C3 disc disruption.Grade IV in their scheme had more than 3.5 mm of C2–C3disruption, but less than half of C3 vertebral width with morethan 11 degrees of C2–C3 angulation. Grades II and III wereinjury types graded between Grades I and IV.
Effendi et al. (25) described three types of fractures of thering of the axis based on a series of 131 patients. Their clas-sification scheme was based on the mechanism of injury: TypeI, axial loading and hyperextension; Type II, hyperextensionand rebound flexion; Type III, primary flexion and reboundextension. Type I fractures were defined as isolated hairlinefractures of the ring of the axis with minimal displacement ofthe body of C2. Type II fractures were defined as displace-ment of the anterior fragment with disruption of the discspace below the axis. Type III fractures were defined as dis-placement of the anterior fragment with the body of the axisin a flexed position in conjunction with C2–C3 facet disloca-tion. This Type III fracture is associated with a flexed forwardposition of the axis body. The incidence of Type I, II, and IIIfracture injury in their series was 65, 28, and 7%, respectively.Levine and Edwards (47) modified Effendi’s classificationscheme in 1985. They added flexion-distraction as a mecha-nism of injury (Type IIA) and offered a tailored treatmentstrategy for each of the four injury types. In the largest seriesof axis fractures yet described, Greene et al. (36) used theclassification schemes of both Effendi et al. and Francis et al.to characterize 74 hangman’s fractures. The most commonfracture pattern identified was the Effendi Type I (72%) andthe Francis Grade I (65%). The investigators found a strongcorrelation between Effendi Types I and III and FrancisGrades I and IV, respectively.
Not all authors think that all hangman’s fractures fit intoone or both of these classification schemes. In the review byBurke and Harris (13) of 165 acute injuries of the axis vertebra,62 (38%) were traumatic spondylolisthesis of the axis, includ-
ing 13 Effendi I, 35 Effendi II, and 3 Effendi III injuries. Elevenpatients (18%) had a fracture pattern not previously describedin which one or both fractures involved a portion of theposterior cortex of the body of the axis.
Incidence of traumatic spondylolisthesis andassociated injuries
In Greene et al.’s (36) series of 1820 cervical fractures, 340(19%) were fractures of the axis and 74 (4%) were hangman’stype. In the series of acute fractures of the axis vertebradescribed by Burke and Harris (13), injuries of the axis wereassociated with other fractures of the cervical vertebra in 8%of cases. Ryan and Henderson (61) studied 657 patients withcervical spine fractures over a 13-year period. Hangman’s-type fractures occurred as isolated fractures in 74% of theirseries. Only 9% were associated with fractures of C1. Anadditional 9% were associated with subaxial cervical spinefractures. In the series of Guiot and Fessler (37) of 10 complexcombined atlantoaxial fractures, only one involved a hang-man’s injury. Although the incidence of spinal cord and nerveroot injury as a result of a hangman’s fracture is reportedlylow, unstable hangman’s injuries do occur with some fre-quency (12, 57). If the patient survives the initial injury, it hasbeen proposed that the relatively spacious intracanicular di-ameter affords some protection against spinal cord compres-sion (54). Starr and Eismont (68) described an atypical fracturepattern occurring through the posterior aspect of the vertebralbody, with continuity of the posterior cortex or pedicle andnarrowing of the spinal canal as a result of the associatedsubluxation. In their series of 19 patients, this hangman’sfracture variant occurred in six patients, including two pa-tients with resultant paralysis. In the series described by Fran-cis et al. (29), 8 (6%) of 123 patients they managed had neu-rological deficits. Tan and Balanchandran’s (70) retrospectiveseries of 33 hangman’s fractures included 14 patients withno neurological deficit at admission. The other 19 (57%) hadneurological deficits ranging from quadriparesis to urinaryretention. Twenty-eight patients (85%) had returned to em-ployment at the 1-year follow-up. Mirvis et al.’s (53) seriesof 27 patients had associated neurological findings in 26% ofpatients with hangman’s fractures. Combination fracturesof C1 and C2 in association with a hangman’s-type C2 injuryseem to have a higher incidence of associated neurologicalinjury, likely because of increased instability and a moresevere traumatic injury pattern (23, 37).
Treatment
Most patients with traumatic hangman’s fractures reportedin the reviewed literature were treated with cervical immobi-lization with good results. The three largest experiences re-ported are the multi-institutional series of Effendi et al. (25)and Francis et al. (29) and the single institutional experiencedescribed by Greene et al. (36). Management strategies andsurgical indications vary somewhat among investigators.
In the series reported by Effendi et al. (25) in 1981, therewere 85 Type I fractures, 62 of which were managed withexternal immobilization. They reported 37 Type II fractures;
Isolated Fractures of the Axis in Adults S135
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
17 of the patients were treated with bracing and 15 withsurgical fusion. Of the patients managed surgically, 4 weretreated with a C2–C3 anterior fusion and 11 were treated withdorsal internal fixation and fusion. Nine patients had EffendiType III fractures. Three died without definitive treatment,one was managed in a brace, and five were treated surgicallywith fusion, one anterior and four posterior. The authorsconcluded that most hangman’s fractures were best managednonoperatively. They commented that they might have over-treated patients early in their series, offering surgery whenexternal immobilization might well have been successful.They decided that surgery should be reserved for unusualType III fractures and those patients with failure of fusiondespite 3 months of halo immobilization.
In Pepin and Hawkins’ (57) series, also reported in 1981, all42 patients with hangman’s fractures they treated healed suc-cessfully with external immobilization alone. Francis et al. (29)described and classified hangman’s fracture injuries in 123patients from four institutions. Injuries were categorized intoGrades I through V on the basis of displacement and angula-tion. There were 19 Grade I, 9 Grade II, 46 Grade III, 42 GradeIV, and 7 Grade V fractures. All patients were initially man-aged with traction with conversion to a halo orthosis or weretreated in a halo vest without traction. Healing occurred in116 patients (95%) with halo immobilization alone. Sevenpatients received surgical management with fusion for non-union despite halo immobilization (four had an anteriorC2–C3 fusion, two had a posterior C1–C3 fusion, and one hada posterior C2–C4 fusion). The authors assessed the injurytype with respect to success with nonoperative management.Three (33%) of 9 Grade II injury patients and 2 (28%) of 7Grade V injury patients developed nonunion despite halomanagement and required subsequent surgical treatment.Halo treatment failed in none of the Grade I and Grade IIIinjury patients and in only one Grade IV injury patient. Theauthors concluded that primary surgical treatment for hang-man’s fracture injuries is not indicated. All patients should beprovided late follow-up to assess for nonunion, particularlyGrade II and Grade V injury patients. When surgical manage-ment is considered, the authors recommended an anteriorC2–C3 fusion.
In Levine and Edwards’ (47) series of 52 patients withhangman’s fractures, all isolated Type I, II, and IIa injurieswere managed nonoperatively. Three of five Type III patientsunderwent surgical stabilization and fusion for failure to ob-tain or maintain fracture reduction in a halo. The authorssingled out the subgroup of the Effendi Type II fracture thatsignificantly distracted with the application of craniocervicaltraction. They thought that Type II injuries were likely theresult of flexion-distraction forces. The three patients withType II fractures in their series were successfully treated withmild compression-extension in a halo vest under fluoroscopiccontrol (47).
Greene et al. (36) noted a strong correlation between EffendiType I and Francis Grade I hangman’s injury and betweenEffendi Type III and Francis Grade IV fractures in their seriesof 74 patients. Sixty-five of 74 patients were treated nonop-eratively with external immobilization for a median of 12
weeks. There were two early deaths. Seven patients requiredearly surgical fixation and fusion for inability to maintainfracture alignment in a halo brace. All seven early surgicalpatients were either Effendi Grade II or III, and five of theseven were Francis Grade III or IV. Overall, 33% of patientswith unstable Effendi Types II and III or 36% of FrancisGrades III, IV, and V injuries required early surgical treat-ment. Eventually, all seven patients achieved solid fusionwithout evidence of instability. The authors compared theirexperiences with those of Effendi et al. and Francis et al. andconcluded that conservative management (external immobi-lization) should be the initial treatment in virtually everypatient with a hangman’s fracture. Early surgical manage-ment should be reserved for unstable injuries that are ineffec-tively immobilized in a halo device.
In a combined clinical and cadaveric anatomic study, Mest-dagh et al. (52) described their experience with 41 fractures ofthe posterior neural arch of the axis. Eleven cases were treatedsurgically with anterior C2–C3 interbody fusion, and 30 pa-tients were treated with external immobilization. Thirty pa-tients were available for follow-up. Cervical mobility wasbetter in patients managed conservatively. Displacement ofup to 5 mm at the hangman’s fracture site in a cadaveric studywas compatible with stability without disruption of the liga-ments or the C2–C3 disc space. The authors concluded thatconservative management was the ideal treatment for hang-man’s fractures, except in cases of marked instability or fail-ure of union. Grady et al. (35) reported their experience with27 patients with hangman’s fractures. Sixteen were managedin a halo device, eight in a rigid collar, and three with bed restonly. All achieved fusion with no residual symptoms. Theauthors concluded that use of a Philadelphia collar alone forhangman’s fractures is a reasonable alternative to halo immo-bilization, particularly for injuries with minimal C2–C3 dis-placement. In 1987, Govender and Charles (33) prospectivelystudied 39 patients with traumatic spondylolisthesis of theaxis. Injuries were classified for stability by the criteria ofWhite and Panjabi (76). All patients were successfully treatedwith collar immobilization regardless of assessment of stabil-ity. The authors argue against basing treatment on dynamicimaging, as advocated by Effendi et al. (25) and Levine andEdwards (47). A number of other reports favor nonoperativemanagement of hangman’s fractures (4, 11, 17, 27, 33, 47, 50,52, 66, 71).
Surgical management
Surgical options for unstable hangman’s fracture injuries,particularly those that fail to heal despite external immobili-zation, include anterior C2–C3 interbody fusion and dorsalC1–C3 fusion procedures. In the series of Effendi et al. (25), 42of 131 patients with hangman’s fractures were treated surgi-cally. Ten were treated with an anterior C2–C3 fusion, and 32underwent a posterior fusion. All were successfully stabilizedat last follow-up. In the Francis et al. (29) series of 123 hang-man’s fracture patients, only 7 patients were treated surgi-cally. Four underwent anterior C2–C3 fusion, two had poste-rior C1–C3 fusion, and one underwent posterior C2–C4
S136 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
fusion. The authors noted that 6 of the 7 patients requiringfusion for nonunion had C2–C3 angulation of more than 11degrees. All seven patients achieved bony stability.
A number of case series of hangman’s fractures offer similarexperiences with surgical management. McLaurin et al. (51)described their experience with early fusion in two patientswith hangman’s fractures to allow early mobilization. Theauthors acknowledged that both injuries would likely havehealed with external immobilization alone. Salmon (64) de-scribed 20 patients with hangman’s fractures treated withposterior interlaminar wiring and fusion with no morbidity.Verheggen and Jansen (73), in their 1998 report, arguedstrongly for surgical fixation and fusion of Effendi Type II andIII hangman’s fractures. In their opinion, the optimal manage-ment of these injuries remains controversial. They described16 patients with hangman’s fractures they treated with surgi-cal fixation of the posterior arch of the axis with screw fixa-tion. They found that this fixation technique resulted in su-perior functional results as compared with historical controls.They favor this management strategy in the setting of theLevine and Edwards (47) Type IIa fracture. The viewpoint ofVerheggen and Jansen (73) is challenged by Sypert (69a) in hiscomments that accompany their article. Borne et al. (10), in1984, published their approach to the management of pedic-ular fractures of the axis. They used a technique of bilateralposterior screw fixation. They described excellent results anda 100% fusion rate. Despite this, their technique has notgained widespread acceptance.
Fractures of the axis body
A number of authors have addressed the management ofnon-odontoid, non-hangman’s fractures of the axis. They havebeen labeled as miscellaneous fractures of the axis, non-odontoid non-hangman’s fractures, or simply axis body frac-tures (6, 31, 36, 40). Several attempts have been made toclassify the various fracture types within this diverse group.Benzel et al. (6) reported on 15 patients with fracture of theaxis body and divided them into three types: coronal, sagittal,and transverse oriented. The latter group was thought torepresent the same group as the Anderson and D’AlonzoType III odontoid fracture. The authors proposed that theType III odontoid fracture classification be discarded becauseit is misleading (6). The original authors had the same thought(1). Benzel et al. (6) offered a mechanism of injury for each ofthe three fracture types they described. No treatment or out-come data were included in their report. Greene et al. (36)described 67 patients with miscellaneous axis fractures of alltypes. Of the 61 patients available for follow-up (medianfollow-up, 14 mo), all but one was successfully managed witha variety of nonoperative means. The authors note that this isa diverse injury group and describe a treatment algorithmbased on features of fracture stability. Only one patient witha miscellaneous axis fracture required surgical interventionfor delayed nonunion. Fujimura et al. (31) classified 31 axisbody fractures on the basis of their radiographic injury pat-tern: avulsion, transverse, burst, or sagittal. In their series, allnine cases of avulsion fracture and the two cases of transverse
fracture healed with external immobilization. Two of the threeburst fractures were treated with C2–C3 anterior interbodyfusion. Of the 17 sagittal fractures, 15 healed with nonopera-tive treatment. The remaining two patients required surgicalfusion. The authors recommend initial nonoperative treat-ment for all non-odontoid, non-hangman’s axis fractures.Craig and Hodgson (19) added nine cases of axis fracturesinvolving the superior articular facet. In seven patients, therewas an associated odontoid fracture. This fracture occurred ineither the coronal or sagittal plane, resulting in shearing of theanterior or lateral portion of the facet complex. The lateralmass of the atlas was noted to occasionally sublux into thefacet fracture. The authors recommended immobilization fornondisplaced fractures and the consideration of surgical re-duction, fixation, and fusion for fractures that are difficult toreduce. Bohay et al. (8) described three unusual fractures ofthe posterior body of C2, all of which responded to nonop-erative management. Jakim and Sweet (42) contributed a sin-gle case. Korres et al. (45) described 14 patients with avulsionfractures of the anteroinferior portion of the axis that theybelieved to be extension-type injuries. These cases repre-sented only 3% of the cervical spine fractures they managedover a 12-year period. All 14 of these body fracture types weresuccessfully managed with cervical immobilization (meanfollow-up, 8.5 yr).
SUMMARY
Fractures of the odontoid
There is no Class I medical evidence addressing the issue ofmanagement of acute traumatic odontoid fractures. A singleClass II evidence paper reviews the management of Type IIodontoid fractures in halo immobilization devices. This studydemonstrated a 21-fold increase in risk of nonunion with haloimmobilization in patients older than 50 years. All other arti-cles reviewed contain Class III evidence that supports severaltreatments.
Type II odontoid fractures in patients 50 years and oldershould be considered for surgical stabilization and fusion.Type I, Type II, and Type III fractures may be managedinitially with external cervical immobilization. Type II andType III odontoid fractures should be considered for surgicalfixation in cases of dens displacement of 5 mm or more,comminution of the odontoid fracture (Type IIA), and/orinability to achieve or maintain fracture alignment with ex-ternal immobilization. Isolated Type I and Type III odontoidfractures may be treated with cervical immobilization, result-ing in fusion rates of 100 and 84%, respectively. Anteriorsurgical fixation of Type III fractures has been associated witha 100% fusion rate. Type II odontoid fractures may be treatedwith external immobilization or surgical fixation and fusion.Halo immobilization and posterior fixation have both beenused successfully for these injuries. Anterior odontoid-screwfixation has been reported with an up to 90% fusion successrate, except in older patients. Treatment of Type II odontoidfracture with a cervical collar alone or traction and thencervical collar immobilization may also be undertaken, butthese methods have lower success rates.
Isolated Fractures of the Axis in Adults S137
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Traumatic spondylolisthesis of the axis
There is no Class I or Class II medical evidence addressingthe management of traumatic spondylolisthesis of the axis.All articles reviewed contain Class III evidence that supportsa variety of treatments. Most hangman’s fractures heal with12 weeks of cervical immobilization with either a rigid cervi-cal collar or a halo immobilization device. Surgical stabiliza-tion is an option in cases of severe angulation (Francis GradeII and IV, Effendi Type II), disruption of the C2–C3 disc space(Francis Grade V, Effendi Type III), or the inability to establishor maintain fracture alignment with external immobilization.
Fractures of the axis body (miscellaneousaxis fractures)
There is no Class I or Class II medical evidence addressing themanagement of traumatic fractures of the axis body. All articlesreviewed contain Class III evidence that supports the use ofexternal immobilization as the initial treatment strategy.
KEY ISSUES FOR FUTURE INVESTIGATION
More data are necessary to determine treatment standardsand/or guidelines for the definitive management of odontoidfractures. For Type I and Type III fractures, the available ClassIII evidence suggests that a well-designed multicenter case-control study could provide sufficient evidence to define theirappropriate management in the early postinjury period. ForType II fractures, the literature suggests that both operativeand nonoperative management remain treatment options. Arandomized or case-control study would be of benefit inestablishing definitive treatment recommendations for thisfracture type. Traumatic spondylolisthesis of the axis andmiscellaneous axis fractures are treated successfully with ex-ternal immobilization in most cases. A multicenter case-control study of patients with these injury types would help todefine optimal treatment of each specific fracture subtype.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Anderson LD, D’Alonzo RT: Fractures of the odontoid process ofthe axis. J Bone Joint Surg Am 56A:1663–1674, 1974.
2. Andersson S, Rodrigues M, Olerud C: Odontoid fractures: Highcomplication rate associated with anterior screw fixation in theelderly. Eur Spine J 9:56–60, 2000.
3. Apfelbaum RI, Lonser RR, Veres R, Casey A: Direct anteriorscrew fixation for recent and remote odontoid fractures.J Neurosurg 93[Suppl 2]:227–236, 2000.
4. Barros TE, Bohlman HH, Capen DA, Cotler J, Dons K, Biering-Sorensen F, Marchesi DG, Zigler JE: Traumatic spondylolisthesisof the axis: Analysis of management. Spinal Cord 37:166–171,1999.
5. Bednar DA, Parikh J, Hummel J: Management of Type II odon-toid process fractures in geriatric patients: A prospective studyof sequential cohorts with attention to survivorship. J SpinalDisord 8:166–169, 1995.
6. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW,Espinosa MC: Fractures of the C-2 vertebral body. J Neurosurg81:206–212, 1994.
7. Berlemann U, Schwarzenbach O: Dens fractures in the elderly:Results of anterior screw fixation in 19 elderly patients. ActaOrthop Scand 68:319–324, 1997.
8. Bohay D, Gosselin RA, Contreras DM: The vertical axis fracture:A report on three cases. J Orthop Trauma 6:416–419, 1992.
9. Bohler J: An approach to non-union of fractures. Surg Annu14:299–315, 1982.
10. Borne GM, Bedou GL, Pinaudeau M: Treatment of pedicularfractures of the axis: A clinical study and screw fixation tech-nique. J Neurosurg 60:88–93, 1984.
11. Brashear R Jr, Venters G, Preston ET: Fractures of the neural archof the axis: A report of twenty-nine cases. J Bone Joint Surg Am57A:879–887, 1975.
12. Bucholz RW: Unstable hangman’s fractures. Clin Orthop 154:119–124, 1981.
13. Burke JT, Harris JH Jr: Acute injuries of the axis vertebra. Skel-etal Radiol 18:335–346, 1989.
14. Campanelli M, Kattner KA, Stroink A, Gupta K, West S: Poste-rior C1–C2 transarticular screw fixation in the treatment of dis-placed Type II odontoid fractures in the geriatric population:Review of seven cases. Surg Neurol 51:596–601, 1999.
15. Chiba K, Fujimura Y, Toyama Y, Fujii E, Nakanishi T,Hirabayashi K: Treatment protocol for fractures of the odontoidprocess. J Spinal Disord 9:267–276, 1996.
16. Clark CR, White AA III: Fractures of the dens: A multicenterstudy. J Bone Joint Surg Am 67A:1340–1348, 1985.
17. Coric D, Wilson JA, Kelly DL Jr: Treatment of traumatic spon-dylolisthesis of the axis with nonrigid immobilization: A reviewof 64 cases. J Neurosurg 85:550–554, 1996.
18. Coyne TJ, Fehlings MG, Wallace MC, Bernstein M, Tator CH:C1–C2 posterior cervical fusion: Long-term evaluation of resultsand efficacy. Neurosurgery 37:688–693, 1995.
19. Craig JB, Hodgson BF: Superior facet fractures of the axis verte-bra. Spine 16:875–877, 1991.
20. Dai LY, Yuan W, Ni B, Liu HK, Jia LS, Zhao DL, Xu YK: Surgicaltreatment of nonunited fractures of the odontoid process, withspecial reference to occipitocervical fusion for unreducible atlan-toaxial subluxation or instability. Eur Spine J 9:118–122, 2000.
21. Davis D, Bohlman H, Walker AE, Fisher R, Robinson R: Thepathological findings in fatal craniospinal injuries. J Neurosurg34:603–613, 1971.
22. Dickman CA, Sonntag VKH: Surgical management of atlantoax-ial nonunions. J Neurosurg 83:248–253, 1995.
23. Dickman CA, Foley KT, Sonntag VKH, Smith MM: Cannulatedscrews for odontoid fixation and atlantoaxial transarticularscrew fixation. J Neurosurg 83:1095–1100, 1995.
24. Dunn ME, Seljeskog EL: Experience in the management of odon-toid process injuries: An analysis of 128 cases. Neurosurgery18:306–310, 1986.
25. Effendi B, Roy D, Cornish B, Dussault RG, Laurin CA: Fracturesof the ring of the axis: A classification based on the analysis of131 cases. J Bone Joint Surg Br 63B:319–327, 1981.
26. Ekong CE, Schwartz ML, Tator CH, Rowed DW, Edmonds VE:Odontoid fracture: Management with early mobilization usingthe halo device. Neurosurgery 9:631–637, 1981.
27. Ersmark H, Lowenhielm P: Factors influencing the outcome ofcervical spine injuries. J Trauma 28:407–410, 1988.
28. Fielding JW, Francis WR Jr, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. Clin Orthop 239:47–52,1989.
S138 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
29. Francis WR, Fielding JW, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. J Bone Joint Surg Br63B:313–318, 1981.
30. Fujii E, Kobayashi K, Hirabayashi K: Treatment in fractures ofthe odontoid process. Spine 13:604–609, 1988.
31. Fujimura Y, Nishi Y, Kobayashi K: Classification and treatmentof axis body fractures. J Orthop Trauma 10:536–540, 1996.
32. Garber J: Abnormalities of the atlas ans axis vertebrae:Congential and traumatic. J Bone Joint Surgery Am 46A:1782–1791, 1964.
33. Govender S, Charles RW: Traumatic spondylolisthesis of theaxis. Injury 18:333–335, 1987.
34. Govender S, Grootboom M: Fractures of the dens: The results ofnon-rigid immobilization. Injury 19:165–167, 1988.
35. Grady MS, Howard MA, Jane JA, Persing JA: Use of the Phila-delphia collar as an alternative to the halo vest in patients withC-2, C-3 fractures. Neurosurgery 18:151–156, 1986.
36. Greene KA, Dickman CA, Marciano FF, Drabier JB, Hadley MN,Sonntag VKH: Acute axis fractures: Analysis of managementand outcome in 340 consecutive cases. Spine 22:1843–1852, 1997.
37. Guiot B, Fessler RG: Complex atlantoaxial fractures. J Neurosurg91:139–143, 1999.
38. Hadley MN, Browner CM, Liu SS, Sonntag VKH: New subtypeof acute odontoid fractures (type IIA). Neurosurgery 22:67–71,1988.
39. Hadley MN, Browner C, Sonntag VKH: Axis fractures: A com-prehensive review of management and treatment in 107 cases.Neurosurgery 17:281–290, 1985.
40. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acuteaxis fractures: A review of 229 cases. J Neurosurg 71:642–647,1989.
41. Hanigan WC, Powell FC, Elwood PW, Henderson JP: Odontoidfractures in elderly patients. J Neurosurg 78:32–35, 1993.
42. Jakim I, Sweet MB: Transverse fracture through the body of theaxis. J Bone Joint Surg Br 70B:728–729, 1988.
43. Jenkins JD, Coric D, Branch CL Jr: A clinical comparison of one-and two-screw odontoid fixation. J Neurosurg 89:366–370, 1998.
44. Julien TD, Frankel B, Traynelis VC, Ryken TC: Evidence-basedanalysis of odontoid fracture management. Neurosurg Focus8:Article 1, 2000.
45. Korres DS, Zoubos AB, Kavadias K, Babis GC, Balalis K: The“tear drop” (or avulsed) fracture of the anterior inferior angle ofthe axis. Eur Spine J 3:151–154, 1994.
46. Lennarson PJ, Mostafavi H, Traynelis VC, Walters BC: Manage-ment of type II dens fractures: A case-control study. Spine25:1234–1237, 2000.
47. Levine AM, Edwards CC: The management of traumatic spon-dylolisthesis of the axis. J Bone Joint Surg Am 67A:217–226,1985.
48. Lind B, Nordwall A, Sihlbom H: Odontoid fractures treated withhalo-vest. Spine 12:173–177, 1987.
49. Maiman DJ, Larson SJ: Management of odontoid fractures.Neurosurgery 11:820, 1982 (letter).
50. Marar BC, Tay CK: Fracture of the odontoid process. Aust N ZJ Surg 46:231–236, 1976.
51. McLaurin RL, Vernal R, Salmon JH: Treatment of fractures of theatlas and axis by wiring without fusion. J Neurosurg 36:773–780,1972.
52. Mestdagh H, Letendart J, Sensey JJ, Duquennoy A: Treatment offractures of the posterior axial arch: Results of 41 cases [inFrench]. Rev Chir Orthop Reparatrice Appar Mot 70:21–28,1984.
53. Mirvis SE, Young JW, Lim C, Greenberg J: Hangman’s fracture:Radiologic assessment in 27 cases. Radiology 163:713–717, 1987.
54. Mollan RA, Watt PC: Hangman’s fracture. Injury 14:265–267,1982.
55. Morandi X, Hanna A, Hamlat A, Brassier G: Anterior screwfixation of odontoid fractures. Surg Neurol 51:236–240, 1999.
56. Muller EJ, Wick M, Russe O, Muhr G: Management of odontoidfractures in the elderly. Eur Spine J 8:360–365, 1999.
57. Pepin JW, Hawkins RJ: Traumatic spondylolisthesis of the axis:Hangman’s fracture. Clin Orthop 133–138, 1981.
58. Pepin JW, Bourne RB, Hawkins RJ: Odontoid fractures, withspecial reference to the elderly patient. Clin Orthop 193:178–183,1985.
59. Pitzen T, Caspar W, Steudel WI, Barbier D: Dens fracture inelderly patients and surgical management [in German].Aktuelle Traumatol 24:56–59, 1994.
60. Polin RS, Szabo T, Bogaev CV, Replogle RE, Jane JA: Nonopera-tive management of types II and III odontoid fractures: ThePhiladelphia collar versus the halo vest. Neurosurgery 38:450–456, 1996.
61. Ryan MD, Henderson JJ: The epidemiology of fractures andfracture-dislocations of the cervical spine. Injury 23:38–40, 1992.
62. Ryan MD, Taylor TK: Odontoid fractures: A rational approach totreatment. J Bone Joint Surg Br 64B:416–421, 1982.
63. Ryan MD, Taylor TK: Odontoid fractures in the elderly. J SpinalDisord 6:397–401, 1993.
64. Salmon JH: Fractures of the second cervical vertebra: Internalfixation by interlaminar wiring. Neurosurgery 1:125–127, 1977.
65. Schneider R: “Hangman’s fracture” of the cervical spine.J Neurosurg 22:141–154, 1965.
66. Seljeskog EL, Chou SN: Spectrum of the hangman’s fracture.J Neurosurg 45:3–8, 1976.
67. Seybold EA, Bayley JC: Functional outcome of surgically andconservatively managed dens fractures. Spine 23:1837–1846,1998.
68. Starr JK, Eismont FJ: Atypical hangman’s fractures. Spine 18:1954–1957, 1993.
69. Subach BR, Morone MA, Haid RW Jr, McLaughlin MR, RodtsGR, Comey CH: Management of acute odontoid fractures withsingle-screw anterior fixation. Neurosurgery 45:812–820, 1999.
69a. Sypert GW: Hangman’s fracture: Arguments in favor of surgicaltherapy for type II and III according to Edwards and Levine.Surg Neurol 49:262, 1998 (comment).
70. Tan ES, Balachandran N: Hangman’s fracture in Singapore(1975–1988). Paraplegia 30:160–164, 1992.
71. Torreman M: Long-term prognosis of the hangman’s fracture [inDutch]. Ned Tijdschr Geneeskd 134:1173–1176, 1990.
72. Traynelis VC: Evidence-based management of type II odontoidfractures. Clin Neurosurg 44:41–49, 1997.
73. Verheggen R, Jansen J: Hangman’s fracture: Arguments in favorof surgical therapy for type II and III according to Edwards andLevine. Surg Neurol 49:253–262, 1998.
74. Waddell JP, Reardon GP: Atlantoaxial arthrodesis to treat odon-toid fractures. Can J Surg 26:255–258, 1983.
75. Wang GJ, Mabie KN, Whitehill R, Stamp WG: The nonsurgicalmanagement of odontoid fractures in adults. Spine 9:229–230,1984.
76. White AA, Panjabi MM: Clinical Biomechanics of the Spine. Phila-delphia, J.B. Lippincott, 1990, ed 2.
77. Williams TG: Hangman’s fracture. J Bone Joint Surg Br 57B:82–88, 1975.
78. Wood-Jones F: The ideal lesion produced by judicial hanging.Lancet 1:53, 1913.
Isolated Fractures of the Axis in Adults S139
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
CHAPTER 18
Management of Combination Fractures of the Atlas and Axisin Adults
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS: Treatment of atlas-axis combination fractures based primarily on the specific characteristics of the
axis fracture is recommended. External immobilization of most C1–C2 combination fractures is recom-mended. C1–Type II odontoid combination fractures with an atlantodens interval of 5 mm or more andC1–hangman’s combination fractures with C2–C3 angulation of 11 degrees or more should be consideredfor surgical stabilization and fusion. In some cases, the surgical technique must be modified as a result ofloss of the integrity of the ring of the atlas.
RATIONALE
Combined fractures of the atlas and axis often presentmanagement challenges owing to the unique anatomyand biomechanics of the atlantoaxial complex and the
untoward stresses applied to the atlantoaxial region duringtrauma. Although most isolated atlas and axis fractures havebeen managed with cervical immobilization, the occurrence ofthe two fractures in combination often implies a more signif-icant structural and mechanical injury. Although reports ofcombination C1–C2 fractures are relatively infrequent, suffi-cient evidence exists to allow a review of the management ofa variety of combinations of atlas and axis fractures. Thepurpose of this chapter is to examine the available literature todetermine successful treatment strategies for individualC1–C2 combination fracture types.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “vertebral fracture”in combination with “atlas,” “axis,” and “human” yielded1071 references. The abstracts were reviewed, and articlesfocusing on clinical management and follow-up of combina-tion fractures of the atlas and axis were selected for inclusion.The relative infrequency of these fractures, the small numberof case series, and the numerous case reports with pertinentinformation necessitated rather broad inclusion criteria. Sev-eral papers addressing relevant biomechanics and radiologywere included. The bibliographies of the selected papers werereviewed to provide additional references. These efforts re-sulted in 49 articles describing the clinical features and man-agement of acute traumatic atlas and axis combination frac-tures. Forty-eight of the articles are summarized in Tables 18.1and 18.2. No Class I or II evidence has been generated on the
management of these fractures. Treatment options have beenformulated on the basis of Class III medical evidence.
SCIENTIFIC FOUNDATION
Overview
In 1920, Sir Geoffrey Jefferson (30) reviewed 46 cases ofatlas fractures. Although his article is best known for thecharacterization of the C1 burst fracture or “Jefferson frac-ture,” Jefferson’s series included 19 fractures that were de-scribed as “combination fractures” of the atlas and the axis.He noted increased morbidity and mortality for patients withcombination injuries. Eleven of the 19 patients he describedwith C1–C2 combination injuries had significant neurologicalinjuries. In 1986, Levine and Edwards (34) reported theirapproach to the management of C1–C2 traumatic fractureinjuries. They suggested that if an atlas or axis injury wasidentified, a careful search for other related injuries was indi-cated. They stressed that each patient and each injury neededto be evaluated independently. They described staged treat-ment for certain injuries to allow healing of one fracture(usually the atlas) before definitively managing the combina-tion injury (typically the axis fracture). Several of their obser-vations are worthy of consideration in the management ofcombination fracture injuries of the atlas and axis today.
Incidence
Combination fractures of the C1–C2 complex are relativelycommon. In reports focusing primarily on odontoid fractures,the occurrence of a concurrent C1 fracture in the presence of aType II or Type III odontoid fracture has been reported in 5 to53% of cases (4, 12, 23, 25, 26, 28, 36, 39, 40, 43–45, 47, 49).Odontoid fractures have been identified in 24 to 53% of patientswith atlas fractures (18, 32, 35, 45). In the presence of a hang-
S140 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
man’s fracture, the reported incidence of a C1 fracture rangesfrom 6 to 26% (13, 17, 33, 38, 41, 45). Greene et al. (23) reportedon 340 axis fractures and found 48 concurrent atlas fractures(combination injuries), for an incidence of 14%. Ryan and Hen-derson (45) reviewed 717 spine fractures and found combinationatlas-axis fractures in 15% of odontoid fractures and in 9% ofhangman’s fractures. Gleizes et al. (21) reviewed 784 patientswith proximal cervical spine injuries in 2000. One hundred six-teen patients had injuries to C1 and/or C2. Thirty-one patientshad C1 fractures in association with a C2 fracture (combinationinjury), representing 4% of the total cervical spine fracture pop-ulation and 27% of all C1–C2 fracture injuries.
Morbidity and mortality
Various authors have suggested that the morbidity and mor-tality of C1–C2 combination fractures is higher than that associ-ated with isolated fractures of either the atlas or the axis (12, 18,19, 25, 26, 31, 49). Fujimura et al. (19) observed neurologicaldeficits in 82 (34%) of 247 patients with injuries to the C1–C2complex. Those patients with deficit had either burst fractures orfractures of the posterior arch of C1 or a fracture of the C2 bodycoupled with an odontoid or hangman’s fracture. Several au-thors have described a high mortality rate with combinationfractures, in particular C1 fractures combined with Type II odon-toid fractures (18, 25, 26, 49). Fowler et al. (18) found that 6 (86%)of 7 patients with C1–Type II odontoid combination fracturesdied in the early treatment period. Similarly, Hanssen and Ca-banela (26) observed that 5 (83%) of 6 patients with this samecombination fracture pattern died within the first 40 days ofinjury. Both Hanigan et al. (25) and Zavanone et al.(49) reportedearly deaths associated with C1–Type II odontoid fractures. Inother reports on C1–C2 combination fractures, the description ofmorbidity and mortality has been less remarkable (12, 18).Dickman et al. (12) suggested a 12% incidence of neurologicaldeficit for C1–C2 combination fractures compared with a 0% (0of 32) and a 2% (2 of 125) incidence for isolated atlas and axisfractures, respectively. Kesterson et al. (31) described four pa-tients with C1–C2 combination fractures. Only 1 patient (25%)had a neurological deficit. Irrespective of the author, the de-scribed incidence of morbidity and mortality associated withcombination C1–C2 fractures seems to be more than that asso-ciated with isolated atlas and axis fractures.
Treatment
Since Jefferson published the original description of C1–C2combination fracture injuries, nearly every series reviewingeither isolated fractures of the atlas or the axis includes somemention of C1–C2 combination fractures. It is difficult todetermine the specific treatment provided to and outcome formost of those patients. Several authors have focused theirreports specifically on combination C1–C2 fractures and theirmanagement (12, 21, 24).
In 1989, Dickman et al. (12) identified 25 cases of acuteatlas-axis combination fractures in an overall series of 860patients with acute cervical fracture injuries. In their experi-ence, C1–C2 combination fractures represented 3% of theirtotal cervical fracture population. Combination injuries rep-resented 43% of acute atlas fractures (25 of 58 patients) and16% of acute axis fractures (25 of 150 patients). The fracturesof C1 and C2 were identified using plain film x-rays in 76 and92% of the cases, respectively. Computed tomography char-acterized the combination fracture patterns in all cases.Twelve percent of patients (3 of 25 patients) had neurologicaldeficits at admission. Two patients had acute central cordsyndrome, and one patient had a complete neurological in-jury. The etiology of the injury was a motor vehicle accident in60% of cases and a fall in 28%. Four main types of atlas-axisfracture combination were identified: C1–Type II odontoid (10cases, 40%), C1–miscellaneous axis fracture (7 cases, 28%),C1–Type III odontoid (5 cases, 20%), and C1–hangman’s-typefracture (3 cases, 12%). The distribution of the atlas fractureswas reported as multiple ring fractures in 40%, posterior ringfracture in 28%, unilateral ring fracture in 24%, and lateralmass fracture in 8%. Nonoperative therapy was the initialmanagement strategy in 20 (84%) of 25 of patients. Eighteenpatients were placed in a halo orthosis and two in asuboccipital-mandibular immobilizer (SOMI) brace, for a me-dian duration of 12 weeks (range, 10–22 wk). Four patientswere treated with early surgical stabilization and fusion basedon an atlantoaxial interval of 6 mm or more. Three weretreated with posterior C1–C2 wiring and fusion. Follow-upwas accomplished in 23 (92%) of 25 patients. Nineteen (95%)of the 20 patients treated with either a halo or SOMI orthosisachieved stability and fusion. Halo immobilization failed inone patient with an initial atlantoaxial interval of 5 mm, andthe patient was treated with posterior C1–C2 fusion. All pa-tients treated surgically achieved stability using a posteriorfusion technique, four early and one delayed (100%). No
TABLE 18.1. Initial Management of CombinationAxis–Atlas Fracturesa
Combination Fracture Type Treatment Options
C1–Type II odontoid fractureStable Collar, halo, surgical fixation/
fusionUnstable (ADI �5 mm) Halo, surgical fixation/fusion
C1–Type III odontoid fracture Halo
C1–miscellaneous axis Collar, halo
C1–hangman’s fractureStable Collar, haloUnstable (C2–C3angulation �11 degrees)
Halo, surgical fixation/fusion
a ADI, atlantodens interval.
Combination Fractures of the Atlas and Axis
S141Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE18
.2.
Sum
mar
yof
Rep
orts
onFr
actu
res
ofth
eA
tlas
and
Axi
sa
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ce
Cla
ssC
oncl
usio
ns
And
erss
onet
al.,
2000
(2)
Cas
ese
ries
ofpa
tient
s�
65yr
with
odon
toid
frac
ture
s.III
Incl
udes
3pa
tient
sw
ithC
1–Ty
peII
odon
toid
frac
ture
s.Tr
eatm
ent:
Hal
o,2
patie
nts.
Post
erio
rce
rvic
al
fusi
on,
1pa
tient
.
Gle
izes
etal
.,20
00(2
1)R
etro
spec
tive
epid
emio
logi
cal
revi
ewof
coin
cide
nce
offr
actu
res
inth
eup
per
cerv
ical
spin
e.
III78
4ce
rvic
alsp
ine
inju
ries
.11
6up
per
cerv
ical
spin
ein
juri
es(C
1–C
2)(1
5%).
31/1
16(2
6%)
com
bina
tion
of
C1
and
C2.
70%
ofal
lat
las
frac
ture
soc
curr
edin
com
bina
tion
with
anot
her
frac
ture
.30
%of
all
hang
man
’s
and
odon
toid
frac
ture
soc
curr
edin
com
bina
tion
with
anot
her
frac
ture
.41
.9%
ofpa
tient
sw
ithco
mbi
natio
n
frac
ture
sof
the
uppe
rce
rvic
alsp
ine
unde
rwen
tsu
rgic
alfix
atio
nve
rsus
21.7
%of
thos
ew
ithis
olat
ed
inju
ries
.
Mul
ler
etal
.,20
00(4
1)C
ase
seri
esof
39ca
ses
ofha
ngm
an’s
frac
ture
s.III
Incl
udes
2pa
tient
sw
ithC
1ri
ngfr
actu
res
(5.1
%).
Gui
otan
dFe
ssle
r,19
99(2
4)R
etro
spec
tive
revi
ewof
10pa
tient
sun
derg
oing
surg
ical
fixat
ion
for
com
bina
tion
C1–
C2
frac
ture
s.5/
10re
ferr
edsp
ecifi
cally
for
surg
ical
fixat
ion
afte
rfa
iled
exte
rnal
imm
obili
zatio
n.A
vera
ge
follo
w-u
p,28
.5m
o.
IIITy
pe:
C1–
Type
IIod
onto
id,
9pa
tient
s(9
0%).
C1–
Type
IIIod
onto
idan
dha
ngm
an’s
,1
patie
nt(1
0%).
Tech
niqu
e:O
dont
oid
scre
w,
6(6
0%).
Odo
ntoi
dsc
rew
plus
C2
pedi
cle
scre
ws,
1pa
tient
(10%
).C
1–C
2
tran
sart
icul
arsc
rew
s(p
oste
rior
),2
patie
nts
(20%
).C
1–C
2tr
ansa
rtic
ular
scre
ws
(ant
erio
r),
1pa
tient
(10%
).
Out
com
e:1
unre
late
dde
ath.
All
othe
rsfu
sed
succ
essf
ully
with
out
othe
rco
mpl
icat
ion.
Hen
ryet
al.,
1999
(28)
Cas
ese
ries
of61
case
sof
Type
IIod
onto
idfr
actu
res
trea
ted
with
ante
rior
scre
wfix
atio
nin
whi
chfo
llow
-up
was
avai
labl
e.
IIIIn
clud
es10
com
bina
tion
frac
ture
sof
C1–
C2
(16%
).C
1bu
rst
(Jeffe
rson
)-Ty
peII
odon
toid
,3
patie
nts
(5%
).
C1
ante
rior
arch
-Typ
eII
odon
toid
,3
patie
nts
(5%
).C
1po
ster
ior
arch
-Typ
eII
odon
toid
,4
patie
nts
(6%
).A
ll
patie
nts
inth
ese
ries
wer
etr
eate
dw
ithan
teri
orod
onto
idsc
rew
fixat
ion
with
a92
%su
cces
sra
te.
No
prob
lem
sat
trib
uted
dire
ctly
toth
epr
esen
ceof
the
atla
sfr
actu
re.
Mor
andi
etal
.,19
99(4
0)C
ase
seri
esin
clud
ing
17od
onto
idfr
actu
res
trea
ted
with
ante
rior
scre
wfix
atio
n.
IIIIn
clud
es2
case
sof
C1–
post
erio
rar
chfr
actu
repl
usa
post
erio
rly
disp
lace
dTy
peII
odon
toid
.
Lee
etal
.,19
98(3
2)R
etro
spec
tive
revi
ewof
16ca
ses
ofat
las
frac
ture
.III
Incl
udes
8pa
tient
sw
ithco
mbi
natio
nC
1an
dC
2fr
actu
res.
C1–
Type
IIod
onto
id,
3ca
ses.
1tr
eate
dw
ith
halo
imm
obili
zatio
n.2
trea
ted
with
post
erio
rC
1–C
2fu
sion
.C
1–H
angm
an’s
,2
case
s.B
oth
trea
ted
with
cerv
ical
colla
r.C
1–C
2bo
dyfr
actu
re,
3ca
ses.
All
3tr
eate
dw
ithce
rvic
alco
llar.
Aut
hors
conc
lude
that
the
man
agem
ent
ofth
eco
mbi
natio
nfr
actu
resh
ould
beba
sed
onth
eC
2fr
actu
rean
dth
atha
loim
mob
iliza
tion
isno
tal
way
sre
quir
ed.
Seyb
old
and
Bay
ley,
1998
(47)
Cas
ese
ries
of57
odon
toid
frac
ture
s.III
Incl
udes
3ca
ses
ofC
1ri
ngfr
actu
repl
usTy
peII
odon
toid
.Th
eau
thor
ssu
cces
sful
lym
anag
edtw
opa
tient
s
with
aha
lo.
One
patie
ntw
astr
eate
din
aco
llar
with
a“p
oor”
resu
lt.Th
eov
eral
lfu
sion
rate
for
the
Type
II
odon
toid
frac
ture
sin
this
seri
esw
as65
%.
No
spec
ific
effe
ctfr
omth
eC
1fr
actu
rew
asno
ted.
Apo
stol
ides
etal
.,19
97(3
)C
ase
repo
rt.
IIIPa
tient
with
ante
rior
ring
ofC
1fr
actu
rean
da
Type
IIod
onto
idin
who
mha
loim
mob
iliza
tion
faile
d.
Trea
ted
succ
essf
ully
with
ante
rior
C1–
C2
tran
sart
icul
arfix
atio
nan
dan
odon
toid
scre
w.
Ber
lem
ann
and
Schw
arze
nbac
h,19
97(5
)R
etro
spec
tive
revi
ewof
19pa
tient
sag
e�
65w
ithod
onto
id
frac
ture
s.
IIIIn
clud
es4
patie
nts
with
C1
frac
ture
san
dTy
peII
odon
toid
frac
ture
sal
ltr
eate
dw
ithan
teri
orod
onto
idsc
rew
fixat
ion.
Gre
ene
etal
.,19
97(2
3)La
rge
revi
ewof
340
axis
frac
ture
s.III
48pa
tient
sw
ithan
axis
frac
ture
also
had
anat
las
frac
ture
(14%
).Sp
ecifi
cson
man
agem
ent
are
not
pres
ente
d,bu
tth
eau
thor
sin
dica
teth
atth
em
anag
emen
tin
thes
eca
ses
was
base
don
the
C2
frac
ture
.
Cas
tillo
and
Muk
herj
i,19
96(8
)C
ase
repo
rt.
IIIIn
clud
es1
case
ofJe
ffers
onfr
actu
repl
usTy
peII
odon
toid
trea
ted
with
halo
.
Cor
icet
al.,
1996
(9)
Cas
ese
ries
of57
patie
nts
with
hang
man
’sfr
actu
res.
IIIIn
clud
es7
case
sof
com
bina
tion
frac
ture
(C1–
hang
man
’s).
All
wer
etr
eate
dba
sed
onde
gree
of
disp
lace
men
t.If
disp
lace
men
tw
as�
6m
m,
they
wer
etr
eate
dw
ithno
nrig
idim
mob
iliza
tion.
Fujim
ura
etal
.,19
95(1
9)C
ase
seri
esof
axis
body
frac
ture
s.III
Des
crib
es3
patie
nts
with
C1–
mis
cella
neou
sbo
dyfr
actu
real
ltr
eate
dw
ithce
rvic
alim
mob
iliza
tion.
Aut
hors
reco
mm
end
nono
pera
tive
trea
tmen
tex
cept
inca
ses
ofse
vere
angu
latio
n.Ph
ilade
lphi
aco
llar
used
if
min
imal
angu
latio
n.
Polin
etal
.,19
96(4
4)C
ase
seri
esof
62pa
tient
sw
ithod
onto
idfr
actu
res.
IIIIn
clud
es5
case
sof
com
bina
tion
C1–
C2
frac
ture
(8%
).C
1–Je
ffers
on-T
ype
IIod
onto
id,
4ca
ses.
C1–
mis
cella
neou
sC
2bo
dyfr
actu
re,
1ca
se.
All
patie
nts
inse
ries
man
aged
with
eith
erha
loor
colla
r.
Coy
neet
al.,
1995
(10)
Ret
rosp
ectiv
ere
view
of32
patie
nts
with
odon
toid
frac
ture
s
incl
udes
1co
mbi
natio
nfr
actu
re.
III1
case
ofJe
ffers
on-T
ype
IIod
onto
idtr
eate
dw
ithG
allie
fusi
on.
Fujim
ura
etal
.,19
95(1
9)R
etro
spec
tive
revi
ewof
247
adm
issi
ons
with
uppe
rce
rvic
alsp
ine
frac
ture
s.Fo
cuse
son
82pa
tient
sw
ithne
urol
ogic
alde
ficit.
IIIIn
patie
nts
with
com
bine
din
jury
ofC
1–C
2,tw
one
urol
ogic
alde
ficits
occu
rred
inpa
tient
sw
ithpo
ster
ior
arch
frac
ture
,bu
rst
frac
ture
ofth
eat
las,
orbo
dyfr
actu
reof
the
axis
asso
ciat
edw
ithei
ther
anod
onto
id
frac
ture
orha
ngm
an’s
frac
ture
.
S142 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE18
.2.
Con
tinu
ed
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ce
Cla
ssC
oncl
usio
ns
Ben
zel
etal
.,19
94(4
)C
ase
seri
es.
IIIIn
clud
es1
case
ofC
1,ve
rtic
ally
orie
nted
C2
mis
cella
neou
sbo
dyfr
actu
re(tr
eatm
ent
not
desc
ribe
d).
The
auth
ordi
scus
ses
the
poss
ible
mec
hani
sms,
incl
udin
ghy
pere
xten
sion
and
axia
llo
adin
g.
Pede
rsen
and
Kos
tuik
,19
94(4
2)C
ase
repo
rtof
70-y
r-ol
dm
anw
ithfr
actu
redi
sloc
atio
nof
C1–
C2
with
20-m
mat
lant
oaxi
aldi
spla
cem
ent.
IIISu
cces
sful
lytr
eate
dw
ithO
–C4
deco
mpr
essi
onan
dpo
ster
ior
fusi
onw
ithco
mpl
ete
reco
very
.
Han
igan
etal
.,19
93(2
5)C
ase
seri
esof
19pa
tient
s�
80yr
ofag
ew
ithod
onto
idfr
actu
res.
IIIIn
clud
es2
patie
nts
with
aC
1–Je
ffers
on-T
ype
IIod
onto
id.
1pa
tient
died
inth
eho
spita
laf
ter
bein
gpl
aced
in
trac
tion.
1pa
tient
had
ast
able
fibro
usno
nuni
onaf
ter
trea
tmen
tin
aha
lo.
Boh
ayet
al.,
1992
(6)
Cas
ese
ries
.III
Incl
udes
aca
seof
aC
1bu
rst
frac
ture
plus
ave
rtic
alC
2bo
dyfr
actu
resu
cces
sful
lytr
eate
din
ace
rvic
al
colla
ral
one.
Hay
san
dB
ernh
ang,
1992
(27)
Cas
ese
ries
ofun
usua
lfr
actu
res
ofth
eat
las.
IIIIn
clud
es2
case
sof
com
bina
tion
frac
ture
s.C
1(a
nter
ior
arch
)–Ty
peII
odon
toid
frac
ture
faile
dha
lo
trea
tmen
t,re
sulti
ngin
anO
–C2
fusi
on.
Jean
nere
tan
dM
ager
l,19
92(2
9)C
ase
seri
esof
59pa
tient
sw
ithod
onto
idfr
actu
res,
30of
whi
ch
wer
etr
eate
dsu
rgic
ally
.
IIIIn
clud
es2
case
sin
whi
chth
epo
ster
ior
arch
ofC
1w
asno
tin
tact
.C
1–Je
ffers
on-T
ype
IIod
onto
id,
1pa
tient
.
C1–
post
erio
rar
ch-T
ype
IIIod
onto
id,
1pa
tient
.A
utho
rsfe
elst
rong
lyth
at,
ifth
epo
ster
ior
arch
ofC
1is
not
inta
ct,
C1–
C2
tran
sart
icul
arfix
atio
nis
indi
cate
d.In
the
com
men
tth
atfo
llow
sth
ear
ticle
,th
epo
int
ism
ade
that
anon
lay
graf
tbe
twee
nC
1an
dC
2po
ster
iorl
yw
ithou
tw
irin
gof
C1
follo
wed
byha
loim
mob
iliza
tion
has
been
used
inth
issi
tuat
ion.
Rya
nan
dH
ende
rson
,19
92(4
5)Ep
idem
iolo
gica
lre
port
of71
7sp
ine
frac
ture
s.III
Atla
sfr
actu
res
occu
rred
with
odon
toid
frac
ture
s(5
3%)
and
with
hang
man
’sfr
actu
res
(24%
).O
dont
oid
frac
ture
soc
curr
edw
ithat
las
frac
ture
s(1
5%).
Han
gman
’sfr
actu
reoc
curr
edw
ithat
las
frac
ture
(9%
).
Cra
igan
dH
odgs
on,
1991
(11)
Cas
ere
port
.III
Jeffe
rson
plus
supe
rior
face
tof
axis
trea
ted
with
colla
r.
Esse
san
dB
edna
r,19
91(1
6)C
ase
repo
rt.
Atla
san
dod
onto
idfr
actu
re.
IIIJe
ffers
onpl
usTy
peII
odon
toid
ina
34-y
r-ol
dm
antr
eate
dsu
cces
sful
lyw
ithco
llar
only
(see
naf
ter
a1-
mo
dela
yin
diag
nosi
s).
Kes
ters
onet
al.,
1991
(31)
Cas
ese
ries
,re
tros
pect
ive
revi
ew.
IIIIn
clud
es4
patie
nts
with
com
bina
tion
frac
ture
ofth
eat
las
and
Type
IIod
onto
idtr
eate
dw
ithO
–C2
fusi
on.
1
ofth
ese
4pa
tient
sha
da
sign
ifica
ntne
urol
ogic
alde
ficit
(25%
).Th
eau
thor
ssu
gges
tsu
rger
yif
unst
able
and
defin
ein
stab
ility
asat
lant
oaxi
alin
terv
alof
�5
mm
orla
tera
lm
ass
disp
lace
men
t�
7m
m.
Levi
nean
dEd
war
ds,
1991
(35)
Cas
ese
ries
of34
patie
nts
with
atla
sfr
actu
res.
IIIIn
clud
es15
patie
nts
with
aco
mbi
natio
nC
1–C
2fr
actu
re(4
4%).
C1–
Type
IIor
Type
IIIod
onto
id,
8pa
tient
s
(24%
).C
1–ha
ngm
an’s
,7
patie
nts
(21%
).D
escr
ibes
2ca
ses
inth
eC
1–od
onto
idfr
actu
regr
oup
inw
hich
the
post
erio
rC
1ar
chal
tere
dth
etr
eatm
ent
plan
.In
1ca
se,
aG
allie
fusi
onfa
iled,
and
inth
ese
cond
,no
wir
ing
was
used
,ju
ston
lay
bone
graf
t.
Mon
tesa
noet
al.,
1991
(39)
Cas
ese
ries
of14
Type
IIod
onto
idfr
actu
res
trea
ted
with
ante
rior
odon
toid
scre
wfix
atio
n.Fo
llow
-up,
24m
o.
III7
patie
nts
had
aC
1fr
actu
re(5
0%).
The
over
all
fusi
onra
tew
as93
%.
No
prob
lem
sat
trib
uted
dire
ctly
toth
e
C1
frac
ture
.
Zav
onon
eet
al.,
1991
(49)
Cas
ese
ries
of23
C1–
C2
frac
ture
s.III
Incl
udes
2co
mbi
natio
nfr
actu
res
(9%
).C
1–Ty
peII
odon
toid
:th
epa
tient
died
.C
1–ha
ngm
an’s
:tr
eate
d
succ
essf
ully
with
trac
tion
redu
ctio
nan
dM
iner
va.
Fow
ler
etal
.,19
90(1
8)C
ase
seri
esof
48at
las
frac
ture
sfr
omse
ries
of86
7C
-spi
ne
frac
ture
s(5
.5%
).
IIIIn
clud
es18
case
sw
itha
com
bina
tion
C1–
C2
frac
ture
(38%
ofto
tal
seri
es).
C1
burs
t(Je
ffers
on)-
Type
II
odon
toid
,6
patie
nts
(33%
).C
1bu
rst
(Jeffe
rson
)-Ty
peIII
odon
toid
,1
patie
nt(6
%).
C1
burs
t(Je
ffers
on)-
mis
cella
neou
sax
is,
2pa
tient
s(1
1%).
C1
burs
t(Je
ffers
on)-
hang
man
’s,
0pa
tient
s(0
%).
C1
arch
-Typ
eII
odon
toid
,8
patie
nts
(44%
).C
1ar
ch-T
ype
IIIod
onto
id,
1pa
tient
(6%
).C
1ar
ch-m
isce
llane
ous
axis
,1
patie
nt(6
%).
C1
arch
-han
gman
’s,
3pa
tient
s(1
6%).
Thes
eau
thor
spr
esen
tda
tasu
ppor
ting
the
incr
ease
d
mor
talit
yas
soci
ated
with
com
bina
tion
C1–
C2
frac
ture
s.6
(86%
)of
the
7ea
rly
deat
hsha
da
C1
frac
ture
asso
ciat
edw
ithei
ther
aTy
peII
orTy
peIII
odon
toid
frac
ture
.
Dic
kman
etal
.,19
89(1
2)R
etro
spec
tive
revi
ewof
25pa
tient
sw
ithfr
actu
res
ofbo
thC
1an
d
C2.
Com
pris
es3%
ofth
eov
eral
lce
rvic
alsp
ine
inju
ryco
hort
(25/
860)
.
IIIFo
urty
pes
note
d:C
1–Ty
peII
odon
toid
,10
patie
nts
(40%
).C
1–m
isce
llane
ous
axis
,7
patie
nts
(28%
).C
1–
Type
IIIod
onto
id,
5pa
tient
s(2
0%).
C1–
hang
man
’s,
3pa
tient
s(1
2%).
Neu
rolo
gica
lde
ficit
in3/
25pa
tient
s
(12%
).Tr
eatm
ent
dete
rmin
edby
type
ofC
2fr
actu
re.
Non
oper
ativ
e,84
%.
Hal
o,18
patie
nts.
SOM
I,2
patie
nts.
In1
ofth
eC
1–Ty
peII
patie
nts,
halo
faile
d,an
dpa
tient
requ
ired
C1–
C2
fusi
on.
Ope
rativ
e(in
itial
man
agem
ent).
C1–
Type
IIod
onto
idw
ith6
mm
disp
lace
men
t,3
patie
nts
trea
ted
with
post
erio
rC
1–C
2
fusi
on;
1pa
tient
trea
ted
with
O–C
2fu
sion
beca
use
ofm
ultip
lefr
actu
res
inC
1.
Combination Fractures of the Atlas and Axis S143
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE18
.2.
Con
tinu
ed
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ce
Cla
ssC
oncl
usio
ns
Fiel
ding
etal
.,19
89(1
7)C
ase
seri
esof
123
hang
man
’sfr
actu
res.
IIIIn
clud
es:
C1
arch
-han
gman
’s,
10ca
ses
(8%
).C
1bu
rst
(Jeffe
rson
)-ha
ngm
an’s
,2
case
s(2
%).
C1
late
ral
mas
s-
hang
man
’s,
3ca
ses
(3%
).Sp
ecifi
csno
tgi
ven
for
each
subt
ype,
but
over
all
the
auth
ors
reco
mm
end
trea
tmen
tba
sed
onth
eC
2fr
actu
rede
spite
the
pres
ence
ofth
eC
1fr
actu
re.
Reg
ardl
ess
ofth
eC
1fr
actu
re,
the
auth
ors
favo
ran
ante
rior
C2–
C3
fusi
onfo
rth
ose
patie
nts
with
angu
latio
n�
11de
gree
s,as
this
grou
p
had
an85
%no
nuni
onra
tew
ithce
rvic
alim
mob
iliza
tion.
Gov
ende
ran
dC
harl
es,
1987
(22)
Cas
ese
ries
ofup
per
cerv
ical
frac
ture
s.III
Incl
udes
2ca
ses
ofco
mbi
natio
nC
1po
ster
ior
arch
frac
ture
-han
gman
’sfr
actu
re,
trea
ted
succ
essf
ully
with
a
cerv
ical
colla
r(n
onri
gid
cerv
ical
imm
obili
zatio
n).
Han
ssen
and
Cab
enel
a,19
87(2
6)C
ase
seri
esof
42od
onto
idfr
actu
res.
IIIIn
clud
es7
com
bina
tion
frac
ture
s(1
7%).
C1–
Jeffe
rson
-Typ
eII
odon
toid
,6
patie
nts.
5/6
(83%
)di
edw
ithin
first
40d.
1/6
deve
lope
da
stab
leno
nuni
on.
C1–
post
erio
rar
ch-T
ype
IIod
onto
id,
1pa
tient
.H
eale
dw
ith
halo
imm
obili
zatio
n.
Lind
etal
.,19
87(3
6)C
ase
seri
esof
14od
onto
idfr
actu
res
man
aged
inha
loor
thos
es.
IIIIn
clud
es1
case
ofC
1–Je
ffers
on-T
ype
IIod
onto
idm
anag
edin
aha
love
stfo
r12
wk.
Mir
vis
etal
.,19
87(3
8)R
adio
grap
hic
revi
ewof
27C
2fr
actu
res.
IIIN
oted
9as
soci
ated
C1
frac
ture
s(2
6%).
Sega
let
al.,
1987
(46)
Cas
ese
ries
of18
patie
nts
with
atla
sfr
actu
res.
III6
case
sw
ere
com
bina
tion
C1–
C2
frac
ture
s.C
1–Je
ffers
on-o
dont
oid
frac
ture
,5
case
s.3
trea
ted
with
halo
,2
with
trac
tion
follo
wed
byha
lo.
C1–
Jeffe
rson
-han
gman
’s,
1ca
setr
eate
dw
itha
colla
r.
Levi
nean
dEd
war
ds,
1986
(34)
Rev
iew
artic
leon
man
agem
ent
ofC
1–C
2tr
aum
a.III
Com
men
tson
com
bine
din
juri
es:
1.Th
epr
esen
ceof
thre
ein
juri
esto
the
C1–
C2
com
plex
isas
soci
ated
with
ahi
ghlik
elih
ood
ofne
urol
ogic
alin
jury
.2.
If1
inju
ryor
frac
ture
isfo
und,
one
shou
ldlo
okca
refu
lly
for
anot
her.
3.M
echa
nism
ofin
jury
usua
llyis
cons
iste
ntw
ithth
ein
jury
obse
rved
.4.
Each
inju
ryne
eds
to
beev
alua
ted
indi
vidu
ally
;fo
rex
ampl
e,th
epr
esen
ceof
2fr
actu
res
does
not
alw
ays
indi
cate
inst
abili
ty
(pos
teri
orar
chof
C1
plus
ano
ndis
plac
edha
ngm
an’s
frac
ture
).5.
Stag
ing
oftr
eatm
ent
may
bere
quir
ed(a
s
desc
ribe
dby
Lips
onet
al.
belo
w)
with
allo
wan
ceof
one
frac
ture
tohe
albe
fore
trea
ting
defin
itive
ly.
Levi
nean
dEd
war
ds,
1985
(33)
Cas
ese
ries
of53
patie
nts
with
hang
man
’sfr
actu
re.
Des
crib
es
stab
le(T
ype
Iha
ngm
an’s
)an
dun
stab
le(T
ype
IIha
ngm
an’s
)
grou
ps.
IIIIn
clud
es9
case
sof
Type
Iha
ngm
an’s
(sta
ble)
plus
C2
frac
ture
:Ty
peII
odon
toid
,2
case
s.Ty
peIII
odon
toid
,
3ca
ses.
Post
erio
rar
ch,
1ca
se.
Bur
st(Je
ffers
on),
2ca
ses.
Late
ral
mas
s,1
case
.O
nly
1ca
seTy
peII
hang
man
’s(u
nsta
ble)
with
C2
frac
ture
.Po
ster
ior
arch
,1
case
.O
nly
1ca
setr
eate
dsu
rgic
ally
:Ty
peI
hang
man
’spl
usTy
peII
odon
toid
trea
ted
with
post
erio
rC
1–C
2fu
sion
.
Pepi
nan
dH
awki
ns,
1981
(43)
Cas
ese
ries
of41
odon
toid
frac
ture
s.III
Incl
udes
9ca
ses
ofod
onto
idfr
actu
rein
com
bina
tion
with
anot
her
spin
alfr
actu
re,
ofw
hich
the
C1–
Jeffe
rson
-Typ
eII
odon
toid
was
the
mos
tco
mm
on.
All
trea
ted
with
eith
erC
1–C
2fu
sion
orha
lo.
Aut
hor
reco
mm
ends
fusi
onin
the
elde
rly.
Effe
ndi
etal
.,19
81(1
3)C
ase
seri
esof
131
hang
man
’sfr
actu
rew
ithcl
assi
ficat
ion.
IIIIn
clud
esco
mbi
natio
nfr
actu
res:
C1
post
erio
rar
ch-h
angm
an’s
,8
patie
nts
(8/1
31,
6%).
Odo
ntoi
dfr
actu
re-
hang
man
’s,
2pa
tient
s(2
/131
,2%
).Sp
ecifi
cou
tcom
esno
tpr
esen
ted,
but
all
fuse
dw
ithei
ther
ante
rior
or
post
erio
rC
1–C
2fu
sion
orha
lo.
Ove
rall
mor
talit
yw
as9%
.
Ekon
get
al.,
1981
(14)
Cas
ese
ries
of22
patie
nts
with
odon
toid
frac
ture
s.III
Incl
udes
:C
1–Je
ffers
on-T
ype
IIod
onto
id,
1pa
tient
.C
1–Je
ffers
on-T
ype
IIIod
onto
id,
2pa
tient
s.A
lltr
eate
d
with
halo
.H
alo
faile
din
1of
the
C1
Jeffe
rson
-Typ
eIII
odon
toid
patie
nts,
requ
irin
gC
1–C
2po
ster
ior
fusi
on.
Lips
on,
1977
(37)
Cas
ese
ries
of3
case
sof
atla
sfr
actu
repl
usTy
peII
odon
toid
.III
The
auth
ors
reco
mm
end
com
bina
tion
ther
apy
ofha
loim
mob
iliza
tion
for
10–1
2w
kun
tilth
epo
ster
ior
arch
ofth
eat
las
frac
ture
has
heal
ed,
follo
wed
byat
lant
oaxi
alfu
sion
(Gal
liety
pe)
tode
finiti
vely
man
age
the
odon
toid
frac
ture
.
Bra
shea
ret
al.,
1975
(7)
Cas
ese
ries
ofha
ngm
an’s
frac
ture
.III
Incl
udes
2pa
tient
sw
ithC
1po
ster
ior
arch
frac
ture
plus
hang
man
’str
eate
dw
ithre
duct
ion
and
Min
erva
for
3–6
mo.
And
erso
nan
dD
’Alo
nzo,
1974
(1)
Cas
ese
ries
ofod
onto
idfr
actu
res.
IIIIn
clud
es1
patie
ntw
ithco
mbi
ned
C1–
Type
IIod
onto
idfr
actu
retr
eate
dw
ithO
–C2
fusi
on.
Ellio
tet
al.,
1972
(15)
Cas
ese
ries
.III
C1
post
erio
rar
ch-h
angm
an’s
,2
case
str
eate
dw
ithim
mob
iliza
tion.
Sher
kan
dN
icho
lson
,19
70(4
8)C
ase
repo
rt.
III1
case
each
ofa
com
bina
tion
C1–
Type
IIod
onto
idan
da
C1–
hang
man
’s.
Bot
hw
ere
trea
ted
with
imm
obili
zatio
n(r
educ
tion
intr
actio
nfo
llow
edby
aM
iner
vabr
ace)
succ
essf
ully
.
aO
,oc
cipi
tal;
C-s
pine
,ce
rvic
alsp
ine;
SOM
I,su
bocc
ipit
al-m
andi
bula
rim
mob
ilize
r.
S144 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
patient deteriorated during or as a result of treatment. Sixpatients complained of persistent neck pain or limitation ofneck motion. The authors offered a treatment algorithm basedon the type and displacement of the axis fracture. They be-lieve that every patient with a C1 or C2 fracture should bestudied with computed tomography to rule out a combinationinjury. When present, atlas fractures in combination withType II or Type III odontoid fractures with an atlantoaxialinterval of 5 mm or more should be considered for earlysurgical management. The authors stressed that the integrityof the C1 ring must be assessed to determine whether C1–C2wiring techniques can be used. Their perspectives were of-fered before the popularization of C1–C2 transarticular screwfixation techniques.
Guiot and Fessler (24), in 1999, described a series of 10patients with combination atlas-axis fractures treated withsurgical stabilization and fusion. In five (50%) of these pa-tients, halo immobilization had failed, and the patients werereferred specifically for operative intervention. Ninety per-cent were patients with C1–Type II odontoid fractures, andthe remaining patient had a C1–Type III odontoid combina-tion fracture injury. One patient died of unrelated causes inthe follow-up period. There were no other significant compli-cations in a follow-up period of 28.5 months. All nine otherpatients accomplished successful fusion. An odontoid screwalone was used in five patients, an odontoid screw plus C2pedicle screws in one, posterior transarticular screws in two,and anterior transarticular screws in one patient. The authors’indications for surgery included patients with fractures thatcould not be reduced or maintained with external immobili-zation and unstable fractures with a high likelihood of non-union (including evidence of disruption of the transverseatlantal ligament).
Treatment of C1–Type II odontoidcombination fractures
The treatment of specific fracture combinations has beenthe subject of numerous reports. The C1–Type II odontoidfracture combination seems to be the most frequent and thesubject of the most variability in treatment strategy. As notedwith the management of isolated Type II odontoid fractures,optimal treatment remains controversial (see Chapter 17).Management techniques for C1–Type II odontoid combina-tion fractures include semirigid immobilization (collar), trac-tion and then immobilization in a brace, rigid immobilization(halo, Minerva, SOMI), posterior fusion with and withoutinstrumentation, and anterior odontoid screw fixation. Al-though Esses and Bednar (16) describe a single cases of C1–Type II odontoid combination fracture managed successfullyin a cervical collar, the lower fusion rate described for Type IIodontoid fractures managed in a collar alone should be con-sidered when electing this treatment option (see Chapter 17).Sherk and Nicholson (48) described a single patient success-fully treated with traction reduction and then immobilizationin a Minerva brace. Segal et al. (46) treated two patients withtraction and then rigid immobilization. Some authors havedescribed the treatment of C1–Type II odontoid combination
fractures with rigid immobilization (halo, SOMI, Minerva) (8,12, 14, 26, 32, 36, 46). Dickman et al. (12) described five of sixpatients successfully treated in this way (83% success rate).All six patients had an atlantoaxial interval of less than 6 mm.Halo immobilization failed in one patient with an atlantoaxialinterval of 5 mm, and the patient required posterior C1–C2fusion at 12 weeks postinjury. Segal et al. (46) described threecases of C1–Type II odontoid combination fracture success-fully treated with halo immobilization. Andersson et al. (2)described two patients older than 65 years with this combi-nation fracture injury who were successfully treated with ahalo device. Seybold and Bayley (47) added two more patientstreated with a halo resulting in successful union. Additionalsingle cases managed with halo immobilization have beendescribed (8, 14, 26, 32, 36, 43).
The C1–Type II odontoid combination fracture has beensuccessfully managed with surgical stabilization and fusion.Dickman et al. (12) treated four patients with C1–Type IIodontoid combination fractures with early surgical fusionbased on an atlantoaxial interval of 6 mm or more. Threepatients had posterior C1–C2 fusion, and one patient under-went occipitocervical fusion for multiple fractures of the pos-terior atlantal arch. Andersson et al. (2) treated one patientwith C1–Type II odontoid combination fracture with posteriorC1–C2 fusion in a series of elderly patients. Coyne et al. (10)also treated one patient with this injury pattern with a C1–C2posterior fusion. Several authors have suggested that the C1arch fracture be allowed to heal before undertaking definitiveatlantoaxial arthrodesis for this subtype of combination frac-tures. Other authors have suggested using onlay bone graftfor C1–C2 fusion and then halo immobilization in the settingof posterior C1 arch incompetence (29, 34, 37). Lee et al. (32)described the surgical management of two patients with C1–Type II odontoid combination fractures in whom posteriorC1–C2 fusion was performed. Guiot and Fessler (24) de-scribed two patients with this combination injury patterntreated posteriorly with C1–C2 transarticular screw fixationand fusion. Some investigators have used anterior odontoidscrew fixation in the treatment of C1–Type II odontoid com-bination fractures. Montesano et al. (39), in 1991, describedfour cases successfully managed in this fashion. Berlemannand Schwarzenbach (5) published an additional four cases.The report by Guiot and Fessler (24) included six patients inwhom odontoid screw fixation was accomplished. These au-thors added anterior transarticular fixation in one patient. In1999, Henry et al. (28) described a fusion success rate of 90%in 10 patients with C1–Type II odontoid combination fracturestreated with anterior odontoid screw fixation. Apostolides etal. (3) described a single case in which three screws wereplaced, all from an anterior trajectory, to stabilize the C1–C2articulation bilaterally and the odontoid fracture. Occipitocer-vical fusion has been reported in the management of C1–TypeII odontoid combination fractures (1, 2, 12, 27, 31, 42). It seemsthat this approach is reserved for patients with disruption ofthe C1 arch and gross C1–C2 instability.
In summary, a variety of treatment options have been ef-fective in C1–Type II odontoid combination fractures. Exter-nal orthoses have been successfully used in the management
Combination Fractures of the Atlas and Axis S145
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
of most of these injuries. Combination fractures of this sub-type with C1–C2 instability as defined by an atlantodensinterval of 5 mm or more have a high failure rate with externalimmobilization alone and have been successfully managedwith operative reduction, internal fixation, and fusion.
Treatment of C1–Type III odontoidcombination fractures
Dickman et al. (12) described five patients with C1–Type IIIodontoid combination fractures. All were successfully treatedwith halo immobilization for an average of 12 weeks. Ekong etal. (14) identified two similar cases. One was managed suc-cessfully in a halo. In the second patient, halo immobilizationfailed, and a delayed posterior C1–C2 fusion was required.Guiot and Fessler (24) reported a patient with a C1–Type IIIodontoid-hangman’s combination fracture, which they suc-cessfully treated with ventral odontoid screw fixation andthen posterior pedicle screw fixation and fusion. It seems thatexternal immobilization is effective in the management ofthese injuries in most patients.
Treatment of C1–hangman’s combination fractures
Most reported combination injuries of the atlas and theposterior elements of the axis have been successfully man-aged with semirigid or rigid external immobilization (with orwithout initial traction) (7, 9, 12, 15, 32, 46, 49). Coric et al. (9)and Lee et al. (32) described the successful treatment of ninepatients with this combination fracture type with a cervicalcollar only. Dickman et al. (12) reported three patients withC1–hangman’s combination fractures successfully treatedwith either a halo or SOMI device. The reports of Elliott et al.(15), Brashear et al. (7), Segal et al. (46), Govender and Charles(22), and Zavanone et al. (49) each describe patients withsimilar injuries successfully treated with nonoperative tech-niques. As with an isolated unstable hangman’s fracture, sur-gical fixation may be an option. The report by Fielding et al.(17) included 15 patients with C1–hangman’s combinationfractures. These authors recommended that fractures withangulation between C2 and C3 of 11 degrees or more betreated surgically. These combination fractures with angula-tion of more than 11 degrees were associated with an 85%nonunion rate with nonoperative management, in their expe-rience. This combination injury subtype seems to be managedeffectively with external immobilization alone. Unstable inju-ries, as defined by C2–C3 angulation of 11 degrees or more,may require surgical management.
Treatment of C1-miscellaneous C2 bodycombination fractures
Combination fractures of the atlas associated with miscel-laneous axis body fractures have been treated with both rigidand nonrigid immobilization (6, 11, 12, 20, 32, 44). Dickman etal. (12) reported seven cases of this combination fracturesubtype treated successfully with either a halo or SOMI brace.The cases described by Fujimura et al. (20), Lee et al. (32),Craig and Hodgson (11), and Bohay et al. (6) were all man-aged successfully with a cervical collar alone. A single case
described by Polin et al. (44) was treated with traction andsubsequent halo immobilization. Nonoperative managementof this combination injury subtype is effective.
SUMMARY
Combination fractures involving fractures of both the atlasand axis occur relatively frequently. A higher incidence ofneurological deficit is associated with C1–C2 combinationfractures compared with either C1 or C2 fractures in isolation.The C1–Type II odontoid combination fracture seems to be themost common combination injury subtype, and then C1–miscellaneous axis, C1–Type III odontoid, and C1–hangman’scombination fractures. No Class I or Class II evidence ad-dressing the management of patients with combination atlasand axis fractures is available. All of the articles revieweddescribe case series or case reports containing Class III evi-dence supporting a variety of treatment strategies for theseunique fracture injuries.
In most circumstances, the specifics of the axis fracture willdictate the most appropriate management of the combinationfracture injury. As reported for isolated atlas and axis frac-tures, most atlas-axis combination fractures can be effectivelytreated with rigid external immobilization. Combinationatlas-axis fractures with an atlantoaxial interval of 5 mm ormore or angulation of C2–C3 of 11 degrees or more may beconsidered for surgical fixation and fusion. The integrity ofthe ring of the atlas must often be taken into account whenplanning a specific surgical strategy using instrumentationand fusion techniques. If the posterior arch of C1 is inade-quate, both incorporation of the occiput into the fusion con-struct (occipitocervical fusion) and posterior C1–C2 transar-ticular screw fixation and fusion have been successful.
KEY ISSUES FOR FUTURE INVESTIGATION
The identification of which of the atlas-axis combinationfracture subtypes are at greatest risk for nonunion and sub-sequent instability would be useful in determining appropri-ate management for combination fracture injuries. A uniformand clinically useful definition of cranial, C1, and C2 instabil-ity in association with these fractures would be of benefit.Prospective data collection and case-control studies at manyinstitutions would provide meaningful data addressing theseissues. The relative infrequency of combined atlas-axis frac-tures would make a randomized study difficult. Patients witha C1–Type II odontoid combination fracture should be stud-ied to compare operative and nonoperative management andshould be evaluated in terms of management morbidity, long-term success, economic benefit, patient satisfaction, and re-turn to preinjury activities.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 615 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Anderson LD, D’Alonzo RT: Fractures of the odontoid process ofthe axis. J Bone Joint Surg Am 56A:1663–1674, 1974.
S146 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
2. Andersson S, Rodrigues M, Olerud C: Odontoid fractures: Highcomplication rate associated with anterior screw fixation in theelderly. Eur Spine J 9:56–60, 2000.
3. Apostolides PJ, Theodore N, Karahalios DG, Sonntag VKH: Tripleanterior screw fixation of an acute combination atlas-axis fracture:Case report. J Neurosurg 87:96–99, 1997.
4. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW, EspinosaMC: Fractures of the C-2 vertebral body. J Neurosurg 81:206–212,1994.
5. Berlemann U, Schwarzenbach O: Dens fractures in the elderly:Results of anterior screw fixation in 19 elderly patients. ActaOrthop Scand 68:319–324, 1997.
6. Bohay D, Gosselin RA, Contreras DM: The vertical axis fracture:A report on three cases. J Orthop Trauma 6:416–419, 1992.
7. Brashear R Jr, Venters G, Preston ET: Fractures of the neural archof the axis: A report of twenty-nine cases. J Bone Joint Surg Am57A:879–887, 1975.
8. Castillo M, Mukherji SK: Vertical fractures of the dens. AJNRAm J Neuroradiol 17:1627–1630, 1996.
9. Coric D, Wilson JA, Kelly DL Jr: Treatment of traumatic spon-dylolisthesis of the axis with nonrigid immobilization: A reviewof 64 cases. J Neurosurg 85:550–554, 1996.
10. Coyne TJ, Fehlings MG, Wallace MC, Bernstein M, Tator CH:C1–C2 posterior cervical fusion: Long-term evaluation of resultsand efficacy. Neurosurgery 37:688–693, 1995.
11. Craig JB, Hodgson BF: Superior facet fractures of the axis verte-bra. Spine 16:875–877, 1991.
12. Dickman CA, Hadley MN, Browner C, Sonntag VKH: Neurosur-gical management of acute atlas-axis combination fractures: Areview of 25 cases. J Neurosurg 70:45–49, 1989.
13. Effendi B, Roy D, Cornish B, Dussault RG, Laurin CA: Fracturesof the ring of the axis: A classification based on the analysis of 131cases. J Bone Joint Surg Br 63B:319–327, 1981.
14. Ekong CE, Schwartz ML, Tator CH, Rowed DW, Edmonds VE:Odontoid fracture: Management with early mobilization usingthe halo device. Neurosurgery 9:631–637, 1981.
15. Elliott JM Jr, Rogers LF, Wissinger JP, Lee JF: The hangman’sfracture: Fractures of the neural arch of the axis. Radiology 104:303–307, 1972.
16. Esses SI, Bednar DA: Screw fixation of odontoid fractures andnonunions. Spine 16[Suppl 10]:S483–S485, 1991.
17. Fielding JW, Francis WR Jr, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. Clin Orthop 239:47–52,1989.
18. Fowler JL, Sandhu A, Fraser RD: A review of fractures of the atlasvertebra. J Spinal Disord 3:19–24, 1990.
19. Fujimura Y, Nishi Y, Chiba K, Kobayashi K: Prognosis of neuro-logical deficits associated with upper cervical spine injuries. Para-plegia 33:195–202, 1995.
20. Fujimura Y, Nishi Y, Kobayashi K: Classification and treatment ofaxis body fractures. J Orthop Trauma 10:536–540, 1996.
21. Gleizes V, Jacquot FP, Signoret F, Feron JM: Combined injuries inthe upper cervical spine: Clinical and epidemiological data over a14-year period. Eur Spine J 9:386–392, 2000.
22. Govender S, Charles RW: Traumatic spondylolisthesis of the axis.Injury 18:333–335, 1987.
23. Greene KA, Dickman CA, Marciano FF, Drabier JB, Hadley MN,Sonntag VKH: Acute axis fractures: Analysis of management andoutcome in 340 consecutive cases. Spine 22:1843–1852, 1997.
24. Guiot B, Fessler RG: Complex atlantoaxial fractures. J Neurosurg91[Suppl 2]:139–143, 1999.
25. Hanigan WC, Powell FC, Elwood PW, Henderson JP: Odontoidfractures in elderly patients. J Neurosurg 78:32–35, 1993.
26. Hanssen AD, Cabanela ME: Fractures of the dens in adult pa-tients. J Trauma 27:928–934, 1987.
27. Hays MB, Bernhang AM: Fractures of the atlas vertebra: A three-part fracture not previously classified. Spine 17:240–242, 1992.
28. Henry AD, Bohly J, Grosse A: Fixation of odontoid fractures by ananterior screw. J Bone Joint Surg Br 81B:472–477, 1999.
29. Jeanneret B, Magerl F: Primary posterior fusion C1/2 in odontoidfractures: Indications, technique, and results of transarticularscrew fixation. J Spinal Disord 5:464–475, 1992.
30. Jefferson G: Fractures of the atlas vertebra: Report of four casesand a review of those previously reported. Br J Surg 7:407–422,1920.
31. Kesterson L, Benzel EC, Orrison W, Coleman J: Evaluation andtreatment of atlas burst fractures (Jefferson fractures).J Neurosurg 75:213–220, 1991.
32. Lee TT, Green BA, Petrin DR: Treatment of stable burst fracture ofthe atlas (Jefferson fracture) with rigid cervical collar. Spine 23:1963–1967, 1998.
33. Levine AM, Edwards CC: The management of traumatic spon-dylolisthesis of the axis. J Bone Joint Surg Am 67A:217–226, 1985.
34. Levine AM, Edwards CC: Treatment of injuries in the C1–C2complex. Orthop Clin North Am 17:31–44, 1986.
35. Levine AM, Edwards CC: Fractures of the atlas. J Bone Joint SurgAm 73A:680–691, 1991.
36. Lind B, Nordwall A, Sihlbom H: Odontoid fractures treated withhalo-vest. Spine 12:173–177, 1987.
37. Lipson SJ: Fractures of the atlas associated with fractures of theodontoid process and transverse ligament ruptures. J Bone JointSurg Am 59A:940–943, 1977.
38. Mirvis SE, Young JW, Lim C, Greenberg J: Hangman’s fracture:Radiologic assessment in 27 cases. Radiology 163:713–717, 1987.
39. Montesano PX, Anderson PA, Schlehr F, Thalgott JS, Lowrey G:Odontoid fractures treated by anterior odontoid screw fixation.Spine 16[Suppl 3]:S33–S37, 1991.
40. Morandi X, Hanna A, Hamlat A, Brassier G: Anterior screwfixation of odontoid fractures. Surg Neurol 51:236–240, 1999.
41. Muller EJ, Wick M, Muhr G: Traumatic spondylolisthesis of theaxis: Treatment rationale based on the stability of the differentfracture types. Eur Spine J 9:123–128, 2000.
42. Pedersen AK, Kostuik JP: Complete fracture-dislocation of theatlantoaxial complex: Case report and recommendations for anew classification of dens fractures. J Spinal Disord 7:350–355,1994.
43. Pepin JW, Hawkins RJ: Traumatic spondylolisthesis of the axis:Hangman’s fracture. Clin Orthop 157:133–138, 1981.
44. Polin RS, Szabo T, Bogaev CA, Replogle RE, Jane JA: Nonopera-tive management of Types II and III odontoid fractures: ThePhiladelphia collar versus the halo vest. Neurosurgery 38:450–457, 1996.
45. Ryan MD, Henderson JJ: The epidemiology of fractures andfracture-dislocations of the cervical spine. Injury 23:38–40, 1992.
46. Segal LS, Grimm JO, Stauffer ES: Non-union of fractures of theatlas. J Bone Joint Surg Am 69A:1423–1434, 1987.
47. Seybold EA, Bayley JC: Functional outcome of surgically andconservatively managed dens fractures. Spine 23:1837–1846,1998.
48. Sherk HH, Nicholson JT: Fractures of the atlas. J Bone Joint SurgAm 52A:1017–1024, 1970.
49. Zavanone M, Guerra P, Rampini P, Crotti F, Vaccari U: Traumaticfractures of the craniovertebral junction: Management of 23 cases.J Neurosurg Sci 35:17–22, 1991.
Combination Fractures of the Atlas and Axis S147
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
CHAPTER 19
Os Odontoideum
RECOMMENDATIONSDIAGNOSIS:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options: Plain x-rays of the cervical spine (anteroposterior, open-mouth odontoid, and lateral) and plain
dynamic lateral x-rays performed in flexion and extension are recommended. Tomography (computed orplain) and/or magnetic resonance imaging of the craniocervical junction may be considered.
MANAGEMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• Patients with os odontoideum, either with or without C1–C2 instability, who have neither symptoms nor
neurological signs may be managed with clinical and radiographic surveillance.• Patients with os odontoideum, particularly with neurological symptoms and/or signs, and C1–C2 instability
may be managed with posterior C1–C2 internal fixation and fusion.• Postoperative halo immobilization as an adjunct to posterior internal fixation and fusion is recommended
unless successful C1–C2 transarticular screw fixation and fusion can be accomplished.• Occipitocervical fusion with or without C1 laminectomy may be considered in patients with os odontoi-
deum who have irreducible cervicomedullary compression and/or evidence of associated occipitoatlantalinstability.
• Transoral decompression may be considered in patients with os odontoideum who have irreducible ventralcervicomedullary compression.
RATIONALE
The definition of os odontoideum is uniform throughoutthe literature: an ossicle with smooth circumferentialcortical margins representing the odontoid process that
has no osseous continuity with the body of C2 (16, 22). Theetiology of os odontoideum remains debated in the literaturewith evidence for both acquired and congenital causes (18, 23,25). The etiology of os odontoideum, however, does not playan important role in its diagnosis or subsequent management.
Diagnosis
Os odontoideum can present with a wide range of clinicalsymptoms and signs; it can also be an incidental finding onimaging. The literature has focused on three groups of pa-tients with os odontoideum: 1) those with occipitocervicalpain alone, 2) those with myelopathy, and 3) those withintracranial symptoms or signs from vertebrobasilar ischemia(4). Patients with os odontoideum and myelopathy have beensubcategorized further into those with 1) transient myelopa-thy (commonly after trauma), 2) static myelopathy, and 3)progressive myelopathy (10). Because patients with occipito-
cervical pain, myelopathy, or vertebrobasilar ischemia likelywill have etiologies other than os odontoideum, the diagnosisof os odontoideum is not usually considered until imaging isobtained. The presence of an os odontoideum is usually firstsuggested after obtaining plain cervical spine x-rays. Mostoften, plain cervical spine x-rays are sufficient to obtain adiagnosis (15).
Os odontoideum has been classified into two anatomictypes, orthotopic and dystopic. Orthotopic defines an ossiclethat moves with the anterior arch of C1, whereas dystopicdefines an ossicle that is functionally fused to the basion. Thedystopic os odontoideum may sublux anterior to the arch ofC1 (10). Tomography and computed tomography (CT) havebeen used to better define the bony anatomy of the os odon-toideum and the odontoid process. Plain dynamic x-rays inflexion/extension have been used to depict the degree ofabnormal motion between C1 and C2. Most often, there isanterior instability, with the os odontoideum subluxing for-ward in relation to the body of C2. However, at times one willsee either no discernible instability or “posterior instability,”with the os odontoideum moving posteriorly into the spinalcanal during neck extension (10, 20).
S148 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
With regard to diagnosis, there are two issues of concern inthe imaging of os odontoideum. First, although plain x-raysare often diagnostic for os odontoideum, the sensitivity andspecificity of plain cervical x-rays for os odontoideum havenot been reported. The usefulness of confirmatory studies,such as CT and plain tomography and magnetic resonanceimaging (MRI), has not been well defined. Second, after thediagnosis of os odontoideum on plain cervical x-rays, insta-bility and osseous anomalies associated with os odontoideumcan influence clinical management. The best methods of fur-ther evaluating or excluding these complicating factors de-serve definition.
Management
The natural history of untreated os odontoideum covers awide spectrum. The literature provides many examples ofboth asymptomatic and symptomatic patients with known osodontoideum who have never been treated and who have hadno reported new problems in follow-up over many years (22).Conversely, examples of sudden spinal cord injury in associ-ation with os odontoideum after minor trauma have also beenreported (17). The natural history of os odontoideum is vari-able, and predictive factors for deterioration, particularly inthe asymptomatic patient, have not been identified. Indica-tions for surgical stabilization include the simple existence ofan os odontoideum, os odontoideum in association with oc-cipitocervical pain alone, and/or os odontoideum in associa-tion with neurological deficit (10, 22). Other factors that mayassist in determining the need for stabilization and/or decom-pression include C1–C2 instability, associated deformities,and spinal cord compression. A variety of techniques havebeen used to stabilize C1 and C2 in patients with os odontoi-deum (2, 3, 5, 6, 10, 20, 22, 26, 27). Fusion success rates andcomplication rates for these various procedures may provideevidence as to whether a preferred method of C1–C2 arthro-desis is supported by the literature.
Finally, neural compression is an important considerationin patients with os odontoideum. Neural compression may beanterior from a combination of bone and soft tissue, or pos-terior from the dorsal arch of C1. Surgical techniques to sta-bilize and fuse across the craniocervical junction with or with-out C1 laminectomy and techniques that provide ventraldecompression have been reported in the treatment of os odon-toideum with irreducible neural compression (6, 24). The litera-ture will be examined in light of the risks and benefits thesetechniques may provide to patients with os odontoideum.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The key phrase “os odontoideum” identified 121articles. Articles written in English were reviewed. Twenty-seven articles that described the clinical aspects and manage-
ment of patients with os odontoideum were identified andused to generate these guidelines. None of the articles meet-ing selection criteria represented Class I or Class II studies. All27 provided Class III evidence regarding the diagnosisand/or management of os odontoideum. These 27 articlesrepresent the basis for this review and are summarized inTable 19.1. In addition, one article germane to the topic but notmeeting criteria for inclusion in the Evidentiary Tables isreferenced in the Scientific Foundation section and includedin the references.
SCIENTIFIC FOUNDATION
Diagnostic evaluation
There is no literature that describes the sensitivity andspecificity of imaging studies for os odontoideum. Dai et al.(6), in their review of 44 patients with os odontoideum, usedtomography, CT, and MRI in addition to “routine” plaincervical x-rays (anteroposterior, lateral, open-mouth, flexion/extension x-rays) in 39, 27, and 22 patients, respectively. Mat-sui et al. (16) described only the plain x-rays of 12 patientswith os odontoideum. They excluded patients with Down’ssyndrome and Klippel-Feil anomalies. The authors made nomention of any other studies to obtain or confirm the diag-nosis in these 12 patients. Likewise, Watanabe et al. (27) andSpierings and Braakman (22) described the plain x-rays of 34and 37 patients, respectively, with os odontoideum, withoutreference to other imaging studies. Fielding et al. (10) de-scribed 35 patients with os odontoideum in which each pa-tient had extensive roentgenographic investigation, includingmultiple roentgenograms of the cervical spine and oftenflexion-extension lateral roentgenograms and flexion-extension laminagrams (10). No mention was made as towhether additional studies beyond static plain cervical spinex-rays were necessary to confirm the diagnosis of os odontoi-deum in their series of patients.
The literature supports the ability of plain cervical spinex-rays to establish the diagnosis of os odontoideum. There isno compelling evidence in the literature that supports theneed for additional studies to confirm the diagnosis of osodontoideum. Specific characteristics or associated abnormal-ities of os odontoideum, including C1–C2 instability, softtissue masses, spinal canal diameter, associated osseousanomalies, spinal cord appearance, and vertebral artery com-promise have been investigated with a variety of imagingstudies. The imaging of abnormal motion and spinal cordcompression in association with os odontoideum has receivedthe most attention in the reported clinical series.
Instability of C1–C2 in association with os odontoideum hasbeen investigated with multiple imaging modalities. Usingflexion/extension lateral cervical spine x-ray studies in 33patients, Fielding et al. (10) reported 22 patients (67%) withanterior instability who had a mean atlantodens interval of
Os Odontoideum
S149Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE19
.1.
Sum
mar
yof
Rep
orts
onO
sO
dont
oide
uma
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Apf
elba
umet
al.,
2000
(1)
18pa
tient
sw
ithod
onto
idfr
actu
res
incu
rred
�18
mo
befo
retr
eatm
ent
who
wer
e
trea
ted
with
ante
rior
odon
toid
scre
wfix
atio
n.
III16
patie
nts
with
follo
w-u
p.25
%fu
sion
rate
.3
with
scre
wfr
actu
rean
d1
with
scre
w
pullo
ut.
Bro
ckm
eyer
etal
.,20
00(2
)R
evie
wof
tran
sart
icul
arsc
rew
plac
emen
tin
31ch
ildre
n.12
child
ren
with
os
odon
toid
eum
(age
d4–
16yr
).
III55
/62
poss
ible
site
sde
emed
suita
ble
for
tran
sart
icul
arsc
rew
s.A
llch
ildre
nw
ithos
odon
toid
eum
wer
eab
leto
have
2sc
rew
spl
aced
.
Dai
etal
.,20
00(6
)A
revi
ewof
44pa
tient
sag
ed7–
56yr
with
osod
onto
ideu
m.
Mea
nfo
llow
-up,
6.5
yr.
III7
patie
nts
wer
eas
ympt
omat
ic.
5of
thes
e7
wer
etr
eate
dw
itha
cerv
ical
colla
ron
ly
and
have
rem
aine
dst
able
.39
unde
rwen
tfu
sion
succ
essf
ully
(9at
lant
oaxi
alan
d33
occi
pito
cerv
ical
).Sy
mpt
oms
and
sign
sdi
sapp
eare
din
26an
dim
prov
edin
13.
Tagg
ard
etal
.,20
00(2
4)A
revi
ewof
cran
iove
rteb
ral
junc
tion
abno
rmal
ities
in36
Dow
n’s
synd
rom
e
patie
nts.
Os
odon
toid
eum
pres
ent
in12
.
III27
unde
rwen
tsu
rgic
alpr
oced
ures
.O
f11
with
basi
lar
inva
gina
tion,
itw
as
irre
duci
ble
in5,
and
tran
sora
lde
com
pres
sion
was
perf
orm
ed.
Dic
kman
and
Sonn
tag,
1998
(7)
Rev
iew
of12
1pa
tient
str
eate
dw
ithpo
ster
ior
C1–
C2
tran
sart
icul
arsc
rew
san
d
wir
edpo
ster
ior
bone
stru
ts.
Os
odon
toid
eum
was
pres
ent
in9.
This
grou
pw
as
com
pare
dw
ith74
patie
nts
trea
ted
with
post
erio
rw
irin
gte
chni
ques
alon
e.
III2
failu
res
inth
etr
ansa
rtic
ular
grou
p.Th
eet
iolo
gyof
the
C1–
C2
inst
abili
tyw
asno
t
stat
edfo
rth
ese
2fa
ilure
s.1/
8pa
tient
sw
ithos
odon
toid
eum
inth
epo
ster
ior
wir
ing
grou
pha
da
nonu
nion
.O
vera
llfu
sion
rate
for
tran
sart
icul
arw
as98
%ve
rsus
86%
for
post
erio
rw
irin
gte
chni
ques
.
Wan
get
al.,
1999
(26)
16ch
ildre
ntr
eate
dfo
rat
lant
oaxi
alin
stab
ility
.4
ofth
emha
dos
odon
toid
eum
and
wer
etr
eate
dw
ithC
1–C
2tr
ansa
rtic
ular
scre
ws
and
post
erio
rw
irin
gan
dfu
sion
tech
niqu
es.
IIIA
llfu
sed.
No
halo
imm
obili
zatio
n.
Tran
sart
icul
arsc
rew
sw
ere
succ
essf
ully
used
inch
ildre
nas
youn
gas
4yr
.
Kuh
nset
al.,
1998
(14)
4ch
ildre
nw
ithos
odon
toid
eum
unde
rwen
tM
RI
exam
inat
ions
.III
All
4ch
ildre
nha
dch
ange
sin
the
nuch
alco
rdco
nsis
tent
with
inju
ry.
Low
ryet
al.,
1997
(15)
Are
view
of25
child
ren
requ
irin
gup
per
cerv
ical
fusi
ons.
11ch
ildre
nha
dos
odon
toid
eum
.
III10
unde
rwen
ta
Bro
oks-
type
C1–
C2
fusi
on.
2of
thes
ech
ildre
ndi
dno
tfu
se.
1
unde
rwen
ta
Gal
lie-t
ype
fusi
on.
This
child
rem
aine
dun
stab
lean
dw
asre
vise
dto
a
Bro
oks-
type
fusi
on,
whi
chw
assu
cces
sful
.
Mat
sui
etal
.,19
97(1
6)R
evie
wof
the
plai
nra
diog
raph
icm
orph
olog
yof
C2
eval
uate
din
12pa
tient
s(1
5–
71yr
old)
with
osod
onto
ideu
mun
rela
ted
toan
ysy
ndro
me.
III3
conf
igur
atio
nsde
scri
bed
from
anan
tero
post
erio
rvi
ew:
roun
d,co
ne,
and
blun
t-
toot
h.M
yelo
path
yw
asm
ore
seve
rein
the
grou
pw
itha
roun
dco
nfig
urat
ion.
Ver
ska
and
And
erso
n,19
97(2
5)R
epor
tof
apa
irof
iden
tical
twin
s,1
with
osod
onto
ideu
m,
and
1w
ithou
tos
odon
toid
eum
.
IIIH
isto
ryof
trau
ma
inth
etw
inw
ithan
osod
onto
ideu
m.
Fell
atag
e3
yr,
had
tort
icol
lisan
dne
ckpa
info
rse
vera
lm
onth
s.
Wat
anab
eet
al.,
1996
(27)
Rev
iew
of34
patie
nts
with
osod
onto
ideu
m(5
–76
yrol
d).
Div
ided
into
grou
psby
Row
land
clas
sific
atio
n(1
�lo
cal
sym
ptom
s,2
�po
sttr
aum
atic
tran
sien
t
mye
lopa
thy,
3,4
�pr
ogre
ssiv
em
yelo
path
yor
intr
acra
nial
sym
ptom
s).
Late
ral
neut
ral
and
dyna
mic
x-ra
ysob
tain
ed.
Sagi
ttal
plan
ero
tatio
nan
gle,
min
imum
dist
ance
,an
din
stab
ility
inde
xw
ere
mea
sure
d.
IIILo
wco
rrel
atio
nbe
twee
nsa
gitta
lpl
ane
rota
tion
angl
ean
din
stab
ility
inde
x.Sa
gitta
l
plan
ero
tatio
nan
gle
of�
20de
gree
sor
inst
abili
tyin
dex
of�
40%
corr
elat
esw
ith
mye
lopa
thy.
Cle
men
tset
al.,
1995
(4)
Rep
ort
ofno
nope
rativ
etr
eatm
ent
ofan
inci
dent
ally
disc
over
edos
odon
toid
eum
with
out
C1–
C2
inst
abili
tyat
diag
nosi
s.
IIIA
fter
5yr
,pr
ofou
ndC
1–C
2in
stab
ility
and
sym
ptom
sha
dde
velo
ped,
nece
ssita
ting
post
erio
rin
stru
men
tatio
nan
dfu
sion
.
Coy
neet
al.,
1995
(5)
Rev
iew
ofpo
ster
ior
C1–
C2
fusi
onan
din
stru
men
tatio
nte
chni
ques
.5/
32pa
tient
s
had
osod
onto
ideu
m.
III3/
5w
ithos
odon
toid
eum
faile
dw
ithpo
ster
ior
wir
ing
tech
niqu
es.
All
wer
e
imm
obili
zed
inha
los.
2/5
deve
lope
dne
wne
urol
ogic
alde
ficits
asop
erat
ive
com
plic
atio
ns.
Stev
ens
etal
.,19
94(2
3)R
evie
wof
abno
rmal
odon
toid
san
dC
1–C
2in
stab
ility
.24
/62
patie
nts
with
os
odon
toid
eum
.9
child
ren
and
15ad
ults
.
IIIPe
riod
onto
idso
fttis
sue
thic
keni
ngw
aspr
esen
ton
lyin
thos
ew
ithM
orqu
io’s
dise
ase.
Afte
rfu
sion
,th
eod
onto
idw
asno
ted
topa
rtia
llyor
com
plet
ely
rege
nera
te
inca
ses
ofM
orqu
io’s
dise
ase.
Men
ezes
and
Ryk
en,
1992
(17)
Rev
iew
of18
Dow
n’s
synd
rom
epa
tient
sw
ithsy
mpt
omat
icce
rvic
omed
ulla
ry
com
prom
ise.
4ha
dos
odon
toid
eum
.
IIIA
ll4
had
gros
sin
stab
ility
ondy
nam
icx-
rays
.Su
cces
sful
fusi
onw
ithpo
ster
ior
wir
ing
tech
niqu
esan
dfu
ll-th
ickn
ess
rib
graf
ts.
Imm
obili
zed
for
a“m
inim
umof
3m
onth
s.”
Dic
kman
etal
.,19
91(8
)R
evie
wof
36pa
tient
str
eate
dw
ithC
1–C
2po
ster
ior
wir
ing
and
fusi
onfo
rva
riou
s
reas
ons.
4pa
tient
s(a
ged
16,
25,
38,
43yr
)ha
dos
odon
toid
eum
.A
llpa
tient
s
wer
em
aint
aine
din
aha
lofo
r12
wk
afte
rsu
rger
y.
IIIO
fth
e4
with
osod
onto
ideu
m,
3de
velo
ped
osse
ous
unio
nsan
d1
had
ast
able
fibro
usun
ion
(follo
w-u
p,15
–44
mo)
.N
oco
mpl
icat
ions
for
thes
e4
patie
nts.
Hos
ono
etal
.,19
91(1
3)C
iner
adio
grap
hic
eval
uatio
nof
6pa
tient
sw
ithos
odon
toid
eum
.III
2ty
pes
ofC
1po
ster
ior
arch
tran
slat
ion:
stra
ight
(ver
tical
)(n
�4)
and
sigm
oid
(n�
2).
Cor
rela
ted
abno
rmal
mot
ion
with
biom
echa
nics
ofpo
ster
ior
wir
ing
tech
niqu
es.
Smith
etal
.,19
91(2
1)R
evie
wof
17ch
ildre
nop
erat
edon
for
C1–
C2
inst
abili
ty.
11ha
dos
odon
toid
eum
.Po
ster
ior
wir
ing
tech
niqu
es,
auto
logo
usbo
ne,
and
halo
used
inal
l.
III2/
11w
ithos
odon
toid
eum
had
nonu
nion
s.1
cord
inju
ryth
ough
tse
cond
ary
to
subl
amin
arw
ire
pass
age.
S150 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TAB
LE19
.1.
Con
tinu
ed
Seri
es(R
ef.
No.
)D
escr
iptio
nof
Stud
yEv
iden
ceC
lass
Con
clus
ions
Shir
asak
iet
al.,
1991
(20)
9pa
tient
sw
ithos
odon
toid
eum
and
post
erio
rin
stab
ility
had
3ra
diog
raph
ic
para
met
ers
mea
sure
d.D
ista
nce
betw
een
the
osan
dC
2sp
inou
spr
oces
sin
exte
nsio
n(D
ext),
dist
ance
betw
een
the
osan
dpo
ster
ior
C1
arch
(Dat
l),an
d
“deg
ree
ofin
stab
ility
”(In
st).
Thes
efin
ding
sw
ere
com
pare
dto
thei
rne
urol
ogic
al
stat
us.
IIITh
ose
with
out
hist
ory
orev
iden
ceof
mye
lopa
thy
had
aD
ext
of�
16m
m.
Dex
tw
as
�16
mm
inth
ose
with
mye
lopa
thy.
The
pres
ence
orab
senc
eof
mye
lopa
thy
was
not
rela
ted
toth
eIn
st.
Inth
ose
with
mye
lopa
thy
and
Dat
l�
13m
m,
ther
ew
as
reve
rsib
leco
rdco
mpr
essi
onin
exte
nsio
n;in
thos
ew
itha
Dat
lof
�13
mm
,th
eco
rd
rem
aine
dco
mpr
esse
din
flexi
onan
dex
tens
ion.
Mor
gan
etal
.,19
89(1
8)R
epor
tof
3fa
mily
mem
bers
with
C2–
C3
Klip
pel-
Feil
abno
rmal
ities
and
os
odon
toid
eum
.
IIIA
ges
16(in
dex
case
),39
(fath
er),
and
64yr
(pat
erna
lgr
andm
othe
r).
Non
ew
ith
neur
olog
ical
sign
sor
sym
ptom
s.
Yam
ashi
taet
al.,
1989
(28)
Cor
rela
tion
ofcl
inic
alst
atus
,M
RI
scan
s,an
dx-
rays
in29
patie
nts
with
C1–
C2
inst
abili
ty.
4ha
dos
odon
toid
eum
.
IIITh
eA
DI
did
not
corr
elat
ew
ithth
ede
gree
ofm
yelo
path
y,bu
tM
RI
degr
eeof
cord
com
pres
sion
did
corr
elat
ew
ithde
gree
ofm
yelo
path
y.
Fren
chet
al.,
1987
(11)
Rev
iew
ofdy
nam
icce
rvic
alsp
ine
x-ra
ysin
185
patie
nts
with
Dow
n’s
synd
rom
e.III
6ha
dab
norm
alod
onto
ids
cons
iste
ntw
ithos
odon
toid
eum
for
anin
cide
nce
of3%
.
3ha
dpr
evio
usx-
rays
show
ing
noab
norm
ality
.1
had
anex
agge
rate
dA
DI
of6
mm
.
Spie
ring
san
dB
raak
man
,19
82(2
2)37
patie
nts
with
osod
onto
ideu
m.
20tr
eate
dco
nser
vativ
ely.
IIIO
f20
man
aged
cons
erva
tivel
y,1
was
lost
tofo
llow
-up.
15ha
dno
mye
lopa
thy
(med
ian
follo
w-u
p,5
yr),
and
none
deve
lope
dm
yelo
path
y.O
f4
with
mye
lopa
thy
(follo
w-u
p,0.
5,1,
7,an
d14
yr),
1is
dead
from
canc
er,
1ha
sne
ckpa
in,
1ha
sne
ck
pain
and
pare
sthe
sias
,an
d1
has
head
ache
s.
Fiel
ding
etal
.,19
80(1
0)35
patie
nts
(3–6
5yr
old)
with
osod
onto
ideu
m.
25pa
tient
sw
ere
sym
ptom
atic
.III
22pa
tient
sha
dan
teri
orin
stab
ility
with
am
ean
AD
Iof
10.3
mm
.5
had
post
erio
r
inst
abili
ty.
3ha
dno
dete
ctab
lem
otio
n.3
had
�3
mm
ofC
1–C
2m
otio
n.26
unde
rwen
tpo
ster
ior
fusi
onsu
cces
sful
ly.
5w
ere
not
oper
ated
on,
3w
ere
asym
ptom
atic
with
noin
stab
ility
.Th
eyre
mai
ned
wel
lw
ithno
inst
abili
tyat
1,2,
and
3yr
,re
spec
tivel
y.1
patie
ntw
ithin
stab
ility
refu
sed
surg
ery
but
was
wel
lat
2-yr
follo
w-u
p.1
patie
ntdi
edof
rena
lfa
ilure
.
Bro
oks
and
Jenk
ins,
1978
(3)
3ch
ildre
n(8
,11
,an
d12
yrol
d)w
ithos
odon
toid
eum
trea
ted
with
subl
amin
ar
C1–
C2
wir
esan
dau
tolo
gous
iliac
cres
tgr
aft.
Min
erva
cast
imm
obili
zatio
n.
IIIA
llfu
sed.
Spon
tane
ous
exte
nsio
nof
fusi
onto
C3
in1
child
.
Dyc
k,19
78(9
)R
evie
wof
8ch
ildre
n(a
ges
7–17
yr)
with
osod
onto
ideu
m.
6w
ere
trea
ted
with
post
erio
rw
irin
gan
dfu
sion
ofC
1–C
3.Ex
tern
alim
mob
iliza
tion
for
“usu
ally
”3–
4
mo.
III6
child
ren
unde
rwen
tpo
ster
ior
fusi
onby
the
auth
or.
2re
quir
edre
oper
atio
n.
Gri
swol
det
al.,
1978
(12)
4pa
tient
sw
ithos
odon
toid
eum
wer
etr
eate
dw
ithsu
blam
inar
C1–
C2
wir
esan
d
auto
logo
usili
accr
est.
III3
fuse
d.1
did
not
fuse
afte
r3
atte
mpt
s.
aM
RI,
mag
neti
cre
sona
nce
imag
ing;
AD
I,at
lant
oden
sin
terv
al.
Os Odontoideum S151
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
10.3 mm, five patients (15%) with posterior instability (meanposterior translation of the os odontoideum during extension,8.4 mm), three patients (9%) with less than 3 mm of C1–C2motion, and three patients (9%) with no detectable C1–C2 mo-tion. Eight patients (23%) had both anterior and posteriorinstability. The authors noted that cineradiography was help-ful in examining range of motion at C1–C2 in these patients,but it was not of benefit in the measurement of the degree ofinstability. Of note is that almost one-fifth of the patients intheir series manifested no radiographic evidence of C1–C2instability.
Spierings and Braakman (22) studied 21 of their 37 patientswith os odontoideum with flexion/extension cervical spinex-rays or tomograms. They measured the maximal distancethe os odontoideum moved in the sagittal plane, the innerdiameter of the atlas, and the minimal spinal canal diameter(the distance between the posterior aspect of the C2 body andthe dorsal arch of C1 during flexion). They compared thesemeasurements in two groups, those with and without my-elopathy. The degree of C1–C2 instability did not correlatewith neurological status, but the measured minimal spinalcanal diameter was significantly smaller (P � 0.05) in thegroup with myelopathy. They identified 13 mm as the criticalanteroposterior spinal diameter. Watanabe et al. (27) madesimilar measurements in 34 patients using plain lateral cervi-cal x-rays in flexion/extension. As in the Spierings and Braak-man study, the degree of instability in their patients did notcorrelate with the presence of myelopathy. Shirasaki et al. (20)described radiographic findings on lateral flexion/extensionx-rays in nine patients with os odontoideum. They reportedthat a distance of 13 mm or less between the os odontoideumand the dorsal arch of C1 specifically defined severe cervicalmyelopathy (20) in their patients. They, too, found that thedegree of C1–C2 instability did not correlate with the presenceof myelopathy. Yamashita et al. (28) studied atlantoaxial sub-luxation with plain radiography and MRI and correlated theimaging studies with the degree of myelopathy in 29 patients(4 with os odontoideum). They found that the degree ofmyelopathy did not correlate with the distance of subluxationof C1 on C2 on plain x-rays. The degree of cord compressionon MRI did correlate well with the degree of myelopathymeasured clinically. Matsui et al. (16) classified os odontoi-deum into three types according to the shape of the os odon-toideum on plain x-rays. Three types were described: round,cone, and blunt-tooth. They compared these three os odon-toideum types to the degree of clinical myelopathy and foundthat the degree of myelopathy correlated most closely withthe “round” os odontoideum type. Kuhns et al. (14) describedthe MRI appearance of os odontoideum in four children andidentified signal changes within the posterior ligaments con-sistent with trauma. They could not discern whether thesechanges represented a primary or secondary phenomenonwith respect to atlantoaxial instability.
These studies provide two consistent conclusions: 1) thedegree of C1–C2 instability does not seem to correlate withneurological status in patients with os odontoideum; and 2)sagittal spinal canal diameter on plain x-rays of 13 mm or lessis strongly associated with myelopathy.
Beyond plain spine x-rays and flexion-extension x-rays,imaging to assist with operative planning of unstable os odon-toideum receives brief mention in several reports (11, 17, 24,26). Important factors to consider before proceeding withsurgical intervention for this disorder are the ability to reduceC1–C2, spinal cord compression, an assimilated atlas, an in-complete C1 ring, the course of the vertebral arteries at C1 andC2, and the presence of an associated congenital fusion of thecervical spine (e.g., Klippel-Feil). Plain x-rays, tomography,and CT provide information regarding the ability to achieveanatomic alignment of C1 on C2 and the presence or absenceof a congenital fusion. MRI is the best modality for viewingcord compression even after apparent C1–C2 realignment(28). CT can provide important information about the bonyanatomy at the craniocervical junction, including the com-pleteness of the atlas ring and the position of the transverseforamina at C1 and C2 (19). Hosono et al. (13) made interest-ing observations on the different motions of the posterior archof C1 in relation to C2 in patients with os odontoideum. Theyobserved two patterns of motion, linear and sigmoid. Theythought that, in those patients with a sigmoid-shaped motionpattern, posterior wiring techniques may not provide ade-quate stability. The selection of and necessity for additionalimaging studies in the evaluation of os odontoideum seems tobe made on a patient-by-patient basis. The literature providesno convincing evidence as to which patients should undergosupplemental imaging (CT or MRI) after the diagnosis of osodontoideum has been made.
Management
The universal theme of the various management strategiesoffered in the treatment of patients with os odontoideum hasbeen either to confirm or to secure cervical spinal stability atthe C1–C2 level. The earliest reports of os odontoideum de-scribe small pediatric case series treated surgically. In 1978,Griswold et al. (12) described four children with os odontoi-deum who underwent posterior C1–C2 wiring and autolo-gous iliac fusion. Three children had successful arthrodesis.The fourth child did not achieve fusion/stability despite threeattempts. In the same year, Brooks and Jenkins (3) describedtheir technique of C1–C2 wiring and fusion and reportedthree children with os odontoideum who were immobilizedpostoperatively in Minerva jackets. All three patientsachieved successful fusion. In summary, six of the sevenchildren with os odontoideum described in these two reportswere successfully treated.
Two larger series, reported in the early 1980s, includedadults and children with os odontoideum and described bothoperative and nonoperative management strategies for thesepatients. Fielding et al. (10) described 35 patients with osodontoideum, 27 with radiographic evidence of instability.Twenty-six of these 27 patients underwent successful poste-rior C1–C2 fusion (Gallie type). Fusions were noted to be“solid” after 2 months of immobilization in children and 3months in adults. One patient with instability refused surgeryand remained well at the 2-year follow-up examination. Theeight remaining patients with no evidence of C1–C2 instabil-
S152 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
ity managed nonoperatively remained well at the lastfollow-up (1–3 yr). Spierings and Braakman (22) describedtheir management of 37 patients with os odontoideum. Sev-enteen were treated surgically. They provide 20 patients foranalysis of the natural history of os odontoideum. Informa-tion about radiographic stability was provided for only 21 ofthe 37 patients they reported. Sixteen patients in their seriespresented with neck pain only or had an incidentally discov-ered os odontoideum. Nine of these 16 patients had flexion/extension x-rays. Of these nine patients, seven had abnormalmotion of 8 mm or more. With a median follow-up of 7 years,none of these 16 patients developed a neurological deficit.Four additional patients who presented with myelopathywere treated nonoperatively with follow-up from 6 months to14 years. Three of these four patients presented with transientmyelopathy and had no recurrence at last follow-up, despiteabnormal motion of C1 on C2 of 8 to 16 mm. The fourthpatient had a stable monoparesis at last follow-up. Of the17 patients who underwent surgery, one patient had neuro-logical worsening and two died. Eight of these 17 patientstreated surgically had a posterior C1–C2 fusion. Nine patientsunderwent occipitocervical fusion with C1 laminectomy. Theauthors did not report a single failed fusion. They had acombined surgical morbidity and mortality of 18% (3 of 17patients). The authors conclude that patients with os odontoi-deum without C1–C2 instability can be managed withoutsurgical stabilization and fusion with good results. Althoughthey did not provide operative treatment to every os odon-toideum patient with C1–C2 instability, those with myelopa-thy and greater amounts of instability were more likely toundergo surgery. If these two series are considered represen-tative of patients with os odontoideum, the implication is thatminimally symptomatic or asymptomatic patients with osodontoideum without C1–C2 instability can be managed non-operatively with little or no morbidity over time. Althoughpatients with os odontoideum and myelopathy or C1–C2instability have been managed conservatively, most patientswith myelopathy or instability are treated surgically.
Clements et al. (4), in 1995, reported a patient who had adocumented os odontoideum without instability who, at 5years follow-up, developed symptomatic frank C1–C2 insta-bility that required surgical stabilization and fusion. It seemsthat a lack of C1–C2 instability at initial diagnosis does notguarantee that instability will not develop in these patients. Itis recommended, therefore, that clinical and radiographicfollow-up be provided to patients with os odontoideum whoare found to have radiographic C1–C2 stability on initialassessment.
More recent series reported in the literature provide betterdescriptions of the operative procedures and postoperative im-mobilization techniques used for patients with os odontoideum(2, 5–9, 15, 17, 21–26). Smith et al. (21) described 11 children withos odontoideum who underwent posterior wiring and at-tempted fusion. Autologous bone graft and halo immobilizationwere used in all children. Two children had fusion failure withnonunion. One child incurred an intraoperative cord injury sec-ondary to sublaminar wire passage. Lowry et al. (15) also de-scribed 11 children with os odontoideum who were treated with
C1–C2 fusion and posterior wiring. One child treated with aGallie-type procedure had continued instability and fusion fail-ure. The C1–C2 construct was revised successfully with aBrooks-type fusion procedure. The remaining 10 children weresuccessfully treated with Brooks C1–C2 wiring and fusion pro-cedures. Coyne et al. (5), in a review of posterior C1–C2 fixationtechniques, described five patients with os odontoideum. Threeof these five had unsuccessful attempted posterior fusions de-spite halo immobilization. Two developed new neurologicaldeficits after surgery. Dai et al. (6) described 44 patients with osodontoideum with a mean follow-up of 6.5 years. Seven patientswere asymptomatic at presentation. Five of these seven refusedsurgery and were treated with a cervical collar only; they re-mained stable at last follow-up. The remaining 39 patients un-derwent successful fusion procedures after skeletal traction. Theauthors reported that nine patients underwent atlantoaxial fu-sion and 33 required occipitocervical fusion (42 operations in 39patients). Symptoms and signs disappeared in 26 of their oper-ative patients and improved in the remaining 13 at last follow-up. They used occipitocervical constructs with fusion with orwithout C1 laminectomy in those patients with irreducible de-formities because of the concern that sublaminar passage ofwires or cables might result in neurological morbidity.
Wang et al. (26) reported 16 children with C1–C2 instability,four of whom had os odontoideum. These four children weretreated with C1–C2 transarticular screw fixation with poste-rior C1–C2 wiring and fusion. The youngest child was 4 yearsold. All achieved stable fusion arthrodesis without complica-tions. Halo immobilization devices were not used. Brockm-eyer et al. (2) also reported 31 children they treated withC1–C2 transarticular screw fixation and fusion. Twelve ofthese children had os odontoideum. Bilateral screws wereplaced successfully without complication in all children withos odontoideum. The authors did not comment on the type ofpostoperative immobilization devices they used. In 1991,Dickman et al. (8) reviewed their experience with fusion plus12 weeks of halo immobilization in the treatment of C1–C2instability. The authors described 36 patients with C1–C2 in-stability, four of whom had os odontoideum. Three of four osodontoideum patients they treated in this way developedosseous union. One had a stable fibrous union at last follow-up. In a subsequent report in 1998, Dickman and Sonntag (7)compared their series of patients undergoing C1–C2 transar-ticular screw fixation with posterior wiring and fusion withpatients treated with posterior wiring and fusion alone. Thefusion rates in the two groups were 98 and 86%, respectively.No patient with os odontoideum treated with C1–C2 transar-ticular fusion techniques experienced failure to fuse. Only oneof eight patients with os odontoideum in the posterior wiringand fusion group developed a nonunion (previously de-scribed). In contrast to the patients treated with posteriorwiring and fusion only, no patient treated with transarticularscrew fixation required postoperative halo immobilization.Menezes and Ryken (17) described four children with osodontoideum and Down’s syndrome whom they successfullytreated with posterior wiring and fusion, using full-thicknessautograft rib and at least 3 months of postoperative haloimmobilization. Dyck (9) reported eight children with os
Os Odontoideum S153
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
odontoideum, six of whom were treated with posterior C1–C3wiring and fusion techniques. All were externally immobi-lized in a four-poster brace “usually” for 3 to 4 months post-operatively. Two of six children required reoperation fornonunions.
Apfelbaum et al. (1) described their experience in treatingrecent and remote (�18 mo after injury) odontoid injurieswith anterior screw fixation. They reported a fusion rate of25% in 16 “remote” odontoid injuries. If an os odontoideumwere considered anatomically similar to a “remote” odontoidfracture, then the rate of fusion for os odontoideum treatedwith an odontoid screw fixation would also be expected to bepoor. Anterior C1–C2 transfacetal fixation techniques mayhave merit in the surgical treatment of os odontoideum, butthere are no descriptions of their application for os odontoi-deum in the literature.
The surgical treatment of patients with C1–C2 instability inassociation with os odontoideum has been demonstrated to besuccessful when combined fusion and internal fixation tech-niques are used, usually in conjunction with postoperativehalo immobilization. Fusion success rates and reports of op-erative morbidity varied considerably among the clinical caseseries reported in the literature. Although the numbers aresmall, transarticular C1–C2 screw fixation and fusion has beenassociated with higher rates of fusion than posterior wiringand fusion techniques alone. Of note is that patients treatedwith transarticular screw fixation have been managed in hardcollars postoperatively, obviating the need for halo immobi-lization devices. If transarticular screw fixation is not used inthe treatment of unstable os odontoideum, postoperative haloimmobilization as an adjunct to dorsal internal fixation andfusion is recommended.
Ventral or transoral decompression for irreducible ventralcervicomedullary compression in association with os odon-toideum has been suggested (24). Reports of the managementof ventral compression and os odontoideum are scant. In areview of 36 patients with Down’s syndrome and craniover-tebral junction abnormalities, Taggard et al. (24) described 12patients with os odontoideum. Eleven of the 36 patients re-ported were noted to have basilar invagination. Five of these11 patients with basilar invagination had irreducible ventralspinal cord compression and were treated with transoral de-compression. The authors reported stable to excellent out-comes without complications after transoral decompression inall five patients; however, the total number of patients whohad basilar invagination due to os odontoideum was notdescribed. The report implies, however, that selected patientswith atlantoaxial instability and irreducible symptomatic ven-tral cervicomedullary compression may benefit from ventraldecompression. On the other hand, Dai et al. (6) reported thesuccessful use of occipitocervical fusion with or without C1laminectomy in cases of irreducible deformity with cervi-comedullary neural compression in 33 patients with os odon-toideum. These authors described improvement in all patientsand no complications related to their dorsal-only approach.Although it may seem intuitive to remove ventral neuralcompression in association with os odontoideum, the litera-
ture suggests that dorsal stabilization and fusion withoutventral decompression is an effective management option.
SUMMARY
Plain cervical spine x-rays seem to be adequate for makinga diagnosis of os odontoideum in most patients with thisdisorder. Lateral flexion/extension x-rays can provide usefulinformation regarding C1–C2 instability. Tomography (com-puted or plain) may be helpful to define the osseous relation-ships at the cranial base, C1, and C2 in patients in whom thecraniovertebral junction is not well visualized on plain x-rays.The degree of C1–C2 instability identified on cervical x-raysdoes not correlate with the presence of myelopathy. A sagittaldiameter of less than 13 mm in the spinal canal at the C1–C2level does correlate with myelopathy detected on clinicalexamination. MRI can depict spinal cord compression andsignal changes within the cord that correlate with the pres-ence of myelopathy.
Surgical treatment is not required for every patient inwhom os odontoideum is identified. Patients who have noneurological deficit and have no instability at C1–C2 onflexion/extension studies can be managed without operativeintervention. Even patients with documented C1–C2 instabil-ity and neurological deficit have been managed nonopera-tively without clinical consequence during finite follow-upperiods. Most investigators of this disorder favor operativestabilization and fusion of C1–C2 instability associated withos odontoideum. The concern exists that patients with osodontoideum who have C1–C2 instability have an increasedlikelihood of future spinal cord injury. Although not sup-ported by Class I or Class II evidence from the literature,multiple case series (Class III evidence) suggest that stabili-zation and fusion of C1–C2 is meritorious in this circumstance(6, 15, 24, 26). Because a patient with an initially stable osodontoideum has been reported to develop delayed C1–C2instability, and because there are rare examples of patientswith stable os odontoideum who have developed neurologi-cal deficits after minor trauma, longitudinal clinical and ra-diographic surveillance of patients with os odontoideumwithout instability is recommended (4, 10).
Posterior C1–C2 arthrodesis in the treatment of os odontoi-deum provides effective stabilization of the atlantoaxial jointin most patients. Posterior wiring and fusion techniques sup-plemented with postoperative halo immobilization providedsuccessful fusion in 40 to 100% of cases reported (3, 5, 6, 22,26). Atlantoaxial transarticular screw fixation and fusionseems to have merit in the treatment of C1–C2 instability inassociation with os odontoideum and seems to obviate theneed for postoperative halo immobilization. Neural compres-sion in association with os odontoideum has been treated withreduction of deformity, dorsal decompression of irreducibledeformity, and ventral decompression of irreducible defor-mity, each in conjunction with C1–C2 or occipitocervical fu-sion and internal fixation. Each of these combined approacheshas provided satisfactory results. Odontoid screw fixation hasno role in the treatment of this disorder.
S154 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
KEY ISSUES FOR FUTURE INVESTIGATION
A cooperative multi-institutional natural history study ofpatients with os odontoideum without C1–C2 instabilitywould provide demographic and clinical information thatmight indicate predictive factors for the development of sub-sequent instability. In a related study, the prevalence of osodontoideum as an incidental finding should be established.The literature supports essentially no treatment for os odon-toideum, even with C1–C2 subluxation. Whether activity re-striction is called for in these patients would be helpful andshould be studied. A cooperative multi-institutional prospec-tive randomized trial of posterior wiring and fusion tech-niques with and without C1–C2 transarticular screw fixationfor patients with os odontoideum and C1–C2 instabilitywould help to definitively identify the risks and merits of eachof the two procedures in this patient population.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Apfelbaum RI, Lonser RR, Veres R, Casey A: Direct anterior screwfixation for recent and remote odontoid fractures. J Neurosurg93[Suppl 2]:227–236, 2000.
2. Brockmeyer DL, York JE, Apfelbaum RI: Anatomic suitability ofC1–2 transarticular screw placement in pediatric patients.J Neurosurg 92[Suppl 1]:7–11, 2000.
3. Brooks AL, Jenkins EB: Atlanto-axial arthrodesis by the wedgecompression method. J Bone Joint Surg Am 60A:279–284, 1978.
4. Clements WD, Mezue W, Mathew B: Os odontoideum: Congen-ital or acquired? That’s not the question. Injury 26:640–642, 1995.
5. Coyne TJ, Fehlings MG, Wallace MC, Bernstein M, Tator CH:C1–C2 posterior cervical fusion: Long-term evaluation of resultsand efficacy. Neurosurgery 37:688–693, 1995.
6. Dai L, Yuan W, Ni B, Jai L: Os odontoideum: Etiology, diagnosis,and management. Surg Neurol 53:106–109, 2000.
7. Dickman CA, Sonntag VKH: Posterior C1–C2 transarticular screw fix-ation for atlantoaxial arthrodesis. Neurosurgery 43:275–281, 1998.
8. Dickman CA, Sonntag VK, Papadopoulos SM, Hadley MN: Theinterspinous method of posterior atlantoaxial arthrodesis.J Neurosurg 74:190–198, 1991.
9. Dyck P: Os odontoideum in children: Neurological manifesta-tions and surgical management. Neurosurgery 2:93–99, 1978.
10. Fielding JW, Hensinger RN, Hawkins RJ: Os odontoideum.J Bone Joint Surg Am 62A:376–383, 1980.
11. French HG, Burke SW, Roberts JM, Johnston CE II, Whitecloud T,Edmunds JO: Upper cervical ossicles in Down syndrome.J Pediatr Orthop 7:69–71, 1987.
12. Griswold DM, Albright JA, Schiffman E, Johnson R, SouthwickW: Atlanto-axial fusion for instability. J Bone Joint Surg Am60A:285–292, 1978.
13. Hosono N, Yonenobu K, Ebara S, Ono K: Cineradiographic mo-tion analysis of atlantoaxial instability in os odontoideum. Spine16[Suppl 10]:S480–S482, 1991.
14. Kuhns LR, Loder RT, Farley FA, Hensinger RN: Nuchal cordchanges in children with os odontoideum: Evidence for associ-ated trauma. J Pediatr Ortho 18:815–819, 1998.
15. Lowry DW, Pollack IF, Clyde B, Albright AL, Adelson PD: Uppercervical spine fusion in the pediatric population. J Neurosurg87:671–676, 1997.
16. Matsui H, Imada K, Tsuji H: Radiographic classification of osodontoideum and its clinical significance. Spine 22:1706–1709,1997.
17. Menezes AH, Ryken TC: Craniovertebral abnormalities inDown’s syndrome. Pediatr Neurosurg 18:24–33, 1992.
18. Morgan MK, Onofrio BM, Bender CE: Familial os odontoideum:Case report. J Neurosurg 70:636–639, 1989.
19. Paramore CG, Dickman CA, Sonntag VKH: The anatomical suit-ability of the C1–2 complex for transarticular screw fixation.J Neurosurg 85:221–224, 1996.
20. Shirasaki N, Okada K, Oka S, Hosono N, Yonenobu K, Ono K: Osodontoideum with posterior atlantoaxial instability. Spine 16:706–715, 1991.
21. Smith MD, Phillips WA, Hensinger RN: Fusion of the uppercervical spine in children and adolescents: An analysis of 17patients. Spine 16:695–701, 1991.
22. Spierings EL, Braakman R: The management of os odontoideum:Analysis of 37 cases. J Bone Joint Surg Br 64B:422–428, 1982.
23. Stevens JM, Chong WK, Barber C, Kendall BE, Crockard HA: Anew appraisal of abnormalities of the odontoid process associatedwith atlantoaxial subluxation and neurological disability. Brain117:133–148, 1994.
24. Taggard DA, Menezes AH, Ryken TC: Treatment of Downsyndrome-associated craniovertebral junction abnormalities.J Neurosurg 93[Suppl 2]:205–213, 2000.
25. Verska JM, Anderson PA: Os odontoideum: A case report of oneidentical twin. Spine 22:706–709, 1997.
26. Wang J, Vokshoor A, Kim S, Elton S, Kosnik E, Bartkowski H:Pediatric atlantoaxial instability: Management with screw fixa-tion. Pediatr Neurosurg 30:70–78, 1999.
27. Watanabe M, Toyama Y, Fujimura Y: Atlantoaxial instability in osodontoideum with myelopathy. Spine 21:1435–1439, 1996.
28. Yamashita Y, Takahashi M, Sakamoto Y, Kojima R: Atlantoaxialsubluxation: Radiography and magnetic resonance imaging cor-related to myelopathy. Acta Radiol 30:135–140, 1989.
Plate from Gautier D, Duverney M: Essai D’anatomie, en TableauxImprimés. . . . Paris, 1745. Courtesy, Dr. Irwin J. Pincus, Los Angeles,California.
Os Odontoideum S155
CHAPTER 20
Treatment of Subaxial Cervical Spinal Injuries
RECOMMENDATIONSSUBAXIAL CERVICAL FACET DISLOCATION INJURIES:Standards: There is insufficient evidence to recommend treatment standards.Guidelines: There is insufficient evidence to recommend treatment guidelines.Options:• Closed or open reduction of subaxial cervical facet dislocation injuries is recommended.• Treatment of subaxial cervical facet dislocation injuries with rigid external immobilization, anterior arth-
rodesis with plate fixation, or posterior arthrodesis with plate or rod or interlaminar clamp fixation isrecommended.
• Treatment of subaxial cervical facet dislocation injuries with prolonged bedrest in traction is recommendedif more contemporary treatment options are not available.
SUBAXIAL CERVICAL INJURIES EXCLUDING FACET DISLOCATION INJURIES:Standards: There is insufficient evidence to recommend treatment standards.Guidelines: There is insufficient evidence to recommend treatment guidelines.Options:• Closed or open reduction of subluxations or displaced subaxial cervical spinal fractures is recommended.• Treatment of subaxial cervical spinal injuries with external immobilization, anterior arthrodesis with plate
fixation, or posterior arthrodesis with plate or rod fixation is recommended.
RATIONALE
Subaxial cervical vertebral fracture-dislocation injuriesare common after nonpenetrating cervical trauma andare often associated with neurological injury. Before the
advent of spinal instrumentation, many of these injuries weremanaged with traction, postural reduction, or external ortho-ses with frequent success. However, the morbidity and mor-tality associated with prolonged immobilization for 3 monthsor more prompted surgeons to investigate the usefulness ofinternal fixation in the management of these injuries. To de-velop treatment recommendations for closed subaxial cervicalspinal injuries, an analysis of the articles examining theirmanagement is undertaken in this chapter. In particular, thisfocused review assessed the usefulness of closed reductionwith or without external immobilization compared with arth-rodesis with or without internal fixation.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The search terms “cervical vertebrae,” “spinalfractures,” and “dislocations,” limited to English languagearticles, yielded 15,124, 3,010, and 17,811 citations, respec-tively. “Cervical vertebrae” combined with “spinal fractures”yielded a subset of 688 citations, and “cervical vertebrae”combined with “dislocations” yielded a subset of 1,159 cita-
tions. An exploded search of “therapeutics” or “treatment,”limited to English language articles, yielded 1,566,596 cita-tions. This search term was combined with each of the twoearlier subsets, resulting in 198 and 287 citations, respectively.The abstracts were reviewed, and only those containing 10 ormore cases of subaxial cervical injury after nonpenetratingcervical trauma were included. An exception was made forankylosing spondylitis because so few reports included morethan 10 patients with this disorder. Sixty-three articles met theselection criteria and provide the basis for this review. Sixtyarticles providing Class III medical evidence are summarizedin Tables 20.1 to 20.6.
SCIENTIFIC FOUNDATION
The variety and heterogeneity of subaxial cervical spinalinjuries require accurate characterization of the mechanicsand type of injury to enable a comparison of the efficacy ofoperative and nonoperative treatment strategies. The absenceof a uniformly accepted classification scheme for cervicalvertebral injuries limits the ability to compare the effects oftreatment reported in many clinical studies. In 26 articlesdescribing series of patients with cervical injuries, sub-axial cervical injuries are not differentiated. The Allen andFerguson classification system (3) has been the most com-monly used scheme to differentiate and characterize subaxialvertebral injuries. Although few authors reported injuries by
S156 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
subtype, many of the reports described cervical injuries thatcould be grouped into the following broad categories as de-scribed by Allen et al.: distractive flexion, compressiveflexion/vertical compression, extension, and subluxation. Theeffectiveness of treatment of subaxial cervical spinal facetdislocation injuries, a subset of distractive flexion injuries,may be different from that of other subaxial cervical spinalfracture-dislocation injuries and is considered separately (53).Finally, four articles described unique characteristics of themanagement of subaxial cervical injuries in patients withankylosing spondylitis and are included in this review.
Several general principles can be formulated from the re-view of 26 articles that describe the treatment of subaxialcervical injuries without accurate differentiation into specificinjury types. Although closed reduction was successful in 64to 91% of patients with traumatic subaxial cervical malalign-ment (5, 28, 34a, 35, 43), patients with delayed treatment ofinjuries had a higher failure rate of closed reduction (22.5%)compared with those treated early (5). Halo vest applicationwas used successfully to immobilize patients with subaxialcervical injuries on arrival at the hospital to facilitate transportand workup; none had neurological worsening (32). Orthosesfailed to maintain reduction of subaxial cervical fracture-dislocation injuries in 7 to 56% of patients (10, 16, 18, 21–23,24a, 27, 35, 46, 53). Overall, 30% of these injuries (222 of 752injuries) had recurrent displacement or inadequate alignmentduring external immobilization (503 halo vest, 249 traction).Six of these patients were reported to heal with good ultimatealignment after readjustment of the halo device (three pa-tients) or continued postural reduction (three patients) (27,46). Nineteen percent (140 of 752 patients) were maintained inexternal immobilization despite displaced injuries and healedin an unreduced, nonanatomic position (10, 16, 18, 21, 23, 24a,27, 35, 46, 53). Eleven percent (82 of 752 patients) underwentsubsequent surgical treatment, typically for correcting cervi-cal malalignment (10, 16, 18, 22, 23, 24a, 27, 35, 46, 53). Severalrisk factors were identified in association with failure of non-operative management of subaxial cervical injuries. Patientswith more than 40% compression of a cervical vertebra, morethan 15 degrees of kyphotic angulation, or more than 20%subluxation of one vertebra on another were more likely toexperience failure of treatment with external immobilization(craniocervical traction alone or traction and then externalorthosis) (35).
In contrast, failure to maintain anatomic reduction of sub-axial cervical fracture-dislocation injuries after operativetreatment ranged from 1 to 18% (7, 10, 23, 28, 36, 46, 48, 51, 56).Anterior cervical fusion procedures (28, 48, 56, 59) were asso-ciated with less frequent failure to maintain reduction (10[5%] of 213 patients) when compared with posterior cervicalfusion procedures (38 [14%] of 280 patients) among all pa-tients with subaxial cervical injuries treated operatively (7, 36,51). Overall, 9% (61 of 704 patients) had recurrent angulationor subluxation despite surgical management (7, 10, 23, 28, 33,
35, 36, 46, 48, 51, 56, 59). A second operation in treatment ofprogressive deformity was rare in these patients. Successfularthrodesis occurred in nearly every patient reported (7, 15,33, 36, 43, 46, 48, 56). Surgical complications were relativelycommon in these series, ranging from 9 to 25% (7, 20, 27, 33,46, 48, 56, 59). In particular, graft extrusion after anteriorcervical surgery without plate fixation was observed in asmany as 10% of patients managed in this way (20, 27). Overall4 (4%) of 104 patients experienced graft displacement (20, 23,27, 35, 52, 59) after anterior fusion without plate fixation,compared with none of 291 patients treated with anteriorfusion with plate fixation (5, 28, 33, 48, 56). Complicationshave been reported using posterior plate or rod fixation aswell (7, 36, 51); radiculopathy occurred in 25% of patients inone report (36) describing these techniques.
Subaxial cervical facet dislocation injuries
Twenty-eight articles provided sufficient information toevaluate patients with subaxial cervical distractive flexioninjuries. Most reports were retrospective series of patientswith subaxial cervical spinal facet dislocation injuries (unilat-eral, bilateral, or both) (8, 11–13, 16, 19, 27, 30, 37, 39, 41, 45, 47,49, 50, 53, 54, 55, 57, 61). Overall, 26% (181 of 701 patients)with cervical spinal facet dislocation injuries had failure toachieve closed reduction with craniocervical traction (8, 11,12, 17, 19, 30, 39, 41, 42, 45, 47, 50, 53–55, 57, 61). The issue ofclosed reduction of cervical fracture dislocation injuries, in-cluding the potential of an associated ventral disc herniation,is described in the chapter “Initial Closed Reduction of Cer-vical Spine Fracture-Dislocation Injuries” (Chapter 6). Reduc-tion, when accomplished, could not be maintained in 28% ofpatients (112 of 393 patients) treated with subsequent externalimmobilization (8, 11–13, 16, 19, 27, 30, 37, 39, 41, 45, 47, 49, 50,53–55, 57, 61). Mortality associated with closed treatment offacet dislocation injuries was 7% (28 of 392 patients) in seriesreporting this complication (8, 11–13, 19, 30, 37, 39, 41, 45, 50,53, 57, 61). Prolonged bedrest and cervical traction alone for 12to 16 weeks’ duration was associated with the highest mor-tality of all treatment strategies reported for these injuries,27% in one series of 41 patients managed in this way (13).
Vertebral subluxation, facet injury (ligamentous or frac-ture), and a locked/perched facet on the initial x-rays orsubsequent computed tomography or magnetic resonance im-aging studies have been cited as factors associated with fail-ure of nonoperative treatment (8, 11, 30, 31, 57). Facet frac-tures associated with cervical spinal facet dislocation injuriesmay preclude successful closed reduction (30, 57). They havealso been associated with a high rate of arthrodesis withexternal immobilization alone (halo device) if closed reduc-tion can be accomplished, 97% in one report on this issue (30).Ligamentous disruption without facet fracture is associatedwith an increased likelihood of failure of external immobili-zation (halo device, Minerva cast) in the treatment of these
Subaxial Cervical Spinal Injuries
S157Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TABLE 20.1. Summary of Reports on Subaxial Cervical Spinal Injuries, Mixed Typesa
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Kalff et al., 1993 (33) Retrospective study. 97 cervical injuries (79 anterior and 18 anterior-posterior
fusion/instrumentation procedures). 16 DF, 14 VC, 64 fracture-dislocation
injuries.
III 9% operative complications related to fixation devices, but less than half
required reoperation. All patients achieved fusion.
Lemons and Wagner,
1993 (35)
Retrospective study. 64 cervical fractures: 14 VC, 12 CF, 12 UFD, 16 BFD,
10 extension. Treatment: 38 halo, 38 fusion (12 failed halo, 26 primary), 26
posterior fusion/instrumentation, 4 anterior fusions, 4 anterior-posterior.
III 12/38 halo-treated injuries were unstable and were fused. 4 healed
malaligned. None with extension injuries were unstable. 5/38 treated with
fusion were unstable. Risk for orthosis failure: �40% compression, �15
degree angulation, �20% subluxation.
Cybulski et al., 1992
(20)
Retrospective study. 21 cervical injuries failing posterior wiring treated with
anterior fusion.
III 2/21 had graft extrusions.
Della Torre and
Rinonapoli, 1992 (21)
Retrospective study. 28 cervical injuries.
3 CE, 7 CF, 4 DF treated with halo.
III 4/7 CF injuries were not reducible.
All were stable with immobilization.
Heary et al., 1992 (32) Retrospective study. 78 cervical injuries.
Halo for transport. 49 subaxial fractures, 45 subaxial subluxations.
III No patient worsened neurologically in halo before receiving definitive
treatment.
Levine et al., 1992 (36) Retrospective study. 24 facet fractures.
Posterior fusion with instrumentation.
III 11 complications including 4 who lost correction and 6 with radiculopathy.
All achieved fusion.
Roy-Camille et al.,
1992 (51)
Retrospective study. 221 cervical injuries.
89% posterior fusion, 11% anterior fusion.
III 15% developed kyphosis after surgery.
Nazarian and Louis,
1991 (43)
Retrospective study. 23 cervical injuries.
Posterior fusion with plates. 11 UFD, 4 BFD, 3 subluxations, 5 facet
fractures.
III 3/12 failed closed reduction, and 3 were unstable in an orthosis. All achieved
fusion.
Ripa et al., 1991 (48) Retrospective study. 92 cervical injuries. Anterior fusion instrumentation with
plate. 48 multiple fractures, 20 VC, 13 DF, 6 extension.
III No patient worsened neurologically. 12/15 complications were hardware-
related. 1 patient had pseudarthrosis.
Sears and Fazl, 1990
(53)
Retrospective study. 173 cervical injuries. 103 non-facet-dislocation injuries.
Halo treatment.
Operative procedure unreported.
III Nonoperative treatment failed in 31/103 patients (3 were irreducible, 10 were
neurologically worse, 16 subluxed in halo, 2 had late instability). Subluxation
and angulation predicted failed treatment, whereas fracture did not.
Benzel and Kesterson,
1989 (7)
Retrospective study. 50 cervical injuries. 25 fracture-subluxation. Posterior
fusion/instrumentation with wire.
III 1 patient with complete injury of 25 patients died.
Remainder healed.
Goffin et al., 1989 (28) Retrospective study. 41 cervical injuries. Anterior fusion with plate. III 2/41 subluxed, requiring surgery. 3/12 dislocations were irreducible. All 4
deaths were in quadriplegics.
Shoung and Lee, 1989
(56)
Retrospective study. 37 cervical injuries. Anterior fusion with plate. III All 37 healed. No graft extrusion. 1 death, 1 infection, 2 screw loosening.
Argenson et al., 1988
(5)
Retrospective study. 47 cervical injuries. 7 posterior fusion, 40 anterior
fusion.
III 17/22 were reducible, but 5 old dislocations were irreducible. 1 died of
vertebrobasilar thrombosis.
Bucci et al., 1988 (10) Retrospective study. 49 cervical injuries. Treatment: 20 halo alone (1
refused), 28 fusion with immobilization, procedure unreported.
III 12/20 with halo stable.
26/28 fused were stable (P � 0.01).
2 in each group lost reduction. 1 in each group were neurologically worse.
Donovan et al., 1987
(23)
Retrospective study. 61 cervical injuries. Treatment: 17 fusion with
immobilization (4 anterior, 13 posterior), 43 6-wk treatment in halo, 1
laminectomy without fusion.
III 18/43 had alignment in halo.
3/9 DF unstable: 2 surgery/1 asymptomatic.
All patients treated with fusion were stable, but 3 developed kyphosis.
Savini et al., 1987 (52) Retrospective study. 12 cervical injuries with late instability after closed
treatment.
III No grafts dislodged when anterior fusion was performed before posterior
reduction.
Ersmark and Kalen,
1986 (24a)
Retrospective study. 64 cervical injuries. Treatment with halo vest (36
subaxial).
III 29 dislocations and 5 VC injuries were stable after halo vest treatment.
Glaser et al., 1986 (27) Retrospective study. 245 cervical injuries, 125 complex fractures. Halo
treatment. Fusion posteriorly with wire or anteriorly without plate.
III 17/86 lost alignment in the halo. 2 interbody grafts displaced after anterior
surgery without plate.
De Smet et al., 1984
(22)
Retrospective study. 28 cervical injuries. Traction. III Early reduction failed in 4/28 patients. 2/24 had late instability.
Cahill et al., 1983 (15) Retrospective study of 25 DF or CF injuries. Treated with posterior fusion
with wiring.
III 18/18 with 3-mo follow-up were stable, and none had complications.
Chan et al., 1983 (16) Retrospective study. 188 cervical injuries. 150 subaxial with follow-up. Halo
treatment.
III 4/55 fracture-dislocation or complex fractures, 13/53 had UFD/BFD, and 2/41
VC were unstable with halo treatment.
Cooper et al., 1979 (18) Retrospective study. 33 cervical injuries treated with halo. III 2/11 “complex” fractures. 1/3 subluxations unstable with halo.
Verbiest, 1969 (59) Retrospective study. 47 cervical injuries. Anterior fusion without plate. III 5 patients died, 4 with complete spinal cord injuries. 6 had residual
malalignment, and 1 other had reoperation for lost alignment.
Paeslack et al., 1973
(46)
Retrospective study. 221 cervical injuries. 68 CF, 114 DF treated with
postural reduction, traction. 31 cervical injuries treated with anterior or
posterior fusion.
III 75 aligned, 67 wedged, 43 partially reduced, 36 failed reduction.
4/221 had late instability, 3 stable with further treatment and 1 with surgery.
2/31 were unstable after surgery.
Koskinen and
Nieminen, 1967 (34a)
Retrospective study.
159 cervical injuries.
Various treatments.
III No difference in pain, neck mobility, radiculopathy, or mortality when
operative and nonoperative treatments were compared.
a DF, distractive flexion; VC, vertical compression; CF, compressive flexion; UFD, unilateral facet dislocation; BFD, bilateral facet dislocation;CE, compressive extension.
S158 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TABLE 20.2. Summary of Reports on Subaxial Cervical Spinal Injuries: Distractive Flexiona
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Ordonez et al., 2000 (44) Retrospective study. 6 UFD, 4 BFD. 9 anterior reduction and fusion,
1 anterior-posterior-anterior.
III 10 with stable fusions, although 1 was incompletely reviewed.
Shapiro et al., 1999 (55) Retrospective study. 51 UFD. 24 SP posterior wiring, 22 SP wiring
and plates. 5 anterior-posterior-anterior.
III 1/24 with wire failed and 13/24 had late kyphosis. All patients with plate
fixation had stable fusions.
Fehlings et al., 1994 (25) Retrospective study. 44 cervical injuries. 19 facet dislocations.
Posterior fusion with plate.
III 5/19 patients had complications, including 2 late failed reductions.
Lieberman and Webb,
1994 (37)
Retrospective study. 41 cervical injuries. 9 facet dislocations.
Patients �65 yr old.
III 5 patients died, 1 treated with traction and 4 with halo. 3/4 survivors treated
with traction healed. All 4 survivors with halo treatment healed.
Lukhele, 1994 (40) Retrospective study. 43 facet dislocations. 12 with laminar fractures.
Posterior fusion with wire.
III 5/12 patients developed kyphosis.
Pasciak and Doniec, 1993
(47)
Retrospective study. 32 UFD. 23 nonoperative (treatment with halo
or plaster vest). 9 operative.
III All 9 treated with surgery healed. 8/23 in whom closed reduction failed were
fused. 7/15 with failure to maintain reduction achieved fusion.
Shapiro, 1993 (54) Retrospective study. 24 UFD. Posterior fusion with wire. III 23/24 patients with surgery healed. 1 with resubluxation healed with ACF.
9/17 in whom closed reduction failed had fractures of laminae.
Hadley et al., 1992 (30) Retrospective study. 31 UFD, 37 BFD. III 18/29 UFD and 20/37 BFD had successful closed reduction. 16 UFD and 15
BFD healed in halo. 7/31 had failed halo treatment. (5/7 without associated
facet fractures). When facet fractures were present, once reduced, 97%
success rate in halo. UFD/BFD results were similar.
Mahale and Silver, 1992
(41)
Retrospective study. 13 missed BFD with neurologically worse and
late treatment.
III All 13 reduced (10 completely). 12/13 healed with traction, 1 needed surgery.
Beyer et al., 1991 (8) Retrospective study. 36 UFD with and without fractures. 24 treated
with halo or orthosis. 10 posterior ORIF.
III 15/24 reduced in halo. 8/10 reduced with surgery. 11/24 had failed treatment
in halo. All 10 healed with surgery. Pain was more frequent despite healing if
unreduced.
Wolf et al., 1991 (61) Retrospective study. 52 BFD. 44 posterior fusion with wire, 3
anterior fusion and plate, 2 both.
III 12/52 had failed closed reduction. All 3 deaths had complete quadriplegia.
Cotler et al., 1990 (19) Retrospective study. 23 UFD (10 nonoperative), 12 BFD (4
nonoperative). 30 fused (21 primary).
III 1/2 had failed halo. 8/12 had failed traction. Complications were not reported.
Rockswold et al., 1990
(49)
Retrospective study. 140 cervical injuries. 8 facet dislocations (6
UFD, 2 BFD). Treated with halo or surgery.
III 1/6 UFD had failed halo. 0/4 had failed surgery. 1/2 BFD had failed halo. 2/9
had failed surgery.
Sears and Fazl, 1990 (53) Retrospective study. 173 cervical injuries. 70 dislocation injuries (38
UFD, 32 BFD).
III 19 healed with halo, 16 in good alignment. 16 had failed reduction and 23
subluxed in halo required surgery. Subluxation and angulation were not
associated with failure of halo. UFD/BFD results similar.
Benzel and Kesterson,
1989 (7)
Retrospective study. 50 cervical injuries. 19 UFD, 6 BFD. Posterior
fusion with wiring.
III All healed with fusion. 1 BFD neurologically worse required ACF, 1 UFD
incomplete patient died.
Bucholz and Cheung,
1989 (11)
Retrospective study. 124 cervical injuries. 20 distractive flexion
injuries. Treated with halo or surgery.
III 9/20 had failed halo. 1 neurologically worse postoperatively, unreported if
distractive flexion or subluxation patient.
Osti et al., 1989 (45) Retrospective study. 167 dislocations. 82 nonoperative (traction), 85
operative (anterior fusion without plate).
III 6/82 who had failed reduction were fused. 14/76 with late instability were
fused. 7 operatively treated within 24 h died (all ASIA A).
Lind et al., 1988 (39) Retrospective study. 83 injuries. Treated with halo. III 4/31 had failed halo. Loose pins common.
Rorabeck et al., 1987 (50) Retrospective study. 26 UFD. III 20/26 had failed closed reduction. 10 healed in halo. 8/10 remained reduced
with surgery. Pain common with failed reduction.
Glaser et al., 1986 (27) Retrospective study. 245 cervical injuries. 17 dislocations. III 3/12 UFD had failed halo. 1/5 BFD had failed halo.
Maiman et al., 1986 (42) Retrospective study. 26 BFD. III 10/18 reduced with closed reduction. 3 died, all complete injuries.
Chan et al., 1983 (16) Retrospective study. 188 cervical injuries. 150 subaxial with follow-
up. 40 halo alone, 20 halo and posterior fusion.
III 27/40 healed with halo. All 20 with primary surgery healed.
Dorr et al., 1982 (24) Retrospective study. 117 cervical injuries. 25 flexion-rotation
injuries.
III 2/3 with ACF had complications (1 graft displaced, 1 kyphosis).
Sonntag, 1981 (57) Retrospective study. 15 BFD. Halo or surgery. III 10/15 reduced with closed reduction. 4 halo healed (2 no follow-up), 8
posterior fusion with wire healed, 1 died (complete).
Stauffer and Kelly, 1977
(58)
Retrospective review. 10 dislocations. 5 fractures. 1 fracture
subluxation. Anterior fusion.
III 16/16 had recurrent angular deformity after ACF without plate. 3/16 fused
angulated.
Burke and Tiong, 1975
(13)
Retrospective review. 175 cervical injuries. Treated with traction,
traction-manipulation, or collar.
III 2/14 UFD and 0/13 BFD had failed nonoperative treatment.
Burke and Berryman, 1971
(12)
Retrospective review. 76 facet dislocations. 41 UFD, 35 BFD. 41
manipulation. 35 traction (3/35 failed manipulation). 3 fusion
primarily.
III 4/41 failed manipulation and 4 of remaining 37 had late instability. 0/32 had
failed traction.
Cheshire, 1969 (17) Retrospective review. 257 cervical injuries. Treated with traction or
collar (33 excluded).
III 3/40 UFD and 2/35 BFD had failed nonoperative treatment.
a UFD, unilateral facet dislocation; BFD, bilateral facet dislocation; SP, spinal process; ACF, anterior cervical fusion; ORIF, open reduction withinterior fixation; ASIA, American Spinal Injury Association scale.
Subaxial Cervical Spinal Injuries S159
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
injuries (30, 31). Laminar fractures have been associated withan increased risk of late kyphosis after surgical treatment ofcervical spinal facet dislocation injuries (40). Although pa-tients with unreduced facet dislocations treated with externalimmobilization often achieve spinal stability once treatmenthas been completed, arthrodesis in a position of malalignmenthas been associated with persistent cervical pain (8, 50, 55).No differences were observed in the success of achievingclosed reduction and/or maintaining cervical spinal align-
ment in patients with unilateral facet dislocations comparedwith patients with bilateral facet dislocation injuries.
In contrast, open reduction was achieved in all but 1 of 24patients treated with anterior fusion procedures (42, 44, 55)and in all but 7 of 167 patients treated with posterior fusionprocedures in series that reported this finding (7, 8, 11, 19, 30,42, 47, 50, 54, 57, 61). Delayed instability occurred in 6 (6%) of101 patients treated with anterior fusion procedures (12, 13,24, 40, 42, 55, 61), and 6 (3%) of 237 patients treated with
TABLE 20.3. Summary of Reports on Subaxial Cervical Spinal Injuries: Compressive Flexion or Vertical Compressiona
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Fehlings et al., 1994 (25) Retrospective study. 44 cervicalinjuries. Posterior fusion with plates.
III Complications in 6/17 including 1residual kyphosis and 1 new kyphosis(reoperation).
Lieberman and Webb, 1994 (37) Retrospective study. 41 cervicalinjuries. Patients �65 yr old.
III 1/4 died. 2 in collar and 1 fusedpatient were stable.
Kiwerski, 1993 (34) Retrospective “crossover” study. 273VC. 1st 70 traction. 2nd 203 anteriorfusion.
III Fewer died and more recoveredneurological function when treatedwith surgery.
Aebi et al., 1991 (1) Retrospective study. 22 cervicalinjuries. Anterior corpectomy withplate.
III All 22 achieved stable fusion. 2 screwcomplications occurred.
Anderson et al., (4) Prospective study. 30 cervicalinjuries. Posterior fusion with plate.
III All 9 achieved stable fusion, although1 had late kyphosis.
Bucholz and Cheung, 1989 (11) Retrospective study. 32 cervicalinjuries. 19 VC and CF injuries.
III 1/19 had failed halo treatment. Patienthad failed posterior fusion with wire.
Cabanela and Ebersold, 1988 (14) Retrospective study. 8 teardropfracture. Anterior fusion with plate.
III All 8 achieved stable fusion with nonedeveloping kyphosis.
Lind et al., 1988 (39) Retrospective study. 83 cervicalinjuries. Halo treatment.
III 2/19 were unstable. Drainage andloose pins were common.
Chan et al., 1983 (16) Retrospective study. 188 cervicalinjuries. 150 subaxial with follow-up.Halo treatment.
III All 22 burst fractures and 17 teardropfractures achieved stable fusion.
Dorr et al., 1982 (24) Retrospective study. 117 cervicalinjuries. 32 VC injuries.
III 1/11 had graft displacement afteranterior fusion without plate.
Burke and Tiong, 1975 (13) Retrospective study. 175 cervicalinjuries. Treated with traction,traction-manipulation, collar.
III 1/46 had failed nonoperativetreatment.
Frankel et al., 1973 (26) Retrospective study. 218 cervicalinjuries. 45 burst, 97 teardrop. Closedtreatment.
III 7/142 had failed postural reduction.103 had residual deformities.
Cheshire, 1969 (17) Retrospective review. 257 cervicalinjuries. Treated with traction orcollar (33 excluded).
III 3/63 had failed nonoperativetreatment.
Beatson, 1963 (6) Retrospective study. 59 cervicalinjuries. All immobilized.
III All 16 were stable withimmobilization.
a VC, vertical compression; CF, compressive flexion.
S160 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
posterior fusion procedures (7, 8, 11, 25, 30, 37, 40, 42, 49, 50,55, 57, 61). None of these six patients who failed to achievestability after anterior fusion was treated with plate fixation inaddition to fusion. Seven (8%) of 85 patients treated withanterior fusion procedures developed kyphosis; none hadbeen treated with anterior plate fixation (7, 8, 11, 25, 30, 37, 40,42, 49, 50, 55, 57, 61). Sixteen patients described by Staufferand Kelly (58) also developed kyphotic angulation after ante-rior cervical fusion without internal fixation. In contrast, 22(13%) of 165 patients developed kyphosis after posterior cer-vical fusion with wiring (25, 40, 54, 55), whereas only 1 (3%)of 40 patients developed kyphosis after posterior fusion withlateral mass plate fixation (25, 55). Alternatively, Halifax in-terlaminar clamps were successfully used in five patients withfacet dislocations treated with posterior arthrodesis (2).
Graft displacement was the most common complicationafter attempted anterior arthrodesis without internal fixation(7 [8%] of 85 patients) (6, 13, 24, 40, 42, 44, 55). Seven percent(8 of 113 patients) died after anterior fusion procedures (13,24, 40, 42, 44, 45, 55, 61); 3% (7 of 268 patients) died afterposterior fusion procedures (7, 11, 19, 25, 30, 42, 47, 49, 50, 55,57, 61). In these reports, all but 1 of the 15 patients who diedafter surgery that was performed in an attempt to correctdeformity and stabilize the spine had complete cervical spinalcord injuries (7, 42, 57, 61).
Subaxial cervical spinal injuries excluding facetdislocation injuries
Fourteen articles provided sufficient information to evalu-ate patients with subaxial cervical spinal compression fracture
injuries (1, 4, 6, 11, 13, 14, 16, 17, 24–26, 34, 37, 39). Althoughsome authors differentiated compressive flexion injuries fromvertical compression injuries, others considered these injuriestogether. Many nonoperative treatment strategies were de-scribed, including traction and external immobilization incollar, plaster jacket, or halo vest. Overall, 5% (17 of 349patients) treated with immobilization for compressive injuriesof the subaxial cervical spine had persistent instability afternonoperative treatment used for 8 to 12 weeks (6, 11, 13, 17,26, 37, 39). In contrast, nearly every patient with these injuriestreated with anterior (22 of 22 patients) or posterior (26 of 27patients) fusion procedures developed a stable union (1, 4, 25,27). Subluxation or kyphosis developed in 2 of 18 patientswho were treated with posterior fusion (11, 25). Operativecomplications were more common in patients treated withposterior fusion procedures (10 [37%] of 27 patients) thananterior fusion procedures (3 [9%] of 33 patients) (1, 4, 24, 25).Graft displacement was the most common complication de-scribed in patients treated with anterior cervical fusion with-out internal fixation (3 [9%] of 33 patients) (1, 24).
Only seven articles reported sufficient data to analyze pa-tients treated for extension injuries of the subaxial cervicalspine (4, 11, 13, 24, 37, 38, 43). Twenty-four percent (19 of 79patients) failed treatment with external immobilization (11,13, 37, 38, 43). In contrast, none of 19 patients failed treatmentwith anterior cervical fusion (37, 38). Two patients had irre-ducible vertebral displacements, and three patients developedkyphotic deformities among 11 patients with cervical spinalextension injuries treated with attempted posterior cervicalfusion (38).
TABLE 20.4. Summary of Reports on Subaxial Cervical Spinal Injuries: Extensiona
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Lifeso and Colucci, 2000 (38) Retrospective and prospective study.32 CE1 injuries (3 lost to follow-up).
III All 18 treated with brace failed (17were unreduced). 9/11 healed withPCF, but 3 had stable kyphosis.2/11 healed without reduction.
Lieberman and Webb, 1994 (37) Retrospective study. 41 cervicalinjuries. Patients �65 yr old.
III 1/3 healed with collar, and 1/3healed with surgery.
Anderson et al., 1991 (4) Prospective study. 30 cervical injuries.Posterior fusion with plates.
III All 30 healed, but 1 had screwloosening.
Rockswold et al., 1990 (49) Retrospective study. 140 cervicalinjuries. Treated with halo or surgery.
III All 3 treated with halo healed. All3 treated with surgery healed.
Bucholz and Cheung, 1989 (11) Retrospective study. 32 cervicalinjuries, 12 extension injuries.
III 1/12 failed halo treatment. 1patient stable after posterior fusionwith wire.
Dorr et al., 1982 (24) Retrospective study. 117 cervicalinjuries, 45 extension injuries.
III 40/45 were treated with brace. Ofall cervical injuries treated withbrace, 5/86 failed.
Burke and Tiong, 1975 (13) Retrospective review. 175 cervicalinjuries. Treated with traction, traction-manipulation, collar (30 excluded).
III All 45 healed without surgery.
a CE, compressive extension; PCF, posterior cervical fusion.
Subaxial Cervical Spinal Injuries S161
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Eight articles reported sufficient data to analyze patientstreated for vertebral subluxation injuries of the cervical spine(4, 6, 11, 13, 17, 18, 25, 49). Sixty-four percent of patients withthese injuries had successful treatment with external immo-bilization; patients with more than 50% subluxation were
twice as likely to maintain anatomic cervical realignment afterclosed reduction (72% versus 36%) (6). Thirty-six percent ofpatients (39 of 108 patients) had failure of external immobili-zation after closed reduction (11, 13, 17, 18, 49), comparedwith 7% of patients in whom these injuries were managed
TABLE 20.5. Summary of Reports on Subluxation
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Fehlings et al., 1994 (25) Retrospective study. 44 cervicalinjuries. Posterior fusion with plates.
III 2/5 lost reduction, including 1 whodied.
Anderson et al., 1991 (4) Prospective study. 30 cervical injuries.Posterior fusion with plates.
III 19/19 healed with fusion. 8/19 hadcomplications, including 2 withincreased kyphosis and 3 requiringadditional levels to be fused.
Rockswold et al., 1990 (49) Retrospective study. 140 cervicalinjuries. Treated with halo or surgery.
III 12/26 had failed halo treatment. 2/10had failed surgical treatment.
Bucholz and Cheung, 1989 (11) Retrospective study. 32 cervicalinjuries, 6 subluxation injuries.
III 2/6 had failed halo treatment. 1 worsepostoperatively, unreported if distractiveflexion or subluxation patient.
Cooper et al., 1979 (18) Retrospective study. 33 cervicalinjuries. Treated with halo.
III 1/3 had failed halo treatment.
Burke and Tiong, 1975 (13) Retrospective study. 175 cervicalinjuries. Treated with traction,traction-manipulation, collar (30excluded).
III 1/14 had failed nonoperative treatment.
Cheshire, 1969 (17) Retrospective study. 257 cervicalinjuries. Treated with traction or collar(33 excluded).
III 4/19 had failed nonoperative treatment.
Beatson, 1963 (6) Retrospective study. 59 cervicalinjuries (3 excluded).
III 8/22 with �50% subluxation reduced.2/14 remaining had surgery. 13/18 with�50% subluxation reduced. 5/5remaining had surgery.
TABLE 20.6. Summary of Reports on Ankylosing Spondylitisa
Series (Ref. No.) Description of StudyEvidence
ClassConclusions
Weinstein et al., 1982 (60) Retrospective study. 13 AS. 7traumatic cervical, 6 quadriplegic.2 central cords without fracture.
III 2 treated with traction died ofpneumonia. 2 treated with traction/bracehealed. 1 worse with halo treatedsurgically. 1 laminectomy/fusion worse, 1laminectomy/fusion had pseudarthrosis.
Bohlman, 1979 (9) Retrospective study. 300 cervicalinjuries. 8 AS.
III 5/8 patients died. 2 healed after bracetreatment and 1 after laminectomy.
Cheshire, 1969 (17) Retrospective study. 257 cervicalinjuries. 1 AS.
III 1 C5–C6 extension injury healed withsurgical fusion.
Grisolia et al., 1967 (29) Retrospective study. 6 AS. III 3/4 healed with brace with or withouttraction. 2 with laminectomy and PCFdied of PE.
a AS, ankylosing spondylititis; PCF, posterior cervical fusion; PE, pulmonary embolism.
S162 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
surgically (4, 49). A kyphotic deformity developed in 4% ofreported patients (3 of 74 patients) treated with posteriorcervical fusion and lateral mass plate or rod fixation proce-dures (4, 25).
Several characteristics of subluxation injuries of the sub-axial cervical spine were associated with failure of nonopera-tive treatment (6, 53). Patients with subluxation or kyphoticangulation frequently failed to achieve a good anatomic resultafter treatment with halo vest immobilization (46 [45%] of 103patients). Combined fractures to all parts of the cervical spinalcolumn and the presence of facet fractures were not associatedwith a higher likelihood of failure of treatment with externalimmobilization (53). Closed reduction was more successfulwith a subluxation of more than 50% of the vertebral bodydiameter (6).
Comparatively few studies examined the specific difficul-ties associated with the management of patients with anky-losing spondylitis who sustained cervical spinal injuries (9, 17,29, 60). In four articles reporting patients with this entity andinjury, 9 of 22 patients died. Four patients managed nonop-eratively died. Two of nine survivors treated with externalimmobilization experienced failure of treatment. One wors-ened neurologically when placed in a halo and subsequentlywas successfully treated with laminectomy and fusion. Theother patient had persistent cervical subaxial spinal instabilitybut refused further therapy. In contrast, five of nine ankylos-ing spondylitis patients with cervical fracture injuries treatedprimarily with surgery died. One patient was neurologicallyworse after surgery. Three patients healed successfully with-out instability.
SUMMARY
In conclusion, closed reduction is successful for most sub-axial cervical spinal fracture-dislocation injuries. Failure ofclosed reduction is more common with facet dislocation inju-ries. Similarly, treatment with external immobilization is fre-quently successful in the management of most subaxial cer-vical spinal injuries, although failure to maintain reduction ismore frequent with facet dislocation injuries as well. Virtuallyall forms of external immobilization have been used in thetreatment of subaxial cervical spinal injuries. More rigid or-thoses (halo, Minerva) seem to have better success rates thanless rigid orthoses (collars, traction only) for fracture-dislocation injuries after reduction has been accomplished.Treatment with traction and prolonged bedrest has been as-sociated with increased morbidity and mortality.
Both anterior and posterior cervical fusion procedures aresuccessful in achieving spinal stability for most patients withsubaxial cervical spinal injuries. Indications for surgical treat-ment offered in the literature include failure to achieve ana-tomic injury reduction (irreducible injury), persistent instabil-ity with failure to maintain reduction, ligamentous injurywith facet instability, spinal kyphotic deformity more than 15degrees, vertebral body fracture compression of 40% or more,vertebral subluxation of 20% or more, and irreducible spinalcord compression. Anterior fusion without plate fixation isassociated with an increased likelihood of graft displacement
and the development of late kyphosis, particularly in patientswith distractive flexion injuries. Similarly, late displacementwith kyphotic angulation is more common in patients treatedfor facet dislocation injuries with posterior fusion and wiringcompared with those treated with posterior fusion and lateralmass plate or rod or interlaminar clamp fixation. Althoughpatients with persistent or recurrent cervical spinal malalign-ment often achieve spinal stability with either external immo-bilization or surgical fusion with or without internal fixation,a higher proportion of these patients have residual cervicalpain than similarly treated patients for whom anatomic spinalalignment is achieved and maintained.
KEY ISSUES FOR FUTURE INVESTIGATION
To better compare the advantages and disadvantages ofnonoperative versus operative treatment strategies for sub-axial cervical injuries, future studies must differentiate be-tween the mechanisms of injury that have resulted in subaxialinjury. Although the Allen and Ferguson classification offers acommonly used framework for stratifying these patients,many investigators find that the number of subtypes in theirscheme precludes obtaining sufficient numbers of patientswith specific injuries. A broader classification of patients intocompressive flexion, distractive flexion, vertical compression,and extension injuries would allow comparison of most pa-tients who sustain subaxial cervical spinal injuries. A multi-center study would allow more rapid accumulation of pa-tients with these various categories of subaxial cervicalinjuries. A prospective examination of the efficacy of rigidexternal immobilization compared with surgical arthrodesiswith internal fixation (anterior and posterior approaches) mayfurther refine the most effective treatment for patients withsubaxial cervical spinal injuries.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue, Birmingham, AL 35294-3295.
REFERENCES
1. Aebi M, Zuber K, Marchesi D: Treatment of cervical spine inju-ries with anterior plating: Indications, techniques, and results.Spine 16[Suppl 3]:S38–S45, 1991.
2. Aldrich EF, Crow WN, Weber PB, Spagnolia TN: Use of MRimaging-compatible Halifax interlaminar clamps for posteriorcervical fusion. J Neurosurg 74:185–189, 1991.
3. Allen BL Jr, Ferguson RL, Lehmann TR, O’Brien RP: A mecha-nistic classification of closed, indirect fractures and dislocationsof the lower cervical spine. Spine 7:1–27, 1982.
4. Anderson PA, Henley MB, Grady MS, Montesano PX, Winn HR:Posterior cervical arthrodesis with AO reconstruction plates andbone graft. Spine 16[Suppl 3]:S72–S79, 1991.
5. Argenson C, Lovet J, Sanouiller JL, de Peretti F: Traumaticrotatory displacement of the lower cervical spine. Spine 13:767–773, 1988.
6. Beatson TR: Fractures and dislocation of the cervical spine.J Bone Joint Surg Br 45B:21–35, 1963.
Subaxial Cervical Spinal Injuries S163
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
7. Benzel EC, Kesterson L: Posterior cervical interspinous compres-sion wiring and fusion for mid to low cervical spinal injuries.J Neurosurg 70:893–899, 1989.
8. Beyer CA, Cabanela ME, Berquist TH: Unilateral facet disloca-tions and fracture-dislocations of the cervical spine. J Bone JointSurg Br 73B:977–981, 1991.
9. Bohlman HH: Acute fractures and dislocations of the cervicalspine: An analysis of three hundred hospitalized patients andreview of the literature. J Bone Joint Surg Am 61A:1119–1142,1979.
10. Bucci MN, Dauser RC, Maynard FA, Hoff JT: Management ofpost-traumatic cervical spine instability: Operative fusion versushalo vest immobilization—Analysis of 49 cases. J Trauma 28:1001–1006, 1988.
11. Bucholz RD, Cheung KC: Halo vest versus spinal fusion forcervical injury: Evidence from an outcome study. J Neurosurg70:884–892, 1989.
12. Burke DC, Berryman D: The place of closed manipulation in themanagement of flexion-rotation dislocations of the cervicalspine. J Bone Joint Surg Br 53B:165–182, 1971.
13. Burke DC, Tiong TS: Stability of the cervical spine after conser-vative treatment. Paraplegia 13:191–202, 1975.
14. Cabanela ME, Ebersold MJ: Anterior plate stabilization for burst-ing teardrop fractures of the cervical spine. Spine 13:888–891,1988.
15. Cahill DW, Bellegarrigue R, Ducker TB: Bilateral facet to spinousprocess fusion: A new technique for posterior spinal fusion aftertrauma. Neurosurgery 13:1–4, 1983.
16. Chan RC, Schweigel JF, Thompson GB: Halo-thoracic brace im-mobilization in 188 patients with acute cervical spine injuries.J Neurosurg 58:508–515, 1983.
17. Cheshire DJ: The stability of the cervical spine following theconservative treatment of fractures and fracture-dislocations.Paraplegia 7:193–203, 1969.
18. Cooper PR, Maravilla KR, Sklar FH, Moody SF, Clark WK: Haloimmobilization of cervical spine fractures: Indications and re-sults. J Neurosurg 50:603–610, 1979.
19. Cotler HB, Cotler JM, Alden ME, Sparks G, Biggs CA: Themedical and economic impact of closed cervical spine disloca-tions. Spine 15:448–452, 1990.
20. Cybulski GR, Douglas RA, Meyer PR Jr, Rovin RA: Complica-tions in three-column cervical spine injuries requiring anterior-posterior stabilization. Spine 17:253–256, 1992.
21. Della Torre P, Rinonapoli E: Halo-cast treatment of fractures anddislocations of the cervical spine. Int Orthop 16:227–231, 1992.
22. De Smet L, Vercauteren M, Verdonk R, Claessens H: Severeacute cervical spine injuries: Conservative treatment. ActaOrthop Belg 50:512–520, 1984.
23. Donovan WH, Kopaniky D, Stolzmann E, Carter RE: The neu-rological and skeletal outcome in patients with closed cervicalspinal cord injury. J Neurosurg 66:690–694, 1987.
24. Dorr LD, Harvey JP Jr, Nickel VL: Clinical review of the earlystability of spine injuries. Spine 7:545–550, 1982.
24a. Ersmark H, Kalen R: A consecutive series of 64 halo-vest-treatedcervical spine injuries. Arch Orthop Trauma Surg 105:243–246,1986.
25. Fehlings MG, Cooper PR, Errico TJ: Posterior plates in the man-agement of cervical instability: Long-term results in 44 patients.J Neurosurg 81:341–349, 1994.
26. Frankel H, Michaelis L, Paeslack V, Ungar G, Walsh JJ: Closedinjuries of the cervical spine and spinal cord: Results of conser-vative treatment of vertical compression injuries of the cervicalspine. Proc Veterans Adm Spinal Cord Inj Conf 19:28–32, 1973.
27. Glaser JA, Whitehill R, Stamp WG, Jane JA: Complications as-sociated with the halo-vest: A review of 245 cases. J Neurosurg65:762–769, 1986.
28. Goffin J, Plets C, Van den Bergh R: Anterior cervical fusion andosteosynthetic stabilization according to Caspar: A prospectivestudy of 41 patients with fractures and/or dislocations of thecervical spine. Neurosurgery 25:865–871, 1989.
29. Grisolia A, Bell RL, Peltier LF: Fractures and dislocations of thespine complicating ankylosing spondylitis: A report of six cases.J Bone Joint Surg Am 49A:339–344, 1967.
30. Hadley MN, Fitzpatrick BC, Sonntag VKH, Browner CM: Facetfracture-dislocation injuries of the cervical spine. Neurosurgery30:661–666, 1992.
31. Halliday AL, Henderson BR, Hart BL, Benzel EC: The manage-ment of unilateral lateral mass/facet fractures of the subaxialcervical spine: The use of magnetic resonance imaging to predictinstability. Spine 22:2614–2621, 1997.
32. Heary RF, Hunt CD, Krieger AJ, Antonio C, Livingston DH:Acute stabilization of the cervical spine by halo/vest applicationfacilitates evaluation and treatment of multiple trauma patients.J Trauma 33:445–451, 1992.
33. Kalff R, Kocks W, Grote W, Schmit-Neuerburg KP: Operativespondylodesis in injuries of the lower cervical spine. NeurosurgRev 16:211–220, 1993.
34. Kiwerski JE: Early anterior decompression and fusion for crushfractures of cervical vertebrae. Int Orthop 17:166–168, 1993.
34a. Koskinen EV, Nieminen R: Fractures and dislocations of thecervical spine: Treatment and results of 159 cases. Int Surg47:472–485, 1967.
35. Lemons VR, Wagner FC Jr: Stabilization of subaxial cervicalspinal injuries. Surg Neurol 39:511–518, 1993.
36. Levine AM, Mazel C, Roy-Camille R: Management of fractureseparations of the articular mass using posterior cervical plating.Spine 17[Suppl 10]:S447–S454, 1992.
37. Lieberman IH, Webb JK: Cervical spine injuries in the elderly.J Bone Joint Surg Br 76B:877–881, 1994.
38. Lifeso RM, Colucci MA: Anterior fusion for rotationally unstablecervical spine fractures. Spine 25:2028–2034, 2000.
39. Lind B, Sihlbom H, Nordwall A: Halo-vest treatment of unstabletraumatic cervical spine injuries. Spine 13:425–432, 1988.
40. Lukhele M: Fractures of the vertebral lamina associated withunifacet and bifacet cervical spine dislocations. S Afr J Surg32:112–114, 1994.
41. Mahale YJ, Silver JR: Progressive paralysis after bilateral facetdislocation of the cervical spine. J Bone Joint Surg Br 74B:219–223, 1992.
42. Maiman DJ, Barolat G, Larson SJ: Management of bilaterallocked facets of the cervical spine. Neurosurgery 18:542–547,1986.
43. Nazarian SM, Louis RP: Posterior internal fixation with screwplates in traumatic lesions of the cervical spine. Spine 16[Suppl3]:S64–S71, 1991.
44. Ordonez BJ, Benzel EC, Naderi S, Weller SJ: Cervical facet dis-location: Techniques for ventral reduction and stabilization.J Neurosurg 92[Suppl 1]:18–23, 2000.
45. Osti OL, Fraser RD, Griffiths ER: Reduction and stabilisation ofcervical dislocations: An analysis of 167 cases. J Bone Joint SurgBr 71B:275–282, 1989.
46. Paeslack V, Frankel H, Michaelis L: Closed injuries of the cervi-cal spine and spinal cord: Results of conservative treatment offlexion fractures and flexion rotation fracture dislocation of thecervical spine with tetraplegia. Proc Veterans Adm Spinal CordInj Conf 19:39–42, 1973.
S164 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
47. Pasciak M, Doniec J: Results of conservative treatment of unilat-eral cervical spine dislocations. Arch Orthop Trauma Surg 112:226–227, 1993.
48. Ripa DR, Kowall MG, Meyer PR Jr, Rusin JJ: Series of ninety-twotraumatic cervical spine injuries stabilized with anterior ASIFplate fusion technique. Spine 16[Suppl 3]:S46–S55, 1991.
49. Rockswold GL, Bergman TA, Ford SE: Halo immobilization and sur-gical fusion: Relative indications and effectiveness in the treatment of140 cervical spine injuries. J Trauma 30:893–898, 1990.
50. Rorabeck CH, Rock MG, Hawkins RJ, Bourne RB: Unilateralfacet dislocation of the cervical spine: An analysis of the resultsof treatment in 26 patients. Spine 12:23–27, 1987.
51. Roy-Camille R, Saillant G, Laville C, Benazet JP: Treatment oflower cervical spinal injuries: C3 to C7. Spine 17[Suppl 10]:S442–S446, 1992.
52. Savini R, Parisini P, Cervellati S: The surgical treatment of lateinstability of flexion-rotation injuries in the lower cervical spine.Spine 12:178–182, 1987.
53. Sears W, Fazl M: Prediction of stability of cervical spine fracturemanaged in the halo vest and indications for surgical interven-tion. J Neurosurg 72:426–432, 1990.
54. Shapiro SA: Management of unilateral locked facet of the cervi-cal spine. Neurosurgery 33:832–837, 1993.
55. Shapiro SA, Snyder W, Kaufman K, Abel T: Outcome of 51 cases ofunilateral locked cervical facets: Interspinous braided cable for lateralmass plate fusion compared with interspinous wire and facet wiringwith iliac crest. J Neurosurg 91[Suppl 1]:19–24, 1999.
56. Shoung HM, Lee LS: Anterior metal plate fixation in the treat-ment of unstable lower cervical spine injuries. Acta Neurochir(Wien) 98:55–59, 1989.
57. Sonntag VKH: Management of bilateral locked facets of thecervical spine. Neurosurgery 8:150–152, 1981.
58. Stauffer ES, Kelly EG: Fracture-dislocations of the cervicalspine: Instability and recurrent deformity following treatmentby anterior interbody fusion. J Bone Joint Surg Am 59A:45–48, 1977.
59. Verbiest H: Anterolateral operations for fractures and dislocations inthe middle and lower parts of the cervical spine: Report of a series offorty-seven cases. J Bone Joint Surg Am 51A:1489–1530, 1969.
60. Weinstein PR, Karpman RR, Gall EP, Pitt M: Spinal cord injury,spinal fracture, and spinal stenosis in ankylosing spondylitis.J Neurosurg 57:609–616, 1982.
61. Wolf A, Levi L, Mirvis S, Ragheb J, Huhn S, Rigamonti D,Robinson WL: Operative management of bilateral facet disloca-tion. J Neurosurg 75:883–890, 1991.
Drawing by Leonardo da Vinci. Courtesy, Dr. Edwin Todd, Pasadena, California.
Subaxial Cervical Spinal Injuries S165
CHAPTER 21
Management of Acute Central Cervical Spinal Cord Injuries
RECOMMENDATIONSSTANDARDS: There is insufficient evidence to support treatment standards.GUIDELINES: There is insufficient evidence to support treatment guidelines.OPTIONS:• Intensive care unit (or other monitored setting) management of patients with acute central cervical spinal
cord injuries, particularly patients with severe neurological deficits, is recommended.• Medical management, including cardiac, hemodynamic, and respiratory monitoring, and maintenance of
mean arterial blood pressure at 85 to 90 mm Hg for the first week after injury to improve spinal cordperfusion is recommended.
• Early reduction of fracture-dislocation injuries is recommended.• Surgical decompression of the compressed spinal cord, particularly if the compression is focal and anterior,
is recommended.
RATIONALE
Central spinal cord injuries are among the most com-mon, well-recognized spinal cord injury patterns iden-tified in neurologically injured patients after acute
trauma. Originally described by Schneider et al. (19) in 1954,this pattern of neurologically incomplete spinal cord injury ischaracterized by disproportionately more motor impairmentof the upper than of the lower extremities, bladder dysfunc-tion and varying degrees of sensory loss below the level of thelesion (19). It has been associated with hyperextension injuriesof the cervical spine, even without apparent damage to thebony spine, but has also been described in association withvertebral body fractures and fracture-dislocation injuries. Thenatural history of acute central cervical spinal cord injuriesindicates gradual recovery of neurological function for mostpatients, albeit usually incomplete and related to the severityof the original injury and the age of the patient (4, 13, 15, 17,19–21). The role of surgery and its timing for patients withacute central spinal cord injuries without fracture compres-sion or dislocation injuries are the subjects of considerabledebate (3, 5–8, 19, 20). The optimal management of patientswho have sustained acute central cervical spinal cord injuriesis the subject of this review.
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject heading “spinal cord injury”combined with “central cord syndrome” and “incompletecervical spinal cord injury” yielded approximately 1450 cita-tions. Non-English language citations were excluded. Titlesand abstracts of the remaining publications were reviewed,
and relevant articles were selected to develop the guidelines.We focused on the specific issues of the natural history, med-ical management, and surgical treatment of human acutecentral cervical spinal cord injuries. These efforts resulted in13 articles (all Class III studies) specifically describing man-agement and outcomes of patients with central cervical spinalcord injuries. The Bibliography includes several articles onmagnetic resonance imaging (MRI) of central cervical spinalcord injuries, many articles (all Class III studies) describingseries of patients with acute spinal cord injuries, most ofwhom had incomplete cervical spinal cord injuries, and sev-eral general review articles that address issues of acute spinalcord injuries, including pathophysiology and treatment. The13 case series describing the management of patients withacute central cervical spinal cord injuries are summarized inTable 21.1.
SCIENTIFIC FOUNDATION
In 1951, Schneider (18) described two patients with acuteneurologically incomplete cervical spinal cord injuries forwhom he suggested that early operation was indicated. Bothpatients presented after trauma with sudden loss of motorfunction in the distal upper extremities, the torso, and thelower extremities, but with preservation of touch and vibra-tion sense. Both patients had anterior spinal cord compressionfrom acute traumatic cervical disc herniations (one had anassociated vertebral endplate fracture). The diagnosis andanatomic localization were based on the clinical examination.Both patients made incomplete but significant neurologicalrecoveries after delayed surgical decompression via laminec-tomy, dentate ligament sectioning, and transdural discec-tomy. Three years later, Schneider et al. (19) described eight
S166 Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
patients they managed along with six other patients culledfrom the available literature. All but two of these patientspresented with disproportionately more motor impairment in
the upper extremities than in the lower extremities, bladderdysfunction with retention, and varying degrees of sensoryloss below the level of the lesion. Two of the six patients
TABLE 21.1. Summary of Reports on Acute Cervical Spinal Cord Injuriesa
Series (Ref. No.) Description of Study Evidence Class Conclusions
Dai and Jia, 2000 (8) Retrospective review of 24 patients with acute discherniation as cause of ACCSCI treated with ACDF.
III Disc herniation common cause. Surgery successful in allpatients, more rapid improvement. Poor outcome with fracturedislocation injuries.
Newey et al., 2000 (15) Retrospective review of 32 patients with ACCSCImanaged conservatively.
III Improvement seen in most patients over time. Older patientshad worse outcome.
Chen et al., 1998 (6) Retrospective review of 37 patients with ACSI withpreexisting spondylosis. Many with central cord injurypattern. MRI assessment of compression, cord injury.16 managed with surgical decompression, 21medically.
III MRI modality of choice to image cord compression/injury.Surgical decompression associated with more rapidimprovement, shorter hospital and rehabilitation stay. Nodifference in outcome at 2-yr follow-up.
Chen et al., 1997 (7) Retrospective review of 114 patients with acute orchronic CCSCI. 28 patients managed with surgery (3chronic patients), 86 medically. No randomization.
III Surgery associated with more rapid and complete recovery,particularly in upper extremities, compared with similar patientsmanaged medically. Patients with long-segment stenosis hadpoor prognosis.
Bridle et al., 1990 (4) Random late assessment of 18 patients with ACCSCI. III Most patients improved over time, although most with long-term deficits, pain, and dysfunction.
Roth et al., 1990 (17) Retrospective review of 81 rehabilitation patients afterACCSCI.
III 2 age groups of patients, marked heterogeneity. In general,most patients improved over time. Outcome related to age andseverity of initial injury.
Merriam et al., 1986 (13) Retrospective review of 77 patients with ACCSCI. Nopatient with surgical decompression, 30 underwentlate stabilization and fusion.
III Marked variation among patients and injury patterns. Mostimproved. Outcome related to age and severity of initial injury.
Bose et al., 1984 (3) Retrospective review of 28 patients with ACCSCI, 14managed with aggressive medical therapy, 14 withmedical therapy and surgical treatment. Norandomization. Follow-up at time of discharge.
III No patient worse with treatment, medical or surgical. Surgeryprovided more rapid, more complete recovery at time ofdischarge.
Brodkey et al., 1980 (5) Retrospective review of 7 patients with ACCSCIoperated on late after injury. All had stable, profounddeficits and myelographic evidence of cordcompression.
III All had accelerated neurological improvement after surgery. 3patients normal at late follow-up. Surgery of benefit in selectedpatients with persistent deficits and evidence of cordcompression.
Shrosbree, 1977 (21) Retrospective review with late follow-up of 99patients with ACCSCI managed conservatively.
III 2 groups identified.Younger patients with flexion rotation injuries.Older patients with hyperextension injuries.Outcome related to severity of initial injury.
Bosch et al., 1971 (2) Retrospective review and long-term follow-up of 42patients with ACCSCI managed conservatively.
III Most patients improved over time. 75% regained ambulatoryskills, 56% regained functional hands. 10/42 patients had latedeterioration after initial gains (“chronic central cordsyndrome”).
Schneider et al., 1958 (20) Retrospective review of 12 additional patients withACCSCI. 11 managed expectantly, 1 managed withsurgical decompression 13 h after injury.
III 2 age groups of patients.Young patients with fracture dislocation injuries.Older patients with hyperextension injuries often without bonyvertebral damage.Most patients improved.Expectant management is ideal treatment.
Schneider et al., 1954 (19) Retrospective review (and first description) of 8patients with ACCSCI they managed (6 expectantly, 2surgically) and discussion of 6 cases from literature.
III Most patients with ACCSCI improved with time and expectantmanagement.Injury and its recovery follows specific pattern. Surgerycontraindicated for this injury.
a ACDF, anterior cervical discectomy with fusion; ACCSCI, acute central cervical spinal cord injury; ACSI, acute cervical spinal cord injury;CCSCI, central cervical spinal cord injury; MRI, magnetic resonance imaging.
Acute Central Cervical Spinal Cord Injuries
S167Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
identified in the literature review had complete motor injuriesin both the upper and lower extremities with some preserva-tion of sensory function below the level of injury. Theseincomplete neurological deficits were related to acute trau-matic central cervical spinal cord injuries, usually—but notexclusively—as a result of hyperextension of the head andneck relative to the torso. In several patients, there was nodamage to the bony spine. In these instances, it was presumedthat hypertrophic changes (spurs, ridges, thickened liga-ments) within the spinal canal caused anterior and posteriorcord compression in the position of hyperextension, resultingin injury to the central substance of the cervical spinal cord.Other patients had cervical compression fractures or fracture-dislocation injuries of the cervical spine that contributed to thecentral spinal cord injury. The authors operated on the firsttwo of the eight patients they treated with this disorder. Bothhad central cervical spinal cord injuries without bony damageor displacement. Both were treated in delayed fashion vialaminectomy with sectioning of the dentate ligaments andthen transdural exploration anterior to the cervical spinalcord. In both cases, anterior bony osteophytes were identifiedbut were not removed. Patient 1 was quadriplegic postoper-atively. Patient 2 was neither better nor worse after surgery.Six additional consecutive patients were managed withoutsurgical decompression (Patient 7 underwent delayed dorsalcervical fusion in treatment of fracture instability). Five of sixpatients were managed expectantly (Patient 4 had progressionof neurological deficits despite immobilization and ultimatelydied 3 wk after injury), and three of six patients from theliterature improved without surgery (75%). Most patients ex-perienced permanent loss of hand function and strength. Oneof six patients from their series and three of six from theliterature died without neurological recovery (25%). Con-versely, one (50%) of two patients treated surgically wasimmediately neurologically worse; the second (50%) made aprogressive albeit incomplete recovery over time, much likethat observed in comparable patients managed withoutsurgery. It was on the basis of this early experience thatSchneider et al. determined that the prognosis after acutecentral cervical spinal cord injury was reasonably good. Sur-gery for these patients, they concluded, was contraindicatedand in fact known to harm rather than improve them (19).
In 1958, Schneider et al. (20) added observations on 12additional patients they managed with acute central cervicalspinal cord injuries. One patient died of pneumonia withoutneurological recovery, one patient had a full neurologicalrecovery, and the remaining 10 improved compared withtheir initial postinjury neurological status but were pro-foundly impaired at last follow-up. The authors noted twodistinct age groups with acute central cervical spinal cordinjuries. They described an older group of patients (mean age,59 yr) without bony vertebral damage but with hypertrophicchanges compromising the cervical spinal canal, and ayounger group (mean age, 31 yr) with fracture or fracture-dislocation injuries of the cervical spine. They reported thatcentral cord edema, venous congestion, and ischemia werecomponents of the pathophysiology of this unique injurytype. They advocated expectant management, including
closed reduction with skeletal traction for all patients withthis syndrome (as for four patients in their study), despiteimportant, near-complete neurological recovery in a 17-year-old patient after operative reduction and decompression of aunilateral facet dislocation injury within 13 hours of injury.
Schneider et al.’s (20) collective reported experience in themanagement of patients with acute central cervical cord inju-ries numbered 20 patients at the time of the report. Of the 20patients, 17 were managed medically: 2 patients died withoutimprovement, 14 patients improved but had profound resid-ual deficits, and 1 patient regained normal function. Threepatients were treated with surgical decompression: one early(hours) and two late (weeks). The patient with early decom-pression improved dramatically. One late decompression pa-tient neither improved nor worsened immediately after sur-gery but showed progressive long-term improvement; theother was quadriplegic after surgery. From this experience,they concluded that accurate diagnosis is stressed, with em-phasis placed on the fact that operation is contraindicated,that the prognosis may be good and that should recoveryoccur it will follow a definite pattern (20). These suggestionshave guided the care of acute central cervical spinal cordinjury patients ever since publication of their report.
In 1971, Bosch et al. (2) described observations made duringtheir management of 42 patients with subacute central cervi-cal spinal cord injuries treated at a rehabilitation hospital,with a follow-up period of 4 months to 26 years. At admis-sion, 19% could walk independently, 14% were partial walk-ers, and 67% could not walk. Twenty-six percent had func-tional hands at admission. At discharge, 57% had functionalwalking skills, 20% were partial walkers, and 42% had func-tional hands. Bladder control improved from 17% at admis-sion to 53% at discharge. A similar improvement in bowelcontrol was documented. Importantly, these authors notedlate deterioration in 24% of patients who showed initial im-provements in neurological function after central cervical spi-nal cord injury. Ten (24%) of 42 patients experienced the lateneurological sequelae of “chronic central cord syndrome” andlost walking, hand, and bladder control skills, as observed inlong-term follow-up. The authors concluded that at leastsome return of neurological function in the immediate postin-jury period could be expected in about 75% of cases, with 56%of patients regaining functional hands. In the long term, only59% of the patients with central cervical spinal cord injuriesthey followed retained functional skills with conventionalmedical management.
In the same year, Turnbull (26) reported his studies on themicrovasculature of the human spinal cord and postulatedmechanisms of vascular insufficiency associated with varioustypes of spinal cord displacement. His work speaks to theanatomic basis of the pathophysiology of acute cervical cen-tral spinal cord injuries, particularly those that occur in olderpatients with underlying cervical spondylosis who sustainacute central spinal cord injury without bony vertebral injury.He found that as the cord becomes compressed, whetherowing to a mass lesion or progressive cervical spondylosis, itbecomes flattened and widened. The vasculature of the cord isaffected by cord distortion. Pial vessels become more tortu-
S168 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
ous. Arteries of the lateral columns are elongated, narrowed,and flattened. Branches from the central arteries that reach thegray matter run laterally and are similarly stretched length-wise and are compressed from side to side. Turnbull reportedthat vessels chronically deformed by cervical spondylosis can-not respond to additional anteroposterior flattening of thecord as would normal arteries in a younger patient. A littleadditional compression would pinch off side branches at theirorigins (26). He concluded that mechanical distortion of thecord and its blood supply plays a major role in the patho-physiology of spinal disease and spinal cord injury (26–28).
In 1977, Shrosbree (21) reported 99 patients with acutecentral cervical spinal cord injuries managed at a South Afri-can spinal cord injury center. Most of the patients were ad-mitted within 72 hours of injury. All patients were treatedconservatively. Fracture/injury reduction was accomplishedvia closed means in 92% of patients with dislocation, either bytraction or by reduction under anesthesia, within 72 hours ofadmission. Two age groups of patients were identified.Younger patients (21–50 yr) had flexion-rotation injuries anda higher incidence of dislocation injuries. Older patients (�50yr) were more likely to have hyperextension injuries super-imposed on preexisting cervical spondylosis. Outcome wasrelated to the severity of the initial neurological deficit. Only5 (22%) of 23 patients with severe motor deficits becameindependent walkers. All of these patients had residual defi-cits in the hands. The author summarized by noting that earlyreduction may well be a factor in promoting more favorableneurological recovery (21) among patients with facet disloca-tion injuries, but he provided no data to support his claim.
In 1977, Maroon (12) reported that “burning hands” (severedysesthesias in the hands and fingers after trauma despitenormal motor function) may indicate acute central spinal cordinjury. He described two football players with dysestheticsymptoms in the hands referable to modest injury to thecentral cervical spinal cord and warned physicians, trainers,and coaches of the importance of this syndrome.
In 1980, Brodkey et al. (5) revisited the management of theacute central cervical spinal cord injury syndrome. They pro-vided operative treatment to seven patients with traumaticcentral cervical spinal cord injuries within 18 to 45 days afteracute injury who had profound residual neurological deficitsafter attentive medical treatment. Myelography revealed sig-nificant defects in all of these patients. Four patients under-went anterior cervical discectomy with fusion (ACDF), onewas treated with multilevel laminectomy, one had multilevelACDF, and one received multilevel laminectomy and thendelayed (4 yr) multilevel ACDF. All patients had an acceler-ated neurological recovery after the surgical procedure, eventhe patient who deteriorated years after laminectomy andrequired late multilevel ACDF. Three patients returned tonormal after severe injuries that persisted until surgical de-compression. The authors concluded that cord compressiondoes play an important role in the pathophysiology of centralcord syndrome and that, when present in the setting of astable poor neurological condition after injury, decompres-sion of the spinal cord may be of benefit.
Bose et al. (3) retrospectively analyzed their management of28 patients with acute central cervical spinal cord injuries. Intheir 1984 report, they noted significantly improved motorscores at the time of discharge in patients managed withcombined medical therapy and surgery, compared with thosemanaged only medically. All were treated aggressively in theintensive care unit setting. Surgical patients had myelo-graphic evidence of cord compression or evidence of cervicalspinal instability. Although selection bias (surgical patientshad cord compression and/or instability or subluxation) andseveral other study flaws precluded direct comparison be-tween the groups, the authors noted that no surgical patientworsened as a result of surgery and all improved neurologi-cally, several substantially. They argued that decompressionof the compressed spinal cord in patients with acute centralcervical spinal cord injury syndrome may be of benefit inselected patients.
Merriam et al. (13), Roth et al. (17), Bridle et al. (4), andNewey et al. (15) described the late outcomes of individualseries of selected groups of patients after central cervicalspinal cord injury. All four groups of investigators notedmarked heterogeneity among patients in the central cervicalspinal cord injury population. All patients were managedmedically. Most patients improved somewhat over time, withmore recovery in the lower extremities than in the upperextremities. All of these authors concluded that outcome wasin general good for patients younger than 70 years. The finalneurological result was influenced by patient age, particularlyage older than 70 years, and the degree of initial neurologicalimpairment. Hand function impairment was the most com-mon long-term disability, even among patients with a “good”outcome. Only Merriam et al. (13) made reference to surgicaltreatment, involving 30 of 77 patients in their series, presum-ably to provide spinal stabilization and fusion. No associationbetween surgical management and outcome was discerned.
Chen et al. (7) reported their experience with 114 patientswith acute and chronic traumatic central cervical cord syn-drome. Twenty-eight patients were managed with surgicaltreatment, and 86 were managed medically. The authors didnot randomize patients to one treatment group or another.Selection criteria for surgical decompression included failureto improve with medical therapy or deterioration in neuro-logical function despite medical treatment with radiographic(MRI or computed tomographic/myelographic) evidence offocal cord compression, or gross instability of the spine. Theyoperated on three patients late (8, 12, and 24 mo after injury)for “chronic” central cord syndrome. Their 1997 retrospectivereview found that younger patients had better long-term re-sults than did older patients (in both management groups)and that surgical decompression was associated with morerapid and complete motor improvement compared with pa-tients managed medically, even if the operation was per-formed a long time after injury. Both management groups hadsimilar outcomes over time with respect to lower extremityand bladder function. Patients selected for surgery had morerapid and more complete recovery of function, particularly inthe upper extremities. The authors noted that patients withstenosis at multiple levels who were managed conservatively
Acute Central Cervical Spinal Cord Injuries S169
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
had a poorer prognosis and a relatively higher chance todevelop late myelopathy. The authors did not describe theoutcome of similar patients with multilevel stenosis managedwith operative decompression.
In 1998, Chen et al. (6) described the management of 37patients with preexisting cervical spondylosis who sustainedacute incomplete neurological cervical spinal cord injuriesafter trauma. Many of these patients had acute central spinalcord injuries. No patient sustained a bony vertebral columninjury. In their retrospective review, patients were treatedwith surgical decompression if they did not improve morethan one motor grade within 9 days of injury (range, 3–14 d).Patients were studied with MRI to document cord compres-sion and/or signal change within the spinal cord. In total, 16patients underwent surgical decompression, 9 anteriorly and7 posteriorly. Twenty-one patients were managed medically.Thirteen (81%) of 16 surgical patients improved “remarkably”immediately after surgery. Thirteen (62%) of 21 patients man-aged medically improved to the same degree over time. Aswith surgical patients, patients with cervical stenosis in morethan three vertebral levels fared less well than did patientswith focal compression or with stenosis in three vertebrallevels or fewer. There were no reported differences in out-come between patients in the two groups at 2-year follow-up.The authors concluded that surgical decompression might beassociated with more rapid neurological improvement, earlymobilization, and shorter periods of hospitalization and reha-bilitation. They consider MRI to be the imaging modality ofchoice to assess the spinal cord in patients with acute centralcervical spinal cord injuries, a conclusion consistent withthose of other investigators of the role of MRI in the assess-ment of patients with spinal cord injuries (9, 14, 16, 25).
In 2000, Dai and Jia (8) described their experience with 24patients with acute traumatic disc herniation as the cause ofacute central cervical spinal cord injuries. Acute disc herniationwas confirmed with preoperative MRI. The authors provided aretrospective assessment of patients operated on anteriorly(ACDF without internal fixation) for cord decompression andspinal stabilization. The timing of surgery relative to injury wasnot described. They noted an inverse correlation between rate ofrecovery and age and found that patients with fracture-dislocation injuries with acute disc herniation were more im-paired preoperatively and fared less well than patients withoutfracture-dislocation injuries at late follow-up. They reported thatsurgical decompression, stabilization, and fusion were success-ful in all patients and described marked improvement in neuro-logical function in most patients treated.
Contemporary reviews confirm early reports that most pa-tients with incomplete cervical spinal cord injuries meetingthe clinical neurological criteria of acute central spinal cordinjury will show neurological improvement over time (2, 13,15, 17, 19–21). Some patients with these injuries will die, andmany will remain profoundly impaired at late follow-up.These patients in general are older, have spinal cord injurieswithout bony vertebral injury, and have medical comorbidity,or they are younger but have fracture-dislocation injuries as acause of their neurological deficits. A large portion of patientswill regain walking skills over time but will not have useful
hands. A smaller portion of patients will demonstrate signif-icant neurological recovery and regain hand function. Thesepatients are typically younger, do not have fracture-dislocation injuries, and have less severe neurological deficitsat the outset. Up to 24% of patients managed nonoperativelywill improve early but decline again years later (“chroniccentral cord injury syndrome”) (2).
Surgery for decompression of the spinal cord in patientswith acute central cervical spinal cord injuries has been de-nounced on the basis of Schneider’s early poor experiencewith a single patient who underwent surgery (19, 20). Thatpatient, quadriplegic after dorsal cervical exploration anddecompression, experienced significant manipulation of theinjured cord during the process of dentate ligament sectioningand transdural anterior cord exploration (19), a procedureunlikely to be performed in similar fashion today. The samegroup had a rewarding experience with early (13 h afterinjury) surgical decompression and facet fracture reduction ina 17-year-old boy with profound early central cord neurolog-ical deficits (20). Many other authors, including those report-ing three contemporary series of patients with this disorder,have described good to excellent outcomes without neurolog-ical complications for surgical decompression of patients withspinal cord compression, particularly focal anterior cord com-pression (3, 5–8). However, no study to date has provided arandomized direct comparison of surgical patients with sim-ilar patients managed without surgery. Nor has any studyadequately assessed the potential role of dorsal spinal decom-pression for multilevel cervical cord compression in patientswith this disorder, particularly those with acute central cer-vical cord syndrome without bony vertebral damage. Surgerymay have a role in the management of patients with acutecentral cervical spinal cord injury, but, as yet, that role has notbeen accurately defined by scientific study.
Schneider et al.’s (20) conclusion that central cord edema,venous congestion, and ischemia were important componentsof the pathophysiology of these injuries, combined with thehypothesis of Turnbull (26, 27) and Turnbull et al. (28) thatvascular compression and distortion attributable to antero-posterior flattening of the cord plays a major role in thepathophysiology of cord injury, suggest several potential op-portunities for treatment. The compression of the cord anddistortion and compression of its blood supply might berelieved by surgical decompression. Ischemia of the cord,caused by either the primary injury or secondary events,might be improved with augmentation of spinal cord perfu-sion. Although Turnbull (26, 27) and Turnbull et al. (28) didnot offer specific strategies, they did offer an anatomic andpathophysiological rationale for the potential of maintenanceof spinal cord perfusion pressures and cervical cord decom-pression for patients who sustain an acute central cervicalcord injury, particularly those with preexisting cervical spon-dylosis. Maintenance or increases in systemic blood pressuremay improve perfusion to the injured, distorted spinal cord(1, 3, 10, 11, 22, 23, 29). Several contemporary series of patientswith spinal cord injuries treated with aggressive medicalmanagement including maintenance of mean arterial bloodpressure at high normal ranges (85–90 mm Hg) have sug-
S170 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
gested improved neurological outcomes with this manage-ment plan (10, 11, 24, 29–31). Decompression of the cord hasthe potential to eliminate both cord compression and vascularcompression and distortion (3, 5–8, 30). Either or both of thesetreatment strategies may improve spinal cord blood flow inthe acute central cervical spinal cord injury setting, whichcould translate into preservation of neurological tissue andrecovery of neurological function. The benefit may or may notbe realized at the site of primary injury, but rather at vulner-able adjacent spinal cord levels fed by sulcal and collateralarteries that pass through the injury site but supply the cordrostral and caudal to the site of injury (1, 22, 23, 27, 28).
SUMMARY
The ideal management strategy for patients with acutecentral cervical spinal cord injuries seems to be multifaceted.As Schneider et al. (20) insisted years ago, a rapid, accuratediagnosis is essential. A detailed clinical examination, cervicalspinal x-rays to assess vertebral column injury (see recom-mendations in Chapter 5), and MRI assessment of the cervicalspinal cord for intrinsic injury and/or compression will ac-complish this goal. Many of these patients may require man-agement in the intensive care unit setting (see recommenda-tions in Chapter 7) for monitoring and treatment of cardiac,pulmonary, and blood pressure disturbances. Blood pressureaugmentation to mean arterial blood pressure levels of 85 to90 mm Hg may be of benefit (see recommendations in Chap-ter 8). Early reduction of fracture or fracture-dislocation inju-ries should be accomplished (see recommendations in Chap-ter 20). Administration of pharmacological agents may be ofbenefit according to specific parameters (see recommenda-tions in Chapter 9). Surgical decompression of the compressedspinal cord, particularly if the compression is focal and ante-rior and is approached anteriorly, seems to be of benefit inselected patients.
KEY ISSUES FOR FUTURE INVESTIGATION:
A prospective, controlled, randomized investigation of pa-tients with acute central cervical spinal cord injuries treatedwith aggressive medical therapy alone (intensive care unitmanagement, blood pressure augmentation, closed fracture-dislocation reduction), compared with patients managed withaggressive medical therapy and early surgical decompressionof the spinal cord, is needed.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Amar AP, Levy ML: Pathogenesis and pharmacological strategiesfor mitigating secondary damage in acute spinal cord injury.Neurosurgery 44:1027–1040, 1999.
2. Bosch A, Stauffer ES, Nickel VL: Incomplete traumatic quadriple-gia: A ten-year review. JAMA 216:473–478, 1971.
3. Bose B, Northrup BE, Osterholm JL, Cotler JM, DiTunno JF:Reanalysis of central cervical cord injury management. Neuro-surgery 15:367–372, 1984.
4. Bridle MJ, Lynch KB, Quesenberry CM: Long term functionfollowing the central cord syndrome. Paraplegia 28:178–185,1990.
5. Brodkey JS, Miller CF Jr, Harmody RM: The syndrome of acutecentral cervical spinal cord injury revisited. Surg Neurol 14:251–257, 1980.
6. Chen TY, Dickman CA, Eleraky M, Sonntag VKH: The role ofdecompression for acute incomplete cervical spinal cord injuryin cervical spondylosis. Spine 23:2398–2403, 1998.
7. Chen TY, Lee, ST, Lui TN, Wong CW, Yeh YS, Tzaan WC, HungSY: Efficacy of surgical treatment in traumatic central cord syn-drome. Surg Neurol 48:435–441, 1997.
8. Dai L, Jia L: Central cord injury complicating acute cervical discherniation in trauma. Spine 25:331–336, 2000.
9. Kalfas I, Wilberger JE, Goldberg A, Prostko ER: Magnetic reso-nance imaging in acute spinal cord trauma. Neurosurgery 23:295–299, 1988.
10. Levi L, Wolf A, Belzberg H: Hemodynamic parameters inpatients with acute cervical cord trauma: Description, interven-tion, and prediction of outcome. Neurosurgery 33:1007–1017,1993.
11. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does thetiming of surgery affect the outcome? Neurosurgery 29:216–222, 1991.
12. Maroon JC: “Burning hands” in football spinal cord injuries.JAMA 238:2049–2051, 1977.
13. Merriam WF, Taylor TK, Ruff SJ, McPhail MJ: A reappraisal ofacute traumatic central cord syndrome. J Bone Joint Surg Br65B:708–713, 1986.
14. Mirvis SE, Geisler FH, Jelinek JJ, Joslyn JN, Gellad F: Acutecervical spine trauma: Evaluation with 1.5-T MR imaging. Radi-ology 166:807–816, 1988.
15. Newey ML, Sen PK, Fraser RD: The long-term outcome af-ter central cord syndrome. J Bone Joint Surg Br 82B:851–855,2000.
16. Quencer RM, Bunge RP, Egnor M, Green BA, Puckett W, NaidichTP, Post MJ, Norenberg M: Acute traumatic central cord syn-drome: MRI-pathological correlations. Neuroradiology 34:85–94,1992.
17. Roth EJ, Lawler MH, Yarkony GM: Traumatic central cord syn-drome: Clinical features and functional outcomes. Arch PhysMed Rehabil 71:18–23, 1990.
18. Schneider RC: A syndrome in acute cervical spine injuries forwhich early operation is indicated. J Neurosurg 8:360–367,1951.
19. Schneider RC, Cherry G, Pantek H: The syndrome of acute centralcervical spinal cord injury. J Neurosurg 546–577, 1954.
20. Schneider RC, Thompson JC, Bebin J: The syndrome of acutecentral cervical spinal cord injury. J Neurol Neurosurg Psychia-try 21:216–227, 1958.
21. Shrosbree RD: Acute central cervical spinal cord syndrome: Ae-tiology, age incidence and relationship to the orthopaedic injury.Paraplegia 14:251–258, 1977.
22. Tator CH: Ischemia as a secondary neuronal injury, in SalzmanSK, Faden AI (eds): Neurobiology of Central Nervous System Trauma.New York, Oxford University Press, 1994, pp 209–215.
Acute Central Cervical Spinal Cord Injuries S171
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
23. Tator CH: Experimental and clinical studies of the pathophysiol-ogy and management of acute spinal cord injury. J Spinal CordMed 19:206–214, 1996.
24. Tator CH, Rowed DW, Schwartz ML, Gertzbein SD, Bharatwal N,Barkin M, Edmonds VE: Management of acute spinal cord inju-ries. Can J Surg 27:289–296, 1984.
25. Tracy PT, Wright RM, Hanigan WC: Magnetic resonance imagingof spinal injury. Spine 14:292–301, 1989.
26. Turnbull IM: Microvasculature of the human spinal cord.J Neurosurg 35:141–147, 1971.
27. Turnbull IM: Blood supply of the spinal cord: Normal and patho-logical considerations. Clin Neurosurg 20:56–84, 1973.
28. Turnbull IM, Brieg A, Hassler O: Blood supply of cervical spinalcord in man: A microangiographic cadaver study. J Neurosurg24:951–965, 1966.
29. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medical andsurgical treatment after acute spinal cord injury: Results of a prospectivepilot study to assess the merits of aggressive medical resuscitation andblood pressure management. J Neurosurg 87:239–246, 1997.
30. Wolf A, Levi L, Mirvis S, Ragheb J, Huhn S, Rigamonti D,Robinson WL: Operative management of bilateral facet disloca-tion. J Neurosurg 75:883–890, 1991.
31. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cord injuriesin the Swiss Paraplegic Centre of Basel. Paraplegia 14:58–65, 1976.
Dissection showing some of the conspicuous differences in passing from the vertebral column to the skull. From, Tilney F,Alsop Riley H: The Form and Functions of the Central Nervous System: An Introduction to the Study of Nervous Diseases.New York, Harper & Brothers, 1938, 3rd ed.
S172 Guidelines for Management of Acute Cervical Spinal Injuries
CHAPTER 22
Management of Vertebral Artery Injuries after NonpenetratingCervical Trauma
RECOMMENDATIONSDIAGNOSTIC:Standards: There is insufficient evidence to support diagnostic standards.Guidelines: There is insufficient evidence to support diagnostic guidelines.Options: Conventional angiography or magnetic resonance angiography is recommended for the diagnosis of
vertebral artery injury after nonpenetrating cervical trauma in patients who have complete cervical spinalcord injuries, fracture through the foramen transversarium, facet dislocation, and/or vertebral subluxation.
TREATMENT:Standards: There is insufficient evidence to support treatment standards.Guidelines: There is insufficient evidence to support treatment guidelines.Options:• Anticoagulation with intravenous heparin is recommended for patients with vertebral artery injury who
have evidence of posterior circulation stroke.• Either observation or treatment with anticoagulation in patients with vertebral artery injuries and evidence
of posterior circulation ischemia is recommended.• Observation in patients with vertebral artery injuries and no evidence of posterior circulation ischemia is
recommended.
RATIONALE
The association of cerebrovascular insufficiency and cer-vical fracture was first described by Suechting andFrench (13) in a patient with Wallenburg’s syndrome
occurring 4 days after a C5–C6 fracture-dislocation injury.Although Schneider et al. (8) implicated vertebral artery in-jury (VAI) at the site of cervical fracture-dislocation as a causeof posterior circulation cerebral ischemia, Gurdjian et al. (4)suggested that unilateral vertebral artery occlusions might beasymptomatic. Subsequent articles described larger series ofpatients with asymptomatic VAIs after blunt cervical spinaltrauma (2, 18). However, Biffl et al. (1), in the largest prospec-tive series to date, consisting of 38 patients with VAI diag-nosed by angiography, reported more frequent strokes inpatients not treated initially with intravenous heparin antico-agulation despite an asymptomatic VAI. Fractures throughthe foramen transversarium, facet fracture-dislocation, or ver-tebral subluxation is almost always seen in patients with VAI(1–3, 5, 17–19). A cadaveric study demonstrated progressivevertebral occlusion with higher degrees of flexion-distractioninjury, confirming this clinical observation (11). To developdiagnostic and treatment recommendations for VAI afterblunt cervical trauma, an analysis of the articles examining itsmanagement is undertaken in this report. Specific issues that
were addressed include the clinical and radiographic criteriaused to prompt diagnostic evaluation, the appropriate diag-nostic tests for identifying VAI, and the management of VAI(observation versus anticoagulation with heparin).
SEARCH CRITERIA
A computerized search of the National Library of Medicinedatabase of the literature published from 1966 to 2001 wasundertaken. The medical subject headings “vertebral artery,”“cervical vertebrae,” “dislocation,” and “wounds and inju-ries” yielded 6,447, 15,667, 24,174, and 459,759 citations, re-spectively. Combination of the first two headings with thethird heading yielded a subset of 61 citations. Combination ofthe first two headings with the fourth heading yielded asubset of 239 citations. Abstracts were reviewed, and onlyarticles in English containing three or more human subjectswith VAI after blunt cervical trauma were included. Fourteenarticles, including two prospective studies that provide ClassII evidence (1–3, 5, 10, 16–18), met these selection criteria andprovided data on 122 patients for this report. These articlesare summarized in Table 22.1. A total of 19 referenced articlesprovide the foundation for this review.
S173Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
TABLE 22.1. Summary of Reports on Vertebral Artery Injuriesa
Series (Ref. No.) Description of Study Evidence Class Conclusions
Schellinger et al., 2001
(7)
Retrospective review of 4 patients with VA dissection
diagnosed with various imaging studies, among 27 patients
with cervical spine injuries.
III 2 with motor complete cord injuries awoke after cervical surgery with stupor
and posterior circulation infarcts; 1 with basilar occlusion by Doppler died,
and 1 with dissection by DSA partially recovered. 1 with C5 foramen
transversarium fracture became comatose 12 h after admission and had fatal
basilar thrombosis by CT hours later. 1 with treated C2 fractures had vertigo
and nystagmus 3 wk later; dissection by angio; symptoms resolved with
warfarin followed by aspirin.
Biffl et al., 2000 (1) Prospective angio screening in blunt craniovertebral injury
identified 38 patients with VA injuries.
III 38 patients with 47 VA injuries of approx. 350 angios performed. 27/38 had
cervical fractures. 6/27 had fractures through the foramen transversarium. 9/38
had a posterior circulation stroke. Stroke not related to occlusion versus
stenosis. 8-h to 12-d delay, most beyond 48 h. 3/21 asymptomatic patients
treated with heparin had stroke versus 6/17 without heparin had stroke (P �
0.13). 1/6 died with heparin. 2/3 died without heparin. 2 treated with heparin
had hemorrhagic strokes.
Weller et al., 1999 (17) Prospective MRA in 12 patients with foramen transversarium
fractures.
III 3/12 had VA occlusion; all remained asymptomatic on aspirin. 1/12 with
stenosis had delayed syncope on aspirin, resolved with brief intravenous
heparin followed by aspirin.
Vaccaro et al., 1998
(16)
Prospective study with follow-up MRA in 6/12 patients (1
excluded) previously reported by Giacobetti et al. (3).
III 1/6 treated without heparin reconstituted by 12 d. 5/6 remained occluded (2/5
with heparin) �1 yr later.
Giacobetti et al., 1997
(3)
Prospective study with MRA in 61 patients with cervical
injuries found 12 patients with VA occlusion.
III 1/4 with transverse foramen fractures had occlusion. 6/15 with facet
dislocation had occlusion. 3/12 with transient blurred vision resolved with 3
mo anticoagulation.
Thibodeaux et al., 1997
(14)
Retrospective review of 3 patients with vertebral dissection on
angio.
III 1 with blindness but no infarct improved with anticoagulation. 1 with C4–C5
facet fracture was asymptomatic without treatment. 1 with ataxia/dizziness 2 d
later with occipital infarct recovered with 6 mo anticoagulation.
Prahbu et al., 1996 (6) Retrospective review of 5 symptomatic patients with 4 VA
occlusions and 1 stenosis on MRA and/or angio.
III 3 with fractures had delayed coma at 3 h, 2 occlusions/1 narrowing, 1 died
with stroke, 2 improved with anticoagulation, but it was discontinued in 1
secondary to rectus hematoma. 1 with fracture had delayed confusion/aphasia
with multiple MCA strokes, improved without anticoagulation. 1 with fracture
was asymptomatic without anticoagulation.
Friedman et al., 1995
(2)
Prospective study of 37 patients with nonpenetrating cervical
trauma found 9 VA injuries by MRA.
II 50% patients with complete cord injuries had VA injury versus 12% patients
with incomplete cord injuries (P � 0.02). 5/13 patients with �3 mm
subluxation had VA injuries versus 4/24 patients with �3 mm subluxation. 1
with bilateral VA injuries died of large cerebellar infarct (bilateral foramen
transversarium fractures) 8 asymptomatic (1/8 with anticoagulation also had
carotid occlusion).
Tulyapronchote et al.,
1994 (15)
Retrospective study of 3 patients with VB ischemia �2 wk after
occult C2 fractures diagnosed by angio.
III 2 occluded, 1 narrowing. Symptoms included syncope, vertigo, dysphagia,
dysarthria, facial numbness, blurred vision. Treatment not reported.
Willis et al., 1994 (18) Prospective study of 26 patients with cervical facet dislocation
or facet fractures through foramen transversarium fractures
revealed 12 with VA injuries on angio.
II 9/14 with normal angio had foramen transversarium fractures versus 7/12 with
abnormal angio. 1/9 with occlusion without anticoagulation died from
unrelated injuries. 1 dissection became an occlusion on heparin. 1 intimal
flap/1 pseudoaneurysm healed with heparin in 7–10 d.
Sim et al., 1993 (10) Prospective delayed duplex sonography of 11 patients with
previously locked facets.
III 1/11 had occlusion (persistent locked facet). 1/11 had a narrow VA.
Woodring et al., 1993
(19)
Retrospective study of 216 patients with cervical fractures
showed 52 with TP fractures. 8 had angio.
III 78% of TP fractures extended into foramen transversarium. 4/8 had occlusion,
3/8 had dissection, 1 of each had stroke that improved with anticoagulation. 3
asymptomatic patients treated with anticoagulation.
Schwarz et al., 1991 (9) Retrospective review of 4 patients with symptomatic VA
injuries by angio.
III 4 had delayed symptoms (3 occlusions, 1 pseudoaneurysm). 1 with coma and
stroke 20 min after reduction of C4–C5 dislocation received streptokinase/
heparin for occlusion but died.
Louw et al., 1990 (5) Prospective study of 12 patients with facet dislocations with
DSA.
III 4/5 unilateral and 5/7 bilateral dislocations had occluded VAs (1 bilateral). 3
occluded at level, 5 within 2 cm. 2/9 symptomatic patients had bilateral C5–
C6 facet dislocation and improved without treatment.
a DSA, digital subtraction angiography; CT, computed tomography; angio, angiography; MRA, magnetic resonance angiography; VA, vertebralartery; MCA, middle cerebral artery; VB, vertebrobasilar artery; TP, transverse process.
S174 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
SCIENTIFIC FOUNDATION
Diagnosis
The diagnosis of VAI can be made with a variety of radio-graphic studies. Angiography is the traditional imaging tech-nique used to diagnose VAI. Angiography was used for allpatients in seven of the studies reviewed (1, 5, 9, 14, 15, 18, 19)and in combination with other modalities in two additionalstudies (6, 7). Three studies prospectively applied conven-tional angiography to patients who had sustained nonpen-etrating cervical trauma and who met certain inclusion crite-ria (1, 5, 18). Similarly, three studies prospectively appliedmagnetic resonance angiography (MRA) to patients who hadsustained nonpenetrating cervical trauma and who met cer-tain inclusion criteria (2, 16, 17).
Biffl et al. (1) reported the largest prospective study usingangiography. Study subjects were selected from among 7205blunt trauma patients by the use of clinical and radiographiccriteria. Patients underwent angiography if they had facialhemorrhage (bleeding from mouth, nose, ears), cervical bruit(in patients �50 yr), expanding cervical hematoma, cerebralinfarction by computed tomography (CT), or lateralizing neu-rological deficit. “Asymptomatic” patients were selected forangiography if they had cervical hyperextension/rotation orhyperflexion injuries, closed head injury with diffuse axonalinjury, near hanging, seat belt or other soft tissue injuries tothe neck, basilar cranial fractures extending into the carotidcanal, and cervical vertebral body fractures or distractioninjuries. Between 350 and 400 angiograms were performed,identifying 38 patients with VAI. However, neither the exactnumber of angiograms performed nor the number of patientswho met the various criteria without VAI were reported. As aresult, sensitivity, specificity, positive predictive value, andnegative predictive value of the selection criteria could not bedetermined. Cervical spine injuries were observed in 27 of 38patients with VAI, including fractures through the foramentransversarium in four, dislocations in six, vertebral subluxa-tions in two, and more than one of these injuries in two.Twenty-nine patients had unilateral VAI (18 left, 11 right);nine had bilateral VAIs. A vascular injury scale was used tostratify patients into five categories: Grade I, arterial dissec-tions with less than 25% luminal narrowing; Grade II, arterialdissections with more than 25% luminal narrowing; Grade III,pseudoaneurysm of the vertebral artery; Grade IV, occlusionof the vertebral artery; and Grade V, vertebral artery transec-tion. Seven patients died; five had bilateral VAIs (Grade I),and two had unilateral VAI (one Grade I, one Grade IV).Three patients with either no neurological deficit or milddeficit had bilateral VAIs. The authors concluded that strokeincidence and neurological outcome appeared to be indepen-dent of the grade of VAI.
Another prospective study by Willis et al. (18) identified 30patients with midcervical fractures and/or dislocation forangiography. However, only 26 patients who met the criteriaagreed to proceed with angiography. Twelve patients sus-tained VAIs demonstrated by angiography (six left occlu-sions, three right occlusions, one left intimal flap, one left
pseudoaneurysm, one left dissection). The authors providedsufficient data regarding the presence of foramen transver-sarium fracture, facet dislocation, and subluxation to deter-mine the usefulness of these radiographic findings in identi-fying patients with VAI. The calculated sensitivity of foramentransversarium fracture as a criterion for identifying VAI inthis study was 58%; the specificity was 36%. The positivepredictive value of foramen transversarium fracture was 44%;the negative predictive value was 50%. The calculated sensi-tivity of facet dislocation as a criterion for identifying VAI was42%; the specificity was 57%. The positive predictive value offacet dislocation was 45%; the negative predictive value was53%. The calculated sensitivity of subluxation as a criterionfor identifying VAI was 67%; the specificity was 29%. Thepositive predictive value of subluxation was 80%; the negativepredictive value was 50%. Any combination of foramen trans-versarium fracture, facet dislocation, and/or vertebral sublux-ation revealed a calculated sensitivity for identifying VAI of92% and a specificity of 0%. The positive predictive value ofthe presence of any of the three criteria and VAI was 44%; thenegative predictive value was 0%.
A prospective study by Louw et al. (5) examined 12 con-secutive patients with cervical spine facet dislocations withdigital subtraction angiography. Five of seven patients withbilateral facet dislocations had vertebral artery occlusion (onebilaterally), and four of five patients with unilateral facetdislocations had unilateral vertebral artery occlusion. Angiog-raphy was not performed in blunt cervical trauma patientswithout facet dislocation. In a retrospective study byWoodring et al. (19), seven of eight patients with transverseprocess fractures who underwent angiography had VAIs (tworight occlusions, two left occlusions, two right dissections, oneleft dissection). Seventy-eight percent of transverse processfractures extended into the foramen transversarium. Angiog-raphy was not performed in 44 other patients with transverseprocess fractures. Alternatively, MRA has been used to diag-nose VAI noninvasively. Weller et al. (17) prospectively ex-amined 12 patients who had experienced nonpenetrating cer-vical trauma and who sustained fractures through theforamen transversarium. Three patients had unilateral verte-bral artery occlusion, and one had focal narrowing, all at thesite of fracture. MRA was not performed on the 26 patientswithout these fractures. Giacobetti et al. (3) prospectivelyevaluated all patients admitted with cervical spine injurieswith MRA. Twelve of 61 patients had vertebral artery occlu-sion demonstrated by MRA; all injuries were unilateral (sixleft, six right). Although 7 of 12 patients with VAI had flexion-distraction injuries with facet dislocations, the types of cervi-cal spinal injuries sustained by the 49 patients with normalMRA results were not reported. Because none of these fourarticles (3, 5, 17, 19) provided sufficient information regardingthe types of injury and results of vertebral artery imaging inthe entire population of patients studied, sensitivity, specific-ity, positive predictive value, and negative predictive value ofthe injury types could not be determined.
Friedman et al. (2) prospectively examined 37 patients ad-mitted with “major” blunt cervical spine injuries identified byMRA and compared these patients with a size-matched con-
Vertebral Artery Injuries in Nonpenetrating Cervical Trauma S175
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
trol group of patients without a history of cervical trauma.Nine patients had VAIs (six unilateral occlusions, two nar-rowing, one bilateral injury). Both vertebral arteries werevisualized in all 37 control subjects. Complete spinal cordinjuries were observed in 12 of 37 patients with cervicaltrauma, 6 of whom had VAIs (P � 0.02; �2 test). More than 3mm of subluxation was observed in 13 of 37 patients, 5 ofwhom had VAIs (P � 0.14; �2 test). The calculated sensitivityof complete spinal cord injury as a criterion for identifyingVAI in this study was 67%; the specificity was 79%. Thepositive predictive value of complete spinal cord injury was50%; the negative predictive value was 88%. The calculatedsensitivity of subluxation as a criterion for identifying VAIwas 56%; the specificity was 71%. The positive predictivevalue of subluxation was 38%; the negative predictive valuewas 83%.
Other diagnostic modalities have also been used to identifyof VAI. CT with intravenous contrast demonstrated a unilat-eral vertebral artery occlusion in one patient with a Jeffersonfracture, which was subsequently confirmed by angiography(12). Duplex sonography has also been used to diagnose VAI(7, 10, 14). Angiography has occasionally been used to confirmthe results of MRA or ultrasonography, but there has not beena study comparing MRA or ultrasonography with angiogra-phy in the diagnosis of VAI.
Treatment
After diagnosis of VAI, treatment options examined by thevarious studies have included observation alone or anticoag-ulation with either intravenously administered heparin ororally administered antiplatelet agents. Some authors treatedasymptomatic patients (1, 2, 18, 19); others did not treat symp-tomatic patients (5, 7).
Several articles retrospectively identified patients with neu-rological complications of VAI. Schellinger et al. (7) describedfour patients with VAI who had delayed onset of neurologicaldysfunction. Two patients awoke after surgery with alteredconsciousness and posterior circulation stroke. One of thesepatients with vertebrobasilar occlusion diagnosed by ultra-sound died; the other with vertebral artery dissection andpseudoaneurysm improved. Neither patient was given anti-coagulation treatment. Two other patients developed delayedsymptoms at 12 hours and 3 weeks after injury. The firstpatient died from basilar artery thrombosis confirmed by CT.The second patient developed vertigo and nystagmus as aresult of a dissection diagnosed by ultrasound and angiogra-phy. This patient was successfully treated with intravenousheparin anticoagulation and recovered within several weeks.Thibodeaux et al. (14) reported one patient who developedblindness immediately after left vertebral artery dissection.No infarction was seen on CT. Treatment included 3 monthsof anticoagulation with sustained improvement at the 4-yearfollow-up examination. A second patient with ataxia anddizziness 2 days after VAI had occipital infarction by CT andbilateral vertebral artery dissections but recovered after 6months of anticoagulation therapy. One patient with dissec-tion remained asymptomatic without anticoagulation. Prabhu
et al. (6) reported three of five patients with VAI who expe-rienced sudden loss of consciousness 3 hours after injury. Allwere anticoagulated. One patient with vertebrobasilar throm-bosis died; one patient with bilateral vertebral occlusionsimproved. The third patient with vertebral artery stenosisimproved, although anticoagulation was stopped severaldays later secondary to a rectus sheath hematoma. Two pa-tients with asymptomatic vertebral artery occlusion were nottreated and remained asymptomatic. Tulyapronchote et al.(15) reported three patients with delayed onset of symptoms2 weeks to 3 months after VAI, including syncope, visualdisturbance, dysarthria, dysphagia, and vertigo. Two patientshad vertebral artery occlusions, and one had a dissection. Notreatment was reported for any of these patients. Woodring etal. (19) reported two strokes in nine patients with transverseprocess fractures after blunt cervical trauma. All nine werestudied with vertebral angiography; seven studies were ab-normal. One patient with vertebral artery occlusion wastreated with anticoagulation and improved. One with dissec-tion improved without treatment. Three of five asymptomaticpatients were anticoagulated (intravenous heparin convertedto warfarin; total treatment of 6 wk); two patients were nottreated. All five remained asymptomatic. Schwarz et al. (9)reported four patients with ischemic vertebrobasilar symp-toms after nonpenetrating cervical trauma. Two patients withatlantoaxial injuries recovered after atlantoaxial stabilization(one treated with halo immobilization, one operatively).Treatment of one patient with delayed symptoms after unrec-ognized facet dislocation was not reported. Streptokinase in-fusion was used in a patient with vertebral artery occlusionwho became comatose shortly after reduction of a facet dis-location injury. This strategy failed to achieve completethrombolysis, and the patient died days later.
Six prospective studies examining the diagnosis of VAIprovided the incidence of neurological complications relatedto VAI. Biffl et al. (1) reported the highest frequency of pos-terior circulation stroke: 24% (9 of 38 patients). Stroke oc-curred 8 hours to 12 days after injury in these patients, occur-ring more than 48 hours after injury in 78%. Three of 21asymptomatic patients treated with intravenous heparin sub-sequently developed stroke; one died and two had mild re-sidual deficits. In contrast, 6 of 17 asymptomatic patients notinitially treated with intravenous heparin had strokes; twodied, three had mild deficits, and one had a severe deficit.Neither of the two patients who died was treated with anti-coagulation when stroke occurred 9 to 12 days after VAI (oneGrade IV, one bilateral Grade I that progressed to right GradeII and left Grade III). All three asymptomatic patients whosubsequently developed stroke and were then treated withheparin had mild residual deficits (one Grade I, one Grade II,one Grade IV). A final patient who had a stroke and wastreated with antiplatelet agents developed severe residualdeficits (Grade I). Two of nine patients with stroke who weretreated with heparin developed hemorrhagic infarction. Twoof the 38 patients had dominant vertebral arteries; both hadinjuries to the dominant artery and neither developed astroke.
S176 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
Weller et al. (17) reported on four of their patients withVAIs treated with aspirin. Three patients with vertebral arteryocclusion remained asymptomatic; one with vertebral ar-tery narrowing developed syncope 17 days after injury. Thispatient was treated with intravenous heparin followed byaspirin without recurrent symptoms. Giacobetti et al. (3) de-scribed 3 of 12 patients with vertebral artery occlusion whodeveloped transient blurred vision; all three were treated with3 months of anticoagulation (intravenous heparin followed bywarfarin), and none had recurrent symptoms. Friedman et al.(2) reported one patient with bilateral VAIs who died after amassive right cerebellar infarct. One patient with vertebralocclusion and concurrent carotid occlusion remained asymp-tomatic on heparin. The remaining seven patients remainedasymptomatic without treatment (five occlusions, two steno-sis). Willis et al. (18) reported no symptoms in 12 patients withVAIs, none of whom were treated with anticoagulation. Louwet al. (5) reported two of nine patients with vertebral arteryocclusion with symptoms, including blurring of vision, thatspontaneously resolved without anticoagulation.
The management and outcome of 106 patients with VAIafter nonpenetrating cervical injury could be determined fromthe articles reviewed. Twelve patients had radiographic evi-dence of a posterior circulation stroke as their first symptom.Four patients were treated with intravenous heparin; one diedand three improved (1, 19). The remaining eight patients werenot treated with intravenous heparin; five died, two im-proved, and one had a severe neurological deficit (1, 2, 7, 19).Fifteen patients developed symptoms of posterior circulationischemia without stroke before treatment was instituted (3,5–7, 9, 14, 17). Eleven of 15 patients were treated with intra-venous heparin; two died (both had strokes), and nine im-proved (one developed a stroke). The remaining four patientswere not treated with intravenous heparin; all improved with-out developing a posterior circulation stroke. Twenty-sevenasymptomatic patients were prophylactically treated with in-travenous heparin (1, 2, 18, 19). Three patients developedposterior circulation strokes; 27 patients remained asymptom-atic. Finally, 52 asymptomatic patients were not prophylacti-cally treated with heparin (8 were treated with aspirin and 1with embolization). Three patients died from injuries unre-lated to the VAI (1, 18). Four patients had severe deficits (threewere treated with aspirin and one with embolization), andthree had mild deficits (one was treated with aspirin); alldefects were related to the associated spinal cord injury (1).The remaining 42 patients (3 of whom were treated withaspirin) remained asymptomatic (1–3, 5, 6, 14, 17–19).
The articles reviewed did not specifically address the risk ofprogressive spinal cord hemorrhage worsening in patientswith VAI and an associated spinal cord injury treated withheparin. One patient with a complete cervical spinal cordinjury and hematomyelia was placed on intravenous heparinprophylactically for left carotid and vertebral artery occlusion;no neurological changes occurred with treatment (2).
At least 13 of 42 patients treated with intravenous heparinhad complications during their treatment; in 6 patients (14%),the complications were significant. Six patients (two of whomdied) developed posterior circulation strokes after treatment
with intravenous heparin was initiated (1, 6, 9, 14). Twopatients developed hemorrhagic posterior circulation strokes;the timing of intravenous heparin relative to the developmentof the posterior circulation stroke was not reported (1). Intra-venous heparin was discontinued in three patients. Onesymptomatic patient developed a rectus sheath hematoma (6);the patient’s symptoms stabilized after intravenous heparintreatment was discontinued. Two asymptomatic patients hadprogression in the grade of VAI; both remained asymptomaticafter intravenous heparin was discontinued (18). Four otherpatients had progression in the grade of VAI; intravenousheparin was continued, and none of the patients developeddeficits related to the progression of injury grade (1).
In several studies, patients were reimaged to determinewhether disease progression or resolution occurred after VAI.Biffl et al. (1) reported follow-up angiography on 21 patients.Of 16 patients treated with heparin, two improved to a lessergrade of vascular injury and four worsened to a higher grade.Of five patients not receiving heparin, one improved andthree had worse vascular injury grades. Vaccaro et al. (16),using MRA, found reconstitution in one of six patients withVAI 12 days after the original diagnosis; this patient was nottreated with anticoagulation. The other five patients still hadvertebral artery occlusion more than 1 year later, includingtwo treated with anticoagulation. Willis et al. (18) describedthe results of follow-up angiography in three patients withVAI. One patient with a pseudoaneurysm received 1 week ofintravenous heparin followed by aspirin; the pseudoaneu-rysm had slightly enlarged 7 days after treatment was begun,but it had disappeared on angiography performed 6 weekslater. One patient treated with intravenous heparin for avertebral artery dissection had an asymptomatic occlusion ofthe artery demonstrated by angiography 2 days later; heparinadministration was discontinued. The third patient wastreated with intravenous heparin for a vertebral artery intimalflap; the patient had a normal vertebral angiogram 10 dayslater. Thibodeaux et al. (14) found a patent vertebral artery 6months after dissection was diagnosed; this patient did notreceive anticoagulation treatment. Sim et al. (10) reporteddelayed duplex sonography in 11 patients with a history offacet dislocation but with unknown vertebral artery status atthe time of the original cervical spine injury. Two of thesestudies demonstrated VAI: one patient with persistent dislo-cation had a vertebral occlusion, and one patient with a re-duced injury had vertebral artery stenosis.
SUMMARY
The incidence of VAI may be as high as 11% after nonpen-etrating cervical spinal trauma in patients with specific clini-cal criteria, including facial hemorrhage (bleeding frommouth, nose, ears), cervical bruit in those younger than 50years, expanding cervical hematoma, cerebral infarction byCT, lateralizing neurological deficit, cervical hyperextension-rotation or hyperflexion injuries, closed head injury with dif-fuse axonal injury, near hanging, seat belt or other soft tissueinjuries to the neck, basilar cranial fractures extending into thecarotid canal, and cervical vertebral body fractures or distrac-
Vertebral Artery Injuries in Nonpenetrating Cervical Trauma S177
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
tion injuries. Many patients with VAI have complete spinalcord injuries, fracture through the foramen transversarium,facet dislocation, and/or vertebral subluxation, but many pa-tients with these injuries have normal vertebral arteries whenimaged, thus compromising the specificity of these injurycriteria. Many patients with VAI are asymptomatic, includingthose with vertebral artery occlusion or dissection. The liter-ature reviewed indicates that patients with posterior circula-tion stroke and VAI have a better outcome when treated withintravenous heparin than patients who do not receive thistreatment. However, others have reported improvementamong patients without anticoagulation (7, 19). The outcomeof patients who develop symptoms of posterior circulationischemia without stroke and are treated with intravenousheparin (3, 6, 7, 14, 17) is similar to that of patients receivingno treatment (5, 9). Although the largest prospective studysuggested a trend toward less frequent stroke in asymptom-atic patients treated with heparin (1), others have not reportedsimilar observations (2, 3, 5, 6, 14, 18, 19). Because the risk ofsignificant complications related to anticoagulation is approx-imately 14% in these studies, there is insufficient evidence torecommend anticoagulation in asymptomatic patients.
KEY ISSUES FOR FUTURE INVESTIGATION
Although several prospective studies examined patients atrisk for VAI, most articles did not provide enough data on thecharacteristics of the patients with normal vertebral arteries tounderstand the clinical or radiographic characteristics thatwould predict which patients may have VAI. A prospectivestudy comparing MRA with conventional angiography innonpenetrating cervical spine trauma may define the role ofnoninvasive imaging studies in these patients. A multicenter,randomized, prospective study comparing anticoagulationwith intravenous heparin versus observation in asymptomaticpatients and in symptomatic patients with posterior circula-tion ischemia but without stroke is recommended to deter-mine whether anticoagulation of these patients is justified.
Reprint requests: Mark N. Hadley, M.D., Division of NeurologicalSurgery, University of Alabama at Birmingham, 516 Medical Educa-tion Building, 1813 6th Avenue South, Birmingham, AL 35294-3295.
REFERENCES
1. Biffl WL, Moore EE, Elliott JP, Ray C, Offner PJ, Franciose RJ,Brega KE, Burch JM: The devastating potential of blunt vertebralarterial injuries. Ann Surg 231:672–681, 2000.
2. Friedman D, Flanders A, Thomas C, Millar W: Vertebral arteryinjury after acute cervical spine trauma: Rate of occurrence asdetected by MR angiography and assessment of clinical conse-quences. AJR Am J Roentgenol 164:443–447, 1995.
3. Giacobetti FB, Vaccaro AR, Bos-Giacobetti MA, Deeley DM,Albert TJ, Farmer JC, Cotler JM: Vertebral artery occlusion asso-ciated with cervical spine trauma: A prospective analysis. Spine22:188–192, 1997.
4. Gurdjian ES, Hardy WG, Lindner DW, Thomas LM: Closed cer-vical trauma associated with involvement of carotid and vertebralarteries. J Neurosurg 20:418–427, 1963.
5. Louw JA, Mafoyane NA, Small B, Neser CP: Occlusion of thevertebral artery in cervical spine dislocations. J Bone Joint SurgBr 72B:679–681, 1990.
6. Prabhu V, Kizer J, Patil A, Hellbusch L, Taylon C, Leibrock L:Vertebrobasilar thrombosis associated with nonpenetrating cervi-cal spine trauma. J Trauma 40:130–137, 1996.
7. Schellinger PD, Schwab S, Krieger D, Fiebach JB, Steiner T, HundEF, Hacke W, Meinck HM: Masking of vertebral artery dissectionby severe trauma to the cervical spine. Spine 26:314–319, 2001.
8. Schneider RC, Crosby EC, Russo RH, Gosch HH: Traumatic spi-nal cord syndromes and their management. Clin Neurosurg 20:424–492, 1973.
9. Schwarz N, Buchinger W, Gaudernak T, Russe F, Zechner W:Injuries to the cervical spine causing vertebral artery trauma: Casereports. J Trauma 31:127–133, 1991.
10. Sim E, Schwarz N, Biowski-Fasching I, Biowski P: Color-codedDuplex sonography of vertebral arteries: 11 cases of blunt cervicalspine injury. Acta Orthop Scand 64:133–137, 1993.
11. Sim E, Vaccaro AR, Berzlanovich A, Pienaar S: The effects ofstaged static cervical flexion-distraction deformities on the pa-tency of the vertebral arterial vasculature. Spine 25:2180–2186,2000.
12. Song WS, Chiang YH, Chen CY, Lin SZ, Liu MY: A simplemethod for diagnosing traumatic occlusion of the vertebral arteryat the craniovertebral junction. Spine 19:837–839, 1994.
13. Suechting RL, French LA: Posterior inferior cerebellar artery syn-drome following a fracture of the cervical vertebra. J Neurosurg12:187–189, 1955.
14. Thibodeaux LC, Hearn AT, Peschiera JL, Deshmukh RM,Kerlakian GM, Welling RE, Nyswonger GD: Extracranial verte-bral artery dissection after trauma: A five year review. Br J Surg84:94, 1997.
15. Tulyapronchote R, Selhorst JB, Malkoff MD, Gomez CR: Delayedsequelae of vertebral artery dissection and occult cervical frac-tures. Neurology 44:1397–1399, 1994.
16. Vaccaro AR, Klein GR, Flanders AE, Albert TJ, Balderston RA,Cotler JM: Long-term evaluation of vertebral artery injuries fol-lowing cervical spine trauma using magnetic resonance angiog-raphy. Spine 23:789–795, 1998.
17. Weller SJ, Rossitch E Jr, Malek AM: Detection of vertebral arteryinjury after cervical spine trauma using magnetic resonance an-giography. J Trauma 46:660–666, 1999.
18. Willis BK, Greiner F, Orrison WW, Benzel EC: The incidence ofvertebral artery injury after midcervical spine fracture or sublux-ation. Neurosurgery 34:435–442, 1994.
19. Woodring JH, Lee C, Duncan V: Transverse process fractures ofthe cervical vertebrae: Are they insignificant? J Trauma 34:797–802, 1993.
S178 Guidelines for Management of Acute Cervical Spinal Injuries
BIBLIOGRAPHY
1. Acland RH, Anthony A, Inglis GS, Walton DI, Xiong X: Meth-ylprednisolone use in acute spinal cord injury. N Z Med J114:99, 2001 (letter).
2. Aebi M, Zuber K, Marchesi D: Treatment of cervical spineinjuries with anterior plating: Indications, techniques, and re-sults. Spine 16[Suppl 3]:S38–S45, 1991.
3. Ahuja A, Glasauer FE, Alker GJ Jr, Klein DM: Radiology insurvivors of traumatic atlanto-occipital dislocation. SurgNeurol 41:112–118, 1994.
4. Ajani AE, Cooper DJ, Scheinkestel CD, Laidlaw J, Tuxen DV:Optimal assessment of cervical spine trauma in critically illpatients: A prospective evaluation. Anaesth Intensive Care26:487–491, 1998.
5. Aldrich EF, Crow WN, Weber PB, Spagnolia TN: Use of MRimaging-compatible Halifax interlaminar clamps for posteriorcervical fusion. J Neurosurg 74:185–189, 1991.
6. Alexander R, Proctor H: Advanced Trauma Life Support Course forPhysicians. Chicago, American College of Surgeons, 1993, pp21–22.
7. Alexander E Jr, Davis CH Jr, Forsyth HF: Reduction and fusionof fracture dislocation of the cervical spine. J Neurosurg 27:588–591, 1967.
8. Allen BL Jr, Ferguson RL, Lehmann TR, O’Brien RP: A mech-anistic classification of closed, indirect fractures and disloca-tions of the lower cervical spine. Spine 7:1–27, 1982.
9. Amar AP, Levy ML: Pathogenesis and pharmacological strate-gies for mitigating secondary damage in acute spinal cordinjury. Neurosurgery 44:1027–1040, 1999.
10. American Academy of Orthopedic Surgeons, Committee onInjuries: Emergency Care and Transportation of the Sick and Injured.Chicago, American Academy of Orthopedic Surgeons, 1971, pp111–115.
11. American College of Surgeons: Advanced Trauma Life Support.Chicago, American College of Surgeons, 1993, p 201.
12. American College of Surgeons: Advanced Trauma Life Support.Chicago, American College of Surgeons, 1993, pp 214–218.
13. American College of Surgeons: Spine and spinal cord trauma,in Advanced Trauma Life Support for Doctors: Student CourseManual. Chicago, American College of Surgeons, 1997, ed 6, pp215–242.
14. American Spinal Injury Association: Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1984.
15. American Spinal Injury Association: Standards for NeurologicalClassification of Spinal Injury Patients. Chicago, American SpinalInjury Association, 1989.
16. American Spinal Injury Association/International Medical So-ciety of Paraplegia: Standards for Neurological and FunctionalClassification of Spinal Cord Injury. Chicago, American SpinalInjury Association, 1992, rev ed.
17. American Spinal Injury Association/International Medical So-ciety of Paraplegia: International Standards for Neurological andFunctional Classification of Spinal Cord Injury. Chicago, AmericanSpinal Injury Association, 1996, rev ed.
18. Anderson LD, D’Alonzo RT: Fractures of the odontoid processof the axis. J Bone Joint Surg Am 56A:1663–1674, 1974.
19. Anderson PA, Montesano PX: Morphology and treatment ofoccipital condyle fractures. Spine 13:731–736, 1988.
20. Anderson PA, Henley MB, Grady MS, Montesano PX, WinnHR: Posterior cervical arthrodesis with AO reconstructionplates and bone graft. Spine 16[Suppl 3]:S72–S79, 1991.
21. Andersson S, Rodrigues M, Olerud C: Odontoid fractures: Highcomplication rate associated with anterior screw fixation in theelderly. Eur Spine J 9:56–60, 2000.
22. Apfelbaum RI, Lonser RR, Veres R, Casey A: Direct anteriorscrew fixation for recent and remote odontoid fractures.J Neurosurg 93[Suppl 2]:227–236, 2000.
23. Apostolides PJ, Theodore N, Karahalios DG, Sonntag VKH:Triple anterior screw fixation of an acute combination atlas-axisfracture: Case report. J Neurosurg 87:96–99, 1997.
24. Argenson C, Lovet J, Sanouiller JL, de Peretti F: Traumaticrotatory displacement of the lower cervical spine. Spine 13:767–773, 1988.
25. Armitage JM, Pyne A, Williams SJ, Frankel H: Respiratoryproblems of air travel in patients with spinal cord injuries. BMJ300:1498–1499, 1990.
26. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support: Advanced Prehospital Care. Englewood Cliffs,Prentice-Hall, 1988, ed 2.
27. Augustine J: Spinal trauma, in Campbell JE (ed): Basic TraumaLife Support for Paramedics and Advanced EMS Providers. UpperSaddle River, Brady/Prentice Hall Health, 1998, ed 3.
28. Bachulis BL, Long WB, Hynes GD, Johnson MC: Clinical indi-cations for cervical spine radiographs in the traumatized pa-tient. Am J Surg 153:473–478, 1987.
29. Balshi JD, Cantelmo NL, Menzoian JO: Complications of cavalinterruption by Greenfield filter in quadriplegics. J Vasc Surg9:558–562, 1989.
30. Banit DM, Grau G, Fisher JR: Evaluation of the acute cervicalspine: A management algorithm. J Trauma 49:450–456, 2000.
31. Banna M, Stevenson GW, Tumiel A: Unilateral atlanto-occipitaldislocation complicating an anomaly of the atlas: A case report.J Bone Joint Surg Am 65A:685–687, 1983.
32. Barboriak JJ, Rooney CB, El Ghatit AZ, Spuda K, Anderson AJ:Nutrition in spinal cord injury patients. J Am Paraplegia Soc6:32–36, 1983.
33. Barker EG Jr, Krumpelman J, Long JM: Isolated fracture of themedial portion of the lateral mass of the atlas: A previouslyundescribed entity. AJR Am J Roentgenol 126:1053–1058, 1976.
34. Barney R, Cordell W: Pain associated with immobilization onrigid spine boards. Ann Emerg Med 18:918, 1989 (abstr).
35. Barros TE, Bohlman HH, Capen DA, Cotler J, Dons K, Biering-Sorensen F, Marchesi DG, Zigler JE: Traumatic spondylolisthesisof the axis: Analysis of management. Spinal Cord 37:166–171,1999.
36. Bauer D, Kowalski R: Effect of spinal immobilization devices onpulmonary function in the healthy, nonsmoking man. AnnEmerg Med 17:915–918, 1988.
37. Baum JA, Hanley EN Jr, Pullekines J: Comparison of halocomplications in adults and children. Spine 14:251–252, 1989.
38. Bayless P, Ray VG: Incidence of cervical spine injuries in associa-tion with blunt head trauma. Am J Emerg Med 7:139–142, 1989.
39. Beatson TR: Fractures and dislocation of the cervical spine.J Bone Joint Surg Br 45B:21–35, 1963.
40. Becker DM, Gonzalez M, Gentili A, Eismont F, Green BA:Prevention of deep venous thrombosis in patients with acutespinal cord injuries: Use of rotating treatment tables. Neuro-surgery 20:675–677, 1987.
S179Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
41. Bednar DA, Parikh J, Hummel J: Management of type II odon-toid process fractures in geriatric patients: A prospective studyof sequential cohorts with attention to survivorship. J SpinalDisord 8:166–169, 1995.
42. Bednarczyk JH, Sanderson DJ: Comparison of functional andmedical assessment in the classification of persons with spinalcord injury. J Rehabil Res Dev 30:405–411, 1993.
43. Bell C: Surgical observations. Middlesex Hosp J 4:469, 1817.44. Belzberg AJ, Tranmer BI: Stabilization of traumatic atlanto-
occipital dislocation: Case report. J Neurosurg 75:478–482,1991.
45. Benzel EC, Kesterson L: Posterior cervical interspinous com-pression wiring and fusion for mid to low cervical spinal inju-ries. J Neurosurg 70:893–899, 1989.
46. Benzel EC, Larson SJ: Functional recovery after decompressivespine operation for cervical spine fractures. Neurosurgery 20:742–746, 1987.
47. Benzel EC, Hadden TA, Saulsbery CM: A comparison of theMinerva and halo jackets for stabilization of the cervical spine.J Neurosurg 70:411–414, 1989.
48. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW,Espinosa MC: Fractures of the C-2 vertebral body. J Neurosurg81:206–212, 1994.
49. Benzel EC, Hart BL, Ball PA, Baldwin NG, Orrison WW,Espinosa MC: Magnetic resonance imaging for the evaluationof patients with occult cervical spine injury. J Neurosurg 85:824–829, 1996.
50. Berlemann U, Schwarzenbach O: Dens fractures in the elderly:Results of anterior screw fixation in 19 elderly patients. ActaOrthop Scand 68:319–324, 1997.
51. Berne JD, Velmahos GC, El-Tawil Q, Demetriades D, AsensioJA, Murray JA, Cornwell EE, Belzberg H, Berne TV: Value ofcomplete cervical helical computed tomographic scanning inidentifying cervical spine injury in the unevaluable blunttrauma patient with multiple injuries: A prospective study.J Trauma 47:896–903, 1999.
52. Bettini N, Malaguti MC, Sintini M, Monti C: Fractures of theoccipital condyles: Report of four cases and review of theliterature. Skeletal Radiol 22:187–190, 1993.
53. Beyer CA, Cabanela ME, Berquist TH: Unilateral facet disloca-tions and fracture-dislocations of the cervical spine. J BoneJoint Surg Br 73B:977–981, 1991.
54. Biffl WL, Moore EE, Elliott JP, Ray C, Offner PJ, Franciose RJ,Brega KE, Burch JM: The devastating potential of blunt verte-bral arterial injuries. Ann Surg 231:672–681, 2000.
54a. Birney TJ, Hanley EN: Traumatic cervical spine injuries inchildhood and adolescence. Spine 14:1277–1282, 1989.
55. Black CA, Buderer NM, Blaylock B, Hogan BJ: Comparativestudy of risk factors for skin breakdown with cervical orthoticdevices: Philadelphia and Aspen. J Trauma Nurs 5:62–66, 1998.
56. Blackmore CC, Emerson SS, Mann FA, Koepsell TD: Cervicalspine imaging in patients with trauma: Determination of frac-ture risk to optimize use. Radiology 211:759–765, 1999.
57. Blackmore CC, Ramsey SD, Mann FA, Deyo RA: Cervical spinescreening with CT in trauma patients: A cost-effectiveness anal-ysis. Radiology 212:117–125, 1999.
58. Blaylock B: Solving the problem of pressure ulcers resultingfrom cervical collars. Ostomy Wound Manage 42:26–33, 1996.
59. Bloom AI, Neeman Z, Floman Y, Gomori J, Bar-Ziv J: Occipitalcondyle fracture and ligament injury: Imaging by CT. PediatrRadiol 26:786–790, 1996.
60. Bloom AI, Neeman Z, Slasky BS, Floman Y, Milgrom M,Rivkind A, Bar-Ziv J: Fracture of the occipital condyles andassociated craniocervical ligament injury: Incidence, CT imag-ing and implications. Clin Radiol 52:198–202, 1997.
61. Bohay D, Gosselin RA, Contreras DM: The vertical axis frac-ture: A report on three cases. J Orthop Trauma 6:416–419, 1992.
62. Bohler J: An approach to non-union of fractures. Surg Annu14:299–315, 1982.
63. Bohlman HH: Acute fractures and dislocations of the cervicalspine: An analysis of three hundred hospitalized patients andreview of the literature. J Bone Joint Surg Am 61A:1119–1142,1979.
64. Bohn D, Armstrong D, Becker L, Humphreys R: Cervical spineinjuries in children. J Trauma 30:463–469, 1990.
65. Bolender N, Cromwell LD, Wendling L: Fracture of the occip-ital condyle. AJR Am J Roentgenol 131:729–731, 1978.
66. Bools JC, Rose BS: Traumatic atlantooccipital dislocation: Twocases with survival. AJNR Am J Neuroradiol 7:901–904, 1986.
67. Borne GM, Bedou GL, Pinaudeau M: Treatment of pedicularfractures of the axis: A clinical study and screw fixation tech-nique. J Neurosurg 60:88–93, 1984.
68. Borock EC, Gabram SG, Jacobs LM, Murphy MA: A prospectiveanalysis of a two-year experience using computed tomographyas an adjunct for cervical spine clearance. J Trauma 31:1001–1006, 1991.
69. Bosch A, Stauffer ES, Nickel VL: Incomplete traumatic quadri-plegia: A ten-year review. JAMA 216:473–478, 1971.
70. Bose B, Northrup BE, Osterholm JL, Cotler JM, DiTunno JF:Reanalysis of central cervical cord injury management. Neuro-surgery 15:367–372, 1984.
71. Botel U, Gläser E, Niedeggen A: The surgical treatment of acutespinal paralysed patients. Spinal Cord 35:420–428, 1997.
72. Botsford DJ, Esses SI: A new scale for the clinical assessment ofspinal cord function. Orthopedics 15:1309–1313, 1992.
73. Boyd CR, Corse KM, Campbell RC: Emergency interhospitaltransport of the major trauma patient: Air versus ground.J Trauma 29:789–794, 1989.
74. Bozboga M, Unal F, Hepgul K, Izgi N, Turantan MI, Turker K:Fracture of the occipital condyle: Case report. Spine 17:1119–1121, 1992.
75. Bracken MB: National Acute Spinal Cord Injury Study of meth-ylprednisolone or naloxone. Neurosurgery 28:628–629, 1991(letter).
76. Bracken MB: Methylprednisolone and spinal cord injury.J Neurosurg 93:175–177, 2000 (letter).
77. Bracken MB: The use of methylprednisolone. J Neurosurg 93:340–341, 2000 (letter).
78. Bracken MB: High dose methylprednisolone must be given for24 or 48 hours after acute spinal cord injury. BMJ 322:862–863,2001 (letter).
78a. Bracken MB: Methylprednisolone and acute spinal cord injury:An update of the randomized evidence. Spine 26(24 Suppl):S47–S54, 2001.
79. Bracken MB: Pharmacological interventions for acute spinalcord injury. Cochrane Database Syst Rev 1:1–32, 2001.
80. Bracken MB, Holford TR: Effects of timing of methylpred-nisolone or naloxone administration on recovery of segmentaland long-tract neurological function in NASCIS 2. J Neurosurg79:500–507, 1993.
81. Bracken MB, Holford TR: Response: Treatment of spinal cordinjury. J Neurosurg 80:954–955, 1994 (letter).
S180 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
82. Bracken MB, Aldrich EF, Herr DL, Hitchon PW, Holford TR,Marshall LF, Nockels RP, Pascale V, Shepard MJ, Sonntag VKH,Winn HR, Young W: Clinical measurement, statistical analysis,and risk-benefit: Controversies from trials of spinal injury.J Trauma 48:558–561, 2000.
83. Bracken MB, Collins WF, Freeman DF, Shepard MJ, WagnerFW, Silten RM, Hellenbrand KG, Ransohoff J, Hunt WE, PerotPL Jr, Grossman RG, Green BA, Eisenberg HM, Rifkinson N,Goodman JH, Meagher JN, Fischer B, Clifton GL, Flamm ES,Rawe SE: Efficacy of methylprednisolone in acute spinal cordinjury. JAMA 251:45–52, 1984.
84. Bracken MB, Shepard MJ, Collins WF, Holford TR, Baskin DS,Flamm E, Eisenberg HM, Leo-Summers L, Maroon JC, MarshallLF, Perot PL Jr, Piepmeier J, Sonntag VKH, Wagner FC Jr,Wilberger JL, Winn HR, Young W: Response: Methylpred-nisolone for spinal cord injury. J Neurosurg 77:325–327, 1992(letter).
85. Bracken MB, Shepard MJ, Collins WF Jr, Holford TR, BaskinDS, Eisenberg HM, Flamm E, Leo-Summers L, Maroon JC,Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, WagnerFC Jr, Wilberger JL, Winn HR, Young W: Methylprednisoloneor naloxone treatment after acute spinal cord injury: 1-yearfollow-up data—Results of the Second National Acute SpinalCord Injury Study. J Neurosurg 76:23–31, 1992.
86. Bracken MB, Shepard MJ, Collins WF, Holford TR, Young W,Baskin DS, Eisenberg HM, Flamm E, Leo-Summers L, MaroonJ, Marshall LF, Perot PL Jr, Piepmeier J, Sonntag VKH, WagnerFC, Wilberger JE, Winn HR: A randomized, controlled trial ofmethylprednisolone or naloxone in the treatment of acutespinal-cord injury: Results of the Second National Acute SpinalCord Injury Study. N Engl J Med 322:1405–1411, 1990.
87. Bracken MB, Shepard MJ, Hellenbrand KG, Collins WF, Leo LS,Freeman DF, Wagner FC, Flamm ES, Eisenberg HM, GoodmanJH, Perot PL Jr, Green BA, Grossman RG, Meagher JN, YoungW, Fischer B, Clifton GL, Hunt WE, Rifkinson N: Methylpred-nisolone and neurological function 1 year after spinal cordinjury: Results of the National Acute Spinal Cord Injury Study.J Neurosurg 63:704–713, 1985.
88. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings MG, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, SonntagVKH, Wagner F, Wilberger JE, Winn HR, Young W: Adminis-tration of methylprednisolone for 24 or 48 hours or tirilazadmesylate for 48 hours in the treatment of acute spinal cordinjury: Results of the Third National Acute Spinal Cord InjuryRandomized Controlled Trial—National Acute Spinal Cord In-jury Study. JAMA 277:1597–1604, 1997.
89. Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, AldrichEF, Fazl M, Fehlings MG, Herr DL, Hitchon PW, Marshall LF,Nockels RP, Pascale V, Perot PL Jr, Piepmeier JM, SonntagVKH, Wagner F, Wilberger JE, Winn HR, Young W: Methyl-prednisolone or tirilazad mesylate administration after acutespinal cord injury: 1-year follow up—Results of the Third Na-tional Acute Spinal Cord Injury randomized controlled trial.J Neurosurg 89:699–706, 1998.
90. Bracken MB, Webb SB Jr, Wagner FC: Classification of theseverity of acute spinal cord injury: Implications for manage-ment. Paraplegia 15:319–326, 1978.
91. Brady WJ, Moghtader J, Cutcher D, Exline C, Young J: ED useof flexion-extension cervical spine radiography in the evalua-tion of blunt trauma. Am J Emerg Med 17:504–508, 1999.
92. Brain Trauma Foundation, AANS: Management and prognosisof severe traumatic brain injury: Part I—Guidelines for themanagement of severe traumatic brain injury. J Neurotrauma17:449–554, 2000.
93. Brashear R Jr, Venters G, Preston ET: Fractures of the neuralarch of the axis: A report of twenty-nine cases. J Bone JointSurg Am 57A:879–887, 1975.
94. Bridgman SA, McNab W: Traumatic occipital condyle fracture,multiple cranial nerve palsies, and torticollis: A case report andreview of the literature. Surg Neurol 38:152–156, 1992.
95. Bridle MJ, Lynch KB, Quesenberry CM: Long-term functionfollowing the central cord syndrome. Paraplegia 28:178–185,1990.
96. Brockmeyer DL, York JE, Apfelbaum RI: Anatomical suitabilityof C1-2 transarticular screw placement in pediatric patients.J Neurosurg 92[Suppl 1]:7–11, 2000.
97. Brodkey JS, Miller CF Jr, Harmody RM: The syndrome of acutecentral cervical spinal cord injury revisited. Surg Neurol 14:251–257, 1980.
98. Brohi K, Wilson-Macdonald J: Evaluation of unstable cervicalspine injury: A 6-year experience. J Trauma 49:76–80, 2000.
99. Brooks AL, Jenkins EB: Atlanto-axial arthrodesis by the wedgecompression method. J Bone Joint Surg Am 60A:279–284, 1978.
100. Brown LH, Gough JE, Simonds WB: Can EMS providers ade-quately assess trauma patients for cervical spinal injury?Prehosp Emerg Care 2:33–36, 1998.
101. Bruce D: Spinal cord injury without radiographic abnormalityin children. Pediatr Neurosci 15:175, 1989 (comment).
102. Brunette DD, Rockswold GL: Neurologic recovery followingrapid spinal realignment for complete cervical spinal cord in-jury. J Trauma 27:445–447, 1987.
103. Bucci MN, Dauser RC, Maynard FA, Hoff JT: Management ofpost-traumatic cervical spine instability: Operative fusion ver-sus halo vest immobilization—Analysis of 49 cases. J Trauma28:1001–1006, 1988.
104. Bucholz RW: Unstable hangman’s fractures. Clin Orthop 154:119–124, 1981.
105. Bucholz RD, Cheung KC: Halo vest versus spinal fusion forcervical injury: Evidence from an outcome study. J Neurosurg70:884–892, 1989.
106. Buhs C, Cullen M, Klein M, Farmer D: The pediatric traumaC-spine: Is the “odontoid” view necessary? J Pediatr Surg35:994–997, 2000.
107. Bulas DI, Fitz CR, Johnson DL: Traumatic atlanto-occipital dis-location in children. Radiology 188:155–158, 1993.
107a. Bullock R, Chesnut RM, Clifton GL, Ghajar J, Marion DW,Narayan RK, Newell DW, Pitts LH, Rosner MJ, Walters BC,Wilberger JE: Guidelines for the management of severe trau-matic brain injury: Intracranial pressure treatment threshold.J Neurotrauma 17:493–495, 2000.
108. Bullock R, Chesnut RM, Clifton G, Ghajar J, Marion DW,Narayan RK, Newell DW, Pitts LH, Rosner MJ, Wilberger JW:Guidelines for the management of severe head injury: BrainTrauma Foundation. Eur J Emerg Med 3:109–127, 1996.
109. Bundschuh CV, Alley JB, Ross M, Porter IS, Gudeman SK:Magnetic resonance imaging of suspected atlanto-occipital dis-location: Two case reports. Spine 17:245–248, 1992.
110. Burke DC, Berryman D: The place of closed manipulation in themanagement of flexion-rotation dislocations of the cervicalspine. J Bone Joint Surg Br 53B:165–182, 1971.
111. Burke JT, Harris JH Jr: Acute injuries of the axis vertebra.Skeletal Radiol 18:335–346, 1989.
Bibliography S181
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
112. Burke DC, Tiong TS: Stability of the cervical spine after conser-vative treatment. Paraplegia 13:191–202, 1975.
113. Burney RE, Waggoner R, Maynard FM: Stabilization of spinalinjury for early transfer. J Trauma 29:1497–1499, 1989.
114. Burns GA, Cohn SM, Frumento RJ, Degutis LC, Hammers L:Prospective ultrasound evaluation of venous thrombosis inhigh-risk trauma patients. J Trauma 35:405–408, 1993.
115. Butman A, Vomacka R: Part 1: Spine immobilization. Emer-gency 23:48–51, 1991.
116. Cabanela ME, Ebersold MJ: Anterior plate stabilization forbursting teardrop fractures of the cervical spine. Spine 13:888–891, 1988.
117. Cahill DW, Bellegarrigue R, Ducker TB: Bilateral facet to spi-nous process fusion: A new technique for posterior spinal fu-sion after trauma. Neurosurgery 13:1–4, 1983.
118. Campanelli M, Kattner KA, Stroink A, Gupta K, West S: Pos-terior C1-C2 transarticular screw fixation in the treatment ofdisplaced type II odontoid fractures in the geriatric population:Review of seven cases. Surg Neurol 51:596–601, 1999.
119. Capaul M, Zollinger H, Satz N, Dietz V, Lehmann D, Schurch B:Analyses of 94 consecutive spinal cord injury patients usingASIA definition and modified Frankel score classification. Para-plegia 32:583–587, 1994.
120. Carey ME, Nance FC, Kirgis HD, Young HF, Megison LC Jr,Kline DG: Pancreatitis following spinal cord injury.J Neurosurg 47:917–922, 1977.
121. Carter VM, Fasen JA, Roman JM Jr, Hayes KW, Petersen CM:The effect of a soft collar, used as normally recommended orreversed, on three planes of cervical range of motion. J OrthopSports Phys Ther 23:209–215, 1996.
122. Casas ER, Sanchez MP, Arias CR, Masip JP: Prophylaxis ofvenous thrombosis and pulmonary embolism in patients withacute traumatic spinal cord lesions. Paraplegia 15:209–214,1977.
123. Castillo M, Mukherji SK: Vertical fractures of the dens. AJNRAm J Neuroradiol 17:1627–1630, 1996.
124. Castling B, Hicks K: Traumatic isolated unilateral hypoglossalnerve palsy: Case report and review of the literature. Br J OralMaxillofac Surg 33:171–173, 1995.
125. Cattell HS, Filtzer DL: Pseudosubluxation and other normalvariations in the cervical spine in children: A study of onehundred and sixty children. J Bone Joint Surg Am 47A:1295–1309, 1965.
126. Catz A, Itzkovich M, Agranov E, Ring H, Tamir A: SCIM:Spinal cord independence measure—A new disability scale forpatients with spinal cord lesions. Spinal Cord 35:850–856, 1997.
127. Chan D, Goldberg RM, Mason J, Chan L: Backboard versusmattress splint immobilization: A comparison of symptomsgenerated. J Emerg Med 14:293–298, 1996.
128. Chan D, Goldberg R, Tascone A, Harmon S, Chan L: The effectof spinal immobilization on healthy volunteers. Ann EmergMed 23:48–51, 1994.
129. Chan RC, Schweigel JF, Thompson GB: Halo-thoracic braceimmobilization in 188 patients with acute cervical spine inju-ries. J Neurosurg 58:508–515, 1983.
130. Chandler DR, Nemejc C, Adkins RH, Waters RL: Emergencycervical-spine immobilization. Ann Emerg Med 21:1185–1188,1992.
131. Chee SG: Review of the role of magnetic resonance imaging inacute cervical spine injuries. Ann Acad Med Singapore 22:757–761, 1993.
132. Chehrazi B, Wagner FC Jr, Collins WF Jr, Freeman DH Jr: Ascale for evaluation of spinal cord injury. J Neurosurg 54:310–315, 1981.
132a. Chen D, Apple DF, Hudson LM, Bode R: Medical complicationsduring acute rehabilitation following spinal cord injury: Cur-rent experience of the Model Systems. Arch Phys Med Rehabil80:1397–1401, 1999.
133. Chen TY, Dickman CA, Eleraky M, Sonntag VKH: The role ofdecompression for acute incomplete cervical spinal cord injuryin cervical spondylosis. Spine 23:2398–2403, 1998.
134. Chen TY, Lee ST, Lui TN, Wong CW, Yeh YS, Tzaan WC, HungSY: Efficacy of surgical treatment in traumatic central cordsyndrome. Surg Neurol 48:435–441, 1997.
135. Cheshire DJ: The stability of the cervical spine following theconservative treatment of fractures and fracture-dislocations.Paraplegia 7:193–203, 1969.
136. Cheshire DJ: A classification of the functional end-results ofinjury to the cervical spinal cord. Paraplegia 8:70–73, 1970.
137. Chesnut RM, Marshall LF, Klauber MR, Blunt BA, Baldwin N,Eisenberg HM, Jane JA, Marmarou A, Foulkes MA: The role ofsecondary brain injury in determining outcome from severehead injury. J Trauma 34:216–222, 1993.
138. Chiba K, Fujimura Y, Toyama Y, Fujii E, Nakanishi T,Hirabayashi K: Treatment protocol for fractures of the odontoidprocess. J Spinal Disord 9:267–276, 1996.
139. Chu DA, Ahn JH, Ragnarsson KT, Helt J, Folcarelli P, RamirezA: Deep venous thrombosis: Diagnosis in spinal cord injuredpatients. Arch Phys Med Rehabil 66:365–368, 1985.
140. Clancy MJ: Clearing the cervical spine of adult victims oftrauma. J Accid Emerg Med 16:208–214, 1999.
141. Clark CR, White AA III: Fractures of the dens: A multicenterstudy. J Bone Joint Surg Am 67A:1340–1348, 1985.
142. Clarke KS: Caloric costs of activity in paraplegic persons. ArchPhys Med Rehabil 47:427–435, 1966.
143. Clements WD, Mezue W, Mathew B: Os odontoideum: Congenital oracquired?—That’s not the question. Injury 26:640–642, 1995.
144. Cline JR, Scheidel E, Bigsby EF: A comparison of methods ofcervical immobilization used in patient extrication and trans-port. J Trauma 25:649–653, 1985.
145. Cloward RB: Reduction of traumatic dislocation of the cervicalspine with locked facets: Technical note. J Neurosurg 38:527–531, 1973.
146. Cohen J: A coefficient of agreement for nominal scales. EducPsychol Meas 20:37–46, 1960.
147. Cohen A, Bosshard R, Yeo JD: A new device for the care ofacute spinal injuries: The Russell Extrication Device (RED).Paraplegia 28:151–157, 1990.
148. Cohen ME, Ditunno JF Jr, Donovan WH, Maynard FM Jr: A testof the 1992 International Standards for Neurological and Func-tional Classification of Spinal Cord Injury. Spinal Cord 36:554–560, 1998.
149. Cohen ME, Sheehan TP, Herbison GJ: Content validity andreliability of the International Standards for Neurological Clas-sification of Spinal Cord Injury. Top Spinal Cord Inj Rehabil1:15–31, 1996.
150. Cohn SM, Lyle WG, Linden CH, Lancey RA: Exclusion of cervicalspine injury: A prospective study. J Trauma 31:570–574, 1991.
151. Coleman WP, Benzel D, Cahill DW, Ducker T, Geisler F, GreenB, Gropper MR, Goffin J, Madsen PW III, Maiman DJ, OndraSL, Rosner M, Sasso RC, Trost GR, Zeidman S: A critical ap-praisal of the reporting of the National Acute Spinal CordInjury Studies (II and III) of methylprednisolone in acute spinalcord injury. J Spinal Disord 13:185–199, 2000.
S182 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
152. Collalto PM, DeMuth WW, Schwentker EP, Boal DK: Traumaticatlanto-occipital dislocation: Case report. J Bone Joint Surg Am68A:1106–1109, 1986.
153. Colnet G, Chabannes J, Commun C, Rigal MC, Alassaf M:Atlanto-occipital luxation and syringomyelia: 2 rare complica-tions of cervical injury—Diagnostic and therapeutic effects:Apropos of a case [in French]. Neurochirurgie 35:58–63, 1989.
154. Cone DC, Wydro GC, Mininger CM: Current practice in clinicalcervical spinal clearance: Implication for EMS. Prehosp EmergCare 3:42–46, 1999.
155. Consortium for Spinal Cord Medicine: Prevention of thrombo-embolism in spinal cord injury. J Spinal Cord Med 20:259–283,1997.
156. Contostavlos DL: Massive subarachnoid hemorrhage due tolaceration of the vertebral artery associated with fracture of thetransverse process of the atlas. J Forensic Sci 16:40–56, 1971.
157. Cooke M: Spinal boards. J Accid Emerg Med 13:433, 1996(letter).
158. Cooper PR, Maravilla KR, Sklar FH, Moody SF, Clark WK: Haloimmobilization of cervical spine fractures: Indications and re-sults. J Neurosurg 50:603–610, 1979.
159. Coric D, Wilson JA, Kelly DL Jr: Treatment of traumatic spon-dylolisthesis of the axis with nonrigid immobilization: A re-view of 64 cases. J Neurosurg 85:550–554, 1996.
160. Cotler HB, Cotler JM, Alden ME, Sparks G, Biggs CA: Themedical and economic impact of closed cervical spine disloca-tions. Spine 15:448–452, 1990.
161. Cotler JM, Herbison GJ, Nasuti JF, Ditunno JF Jr, An H, WolffBE: Closed reduction of traumatic cervical spine dislocationusing traction weights up to 140 pounds. Spine 18:386–390,1993.
162. Cotler HB, Miller LS, DeLucia FA, Cotler JM, Davne SH: Closedreduction of cervical spine dislocations. Clin Orthop 214:185–199, 1987.
163. Cottalorda J, Allard D, Dutour N: Fracture of the occipitalcondyle. J Pediatr Orthop B 5:61–63, 1996.
164. Cox SA, Weiss SM, Posuniak EA, Worthington P, Prioleau M,Heffley G: Energy expenditure after spinal cord injury: Anevaluation of stable rehabilitating patients. J Trauma 25:419–423, 1985.
165. Coyne TJ, Fehlings MG, Wallace MC, Bernstein M, Tator CH:C1-C2 posterior cervical fusion: Long-term evaluation of resultsand efficacy. Neurosurgery 37:688–693, 1995.
166. Craig JB, Hodgson BF: Superior facet fractures of the axisvertebra. Spine 16:875–877, 1991.
167. Cruse JM, Lewis RE, Dilioglou S, Roe DL, Wallace WF, ChenRS: Review of immune function, healing of pressure ulcers, andnutritional status in patients with spinal cord injury. J SpinalCord Med 23:129–135, 2000.
168. Cruse JM, Lewis RE, Roe DL, Dilioglou S, Blaine MC, WallaceWF, Chen RS: Facilitation of immune function, healing of pres-sure ulcers, and nutritional status in spinal cord injury patients.Exp Mol Pathol 68:38–54, 2000.
169. Crutchfield W: Skeletal traction in treatment of injuries to thecervical spine. JAMA 155:29–32, 1954.
170. Curran C, Dietrich AM, Bowman MJ, Ginn-Pease ME, King DR,Kosnik E: Pediatric cervical-spine immobilization: Achievingneutral position? J Trauma 39:729–732, 1995.
171. Curri D, Cervellini P, Zanusso M, Benedetti A: Isolated fractureof occipital condyle: Case report. J Neurosurg Sci 32:157–159,1988.
172. Cybulski GR, Douglas RA, Meyer PR Jr, Rovin RA: Complica-tions in three-column cervical spine injuries requiring anterior-posterior stabilization. Spine 17:253–256, 1992.
173. Dai L, Jia L: Central cord injury complicating acute cervical discherniation in trauma. Spine 25:331–336, 2000.
174. Dai L, Yuan W, Ni B, Jia L: Os odontoideum: Etiology, diagno-sis, and management. Surg Neurol 53:106–109, 2000.
175. Dai LY, Yuan W, Ni B, Liu HK, Jia LS, Zhao DL, Xu YK:Surgical treatment of nonunited fractures of the odontoid pro-cess, with special reference to occipitocervical fusion forunreducible atlantoaxial subluxation or instability. Eur Spine J9:118–122, 2000.
176. D’Alise MD, Benzel EC, Hart BL: Magnetic resonance imagingevaluation of the cervical spine in the comatose or obtundedtrauma patient. J Neurosurg 91[Suppl 1]:54–59, 1999.
177. Davies G, Deakin C, Wilson A: The effect of a rigid collar onintracranial pressure. Injury 27:647–649, 1996.
178. Davis D, Bohlman H, Walker AE, Fisher R, Robinson R: Thepathological findings in fatal craniospinal injuries. J Neurosurg34:603–613, 1971.
179. Davis JW, Parks SN, Detlefs CL, Williams GG, Williams JL,Smith RW: Clearing the cervical spine in obtunded patients:The use of dynamic fluoroscopy. J Trauma 39:435–438, 1995.
180. Davis JW, Phreaner DL, Hoyt DB, Mackersie RC: The etiologyof missed cervical spine injuries. J Trauma 34:342–346, 1993.
181. Davis PC, Reisner A, Hudgins PA, Davis WE, O’Brien MS:Spinal injuries in children: Role of MR. AJNR Am JNeuroradiol 14:607–617, 1993.
182. Davis LA, Warren SA, Reid DC, Oberle K, Saboe LA, GraceMG: Incomplete neural deficits in thoracolumbar and lumbarspine fractures: Reliability of Frankel and Sunnybrookscales. Spine 18:257–263, 1993.
183. Deeb ZL, Rothfus WE, Goldberg AL, Daffner RH: Occult oc-cipital condyle fractures presenting as tumors. J ComputTomogr 12:261–263, 1988.
184. Della Torre P, Rinonapoli E: Halo-cast treatment of fracturesand dislocations of the cervical spine. Int Orthop 16:227–231,1992.
185. De Lorenzo RA: A review of spinal immobilization techniques.J Emerg Med 14:603–613, 1996.
186. De Lorenzo RA, Olson JE, Boska M, Johnston R, Hamilton GC,Augustine J, Barton R: Optimal positioning for cervical immo-bilization. Ann Emerg Med 28:301–308, 1996.
187. Demisch S, Lindner A, Beck R, Zierz S: The forgotten condyle:Delayed hypoglossal nerve palsy caused by fracture of theoccipital condyle. Clin Neurol Neurosurg 100:44–45, 1998.
188. Desai SS, Coumas JM, Danylevich A, Hayes E, Dunn EJ: Frac-ture of the occipital condyle: Case report and review of theliterature. J Trauma 30:240–241, 1990.
189. De Smet L, Vercauteren M, Verdonk R, Claessens H: Severeacute cervical spine injuries: Conservative treatment. ActaOrthop Belg 50:512–520, 1984.
190. DeVivo MJ, Kartus PL, Stover SL, Rutt RD, Fine PR: Cause ofdeath for patients with spinal cord injuries. Arch Intern Med149:1761–1766, 1989.
191. Dibenedetto T, Lee CK: Traumatic atlanto-occipital instability:A case report with follow-up and a new diagnostic technique.Spine 15:595–597, 1990.
192. Dick T: Comparing the short-board technique. Ann Emerg Med18:115–116, 1989 (letter).
193. Dick T, Land R: Spinal immobilization devices: Part 1—Cervi-cal extrication collars. J Emerg Med Serv JEMS 7:26–32, 1982.
Bibliography S183
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
194. Dick T, Land R: Full spinal immobilizers. J Emerg Med ServJEMS 8:34–36, 1983.
195. Dickman CA, Sonntag VKH: Surgical management of atlanto-axial nonunions. J Neurosurg 83:248–253, 1995.
196. Dickman C, Sonntag VKH: Injuries involving the transverseatlantal ligament: Classification and treatment guidelines basedupon experience with 39 injuries. Neurosurgery 40:886–887,1997 (letter).
197. Dickman CA, Sonntag VKH: Posterior C1–C2 transarticularscrew fixation for atlantoaxial arthrodesis. Neurosurgery 43:275–281, 1998.
197a. Dickman CA, Foley KT, Sonntag VKH, Smith MM: Cannulatedscrews for odontoid fixation and atlantoaxial transarticularscrew fixation. J Neurosurg 83:1095–1100, 1995.
198. Dickman CA, Greene KA, Sonntag VKH: Injuries involving thetransverse atlantal ligament: Classification and treatmentguidelines based upon experience with 39 injuries. Neuro-surgery 38:44–50, 1996.
199. Deleted in proof.200. Dickman CA, Papadopoulos SM, Sonntag VKH, Spetzler RF,
Rekate HL, Drabier J: Traumatic occipitoatlantal dislocations.J Spinal Disord 6:300–313, 1993.
201. Dickman CA, Rekate HL, Sonntag VKH, Zabramski JM: Pedi-atric spinal trauma: Vertebral column and spinal cord injuriesin children. Pediatr Neurosci 15:237–256, 1989.
202. Dickman CA, Sonntag VKH, Papadopoulos SM, Hadley MN:The interspinous method of posterior atlantoaxial arthrodesis.J Neurosurg 74:190–198, 1991.
203. Ditunno JF Jr: Functional assessment measures in CNS trauma.J Neurotrauma 9[Suppl 1]:S301–S305, 1992.
204. Ditunno JF Jr: New spinal cord injury standards, 1992. Paraple-gia 30:90–91, 1992.
205. Ditunno JF Jr: American spinal injury standards for neurolog-ical and functional classification of spinal cord injury: Past,present and future—1992 Heiner Sell Lecture of the AmericanSpinal Injury Association. J Am Paraplegia Soc 17:7–11, 1994.
206. Ditunno JF Jr, Ditunno PL, Graziani V, Scivoletto G, Bernardi M,Castellano V, Marchetti M, Barbeau H, Frankel HL, D’AndreaGreve JM, Ko H Y, Marshall R, Nance P: Walking Index for SpinalCord Injury (WISCI): An international multicenter validity andreliability study. Spinal Cord 38:234–243, 2000.
207. Ditunno JF Jr, Young W, Donovan WH, Creasey G: The Inter-national Standards Booklet for Neurological and FunctionalClassification of Spinal Cord Injury: American Spinal InjuryAssociation. Paraplegia 32:70–80, 1994.
208. Dodd FM, Simon E, McKeown D, Patrick MR: The effect of acervical collar on the tidal volume of anaesthetised adult pa-tients. Anaesthesia 50:961–963, 1995.
209. Dodds TA, Martin DP, Stolov WC, Deyo RA: A validation ofthe functional independence measurement and its performanceamong rehabilitation inpatients. Arch Phys Med Rehabil 74:531–536, 1993.
210. Dolan EJ, Tator CH: The effect of blood transfusion, dopamine,and gamma hydroxybutyrate on posttraumatic ischemia of thespinal cord. J Neurosurg 56:350–358, 1982.
211. Domeier RM: Indications for prehospital spinal immobilization:National Association of EMS Physicians Standards and ClinicalPractice Committee. Prehosp Emerg Care 3:251–253, 1999.
212. Domeier RM, Evans RW, Swor RA, Hancock JB, Fales W,Krohmer J, Frederiksen SM, Shork MA: The reliability ofprehospital clinical evaluation for potential spinal injury is notaffected by the mechanism of injury. Prehosp Emerg Care3:332–337, 1999.
213. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Frederiksen SM: High-risk criteria for performing pre-hospitalspinal immobilization in trauma. Ann Emerg Med 25:141–142,1995.
214. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Frederiksen SM: Prehospital clinical findings associated withspinal injury. Prehosp Emerg Care 1:11–15, 1997.
215. Domeier RM, Evans RW, Swor RA, Rivera-Rivera EJ,Frederiksen SM: Prospective validation of out-of-hospital spi-nal clearance criteria: A preliminary report. Acad Emerg Med4:643–646, 1997.
216. Donahue DJ, Muhlbauer MS, Kaufman RA, Warner WC,Sanford RA: Childhood survival of atlantooccipital dislocation:Underdiagnosis, recognition, treatment, and review of the lit-erature. Pediatr Neurosurg 21:105–111, 1994.
217. Donovan WH, Kopaniky D, Stolzmann E, Carter RE: The neu-rological and skeletal outcome in patients with closed cervicalspinal cord injury. J Neurosurg 66:690–694, 1987.
218. Doran SE, Papadopoulos SM, Ducker TB, Lillehei KO: Magneticresonance imaging documentation of coexistent traumaticlocked facets of the cervical spine and disc herniation.J Neurosurg 79:341–345, 1993.
219. Dormans JP, Criscitiello AA, Drummond DS, Davidson RS:Complications in children managed with immobilization in ahalo vest. J Bone Joint Surg Am 77A:1370–1373, 1995.
220. Dorr LD, Harvey JP Jr, Nickel VL: Clinical review of the earlystability of spine injuries. Spine 7:545–550, 1982.
221. Dublin AB, Marks WM, Weinstock D, Newton TH: Traumaticdislocation of the atlanto-occipital articulation (AOA) withshort-term survival: With a radiographic method of measuringthe AOA. J Neurosurg 52:541–546, 1980.
222. Ducker TB, Zeidman SM: Spinal cord injury: Role of steroidtherapy. Spine 19:2281–2287, 1994.
223. Ducker TB, Kindt GW, Kempf LG: Pathological findings inacute experimental spinal cord trauma. J Neurosurg 35:700–708, 1971.
224. Duh MS, Shepard MJ, Wilberger JE, Bracken MB: The effective-ness of surgery on the treatment of acute spinal cord injury andits relation to pharmacological treatment. Neurosurgery 35:240–249, 1994.
225. Dunn ME, Seljeskog EL: Experience in the management ofodontoid process injuries: An analysis of 128 cases. Neuro-surgery 18:306–310, 1986.
226. Dyck P: Os odontoideum in children: Neurological manifesta-tions and surgical management. Neurosurgery 2:93–99, 1978.
227. Effendi B, Roy D, Cornish B, Dussault RG, Laurin CA: Fracturesof the ring of the axis: A classification based on the analysis of131 cases. J Bone Joint Surg Br 63B:319–327, 1981.
228. Eismont FJ, Bohlman HH: Posterior atlanto-occipital disloca-tion with fractures of the atlas and odontoid process. J BoneJoint Surg Am 60A:397–399, 1978.
229. Eismont FJ, Arena MJ, Green BA: Extrusion of an intervertebraldisc associated with traumatic subluxation or dislocation ofcervical facets: Case report. J Bone Joint Surg Am 73A:1555–1560.
230. Ekong CE, Schwartz ML, Tator CH, Rowed DW, Edmonds VE:Odontoid fracture: Management with early mobilization usingthe halo device. Neurosurgery 9:631–637, 1981.
231. Eleraky MA, Theodore N, Adams M, Rekate HL, Sonntag VKH:Pediatric cervical spine injuries: Report of 102 cases and reviewof the literature. J Neurosurg 92[Suppl 1]:12–17, 2000.
S184 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
232. El-Khoury GY, Clark CR, Gravett AW: Acute traumatic rota-tory atlanto-axial dislocation in children: A report of threecases. J Bone Joint Surg Am 66A:774–777, 1984.
233. Elliott JM Jr, Rogers LF, Wissinger JP, Lee JF: The hangman’sfracture: Fractures of the neural arch of the axis. Radiology104:303–307, 1972.
234. El Masri WS, Silver JR: Prophylactic anticoagulant therapy inpatients with spinal cord injury. Paraplegia 19:334–342, 1981.
235. El Masry WS, Tsubo M, Katoh S, El Miligui YH, Khan A:Validation of the American Spinal Injury Association (ASIA)motor score and the National Acute Spinal Cord Injury Study(NASCIS) motor score. Spine 21:614–619, 1996.
236. Emery SE, Pathria MN, Wilber RG, Masaryk T, Bohlman HH:Magnetic resonance imaging of posttraumatic spinal ligamentinjury. J Spinal Disord 2:229–233, 1989.
237. Emery E, Saillant G, Ismail M, Fohanno D, Roy-Camille R:Fracture of the occipital condyle: Case report and review of theliterature. Eur Spine J 4:191–193, 1995.
238. Ersmark H, Kalen R: A consecutive series of 64 halo-vest-treated cervical spine injuries. Arch Orthop Trauma Surg 105:243–246, 1986.
239. Ersmark H, Lowenhielm P: Factors influencing the outcome ofcervical spine injuries. J Trauma 28:407–410, 1988.
240. Esses SI, Bednar DA: Screw fixation of odontoid fractures andnonunions. Spine 16[Suppl 10]:S483–S485, 1991.
241. Evans D: Reduction of cervical dislocations. J Bone Joint SurgBr 43B:552–555, 1961.
242. Evans DL, Bethem D: Cervical spine injuries in children.J Pediatr Orthop 9:563–568, 1989.
243. Evarts CM: Traumatic occipito-atlantal dislocation. J Bone JointSurg Am 52:1653–1660, 1970.
244. Farley FA, Graziano GP, Hensinger RN: Traumatic atlanto-occipital dislocation in a child. Spine 17:1539–1541, 1992.
245. Farmer J, Vaccaro A, Albert TJ, Malone S, Balderston RA, CotlerJM: Neurologic deterioration after cervical spinal cord injury.J Spinal Disord 11:192–196, 1998.
246. Farrington JD: Death in a ditch. Bull Am Coll Surg 52:121–130,1967.
247. Farrington JD: Extrication of victims: Surgical principles.J Trauma 8:493–512, 1968.
248. Farthing JW: Atlantooccipital dislocation with survival: A casereport. N C Med J 9:34–36, 1948.
249. Fehlings MG, Tator CH: An evidence-based review of decom-pressive surgery in acute spinal cord injury: Rationale, indica-tions, and timing based on experimental and clinical studies.J Neurosurg 91[Suppl 1]:1–11, 1999.
250. Fehlings MG, Cooper PR, Errico TJ: Posterior plates in themanagement of cervical instability: Long-term results in 44patients. J Neurosurg 81:341–349, 1994.
251. Fehlings MG, Rao SC, Tator CH, Skaf G, Arnold P, Benzel EC,Dickman C, Cuddy B, Green B, Hitchon P, Northrup B, SonntagVKH, Wagner F, Wilberger JE: The optimal radiologic methodfor assessing spinal canal compromise and cord compression inpatients with cervical spinal cord injury: Part II—Results of amulticenter study. Spine 24:605–613, 1999.
252. Fenstermaker RA: Acute neurologic management of the patientwith spinal cord injury. Urol Clin North Am 20:413–421, 1993.
253. Ferrera PC, Bartfield JM: Traumatic atlanto-occipital dislocation: Apotentially survivable injury. Am J Emerg Med 14:291–296, 1996.
254. Field MJ, Lohr KN (eds): Clinical Practice Guidelines: Directionsfor a New Program—Committee to Advise the Public Health Serviceon Clinical Practice Guidelines: Institute of Medicine. WashingtonDC, National Academy Press, 1990.
255. Field-Fote EC, Fluet GG, Schafer SD, Schneider EM, Smith R,Downey PA, Ruhl CD: The Spinal Cord Injury FunctionalAmbulation Inventory (SCI-FAI). J Rehabil Med 33:177–181,2001.
256. Fielding JW, Hawkins RJ: Atlanto-axial rotatory fixation: Fixedrotatory subluxation of the atlanto-axial joint. J Bone Joint SurgAm 59A:37–44, 1977.
257. Fielding JW, Cochran GV, Lawsing JF III, Hohl M: Tears of thetransverse ligament of the atlas: A clinical and biomechanicalstudy. J Bone Joint Surg Am 56A:1683–1691, 1974.
258. Fielding JW, Francis WR Jr, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. Clin Orthop 239:47–52,1989.
259. Fielding JW, Hensinger RN, Hawkins RJ: Os odontoideum.J Bone Joint Surg Am 62A:376–383, 1980.
260. Finch GD, Barnes MJ: Major cervical spine injuries in childrenand adolescents. J Pediatr Orthop 18:811–814, 1998.
261. Flanders AE, Schaefer DM, Doan HT, Mishkin MM, GonzalezCF, Northrup BE: Acute cervical spine trauma: Correlation ofMR imaging findings with degree of neurologic deficit. Radi-ology 177:25–33, 1990.
262. Flohr H, Pöll W, Brock M: Regulation of spinal cord blood flow,in Russell RWR (ed): Brain and Blood Flow: Proceedings of the 4thInternational Symposium on the Regulation of Cerebral Blood Flow.London, Pitman Medical, 1971, pp 406–409.
263. Fotter R, Sorantin E, Schneider U, Ranner G, Fast C, Schober P:Ultrasound diagnosis of birth-related spinal cord trauma: Neo-natal diagnosis and follow-up and correlation with MRI.Pediatr Radiol 24:241–244, 1994.
264. Fowler JL, Sandhu A, Fraser RD: A review of fractures of theatlas vertebra. J Spinal Disord 3:19–24, 1990.
265. Francis WR, Fielding JW, Hawkins RJ, Pepin J, Hensinger R:Traumatic spondylolisthesis of the axis. J Bone Joint Surg Br63B:313–318, 1981.
266. Frankel HL, Hancock DO, Hyslop G, Melzak J, Michaelis LS,Ungar GH, Vernon JD, Walsh JJ: The value of postural reduc-tion in the initial management of closed injuries of the spinewith paraplegia and tetraplegia: Part I. Paraplegia 7:179–192,1969.
267. Frankel H, Michaelis L, Paeslack V, Ungar G, Walsh JJ: Closedinjuries of the cervical spine and spinal cord: Results of conser-vative treatment of vertical compression injuries of the cervicalspine. Proc Veterans Adm Spinal Cord Inj Conf 19:28–32,1973.
268. Frankenfield DC, Smith JS, Cooney RN: Accelerated nitrogenloss after traumatic injury is not attenuated by achievement ofenergy balance. JPEN J Parenter Enteral Nutr 21:324–329, 1997.
269. Freemyer B, Knopp R, Piche J, Wales L, Williams J: Comparisonof five-view and three-view cervical spine series in the evalu-ation of patients with cervical trauma. Ann Emerg Med 18:818–821, 1989.
270. French HG, Burke SW, Roberts JM, Johnston CE II, WhitecloudT, Edmunds JO: Upper cervical ossicles in Down syndrome.J Pediatr Orthop 7:69–71, 1987.
271. Friedman D, Flanders A, Thomas C, Millar W: Vertebral arteryinjury after acute cervical spine trauma: Rate of occurrence asdetected by MR angiography and assessment of clinical conse-quences. AJR Am J Roentgenol 164:443–449, 1995.
272. Frisbie JH, Sasahara AA: Low dose heparin prophylaxis fordeep venous thrombosis in acute spinal cord injury patients: Acontrolled study. Paraplegia 19:343–346, 1981.
Bibliography S185
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
273. Frisbie JH, Sharma GV: Pulmonary embolism manifesting asacute disturbances of behavior in patients with spinal cordinjury. Paraplegia 32:570–572, 1994.
274. Frohna WJ: Emergency department evaluation and treatment ofthe neck and cervical spine injuries. Emerg Med Clin NorthAm 17:739–791, 1999.
275. Fruin AH, Pirotte TP: Traumatic atlantooccipital dislocation:Case report. J Neurosurg 46:663–666, 1977.
276. Fujii E, Kobayashi K, Hirabayashi K: Treatment in fractures ofthe odontoid process. Spine 13:604–609, 1988.
277. Fujimura Y, Nishi Y, Chiba K, Kobayashi K: Prognosis of neu-rological deficits associated with upper cervical spine injuries.Paraplegia 33:195–202, 1995.
278. Fujimura Y, Nishi Y, Kobayashi K: Classification and treatmentof axis body fractures. J Orthop Trauma 10:536–540, 1996.
279. Gabrielsen TO, Maxwell JA: Traumatic atlanto-occipital dislo-cation: With case report of a patient who survived. Am JRoentgenol Radium Ther Nucl Med 97:624–629, 1966.
280. Galandiuk S, Raque G, Appel S, Polk HC Jr: The two-edgedsword of large-dose steroids for spinal cord trauma. Ann Surg218:419–427, 1993.
281. Garber J: Abnormalities of the atlas and axis vertebrae: Con-genital and traumatic. J Bone Joint Surg Am 46A:1782–1791,1964.
282. Garfin SR, Shackford SR, Marshall LF, Drummond JC: Care ofthe multiply injured patient with cervical spine injury. ClinOrthop 239:19–29, 1989.
283. Garth G: Proposal for the establishment of minimum perfor-mance specifications for cervical extrication collars. Presentedat the 14th Annual Meeting of the American Society for Testingand Materials Skeletal Support Committee, WestConshohocken, PA, 1988.
284. Gaskill SJ, Marlin AE: Custom fitted thermoplastic Minervajackets in the treatment of cervical spine instability in preschoolage children. Pediatr Neurosurg 16:35–39, 1990.
285. Gaufin LM, Goodman SJ: Cervical spine injuries in infants:Problems in management. J Neurosurg 42:179–184, 1975.
286. Geerts WH, Code KI, Jay RM, Chen E, Szalai JP: A prospectivestudy of venous thromboembolism after major trauma. N EnglJ Med 331:1601–1606, 1994.
287. Geisler FH: Commentary on NASCIS-2. J Spinal Disord 5:132–133, 1992 (comment).
288. Geisler FH: GM-1 ganglioside and motor recovery followinghuman spinal cord injury. J Emerg Med 11[Suppl 1]:49–55,1993.
289. Geisler FH, Coleman WP, Grieco G, Dorsey FC, Poonian D, TheSygen Study Group: The GM-1 Ganglioside Multi-center AcuteSpinal Cord Injury Study. Spine 26(24 Suppl):S87–S98, 2001.
290. Geisler FH, Dorsey FC, Coleman WP: Correction: Recovery ofmotor function after spinal-cord injury—A randomized,placebo-controlled trial with GM-1 ganglioside. N Engl J Med325:1659–1660, 1991.
291. Geisler FH, Dorsey FC, Coleman WP: Recovery of motor func-tion after spinal-cord injury: A randomized, placebo-controlledtrial with GM-1 ganglioside. N Engl J Med 324:1829–1838, 1991.
292. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside inhuman spinal cord injury. J Neurotrauma 9[Suppl 2]:S517–S530, 1992.
293. Geisler FH, Dorsey FC, Coleman WP: GM-1 ganglioside forspinal cord injury. N Engl J Med 326:494, 1992 (letter).
294. Geisler WO, Wynne-Jones M, Jousse AT: Early management ofthe patient with trauma to the spinal cord. Med Serv J Can22:512–523, 1966.
295. George ER, Scholten DJ, Buechler CM, Jordan-Tibbs J, MatticeC, Albrecht RM: Failure of methylprednisolone to improve theoutcome of spinal cord injuries. Am Surg 61:659–664, 1995.
296. Georgopoulos G, Pizzutillo PD, Lee MS: Occipito-atlantal in-stability in children: A report of five cases and review of theliterature. J Bone Joint Surg Am 69A:429–436, 1987.
297. Gerhart KA, Johnson RL, Menconi J, Hoffman RE, LammertseDP: Utilization and effectiveness of methylprednisolone in apopulation-based sample of spinal cord injured persons. Para-plegia 33:316–321, 1995.
298. Gerndt SJ, Rodriguez JL, Pawlik JW, Taheri PA, Wahl WL,Micheals AJ, Papadopoulos SM: Consequences of high-dosesteroid therapy for acute spinal cord injury. J Trauma 42:279–284, 1997.
299. Gerrelts BD, Petersen EU, Mabry J, Petersen SR: Delayed diag-nosis of cervical spine injuries. J Trauma 31:1622–1626, 1991.
300. Giacobetti FB, Vaccaro AR, Bos-Giacobetti MA, Deeley DM,Albert TJ, Farmer JC, Cotler JM: Vertebral artery occlusionassociated with cervical spine trauma: A prospective analysis.Spine 22:188–192, 1997.
301. Givens TG, Polley KA, Smith GF, Hardin WD Jr: Pediatriccervical spine injury: A three-year experience. J Trauma 41:310–314, 1996.
302. Glaser JA, Whitehill R, Stamp WG, Jane JA: Complicationsassociated with the halo-vest: A review of 245 cases.J Neurosurg 65:762–769, 1986.
303. Gleizes V, Jacquot FP, Signoret F, Feron JM: Combined injuriesin the upper cervical spine: Clinical and epidemiological dataover a 14-year period. Eur Spine J 9:386–392, 2000.
304. Godard J, Hadji M, Raul JS: Odontoid fractures in the child withneurological injury: Direct anterior osteosynthesis with acortico-spongious screw and literature review. Childs NervSyst 13:105–107, 1997.
305. Goffin J, Plets C, Van den Bergh R: Anterior cervical fusion andosteosynthetic stabilization according to Caspar: A prospectivestudy of 41 patients with fractures and/or dislocations of thecervical spine. Neurosurgery 25:865–871, 1989.
306. Goldstein SJ, Woodring JH, Young AB: Occipital condyle frac-ture associated with cervical spine injury. Surg Neurol 17:350–352, 1982.
307. Gonzalez RP, Fried PO, Bukhalo M, Holevar MR, FalimirskiME: Role of clinical examination in screening for blunt cervicalspine injury. J Am Coll Surg 189:152–157, 1999.
308. Goth P: Spinal Injury: Clinical Criteria for Assessment and Man-agement. Augusta, Medical Care Development Publishing, 1994.
309. Govender S, Charles RW: Traumatic spondylolisthesis of theaxis. Injury 18:333–335, 1987.
310. Govender S, Grootboom M: Fractures of the dens: The results ofnon-rigid immobilization. Injury 19:165–167, 1988.
311. Grabb PA, Pang D: Magnetic resonance imaging in the evalu-ation of spinal cord injury without radiographic abnormality inchildren. Neurosurgery 35:406–414, 1994.
312. Grabb BC, Frye TA, Hedlund GL, Vaid YN, Grabb PA, RoyalSA: MRI diagnosis of suspected atlanto-occipital dissociation inchildhood. Pediatr Radiol 29:275–281, 1999.
313. Grady MS, Howard MA, Jane JA, Persing JA: Use of the Phil-adelphia collar as an alternative to the halo vest in patients withC-2, C-3 fractures. Neurosurgery 18:151–156, 1986.
314. Grahm TW, Zadrozny DB, Harrington T: The benefits of earlyjejunal hyperalimentation in the head-injured patient. Neuro-surgery 25:729–735, 1989.
S186 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
315. Grant GA, Mirza SK, Chapman JR, Winn HR, Newell DW,Jones DT, Grady MS: Risk of early closed reduction in cervicalspine subluxation injuries. J Neurosurg 90[Suppl 1]:13–18,1999.
316. Graziano AF, Scheidel EA, Cline JR, Baer LJ: A radiographiccomparison of prehospital cervical immobilization methods.Ann Emerg Med 16:1127–1131, 1987.
317. Green D, Chen D, Chmiel JS, Olsen NK, Berkowitz M, NovickA, Alleva J, Steinberg D, Nussbaum S, Tolotta M: Prevention ofthromboembolism in spinal cord injury: Role of low molecularweight heparin. Arch Phys Med Rehabil 75:290–292, 1994.
318. Green BA, Eismont FJ, O’Heir JT: Spinal cord injury: A systemsapproach—Prevention, emergency medical services, and emer-gency room management. Crit Care Clin 3:471–493, 1987.
319. Green D, Lee MY, Ito VY, Cohn T, Press J, Filbrandt PR,VandenBerg WC, Yarkony GM, Meyer PR Jr: Fixed- vsadjusted-dose heparin in the prophylaxis of thromboembolismin spinal cord injury. JAMA 260:1255–1258, 1988.
320. Green D, Lee MY, Lim AC, Chmiel JS, Vetter M, Pang T, ChenD, Fenton L, Yarkony GM, Meyer PR Jr: Prevention of throm-boembolism after spinal cord injury using low-molecular-weight heparin. Ann Intern Med 113:571–574, 1990.
321. Greene KA, Dickman CA, Marciano FF, Drabier J, Drayer BP,Sonntag VKH: Transverse atlantal ligament disruption associ-ated with odontoid fractures. Spine 19:2307–2314, 1994.
322. Greene KA, Dickman CA, Marciano FF, Drabier JB, HadleyMN, Sonntag VKH: Acute axis fractures: Analysis of manage-ment and outcome in 340 consecutive cases. Spine 22:1843–1852, 1997.
323. Greenfield LJ: Does cervical spinal cord injury induce a higherincidence of complications after prophylactic Greenfield filterusage? J Vasc Interv Radiol 8:719–720, 1997.
324. Gresham GE, Labi ML, Dittmar SS, Hicks JT, Joyce SZ, StehlikMA: The Quadriplegia Index of Function (QIF): Sensitivity andreliability demonstrated in a study of thirty quadriplegic pa-tients. Paraplegia 24:38–44, 1986.
325. Griffiths SC: Fracture of odontoid process in children. J PediatrSurg 7:680–683, 1972.
326. Grisolia A, Bell RL, Peltier LF: Fractures and dislocations of thespine complicating ankylosing spondylitis: A report of sixcases. J Bone Joint Surg Am 49A:339–344, 1967.
327. Griswold DM, Albright JA, Schiffman E, Johnson R, SouthwickW: Atlanto-axial fusion for instability. J Bone Joint Surg Am60A:285–292, 1978.
328. Grogaard B, Dullerud R, Magnaes B: Acute torticollis in chil-dren due to atlanto-axial rotary fixation. Arch Orthop TraumaSurg 112:185–188, 1993.
329. Grossman MD, Reilly PM, Gillett T, Gillett D: National surveyof the incidence of cervical spine injury and approach to cervi-cal spine clearance in U.S. trauma centers. J Trauma 47:684–690,1999.
330. Gschaedler R, Dollfus P, Molé JP, Molé L, Loeb JP: Reflectionson the intensive care of acute cervical spinal cord injuries in ageneral traumatology centre. Paraplegia 17:58–61, 1979.
331. Guigui P, Milaire M, Morvan G, Lassale B, Deburge A: Trau-matic atlantooccipital dislocation with survival: Case reportand review of the literature. Eur Spine J 4:242–247, 1995.
332. Guiot B, Fessler RG: Complex atlantoaxial fractures.J Neurosurg 91[Suppl 2]:139–143, 1999.
333. Gunby P: New focus on spinal cord injury. JAMA 245:1201–1206, 1981.
334. Gunduz S, Ogur E, Mohur H, Somuncu I, Acjksoz E, UstunsozB: Deep vein thrombosis in spinal cord injured patients. Para-plegia 31:606–610, 1993.
335. Gunn BD, Eizenberg N, Silberstein M, McMeeken JM, Tully EA,Stillman BC, Brown DJ, Gutteridge GA: How should an uncon-scious person with a suspected neck injury be positioned?Prehospital Disaster Med 10:239–244, 1995.
336. Gurdjian ES, Hardy WG, Lindner DW, Thomas LM: Closedcervical trauma associated with involvement of carotid andvertebral arteries. J Neurosurg 20:418–427, 1963.
337. Guttman L: Initial treatment of traumatic paraplegia and tetra-plegia, in Spinal Injuries: Proceedings of the Symposium held in theRoyal College of Surgeons of Edinburgh, June 7–8, 1963. Edinburgh,The Royal College of Surgeons, 1967.
338. Hachen HJ: Anticoagulant therapy in patients with spinal cordinjury. Paraplegia 12:176–187, 1974.
339. Hachen HJ: Emergency transportation in the event of acutespinal cord lesion. Paraplegia 12:33–37, 1974.
340. Hachen HJ: Idealized care of the acutely injured spinal cord inSwitzerland. J Trauma 17:931–936, 1977.
341. Hadley MN: Hypermetabolism after CNS trauma: Arrestingthe “injury cascade.” Nutrition 5:143, 1989.
342. Hadley MN: Hypermetabolism following head trauma: Nutri-tional considerations, in Barrow DL (ed): Complications andSequelae of Head Injury (Neurosurgical Topics series). Park Ridge,AANS, 1992, pp 161–168.
343. Hadley MN, Argires PJ. The acute/emergent management ofvertebral column fracture dislocation injuries, in NeurologicalEmergencies. Park Ridge, AANS, 1994, pp 249–262.
344. Hadley MN, Browner CM, Liu SS, Sonntag VKH: New subtypeof acute odontoid fractures (type IIA). Neurosurgery 22:67–71,1988.
345. Hadley MN, Browner C, Sonntag VKH: Axis fractures: A com-prehensive review of management and treatment in 107 cases.Neurosurgery 17:281–290, 1985.
346. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acutetraumatic atlas fractures: Management and long term outcome.Neurosurgery 23:31–35, 1988.
347. Hadley MN, Dickman CA, Browner CM, Sonntag VKH: Acuteaxis fractures: A review of 229 cases. J Neurosurg 71:642–647,1989.
348. Hadley MN, Fitzpatrick BC, Sonntag VKH, Browner CM: Facetfracture-dislocation injuries of the cervical spine. Neurosurgery30:661–666, 1992.
349. Hadley MN, Grahm TW, Harrington T, Schiller WR,McDermott MK, Posillico DB: Nutritional support andneurotrauma: A critical review of early nutrition in forty-fiveacute head injury patients. Neurosurgery 19:367–373, 1986.
350. Hadley MN, Zabramski JM, Browner CM, Rekate H, SonntagVKH: Pediatric spinal trauma: Review of 122 cases of spinalcord and vertebral column injuries. J Neurosurg 68:18–24, 1988.
351. Hall ED, Wolf DL: A pharmacological analysis of the patho-physiological mechanisms of posttraumatic spinal cord isch-emia. J Neurosurg 64:951–961, 1986.
352. Hall AJ, Wagle VG, Raycroft J, Goldman RL, Butler AR: Mag-netic resonance imaging in cervical spine trauma. J Trauma34:21–26, 1993.
353. Halliday AL, Henderson BR, Hart BL, Benzel EC: The manage-ment of unilateral lateral mass/facet fractures of the subaxialcervical spine: The use of magnetic resonance imaging to pre-dict instability. Spine 22:2614–2621, 1997.
354. Hamilton MG, Myles ST: Pediatric spinal injury: Review of 174hospital admissions. J Neurosurg 77:700–704, 1992.
Bibliography S187
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
355. Hamilton RS, Pons PT: The efficacy and comfort of full-bodyvacuum splints for cervical-spine immobilization. J Emerg Med14:553–559, 1996.
356. Hamilton BB, Granger CV, Sherwin FS, Zielezny M, TashmanJS: A Uniform National Data System for Medical Rehabilitation,in Fuhrer MJ (ed): Rehabilitation Outcomes: Analysis and Measure-ment. Baltimore, Paul H. Brookes Publishing Co., 1987, pp137–147.
357. Hamilton BB, Laughlin JA, Fiedler RC, Granger CV: Interraterreliability of the 7-level Functional Independence Measure(FIM). Scand J Rehabil Med 26:115–119, 1994.
358. Hamilton BB, Laughlin JA, Granger CV, Kayton RM: Interrateragreement of the seven-level Functional Independence Mea-sure (FIM). Arch Phys Med Rehabil 72:790, 1991.
359. Han SY, Witten DM, Mussleman JP: Jefferson fracture of theatlas: Report of six cases. J Neurosurg 44:368–371, 1976.
360. Hanigan WC, Anderson RJ: Commentary on NASCIS-2. J Spi-nal Disord 5:125–133, 1992.
361. Hanigan WC, Powell FC, Elwood PW, Henderson JP: Odontoidfractures in elderly patients. J Neurosurg 78:32–35, 1993.
362. Hanson JA, Blackmore CC, Mann FA, Wilson AJ: Cervical spineinjury: A clinical decision rule to identify high-risk patients forhelical CT screening. AJR Am J Roentgenol 174:713–717, 2000.
363. Hanssen AD, Cabanela ME: Fractures of the dens in adultpatients. J Trauma 27:928–934, 1987.
364. Harding-Smith J, MacIntosh PK, Sherbon KJ: Fracture of theoccipital condyle: A case report and review of the literature.J Bone Joint Surg Am 63A:1170–1171, 1981.
365. Harmanli O, Koyfman Y: Traumatic atlanto-occipital disloca-tion with survival: A case report and review of the literature.Surg Neurol 39:324–330, 1993.
366. Harrington JF, Likavec MJ, Smith AS: Disc herniation in cervi-cal fracture subluxation. Neurosurgery 29:374–379, 1991.
367. Harris JH, Carson GC, Wagner LK: Radiologic diagnosis oftraumatic occipitovertebral dissociation: Part 1—Normal oc-cipitovertebral relationships on lateral radiographs of supinesubjects. AJR Am J Roentgenol 162:881–886, 1994.
368. Harris JH Jr, Carson GC, Wagner LK, Kerr N: Radiologic diag-nosis of traumatic occipitovertebral dissociation: Part 2—Com-parison of three methods of detecting occipitovertebral rela-tionships on lateral radiographs of supine subjects. AJR Am JRoentgenol 162:887–892, 1994.
369. Harris S, Chen D, Green D: Enoxaparin for thromboembolismprophylaxis in spinal injury: Preliminary report on experiencewith 105 patients. Am J Phys Med Rehabil 75:326–327, 1996.
370. Hashimoto T, Watanabe O, Takase M, Koniyama J, Kobota M:Collet-Sicard syndrome after minor head trauma. Neurosur-gery 23:367–370, 1988.
371. Hauswald M, Ong G, Tandberg D, Omar Z: Out-of-hospitalspinal immobilization: Its effect on neurologic injury. AcadEmerg Med 5:214–219, 1998.
372. Hays MB, Alker GJ Jr: Fractures of the atlas vertebra: Thetwo-part burst fracture of Jefferson. Spine 13:601–603, 1988.
373. Hays MB, Bernhang AM: Fractures of the atlas vertebra: Athree-part fracture not previously classified. Spine 17:240–242,1992.
374. Heary RF, Hunt CD, Krieger AJ, Antonio C, Livingston DH:Acute stabilization of the cervical spine by halo/vest applica-tion facilitates evaluation and treatment of multiple traumapatients. J Trauma 33:445–451, 1992.
375. Heller JG, Viroslav S, Hudson T: Jefferson fractures: The role ofmagnification artifact in assessing transverse ligament integ-rity. J Spinal Disord 6:392–396, 1993.
376. Henry AD, Bohly J, Grosse A: Fixation of odontoid fractures byan anterior screw. J Bone Joint Surg Br 81B:472–477, 1999.
377. Herr CH, Ball PA, Sargent SK, Quinton HB: Sensitivity ofprevertebral soft tissue measurement of C3 for detection ofcervical spine fractures and dislocations. Am J Emerg Med16:346–349, 1998.
378. Herzenberg JE, Hensinger RN, Dedrick DK, Phillips WA: Emer-gency transport and positioning of young children who have aninjury of the cervical spine: The standard backboard may behazardous. J Bone Joint Surg Am 71A:15–22, 1989.
379. Highland TR, Salciccioli GG: Is immobilization adequate treat-ment of unstable burst fractures of the atlas? A case report withlong-term follow-up evaluation. Clin Orthop 201:196–200,1985.
379a. Hinchey JJ, Bickel WH: Fracture of the atlas: Review and pre-sentation of data on eight cases. Ann Surg 121:826–830, 1945.
380. Hladky JP, Lejeune JP, Leclercq F, Dhellemmes P, Christiaens JL:Traumatic atlanto-occipital dislocation [in French]. Neurochirurgie37:312–317, 1991.
381. Hoffman JR, Mower WR, Wolfson AB, Todd KH, Zucker MI:Validity of a set of clinical criteria to rule out injury to thecervical spine in patients with blunt trauma: National Emer-gency X-Radiography Utilization Study Group. N Engl J Med343:94–99, 2000.
382. Hoffman JR, Schriger DL, Mower W, Luo JS, Zucker M: Low-risk criteria for cervical-spine radiography in blunt trauma: Aprospective study. Ann Emerg Med 21:1454–1460, 1992.
383. Hoffman JR, Wolfson AB, Todd K, Mower WR: Selective cer-vical spine radiography in blunt trauma: Methodology of theNational Emergency X-Radiography Utilization Study (NEX-US). Ann Emerg Med 32:461–469, 1998.
384. Holliman CJ, Mayer JS, Cook RT Jr, Smith JS Jr: Is the antero-posterior cervical spine radiograph necessary in initial traumascreening? Am J Emerg Med 9:421–425, 1991.
385. Hosono N, Yonenobu K, Ebara S, Ono K: Cineradiographicmotion analysis of atlantoaxial instability in os odontoideum.Spine 16[Suppl 10]:S480–S482, 1991.
386. Hosono N, Yonenobu K, Kawagoe K, Hirayama N, Ono K:Traumatic anterior atlanto-occipital dislocation: A case reportwith survival. Spine 18:786–790, 1993.
387. Huerta C, Griffith R, Joyce SM: Cervical spine stabilization inpediatric patients: Evaluation of current techniques. AnnEmerg Med 16:1121–1126, 1987.
388. Hulsewe KW, van Acker BA, von Meyenfeldt MF, Soeters PB:Nutritional depletion and dietary manipulation: Effects on theimmune response. World J Surg 23:536–544, 1999.
389. Hummel A, Plaue R: Diagnosis and treatment of atlanto-occipital ruptures [in German]. Unfallchirurgie 14:311–319,1988.
390. Hurlbert RJ: Methylprednisolone for acute spinal cord injury:An inappropriate standard of care. J Neurosurg 93[Suppl 1]:1–7, 2000.
391. Hurlbert RJ: The use of methylprednisolone. J Neurosurg93[Suppl 1]:340–341, 2000 (letter).
391a. Hurlbert RJ: The role of steroids in acute spinal cord injury: Anevidence-based analysis. Spine 26(24 Suppl):S39–S46, 2001.
392. Ide C, Nisolle JF, Misson N, Trigaux JP, Gustin T, De Coene B,Gilliard C: Unusual occipitoatlantal fracture dissociation withno neurological impairment: Case report. J Neurosurg 88:773–776, 1998.
393. Iida H, Tachibana S, Kitahara T, Horiike S, Ohwada T, Fujii K:Association of head trauma with cervical spine injury, spinalcord injury, or both. J Trauma 46:450–452, 1999.
S188 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
394. Ireland AJ, Britton I, Forrester AW: Do supine oblique viewsprovide better imaging of the cervicothoracic junction thanswimmer’s views? J Accid Emerg Med 15:151–154, 1998.
395. Jacobs LM, Schwartz R: Prospective analysis of acute cervicalspine injury: A methodology to predict injury. Ann Emerg Med15:44–49, 1986.
396. Jacoby CG: Fracture of the occipital condyle. AJR Am JRoentgenol 132:500, 1979 (letter).
397. Jakim I, Sweet MB: Transverse fracture through the body of theaxis. J Bone Joint Surg Br 70B:728–729, 1988.
398. Jakim I, Sweet MB, Wisniewski T, Gantz ED: Isolated avulsionfracture of the anterior tubercle of the atlas. Arch OrthopTrauma Surg 108:377–379, 1989.
399. Jarrell BE, Posuniak E, Roberts J, Osterholm J, Cotler J, DitunnoJ: A new method of management using the Kim-Ray Greenfieldfilter for deep venous thrombosis and pulmonary embolism inspinal cord injury. Surg Gynecol Obstet 157:316–320, 1983.
400. Jeanneret B, Magerl F: Primary posterior fusion C1/2 in odon-toid fractures: Indications, technique, and results of transartic-ular screw fixation. J Spinal Disord 5:464–475, 1992.
401. Jeanneret B, Magerl F, Ward JC: Overdistraction: A hazard ofskull traction in the management of acute injuries of the cervicalspine. Arch Orthop Trauma Surg 110:242–245, 1991.
402. Jefferson G: Fractures of the atlas vertebra: Report of four casesand a review of those previously reported. Br J Surg 7:407–422,1920.
403. Jenkins JD, Coric D, Branch CL Jr: A clinical comparison of one-and two-screw odontoid fixation. J Neurosurg 89:366–370,1998.
404. Jevtich V: Traumatic lateral atlanto-occipital dislocation withspontaneous bony fusion: A case report. Spine 14:123–124,1989.
405. Johnson DR, Hauswald M, Stockhoff C: Comparison of a vac-uum splint device to a rigid backboard for spinal immobiliza-tion. Am J Emerg Med 14:369–372, 1996.
406. Jones SL: Spinal trauma board. Phys Ther 57:921–922, 1977.407. Jones DN, Knox AM, Sage MR: Traumatic avulsion fracture of
the occipital condyles and clivus with associated unilateralatlantooccipital distraction. AJNR Am J Neuroradiol 11:1181–1183, 1990.
408. Jonsson M, Tollback A, Gonzales H, Borg J: Inter-rater reliabil-ity of the 1992 international standards for neurological andfunctional classification of incomplete spinal cord injury. Spi-nal Cord 38:675–679, 2000.
409. Julien TD, Frankel B, Traynelis VC, Ryken TC: Evidence-basedanalysis of odontoid fracture management. Neurosurg Focus8:Article 1, 2000.
410. Kalfas I, Wilberger JE, Goldberg A, Prostko ER: Magnetic res-onance imaging in acute spinal cord trauma. Neurosurgery23:295–299, 1988.
411. Kalff R, Kocks W, Grote W, Schmit-Neuerburg KP: Operativespondylodesis in injuries of the lower cervical spine.Neurosurg Rev 16:211–220, 1993.
412. Kaneriya PP, Schweitzer ME, Spettell C, Cohen MJ, Karasick D:The cost-effectiveness of oblique radiography in the exclusionof C7-T1 injury in trauma patients. AJR Am J Roentgenol171:959–962, 1998.
413. Katzberg RW, Benedetti PF, Drake CM, Ivanovic M, Levine RA,Beatty CS, Nemzek WR, McFall RA, Ontell FK, Bishop DM,Poirier VC, Chong BW: Acute cervical spine injuries: Prospec-tive MR imaging assessment at a level 1 trauma center. Radi-ology 213:203–212, 1999.
414. Kaufman RA, Dunbar JS, Botsford JA, McLaurin RL: Traumaticlongitudinal atlanto-occipital distraction injuries in children.AJNR Am J Neuroradiol 3:415–419, 1982.
415. Kaufman HH, Rowlands BJ, Stein DK, Kopaniky DR,Gildenberg PL: General metabolism in patients with acute para-plegia and quadriplegia. Neurosurgery 16:309–313, 1985.
416. Kaups K, Davis J: Patients with gunshot wounds to the head donot require cervical spine immobilization and evaluation.J Trauma 44:865–867, 1998.
417. Kawabe N, Hirotani H, Tanaka O: Pathomechanism of atlanto-axial rotatory fixation in children. J Pediatr Orthop 9:569–574,1989.
418. Kearns PJ, Thompson JD, Werner PC, Pipp TL, Wilmot CB:Nutritional and metabolic response to acute spinal-cord injury.JPEN J Parenter Enteral Nutr 16:11–15, 1992.
419. Keiper MD, Zimmerman RA, Bilaniuk LT: MRI in the assess-ment of the supportive soft tissues of the cervical spine in acutetrauma in children. Neuroradiology 40:359–363, 1998.
420. Kesterson L, Benzel EC, Orrison W, Coleman J: Evaluation andtreatment of atlas burst fractures (Jefferson fractures).J Neurosurg 75:213–220, 1991.
421. Key A: Cervical spine dislocations with unilateral facet inter-locking. Paraplegia 13:208–215, 1975.
422. Khansarinia S, Dennis JW, Veldenz HC, Butcher JL, Hartland L:Prophylactic Greenfield filter placement in selected high-risktrauma patients. J Vasc Surg 22:231–236, 1995.
423. Kilburn MP, Smith DP, Hadley MN: The initial evaluation andtreatment of the patient with spinal trauma, in Batjer HH,Loftus CM (eds): Textbook of Neurological Surgery: Principles andPractice. Philadelphia, Lippincott Williams & Wilkins (in press).
424. Kindt GW, Ducker TB, Huddlestone J: Regulation of spinalcord blood flow, in Russell RWR (ed): Brain and Blood Flow:Proceedings of the 4th International Symposium on the Regulation ofCerebral Blood Flow. London: Pitman Medical, 1971, pp 401–405.
425. King BS, Gupta R, Narayan RK: The early assessment andintensive care unit management of patients with severe trau-matic brain and spinal cord injuries. Surg Clin North Am80:855–870, viii–ix, 2000.
426. Kinney TB, Rose SC, Valji K, Oglevie SB, Roberts AC: Doescervical spinal cord injury induce a higher incidence of com-plications after prophylactic Greenfield inferior vena cava filterusage? J Vasc Interv Radiol 7:907–915, 1996.
427. Kiwerski JE: Early anterior decompression and fusion for crushfractures of cervical vertebrae. Int Orthop 17:166–168, 1993.
428. Klein GR, Vaccaro AR, Albert TJ, Schweitzer M, Deely D,Karasick D, Cotler JM: Efficacy of magnetic resonance imagingin the evaluation of posterior cervical spine fractures. Spine24:771–774, 1999.
429. Kleyn PJ: Dislocations of the cervical spine: Closed reductionunder anaesthesia. Paraplegia 22:271–281, 1984.
430. Klose KJ, Green BA, Smith RS, Adkins RH, MacDonald AM:University of Miami Neuro-Spinal Index (UMNI): A quantita-tive method for determining spinal cord function. Paraplegia18:331–336, 1980.
431. Kobrine AI, Doyle TF, Martins AN: Autoregulation of spinalcord blood flow. Clin Neurosurg 22:573–581, 1975.
432. Kobrine AI, Doyle TF, Rizzoli HV: Spinal cord blood flow asaffected by changes in systemic arterial blood pressure.J Neurosurg 44:12–15, 1976.
433. Kolb JC, Summers RL, Galli RL: Cervical collar-inducedchanges in intracranial pressure. Am J Emerg Med 17:135–137,1999.
Bibliography S189
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
434. Koop SE, Winter RB, Lonstein JE: The surgical treatment ofinstability of the upper part of the cervical spine in children andadolescents. J Bone Joint Surg Am 66A:403–411, 1984.
435. Kornberg M: Atypical unstable burst fracture of the atlas:Treated by primary atlantoaxial fusion. Orthop Rev 15:727–729,1986.
436. Korres DS, Zoubos AB, Kavadias K, Babis GC, Balalis K: The“tear drop” (or avulsed) fracture of the anterior inferior angle ofthe axis. Eur Spine J 3:151–154, 1994.
437. Koskinen EV, Nieminen R: Fractures and dislocations of thecervical spine: Treatment and results of 159 cases. Int Surg47:472–485, 1967.
438. Kossuth LC: The removal of injured personnel from wreckedvehicles. J Trauma 5:703–708, 1965.
439. Kossuth LC: The initial movement of the injured. Mil Med132:18–21, 1967.
440. Kowalski HM, Cohen WA, Cooper P, Wisoff JH: Pitfalls in theCT diagnosis of atlantoaxial rotary subluxation. AJR Am JRoentgenol 149:595–600, 1987.
441. Kreipke DL, Gillespie KR, McCarthy MC, Mail JT, Lappas JC,Broadie TA: Reliability of indications for cervical spine films intrauma patients. J Trauma 29:1438–1439, 1989.
442. Kucukdeveci AA, Yavuzer G, Tennant A, Suldur N, Sonel B,Arasil T: Adaptation of the modified Barthel Index for use inphysical medicine and rehabilitation in Turkey. Scand JRehabil Med 32:87–92, 2000.
443. Kuhns LR, Loder RT, Farley FA, Hensinger RN: Nuchal cordchanges in children with os odontoideum: Evidence for associ-ated trauma. J Pediatr Orthop 18:815–819, 1998.
444. Kulkarni JR, Burt AA, Tromans AT, Constable PD: Prophylacticlow dose heparin anticoagulant therapy in patients with spinalcord injuries: A retrospective study. Paraplegia 30:169–172,1992.
445. Laham JL, Cotcamp DH, Gibbons PA, Kahana MD, Crone KR:Isolated head injuries versus multiple trauma in pediatric pa-tients: Do the same indications for cervical spine evaluationapply? Pediatr Neurosurg 21:221–226, 1994.
446. Lam CH, Stratford J: Bilateral hypoglossal nerve injury withoccipital condylar fracture. Can J Neurol Sci 23:145–148, 1996.
447. Lamb GC, Tomski MA, Kaufman J, Maiman DJ: Is chronicspinal cord injury associated with increased risk of venousthromboembolism? J Am Paraplegia Soc 16:153–156, 1993.
448. Landells CD, Van Peteghem PK: Fractures of the atlas: Classi-fication, treatment and morbidity. Spine 13:450–452, 1988.
449. Landi G, Ciccone A: GM-1 ganglioside for spinal-cord injury.N Engl J Med 326:493–494, 1992.
450. Landis JR, Koch GG: The measurement of observer agreementfor categorical data. Biometrics 33:159–174, 1977.
451. Lazar RB, Yarkony GM, Ortolano D, Heinemann AW, PerlowE, Lovell L, Meyer PR: Prediction of functional outcome bymotor capability after spinal cord injury. Arch Phys MedRehabil 70:819–822, 1989.
452. Ledsome JR, Sharp JM: Pulmonary function in acute cervicalcord injury. Am Rev Respir Dis 124:41–44, 1981.
453. Lee C, Woodring JH: Unstable Jefferson variant atlas fractures:An unrecognized cervical injury. AJNR Am J Neuroradiol12:1105–1110, 1991.
454. Lee TT, Green BA, Petrin DR: Treatment of stable burst fractureof the atlas (Jefferson fracture) with rigid cervical collar. Spine23:1963–1967, 1998.
455. Lee AS, MacLean JC, Newton DA: Rapid traction for reductionof cervical spine dislocations. J Bone Joint Surg Br 76B:352–356,1994.
456. Lee C, Woodring JH, Goldstein SJ, Daniel TL, Young AB, TibbsPA: Evaluation of traumatic atlantooccipital dislocations. AJNRAm J Neuroradiol 8:19–26, 1987.
457. Lee C, Woodring JH, Walsh JW: Carotid and vertebral injury insurvivors of atlanto-occipital dislocation: Case reports and lit-erature review. J Trauma 31:401–407, 1991.
458. Legros B, Fournier P, Chiaroni P, Ritz O, Fusciardi J: Basalfracture of the skull and lower (IX, X, XI, XII) cranial nervespalsy: Four case reports including two fractures of the occipitalcondyle—A literature review. J Trauma 48:342–348, 2000.
459. Lehmann KG, Lane JG, Piepmeier JM, Batsford WP: Cardiovas-cular abnormalities accompanying acute spinal cord injury inhumans: Incidence, time course and severity. J Am CollCardiol 10:46–52, 1987.
460. Lemons VR, Wagner FC Jr: Stabilization of subaxial cervicalspinal injuries. Surg Neurol 39:511–518, 1993.
461. Lennarson PJ, Mostafavi H, Traynelis VC, Walters BC: Manage-ment of type II dens fractures: A case-control study. Spine25:1234–1237, 2000.
462. Leone A, Cerase A, Colosimo C, Lauro L, Puca A, Marano P:Occipital condylar fractures: A review. Radiology 216:635–644,2000.
463. Lerner EB, Billittier AJ IV, Moscati RM: The effects of neutralpositioning with and without padding on spinal immobiliza-tion of healthy subjects. Prehosp Emerg Care 2:112–116, 1998.
464. Leventhal MR, Boydston WR, Sebes JI, Pinstein ML, WatridgeCB, Lowrey R: The diagnosis and treatment of fractures of theoccipital condyle. Orthopedics 15:944–947, 1992.
465. Levi L, Wolf A, Belzberg H: Hemodynamic parameters in pa-tients with acute cervical cord trauma: Description, interven-tion, and prediction of outcome. Neurosurgery 33:1007–1017,1993.
466. Levi L, Wolf A, Rigamonti D, Ragheb J, Mirvis S, Robinson WL:Anterior decompression in cervical spine trauma: Does thetiming of surgery affect the outcome? Neurosurgery 29:216–222, 1991.
467. Levine AM, Edwards CC: The management of traumatic spon-dylolisthesis of the axis. J Bone Joint Surg Am 67A:217–226,1985.
468. Levine AM, Edwards CC: Treatment of injuries in the C1-C2complex. Orthop Clin North Am 17:31–44, 1986.
469. Levine AM, Edwards CC: Fractures of the atlas. J Bone JointSurg Am 73A:680–691, 1991.
470. Levine AM, Mazel C, Roy-Camille R: Management of fractureseparations of the articular mass using posterior cervical plat-ing. Spine 17[Suppl 10]:S447–S454, 1992.
471. Levine AM, Nash MS, Green BA, Shea JD, Aronica MJ: Anexamination of dietary intakes and nutritional status of chronichealthy spinal cord injured individuals. Paraplegia 30:880–889,1992.
472. Lewelt W, Jenkins LW, Miller JD: Autoregulation of cerebralblood flow after experimental fluid percussion injury of thebrain. J Neurosurg 53:500–511, 1980.
473. Lewis LM, Docherty M, Ruoff BE, Fortney JP, Keltner RA Jr,Britton P: Flexion-extension views in the evaluation of cervical-spine injuries. Ann Emerg Med 20:117–121, 1991.
474. Lieberman IH, Webb JK: Cervical spine injuries in the elderly.J Bone Joint Surg Br 76B:877–881, 1994.
475. Liew SC, Hill DA: Complication of hard cervical collars inmulti-trauma patients. Aust N Z J Surg 64:139–140, 1994.
476. Lifeso RM, Colucci MA: Anterior fusion for rotationally unsta-ble cervical spine fractures. Spine 25:2028–2034, 2000.
S190 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
477. Linares HA, Mawson AR, Suarez E, Biundo JJ: Associationbetween pressure sores and immobilization in the immediatepost-injury period. Orthopedics 10:571–573, 1987.
478. Lind B, Nordwall A, Sihlbom H: Odontoid fractures treatedwith halo-vest. Spine 12:173–177, 1987.
479. Lind B, Sihlbom H, Nordwall A: Halo-vest treatment of unsta-ble traumatic cervical spine injuries. Spine 13:425–432, 1988.
480. Lindsey RW, Diliberti TC, Doherty BJ, Watson AB: Efficacy ofradiographic evaluation of the cervical spine in emergencysituations. South Med J 86:1253–1255, 1993.
481. Link TM, Schuierer G, Hufendiek A, Horch C, Peters PE: Sub-stantial head trauma: Value of routine CT examination of thecervicocranium. Radiology 196:741–745, 1995.
482. Lipson SJ: Fractures of the atlas associated with fractures of theodontoid process and transverse ligament ruptures. J BoneJoint Surg Am 59A:940–943, 1977.
483. Lockey AS, Handley R, Willett K: “Clearance” of cervical spineinjury in the obtunded patient. Injury 29:493–497, 1998.
484. Louw JA, Mafoyane NA, Small B, Neser CP: Occlusion of thevertebral artery in cervical spine dislocations. J Bone Joint SurgBr 72B:679–681, 1990.
485. Lowry DW, Pollack IF, Clyde B, Albright AL, Adelson PD:Upper cervical spine fusion in the pediatric population.J Neurosurg 87:671–676, 1997.
486. Lu K, Lee TC, Chen HJ: Closed reduction of bilateral lockedfacets of the cervical spine under general anaesthesia. ActaNeurochir (Wien) 140:1055–1061, 1998.
487. Lu K, Lee TC, Liang CL, Chen HJ: Delayed apnea in patientswith mid- to lower cervical spinal cord injury. Spine 25:1332–1338, 2000.
488. Lucas JT, Ducker TB: Motor classification of spinal cord injurieswith mobility, morbidity and recovery indices. Am Surg 45:151–158, 1979.
489. Ludwig SC, Vaccaro AR, Balderston RA, Cotler JM: Immediatequadriparesis after manipulation for bilateral cervical facet sub-luxation: A case report. J Bone Joint Surg Am 79A:587–590,1997.
490. Lui TN, Lee ST, Wong CW, Yeh YS, Tzaan WC, Chen TY, HungSY: C1-C2 fracture dislocations in children and adolescents.J Trauma 40:408–411, 1996.
491. Lukhele M: Fractures of the vertebral lamina associated withunifacet and bifacet cervical spine dislocations. S Afr J Surg32:112–114, 1994.
492. Lyons MK, Partington MD, Meyer FB: A randomized, con-trolled trial of methylprednisolone or naloxone in the treatmentof acute spinal-cord injury. N Engl J Med 323:1207–1208, 1990(letter).
493. Macdonald RL, Schwartz ML, Mirich D, Sharkey PW, NelsonWR: Diagnosis of cervical spine injury in motor vehicle crashvictims: How many x-rays are enough? J Trauma 30:392–397,1990.
494. Mace SE: Emergency evaluation of cervical spine injuries: CTversus plain radiographs. Ann Emerg Med 14:973–975, 1985.
495. Mace SE: Unstable occult cervical-spine fracture. Ann EmergMed 20:1373–1375, 1991.
496. MacKinnon JA, Perlman M, Kirpalani H, Rehan V, Sauve R,Kovacs L: Spinal cord injury at birth: Diagnostic and prognosticdata in twenty-two patients. J Pediatr 122:431–437, 1993.
497. Mahadevan S, Mower WR, Hoffman JR, Peeples N, GoldbergW, Sonner R: Interrater reliability of cervical spine injury crite-ria in patients with blunt trauma. Ann Emerg Med 31:197–201,1998.
498. Mahale YJ, Silver JR: Progressive paralysis after bilateral facetdislocation of the cervical spine. J Bone Joint Surg Br 74B:219–223, 1992.
499. Mahale YJ, Silver JR, Henderson NJ: Neurological complica-tions of the reduction of cervical spine dislocations. J BoneJoint Surg Br 75B:403–409, 1993.
500. Mahoney FI, Barthel DW: Functional evaluation: The BarthelIndex. Md State Med J 14:61–65, 1965.
501. Maiman DJ, Larson SJ: Management of odontoid fractures.Neurosurgery 11:820, 1982 (letter).
502. Maiman DJ, Barolat G, Larson SJ: Management of bilaterallocked facets of the cervical spine. Neurosurgery 18:542–547,1986.
503. Mandabach M, Ruge JR, Hahn YS, McLone DG: Pediatric axisfractures: Early halo immobilization, management and out-come. Pediatr Neurosurg 19:225–232, 1993.
504. Mann FA, Cohen W: Occipital condyle fracture: Significance inthe assessment of occipitoatlantal stability. AJR Am JRoentgenol 163:193–194, 1994.
505. Mansel JK, Norman JR: Respiratory complications and manage-ment of spinal cord injuries. Chest 97:1446–1452, 1990.
506. Marar BC, Tay CK: Fracture of the odontoid process. Aust N ZJ Surg 46:231–236, 1976.
507. Mariani PJ: Occipital condyle fracture presenting asretropharyngeal hematoma. Ann Emerg Med 19:1447–1449,1990.
508. Marino RJ, Huang M, Knight P, Herbison GJ, Ditunno JF Jr,Segal M: Assessing selfcare status in quadriplegia: Comparisonof the Quadriplegia Index of Function (QIF) and the FunctionalIndependence Measure (FIM). Paraplegia 31:225–233, 1993.
509. Marino RJ, Rider-Foster D, Maissel G, Ditunno JF: Superiorityof motor level over single neurological level in categorizingtetraplegia. Paraplegia 33:510–513, 1995.
510. Markenson D, Foltin G, Tunik M, Cooper A, Giordano L, FittonA, Lanotte T: The Kendrick extrication device used for pediatricspinal immobilization. Prehosp Emerg Care 3:66–69, 1999.
511. Marks DS, Roberts P, Wilton PJ, Burns LA, Thompson AG: Ahalo jacket for stabilisation of the paediatric cervical spine.Arch Orthop Trauma Surg 112:134–135, 1993.
512. Maroon JC: “Burning hands” in football spinal cord injuries.JAMA 238:2049–2051, 1977.
513. Marshall LF, Knowlton S, Garfin SR, Klauber MR, EisenbergHM, Kopaniky D, Miner ME, Tabbador K, Clifton GL: Deteri-oration following spinal cord injury: A multicenter study.J Neurosurg 66:400–404, 1987.
514. Massaro F, Lanotte M: Fracture of the occipital condyle. Injury24:419–420, 1993.
515. Matava MJ, Whitesides TE Jr, Davis PC: Traumatic atlanto-occipital dislocation with survival: Serial computerized tomog-raphy as an aid to diagnosis and reduction—A report of threecases. Spine 18:1897–1903, 1993.
516. Matsui H, Imada K, Tsuji H: Radiographic classification of osodontoideum and its clinical significance. Spine 22:1706–1709,1997.
517. Matsumoto T, Tamaki T, Kawakami M, Yoshida M, Ando M,Yamada H: Early complications of high-dose methylpred-nisolone sodium succinate treatment in the follow-up of acutecervical spinal cord injury. Spine 26:426–430, 2001.
518. Maves CK, Souza A, Prenger EC, Kirks DR: Traumatic atlanto-occipital disruption in children. Pediatr Radiol 21:504–507,1991.
Bibliography S191
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
519. Mawson AR, Biundo JJ Jr, Neville P, Linares HA, Winchester Y,Lopez A: Risk factors for early occurring pressure ulcers fol-lowing spinal cord injury. Am J Phys Med Rehabil 67:123–127,1988.
520. Maynard FM Jr, Bracken MB, Creasey G, Ditunno JF Jr,Donovan WH, Ducker TB, Garber SL, Marino RJ, Stover SL,Tator CH, Waters RL, Wilberger JE, Young W: InternationalStandards for Neurological and Functional Classification ofSpinal Cord Injury: American Spinal Injury Association. SpinalCord 35:266–274, 1997.
521. Maynard FM Jr, Reynolds GG, Fountain S, Wilmot C, HamiltonR: Neurological prognosis after traumatic quadriplegia: Three-year experience of California Regional Spinal Cord Injury CareSystem. J Neurosurg 50:611–616, 1979.
522. Mazolewski P, Manix TH: The effectiveness of strapping tech-niques in spinal immobilization. Ann Emerg Med 23:1290–1295, 1994.
523. McCabe JB, Nolan DJ: Comparison of the effectiveness of dif-ferent cervical immobilization collars. Ann Emerg Med 15:50–53, 1986.
524. McGuire RA Jr: Protection of the unstable spine during trans-port and early hospitalization. J Miss State Med Assoc 32:305–308, 1991.
525. McGuire RA Jr, Harkey HL: Primary treatment of unstableJefferson’s fractures. J Spinal Disord 8:233–236, 1995.
526. McGuire RA Jr, Degnan G, Amundson GM: Evaluation ofcurrent extrication orthoses in immobilization of the unstablecervical spine. Spine 15:1064–1067, 1990.
527. McGuire RA Jr, Neville S, Green BA, Watts C: Spinal instabilityand the log-rolling maneuver. J Trauma 27:525–531, 1987.
528. McHugh TP, Taylor JP: Unnecessary out-of-hospital use of fullspinal immobilization. Acad Emerg Med 5:278–280, 1998.
529. McKinley WO, Jackson AB, Cardenas DD, DeVivo MJ: Long-term medical complications after traumatic spinal cord injury:A regional model systems analysis. Arch Phys Med Rehabil80:1402–1410, 1999.
530. McLaurin RL, Vernal R, Salmon JH: Treatment of fractures ofthe atlas and axis by wiring without fusion. J Neurosurg 36:773–780, 1972.
531. McMichan JC, Michel L, Westbrook PR: Pulmonary dysfunc-tion following traumatic quadriplegia: Recognition, prevention,and treatment. JAMA 243:528–531, 1980.
532. McNamara RM, Heine E, Esposito B: Cervical spine injury andradiography in alert, high-risk patients. J Emerg Med 8:177–182, 1990.
533. McSwain NE Jr: Spine management skills, in Pre-HospitalTrauma Life Support. Akron, Educational Direction, 1990, ed 2,pp 225–256.
534. McSwain NE Jr: Proper C-spine immobilization. Emerg Med27:120–121, 1995.
535. Meldon SW, Brant TA, Cydulka RK, Collins TE, Shade BR:Out-of-hospital cervical spine clearance: Agreement betweenemergency medical technicians and emergency physicians.J Trauma 45:1058–1061, 1998.
536. Menezes AH, Ryken TC: Craniovertebral abnormalities inDown’s syndrome. Pediatr Neurosurg 18:24–33, 1992.
537. Menticoglou SM, Perlman M, Manning FA: High cervical spinalcord injury in neonates delivered with forceps: Report of 15cases. Obstet Gynecol 86:589–594, 1995.
538. Merianos P, Tsekouras G, Koskinas A: An unusual fracture ofthe atlas. Injury 22:489–490, 1991.
539. Merli GJ, Crabbe S, Doyle L, Ditunno JF, Herbison GJ: Mechan-ical plus pharmacological prophylaxis for deep vein thrombosisin acute spinal cord injury. Paraplegia 30:558–562, 1992.
540. Merli GJ, Herbison GJ, Ditunno JF, Weitz HH, Henzes JH, ParkCH, Jaweed MM: Deep vein thrombosis: Prophylaxis in acutespinal cord injured patients. Arch Phys Med Rehabil 69:661–664, 1988.
541. Merriam WF, Taylor TK, Ruff SJ, McPhail MJ: A reappraisal ofacute traumatic central cord syndrome. J Bone Joint Surg Br68B:708–713, 1986.
542. Mestdagh H, Letendart J, Sensey JJ, Duquennoy A: Treatmentof fractures of the posterior axial arch: Results of 41 cases [inFrench]. Rev Chir Orthop Reparatrice Appar Mot 70:21–28,1984.
543. Mirvis SE, Diaconis JN, Chirico PA, Reiner BI, Joslyn JN,Militello P: Protocol-driven radiologic evaluation of suspectedcervical spine injury: Efficacy study. Radiology 170:831–834,1989.
544. Mirvis SE, Geisler FH, Jelinek JJ, Joslyn JN, Gellad F: Acutecervical spine trauma: Evaluation with 1.5-T MR imaging. Ra-diology 166:807–816, 1988.
545. Mirvis SE, Young JW, Lim C, Greenberg J: Hangman’s fracture:Radiologic assessment in 27 cases. Radiology 163:713–717,1987.
546. Mody BS, Morris EW: Fracture of the occipital condyle: Casereport and review of the world literature. Injury 23:350–352,1992.
547. Mollan RA, Watt PC: Hangman’s fracture. Injury 14:265–267,1982.
548. Montane I, Eismont FJ, Green BA: Traumatic occipitoatlantaldislocation. Spine 16:112–116, 1991.
549. Montesano PX, Anderson PA, Schlehr F, Thalgott JS, Lowrey G:Odontoid fractures treated by anterior odontoid screw fixation.Spine 16[Suppl 3]:S33–S37, 1991.
550. Morandi X, Hanna A, Hamlat A, Brassier G: Anterior screwfixation of odontoid fractures. Surg Neurol 51:236–240, 1999.
551. Morgan MK, Onofrio BM, Bender CE: Familial osodontoideum: Case report. J Neurosurg 70:636–639, 1989.
552. Deleted in proof.553. Mubarak SJ, Camp JF, Vuletich W, Wenger DR, Garfin SR: Halo
application in the infant. J Pediatr Orthop 9:612–614, 1989.554. Muhr MD, Seabrook DL, Wittwer LK: Paramedic use of a spinal
injury clearance algorithm reduces spinal immobilization in theout-of-hospital setting. Prehosp Emerg Care 3:1–6, 1999.
555. Muller EJ, Wick M, Muhr G: Traumatic spondylolisthesis of theaxis: Treatment rationale based on the stability of the differentfracture types. Eur Spine J 9:123–128, 2000.
556. Muller EJ, Wick M, Russe O, Muhr G: Management of odontoidfractures in the elderly. Eur Spine J 8:360–365, 1999.
557. Myllynen P, Kammonen M, Rokkanen P, Bostman O, Lalla M,Laasonen E: Deep venous thrombosis and pulmonary embo-lism in patients with acute spinal cord injury: A comparisonwith nonparalyzed patients immobilized due to spinal frac-tures. J Trauma 25:541–543, 1985.
558. Naso F: Pulmonary embolism in acute spinal cord injury. ArchPhys Med Rehabil 55:275–278, 1974.
559. Naso WB, Cure J, Cuddy BG: Retropharyngeal pseudomeningoceleafter atlanto-occipital dislocation: Report of two cases. Neurosurgery40:1288–1291, 1997.
560. Nazarian SM, Louis RP: Posterior internal fixation with screwplates in traumatic lesions of the cervical spine. Spine 16[Suppl3]:S64–S71, 1991.
S192 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
561. Neifeld GL, Keene JG, Hevesy G, Leikin J, Proust A, ThistedRA: Cervical injury in head trauma. J Emerg Med 6:203–207,1988.
562. Nesathurai S: Steroids and spinal cord injury: Revisiting theNASCIS 2 and NASCIS 3 trials. J Trauma 45:1088–1093, 1998.
563. Neville S, Watts C: Management of the unstable cervical spinein transport: A re-evaluation. Aeromed J Sept/Oct:32, 1987.
564. Newey ML, Sen PK, Fraser RD: The long-term outcome aftercentral cord syndrome: A study of the natural history. J BoneJoint Surg Br 82B:851–855, 2000.
565. Nischal K, Chumas P, Sparrow O: Prolonged survival afteratlanto-occipital dislocation: Two case reports and review. Br JNeurosurg 7:677–682, 1993.
566. Noble ER, Smoker WR: The forgotten condyle: The appearance,morphology, and classification of occipital condyle fractures.AJNR Am J Neuroradiol 17:507–513, 1996.
567. Nunez DB Jr, Quencer RM: The role of helical CT in the assess-ment of cervical spine injuries. AJR Am J Roentgenol 171:951–957, 1998.
568. Nypaver M, Treloar D: Neutral cervical spine positioning inchildren. Ann Emerg Med 23:208–211, 1994.
569. Oda T, Panjabi MM, Crisco JJ III, Oxland TR: Multidirectionalinstabilities of experimental burst fractures of the atlas. Spine17:1285–1290, 1992.
570. Oda T, Panjabi MM, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part II—Relevance to clin-ical diagnosis and treatment. Spine 16[Suppl 10]:S466–S473,1991.
571. Odent T, Langlais J, Glorion C, Kassis B, Bataille J, PouliquenJC: Fractures of the odontoid process: A report of 15 cases inchildren younger than 6 years. J Pediatr Orthop 19:51–54, 1999.
572. Olerud C, Jonsson H Jr: Compression of the cervical spine cordafter reduction of fracture dislocations: Report of 2 cases. ActaOrthop Scand 62:599–601, 1991.
573. Olson CM, Jastremski MS, Vilogi JP, Madden CM, Beney KM:Stabilization of patients prior to interhospital transfer. Am JEmerg Med 5:33–39, 1987.
574. Olsson R, Kunz R: Fracture of the occipital condyle as anincidental finding during CT-evaluation of a maxillary fracture.Acta Radiol 35:90–91, 1994.
575. Orbay T, Aykol S, Seckin Z, Ergun R: Late hypoglossal nervepalsy following fracture of the occipital condyle. Surg Neurol31:402–404, 1989.
576. Ordonez BJ, Benzel EC, Naderi S, Weller SJ: Cervical facetdislocation: Techniques for ventral reduction and stabilization.J Neurosurg 92[Suppl 1]:18–23, 2000.
577. Orledge JD, Pepe PE: Out-of-hospital spinal immobilization: Isit really necessary? Acad Emerg Med 5:203–204, 1998.
578. Osenbach RK, Menezes AH: Spinal cord injury without radio-graphic abnormality in children. Pediatr Neurosci 15:168–175,1989.
579. Osenbach RK, Menezes AH: Pediatric spinal cord and vertebralcolumn injury. Neurosurgery 30:385–390, 1992.
580. Osterholm JL: The pathophysiological response to spinal cordinjury: The current status of related research. J Neurosurg40:5–33, 1974.
581. Osti OL, Fraser RD, Griffiths ER: Reduction and stabilisation ofcervical dislocations: An analysis of 167 cases. J Bone JointSurg Br 71B:275–282, 1989.
582. Ota T, Akaboshi K, Nagata M, Sonoda S, Domen K, Seki M,Chino N: Functional assessment of patients with spinal cordinjury: Measured by the motor score and the Functional Inde-pendence Measure. Spinal Cord 34:531–535, 1996.
583. Paeslack V, Frankel H, Michaelis L: Closed injuries of thecervical spine and spinal cord: Results of conservative treat-ment of flexion fractures and flexion rotation fracture disloca-tion of the cervical spine with tetraplegia. Proc Veterans AdmSpinal Cord Inj Conf 19:39–42, 1973.
584. Page CP, Story JL, Wissinger JP, Branch CL: Traumaticatlantooccipital dislocation: Case report. J Neurosurg 39:394–397, 1973.
585. Paley MD, Wood GA: Traumatic bilateral hypoglossal nervepalsy. Br J Oral Maxillofac Surg 33:239–241, 1995.
586. Palmer MT, Turney SZ: Tracheal rupture and atlanto-occipitaldislocation: Case report. J Trauma 37:314–317, 1994.
587. Pang D, Hanley EN: Special problems of spinal stabilization inchildren, in Cooper PR (ed): Management of Posttraumatic SpinalInstability (Neurosurgical Topics No. 3). Park Ridge, AANS, 1990,pp 181–206.
588. Pang D, Pollack IF: Spinal cord injury without radiographicabnormality in children: The SCIWORA syndrome. J Trauma29:654–664, 1989.
589. Pang D, Wilberger JE Jr: Traumatic atlanto-occipital dislocationwith survival: Case report and review. Neurosurgery 7:503–508, 1980.
590. Pang D, Wilberger JE Jr: Spinal cord injury without radio-graphic abnormalities in children. J Neurosurg 57:114–129,1982.
591. Panjabi MM, Oda T, Crisco JJ III, Oxland TR, Katz L, Nolte LP:Experimental study of atlas injuries: Part I—Biomechanicalanalysis of their mechanisms and fracture patterns. Spine16[Suppl 10]:S460–S465, 1991.
592. Papadopoulos SM, Dickman CA, Sonntag VKH, Rekate HL,Spetzler RF: Traumatic atlantooccipital dislocation with sur-vival. Neurosurgery 28:574–579, 1991.
593. Paramore CG, Dickman CA, Sonntag VKH: The anatomicalsuitability of the C1-2 complex for transarticular screw fixation.J Neurosurg 85:221–224, 1996.
594. Pasciak M, Doniec J: Results of conservative treatment of uni-lateral cervical spine dislocations. Arch Orthop Trauma Surg112:226–227, 1993.
595. Pasquale M, Fabian TC: Practice Management Guidelines forTrauma from the Eastern Association for the Surgery ofTrauma. J Trauma 44:941–957, 1998.
596. Patton JH, Kralovich KA, Cuschieri J, Gasparri M: Clearing thecervical spine in victims of blunt assault to the head and neck:What is necessary? Am Surg 66:326–331, 2000.
597. Pedersen AK, Kostuik JP: Complete fracture-dislocation of theatlantoaxial complex: Case report and recommendations for anew classification of dens fractures. J Spinal Disord 7:350–355,1994.
598. Peiffer SC, Blust P, Leyson JF: Nutritional assessment of thespinal cord injured patient. J Am Diet Assoc 78:501–505, 1981.
599. Pennecot GF, Leonard P, Peyrot Des Gachons S, Hardy JR,Pouliquen JC: Traumatic ligamentous instability of the cervicalspine in children. J Pediatr Orthop 4:339–345, 1984.
600. Pepin JW, Hawkins RJ: Traumatic spondylolisthesis of the axis:Hangman’s fracture. Clin Orthop 157:133–138, 1981.
601. Pepin JW, Bourne RB, Hawkins RJ: Odontoid fractures, withspecial reference to the elderly patient. Clin Orthop 193:178–183, 1985.
601a. Perkash A: Experience with the management of deep veinthrombosis in patients with spinal cord injury: Part II—A crit-ical evaluation of the anticoagulant therapy. Paraplegia 18:2–14, 1980.
Bibliography S193
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
602. Perkash A, Prakash V, Perkash I: Experience with the manage-ment of thromboembolism in patients with spinal cord injury:Part I—Incidence, diagnosis and role of some risk factors. Para-plegia 16:322–331, 1978.
603. Perry SD, McLellan B, McIlroy WE, Maki BE, Schwartz M,Fernie GR: The efficacy of head immobilization techniquesduring simulated vehicle motion. Spine 24:1839–1844, 1999.
604. Phillips WA, Hensinger RN: The management of rotatoryatlanto-axial subluxation in children. J Bone Joint Surg Am71A:664–668, 1989.
605. Piepmeier JM, Lehmann KB, Lane JG: Cardiovascular instabil-ity following acute cervical spinal cord trauma. Cent Nerv SystTrauma 2:153–160, 1985.
606. Pitzen T, Caspar W, Steudel WI, Barbier D: Dens fracture inelderly patients and surgical management [in German].Aktuelle Traumatol 24:56–59, 1994.
607. Plaisier B, Gabram SG, Schwartz RJ, Jacobs LM: Prospectiveevaluation of craniofacial pressure in four different cervicalorthoses. J Trauma 37:714–720, 1994.
608. Podolsky S, Baraff LJ, Simon RR, Hoffman JR, Larmon B, AblonW: Efficacy of cervical spine immobilization methods. J Trauma23:461–465, 1983.
609. Deleted in proof.610. Pointillart V, Petitjean ME, Wiart L, Vital JM, Lassie P, Thicoipe
M, Dabadie P: Pharmacological therapy of spinal cord injuryduring the acute phase. Spinal Cord 38:71–76, 2000.
611. Polin RS, Szabo T, Bogaev CA, Replogle RE, Jane JA: Nonopera-tive management of types II and III odontoid fractures: ThePhiladelphia collar versus the halo vest. Neurosurgery 38:450–457, 1996.
612. Pollack IF, Pang D, Sclabassi R: Recurrent spinal cord injurywithout radiographic abnormalities in children. J Neurosurg69:177–182, 1988.
613. Powell M, Kirshblum S, O’Connor KC: Duplex ultrasoundscreening for deep vein thrombosis in spinal cord injured pa-tients at rehabilitation admission. Arch Phys Med Rehabil80:1044–1046, 1999.
614. Powers B, Miller MD, Kramer RS, Martinez S, Gehweiler JA Jr:Traumatic anterior atlanto-occipital dislocation. Neurosurgery4:12–17, 1979.
615. Poynton AR, O’Farrell DA, Shannon F, Murray P, McManus F,Walsh MG: An evaluation of the factors affecting neurologi-cal recovery following spinal cord injury. Injury 28:545–548,1997.
616. Prabhu V, Kizer J, Patil A, Hellbusch L, Taylon C, Leibrock L:Vertebrobasilar thrombosis associated with nonpenetrating cer-vical spine trauma. J Trauma 40:130–137, 1996.
617. Prasad VS, Schwartz A, Bhutani R, Sharkey PW, Schwartz ML:Characteristics of injuries to the cervical spine and spinal cordin polytrauma patient population: Experience from a regionaltrauma unit. Spinal Cord 37:560–568, 1999.
618. Priebe MM, Waring WP: The interobserver reliability of therevised American Spinal Injury Association standards for neu-rological classification of spinal injury patients. Am J Phys MedRehabil 70:268–270, 1991.
619. Przybylski GJ, Clyde BL, Fitz CR: Craniocervical junction sub-arachnoid hemorrhage associated with atlanto-occipital dislo-cation. Spine 21:1761–1768, 1996.
620. Putnam WE, Stratton FT, Rohr RJ, Stitzell W, Roat G: Traumaticatlanto-occipital dislocations: Value of the Powers ratio in di-agnosis. J Am Osteopath Assoc 86:798–804, 1986.
621. Quencer RM, Bunge RP, Egnor M, Green BA, Puckett W,Naidich TP, Post MJ, Norenberg M: Acute traumatic centralcord syndrome: MRI-pathological correlations. Neuroradiol-ogy 34:85–94, 1992.
622. Quirke TE, Ritota PC, Swan KG: Inferior vena caval filter use inU.S. trauma centers: A practitioner survey. J Trauma 43:333–337, 1997.
623. Raila FA, Aitken AT, Vickers GN: Computed tomography andthree-dimensional reconstruction in the evaluation of occipitalcondyle fracture. Skeletal Radiol 22:269–271, 1993.
624. Ralston ME, Chung K, Barnes PD, Emans JB, Schutzman SA:Role of flexion-extension radiographs in blunt pediatric cervicalspine injury. Acad Emerg Med 8:237–245, 2001.
625. Ramon S, Dominguez R, Ramirez L, Paraira M, Olona M,Castello T, Garcia Fernandez L: Clinical and magnetic reso-nance imaging correlation in acute spinal cord injury. SpinalCord 35:664–673, 1997.
626. Ramsay AH, Waxman BP, O’Brien JF: A case of traumaticatlanto-occipital dislocation with survival. Injury 17:412–413,1986.
627. Raphael JH, Chotai R: Effects of the cervical collar on cerebro-spinal fluid pressure. Anaesthesia 49:437–439, 1994.
628. Rapp RP, Young B, Twyman D, Bivins BA, Haack D, Tibbs PA,Bean JR: The favorable effect of early parenteral feeding onsurvival in head-injured patients. J Neurosurg 58:906–912,1983.
629. Rathbone D, Johnson G, Letts M: Spinal cord concussion inpediatric athletes. J Pediatr Orthop 12:616–620, 1992.
630. Reid DC, Henderson R, Saboe L, Miller JD: Etiology and clinicalcourse of missed spine fractures. J Trauma 27:980–986, 1987.
631. Reines HD, Harris RC: Pulmonary complications of acute spi-nal cord injuries. Neurosurgery 21:193–196, 1987.
632. Reinges MH, Mayfrank L, Rohde V, Spetzger U, Gilsbach JM:Surgically treated traumatic synchondrotic disruption of theodontoid process in a 15-month-old girl. Childs Nerv Syst14:85–87, 1998.
633. Riley KO, May AK, Hadley MN: Neurological injury and nu-tritional support, in Batjer HH, Loftus CM (eds): Textbook ofNeurological Surgery: Principles and Practice. Philadelphia, Lip-pincott Williams & Wilkins (in press).
634. Rimel RW, Jane JA, Edlich RF: An educational training programfor the care at the site of injury of trauma to the central nervoussystem. Resuscitation 9:23–28, 1981.
635. Ripa DR, Kowall MG, Meyer PR Jr, Rusin JJ: Series of ninety-two traumatic cervical spine injuries stabilized with anteriorASIF plate fusion technique. Spine 16[Suppl 3]:S46–S55, 1991.
636. Rizzolo SJ, Piazza MR, Cotler JM, Balderston RA, Schaefer D,Flanders A: Intervertebral disc injury complicating cervicalspine trauma. Spine 16(6 Suppl):S187–S189, 1991.
637. Roberge RJ, Wears RC, Kelly M, Evans TC, Kenny MA, DaffnerRD, Kremen R, Murray K, Cottington EC: Selective applicationof cervical spine radiography in alert victims of blunt trauma: Aprospective study. J Trauma 28:784–748, 1988.
638. Robertson PA, Ryan MD: Neurological deterioration after re-duction of cervical subluxation: Mechanical compression bydisc tissue. J Bone Joint Surg Br 74B:224–227, 1992.
639. Rockswold GL, Seljeskog EL: Traumatic atlantocranial disloca-tion with survival. Minn Med 62:151–154, 1979.
640. Rockswold GL, Bergman TA, Ford SE: Halo immobilizationand surgical fusion: Relative indications and effectiveness inthe treatment of 140 cervical spine injuries. J Trauma 30:893–898, 1990.
S194 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
641. Rodgers JA, Rodgers WB: Marginal mandibular nerve palsydue to compression by a cervical hard collar. J Orthop Trauma9:177–179, 1995.
642. Rodriguez DJ, Benzel EC: Nutritional support, in Benzel EC(ed): Spine Surgery: Techniques, Complication Avoidance, and Man-agement. New York, Churchill Livingstone, 1999, vol 2, pp1321–1331.
643. Rodriguez DJ, Benzel EC, Clevenger FW: The metabolic re-sponse to spinal cord injury. Spinal Cord 35:599–604, 1997.
644. Roozmon P, Gracovetsky SA, Gouw GJ, Newman N: Examin-ing motion in the cervical spine: Part I—Imaging systems andmeasurement techniques. J Biomed Eng 15:5–12, 1993.
645. Rorabeck CH, Rock MG, Hawkins RJ, Bourne RB: Unilateralfacet dislocation of the cervical spine. An analysis of the resultsof treatment in 26 patients. Spine 12:23–27, 1987.
646. Deleted in proof.647. Rosen PB, McSwain NE Jr, Arata M, Stahl S, Mercer D: Com-
parison of two new immobilization collars. Ann Emerg Med21:1189–1195, 1992.
648. Rosenberg J, Greenberg MK: Practice parameters: Strategies forsurvival into the nineties. Neurology 42:1110–1115, 1992.
649. Rosenfeld JF, Vaccaro AR, Albert TJ, Klein GR, Cotler JM: Thebenefits of early decompression in cervical spinal cord injury.Am J Orthop 27:23–28, 1998.
650. Rosner MJ: National acute spinal cord injury study of methyl-prednisolone or naloxone. Neurosurgery 28:628, 1991 (letter).
651. Rosner MJ: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–327, 1992.
652. Rosner MJ: Treatment of spinal cord injury. J Neurosurg 80:954–955, 1994.
653. Ross SE, O’Malley KF, DeLong WG, Born CT, Schwab CW:Clinical predictors of unstable cervical spinal injury in multiplyinjured patients. Injury 23:317–319, 1992.
654. Rossitch E Jr, Oakes WJ: Perinatal spinal cord injury: Clinical,radiographic and pathologic features. Pediatr Neurosurg 18:149–152, 1992.
655. Roth EJ, Lawler MH, Yarkony GM: Traumatic central cordsyndrome: Clinical features and functional outcomes. ArchPhys Med Rehabil 71:18–23, 1990.
656. Roth BJ, Martin RR, Foley K, Barcia PJ, Kennedy P: Roentgen-ographic evaluation of the cervical spine: A selective approach.Arch Surg 129:643–645, 1994.
657. Roussi J, Bentolila S, Boudaoud L, Casadevall N, Vallee C,Carlier R, Lortat-Jacob S, Dizien O, Bussel B: Contribution ofd-dimer determination in the exclusion of deep venous throm-bosis in spinal cord injury patients. Spinal Cord 37:548–552,1999.
658. Roy-Camille R, Benazet JP, Saillant G, Henry P, Mamoudy P,Leonard P: Traumatic atlanto-occipital luxation: Value of newradiologic signs (apropos of 2 cases) [in French]. Rev ChirOrthop Reparatrice Appar Mot 72:303–309, 1986.
659. Roy-Camille R, Saillant G, Laville C, Benazet JP: Treatment oflower cervical spinal injuries: C3 to C7. Spine 17[Suppl 10]:S442–S446, 1992.
660. Ruge JR, Sinson GP, McLone DG, Cerullo LJ: Pediatric spinalinjury: The very young. J Neurosurg 68:25–30, 1988.
661. Rutledge G, Sumchai A: A safe method for transportation ofpatients with cervical spine injuries. Aeromed J Sept/Oct:33,1987.
662. Ryan MD, Henderson JJ: The epidemiology of fractures andfracture-dislocations of the cervical spine. Injury 23:38–40,1992.
663. Ryan MD, Taylor TK: Odontoid fractures: A rational approachto treatment. J Bone Joint Surg Br 64B:416–421, 1982.
664. Ryan MD, Taylor TK: Odontoid fractures in the elderly. J Spi-nal Disord 6:397–401, 1993.
665. Sabiston CP, Wing PC, Schweigel JF, Van Peteghem PK, Yu W:Closed reduction of dislocations of the lower cervical spine.J Trauma 28:832–835, 1988.
666. Salmon JH: Fractures of the second cervical vertebra: Internalfixation by interlaminar wiring. Neurosurgery 1:125–127, 1977.
667. Sandler AN, Tator CH: Effect of acute spinal cord compressioninjury on regional spinal cord blood flow in primates.J Neurosurg 45:660–676, 1976.
668. Sandler AN, Tator CH: Review of the effect of spinal cordtrauma on the vessels and blood flow in the spinal cord.J Neurosurg 45:638–646, 1976.
669. San Mateo County, California: EMS System Policy Memorandum#F-3A. 1991.
670. Savini R, Parisini P, Cervellati S: The surgical treatment of lateinstability of flexion-rotation injuries in the lower cervicalspine. Spine 12:178–182, 1987.
671. Savolaine ER, Ebraheim NA, Jackson WT, Rusin JJ: Three-dimensional computed tomography in evaluation of occipitalcondyle fracture. J Orthop Trauma 3:71–75, 1989.
672. Scarrow AM, Levy EI, Resnick DK, Adelson PD, Sclabassi RJ:Cervical spine evaluation in obtunded or comatose pediatrictrauma patients: A pilot study. Pediatr Neurosurg 30:169–175,1999.
672a. Schaefer DM, Flanders A, Northrup BE, Doan HT, OsterholmJL: Magnetic resonance imaging of acute cervical spine trauma.Spine 14:1090–1095, 1989.
673. Schafermeyer RW, Ribbeck BM, Gaskins J, Thomason S, HarlanM, Attkisson A: Respiratory effects of spinal immobilization inchildren. Ann Emerg Med 20:1017–1019, 1991.
674. Schellinger PD, Schwab S, Krieger D, Fiebach JB, Steiner T,Hund EF, Hacke W, Meinck HM: Masking of vertebral arterydissection by severe trauma to the cervical spine. Spine 26:314–319, 2001.
675. Schleehauf K, Ross SE, Civil ID, Schwab CW: Computed to-mography in the initial evaluation of the cervical spine. AnnEmerg Med 18:815–817, 1989.
676. Schlicke LH, Callahan RA: A rational approach to burst frac-tures of the atlas. Clin Orthop 154:18–21, 1981.
677. Schneider RC: A syndrome in acute cervical spine injuries forwhich early operation is indicated. J Neurosurg 8:360–367,1951.
678. Schneider R: “Hangman’s fracture” of the cervical spine.J Neurosurg 22:141–154, 1965.
679. Schneider RC, Cherry G, Pantek H: The syndrome of acutecentral cervical spinal cord injury. J Neurosurg 546–577, 1954.
680. Schneider RC, Crosby EC, Russo RH, Gosch HH: Traumaticspinal cord syndromes and their management. Clin Neurosurg20:424–492, 1973.
681. Schneider RC, Thompson JC, Bebin J: The syndrome of acutecentral cervical spinal cord injury. J Neurol Neurosurg Psychi-atry 21:216–227, 1958.
682. Schönhöfer PS: GM-1 ganglioside for spinal-cord injury. N EnglJ Med 326:493–494, 1992.
683. Schriger DL: Immobilizing the cervical spine in trauma: Shouldwe seek an optimal position or an adequate one? Ann EmergMed 28:351–353, 1996.
684. Schriger DL, Larmon B, LeGassick T, Blinman T: Spinal immo-bilization on a flat backboard: Does it result in neutral positionof the cervical spine? Ann Emerg Med 20:878–881, 1991.
Bibliography S195
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
685. Schwarz N: The fate of missed atlanto-axial rotatory subluxa-tion in children. Arch Orthop Trauma Surg 117:288–289, 1998.
686. Schwarz N, Buchinger W, Gaudernak T, Russe F, Zechner W:Injuries to the cervical spine causing vertebral artery trauma:Case reports. J Trauma 31:127–133, 1991.
687. Schwarz N, Genelin F, Schwarz AF: Post-traumatic cervicalkyphosis in children cannot be prevented by non-operativemethods. Injury 25:173–175, 1994.
688. Sears W, Fazl M: Prediction of stability of cervical spine fracturemanaged in the halo vest and indications for surgical interven-tion. J Neurosurg 72:426–432, 1990.
689. Sedlock DA, Laventure SJ: Body composition and resting en-ergy expenditure in long term spinal cord injury. Paraplegia28:448–454, 1990.
690. Sees DW, Rodriguez Cruz LR, Flaherty SF, Ciceri DP: The useof bedside fluoroscopy to evaluate the cervical spine inobtunded trauma patients. J Trauma 45:768–771, 1998.
691. Segal LS, Grimm JO, Stauffer ES: Non-union of fractures of theatlas. J Bone Joint Surg Am 69A:1423–1434, 1987.
692. Seljeskog EL: Non-operative management of acute upper cer-vical injuries. Acta Neurochir (Wien) 41:87–100, 1978.
693. Seljeskog EL, Chou SN: Spectrum of the hangman’s fracture.J Neurosurg 45:3–8, 1976.
694. Senter HJ, Venes JL: Loss of autoregulation and posttraumaticischemia following experimental spinal cord trauma.J Neurosurg 50:198–206, 1979.
695. Seybold EA, Bayley JC: Functional outcome of surgically andconservatively managed dens fractures. Spine 23:1837–1846,1998.
696. Shacked I, Ram Z, Hadani M: The anterior cervical approach fortraumatic injuries to the cervical spine in children. Clin Orthop292:144–150, 1993.
697. Shaffer MA, Doris PE: Limitation of the cross table lateral viewin detecting cervical spine injuries: A retrospective analysis.Ann Emerg Med 10:508–513, 1981.
698. Shah S, Vanclay F, Cooper B: Improving the sensitivity of theBarthel Index for stroke rehabilitation. J Clin Epidemiol 42:703–709, 1989.
699. Shapiro SA: Methylprednisolone for spinal cord injury.J Neurosurg 77:324–327, 1992.
700. Shapiro SA: Management of unilateral locked facet of the cer-vical spine. Neurosurgery 33:832–837, 1993.
701. Shapiro S, Snyder W, Kaufman K, Abel T: Outcome of 51 casesof unilateral locked cervical facets: Interspinous braided cablefor lateral mass plate fusion compared with interspinous wireand facet wiring with iliac crest. J Neurosurg 91[Suppl 1]:19–24,1999.
702. Sharma BS, Mahajan RK, Bhatia S, Khosla VK: Collet-Sicardsyndrome after closed head injury. Clin Neurol Neurosurg96:197–198, 1994.
703. Shaw M, Burnett H, Wilson A, Chan O: Pseudosubluxation ofC2 on C3 in polytraumatized children: Prevalence and signifi-cance. Clin Radiol 54:377–380, 1999.
704. Sherk HH, Nicholson JT: Fractures of the atlas. J Bone JointSurg Am 52A:1017–1024, 1970.
705. Sherk HH, Nicholson JT, Chung SM: Fractures of the odontoidprocess in young children. J Bone Joint Surg Am 60A:921–924,1978.
706. Shirasaki N, Okada K, Oka S, Hosono N, Yonenobu K, Ono K:Os odontoideum with posterior atlantoaxial instability. Spine16:706–715, 1991.
707. Short D: Use of steroids for acute spinal cord injury must bereassessed. BMJ 321:1224, 2000 (letter).
708. Short DJ, El Masry WS, Jones PW: High dose methylpred-nisolone in the management of acute spinal cord injury: Asystematic review from a clinical perspective. Spinal Cord38:273–286, 2000.
709. Shoung HM, Lee LS: Anterior metal plate fixation in the treat-ment of unstable lower cervical spine injuries. Acta Neurochir(Wien) 98:55–59, 1989.
710. Shrosbree RD: Acute central cervical spinal cord syndrome:Aetiology, age incidence and relationship to the orthopaedicinjury. Paraplegia 14:251–258, 1977.
711. Shrosbree RD: Neurological sequelae of reduction of fracturedislocations of the cervical spine. Paraplegia 17:212–221, 1979.
712. Sim E, Schwarz N, Biowski-Fasching I, Biowski P: Color-codedDuplex sonography of vertebral arteries: 11 cases of blunt cer-vical spine injury. Acta Orthop Scand 64:133–137, 1993.
713. Sim E, Vaccaro AR, Berzlanovich A, Pienaar S: The effects of stagedstatic cervical flexion-distraction deformities on the patency of thevertebral arterial vasculature. Spine 25:2180–2186, 2000.
714. Smith MG, Bourn S, et al: Ties that bind: Immobilizing theinjured spine. J Emerg Med Serv JEMS 14:28–35, 1989.
715. Smith MD, Phillips WA, Hensinger RN: Fusion of the uppercervical spine in children and adolescents: An analysis of 17patients. Spine 16:695–701, 1991.
716. Song WS, Chiang YH, Chen CY, Lin SZ, Liu MY: A simplemethod for diagnosing traumatic occlusion of the vertebralartery at the craniovertebral junction. Spine 19:837–839, 1994.
717. Sonntag VKH: Management of bilateral locked facets of thecervical spine. Neurosurgery 8:150–152, 1981.
718. Spence KF Jr, Decker S, Sell KW: Bursting atlantal fractureassociated with rupture of the transverse ligament. J Bone JointSurg Am 52A:543–549, 1970.
719. Spencer JA, Yeakley JW, Kaufman HH: Fracture of the occipitalcondyle. Neurosurgery 15:101–103, 1984.
720. Spierings EL, Braakman R: The management of osodontoideum: Analysis of 37 cases. J Bone Joint Surg Br 64B:422–428, 1982.
721. Sponseller PD, Cass JR: Atlanto-occipital fusion for dislocationin children with neurologic preservation: A case report. Spine22:344–347, 1997.
722. Star AM, Jones AA, Cotler JM, Balderston RA, Sinha R: Imme-diate closed reduction of cervical spine dislocations using trac-tion. Spine 15:1068–1072, 1990.
723. Starr JK, Eismont FJ: Atypical hangman’s fractures. Spine 18:1954–1957, 1993.
724. Stauffer ES: Orthotics for spinal cord injuries. Clin Orthop102:92–99, 1974.
725. Stauffer ES, Kelly EG: Fracture-dislocations of the cervicalspine: Instability and recurrent deformity following treatmentby anterior interbody fusion. J Bone Joint Surg Am 59A:45–48,1977.
726. Stevens JM, Chong WK, Barber C, Kendall BE, Crockard HA: Anew appraisal of abnormalities of the odontoid process associ-ated with atlanto-axial subluxation and neurological disability.Brain 117:133–148, 1994.
727. Stineman MG, Marino RJ, Deutsch A, Granger CV, Maislin G: Afunctional strategy for classifying patients after traumatic spi-nal cord injury. Spinal Cord 37:717–725, 1999.
728. Stover S, Fine PR: Spinal Cord Injury: The Facts & Figures. Bir-mingham, University of Alabama, 1986.
729. Stroobants J, Fidlers L, Storms JL, Klaes R, Dua G, Van Hoye M:High cervical pain and impairment of skull mobility as the onlysymptoms of an occipital condyle fracture: Case report.J Neurosurg 81:137–138, 1994.
S196 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
730. Stroobants J, Seynaeve P, Fidlers L, Klaes R, Brabants K, VanHoye M: Occipital condyle fracture must be considered in thepediatric population: Case report. J Trauma 36:440–441, 1994.
731. Subach BR, McLaughlin MR, Albright AL, Pollack IF: Currentmanagement of pediatric atlantoaxial rotatory subluxation.Spine 23:2174–2179, 1998.
732. Subach BR, Morone MA, Haid RW Jr, McLaughlin MR, RodtsGR, Comey CH: Management of acute odontoid fractures withsingle-screw anterior fixation. Neurosurgery 45:812–820, 1999.
733. Suechting RL, French LA: Posterior inferior cerebellar arterysyndrome following a fracture of the cervical vertebra.J Neurosurg 12:187–189, 1955.
734. Deleted in proof.735. Suter R, Tighe T, et al: Thoracolumbar spinal instability during
variations of the log-roll maneuver. Prehospital Disaster Med7:133–138, 1992.
736. Swain A, Dove J, Baker H: ABCs of major trauma: Trauma ofthe spine and spinal cord—Part I. BMJ 301:34–38, 1990.
737. Sweeney JF, Rosemurgy AS, Gill S, Albrink MH: Is the cervicalspine clear? Undetected cervical fractures diagnosed only atautopsy. Ann Emerg Med 21:1288–1290, 1992.
738. Swischuk LE, John SD, Hendrick EP: Is the open-mouth odon-toid view necessary in children under 5 years? Pediatr Radiol30:186–189, 2000.
738a. Sypert GW: Hangman’s fracture: Arguments in favor of surgi-cal therapy for type II and III according to Edwards and Levine.Surg Neurol 49:262, 1998 (comment).
739. Taggard DA, Menezes AH, Ryken TC: Treatment of Downsyndrome-associated craniovertebral junction abnormalities.J Neurosurg 93[Suppl 2]:205–213, 2000.
740. Tan ES, Balachandran N: Hangman’s fracture in Singapore(1975–1988). Paraplegia 30:160–164, 1992.
741. Tan E, Schweitzer ME, Vaccaro L, Spetell AC: Is computedtomography of nonvisualized C7-T1 cost-effective? J SpinalDisord 12:472–476, 1999.
742. Tator CH: Vascular effects and blood flow in acute spinal cordinjuries. J Neurosurg Sci 28:115–119, 1984.
743. Tator CH: Hemodynamic issues and vascular factors in acuteexperimental spinal cord injury. J Neurotrauma 9:139–141,1992.
744. Tator CH: Ischemia as a secondary neural injury, in SalzmanSK, Faden AI (eds): Neurobiology of Central Nervous SystemTrauma. New York, Oxford University Press, 1994, pp 209–215.
745. Tator CH: Experimental and clinical studies of the pathophys-iology and management of acute spinal cord injury. J SpinalCord Med 19:206–214, 1996.
746. Tator CH: Biology of neurological recovery and functional res-toration after spinal cord injury. Neurosurgery 42:696–708,1998.
747. Tator CH, Fehlings MG: Review of the secondary injury theoryof acute spinal cord trauma with emphasis on vascular mech-anisms. J Neurosurg 75:15–26, 1991.
748. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Comparison of surgical and conservative management in 208patients with acute spinal cord injury. Can J Neurol Sci 14:60–69, 1987.
749. Tator CH, Duncan EG, Edmonds VE, Lapczak LI, Andrews DF:Changes in epidemiology of acute spinal cord injury from 1947to 1981. Surg Neurol 40:207–215, 1993.
750. Tator CH, Rowed DW, Schwartz MI: Sunnybrook Cord InjuryScales for assessing neurological injury and neurologic recov-ery, in Tator CH (ed): Early Management of Acute Spinal CordInjury. New York, Raven Press, 1982, pp 17–24.
751. Tator CH, Rowed DW, Schwartz ML, Gertzbein SD, BharatwalN, Barkin M, Edmonds VE: Management of acute spinal cordinjuries. Can J Surg 27:289–296, 1984.
752. Tehranzadeh J, Bonk RT, Ansari A, Mesgarzadeh M: Efficacy oflimited CT for nonvisualized lower cervical spine in patientswith blunt trauma. Skeletal Radiol 23:349–352, 1994.
753. Thibodeaux LC, Hearn AT, Peschiera JL, Deshmukh RM,Kerlakian GM, Welling RE, Nyswonger GD: Extracranial ver-tebral artery dissection after trauma: A 5-year review. Br J Surg84:94, 1997.
754. Todd JW, Frisbie JH, Rossier AB, Adams DF, Als AV, ArmeniaRJ, Sasahara AA, Tow DE: Deep venous thrombosis in acutespinal cord injury: A comparison of 125I fibrinogen leg scan-ning, impedance plethysmography and venography. Paraple-gia 14:50–57, 1976.
755. Toh E, Arima T, Mochida J, Omata M, Matsui S: Functionalevaluation using motor scores after cervical spinal cord injuries.Spinal Cord 36:491–496, 1998.
756. Tola JC, Holtzman R, Lottenberg L: Bedside placement of inferiorvena cava filters in the intensive care unit. Am Surg 65:833–838, 1999.
756a. Tomaio A, Kirshblum SC, O’Connor KC, Johnston M: Treat-ment of acute deep vein thrombosis in spinal cord injuredpatients with enoxaparin: A cost analysis. J Spinal Cord Med21:205–210, 1998.
757. Torreman M: Long-term prognosis of the hangman’s fracture[in Dutch]. Ned Tijdschr Geneeskd 134:1173–1176, 1990.
758. Toscano J: Prevention of neurological deterioration before admissionto a spinal cord injury unit. Paraplegia 26:143–150, 1988.
759. Totten VY, Sugarman DB: Respiratory effects of spinal immo-bilization. Prehosp Emerg Care 3:347–352, 1999.
760. Tracy PT, Wright RM, Hanigan WC: Magnetic resonance im-aging of spinal injury. Spine 14:292–301, 1989.
761. Traynelis VC: Evidence-based management of type II odontoidfractures. Clin Neurosurg 44:41–49, 1997.
762. Traynelis VC, Marano GD, Dunker RO, Kaufman HH: Trau-matic atlanto-occipital dislocation: Case report. J Neurosurg65:863–870, 1986.
763. Treloar DJ, Nypaver M: Angulation of the pediatric cervicalspine with and without cervical collar. Pediatr Emerg Care13:5–8, 1997.
764. Tuite GF, Veres R, Crockard HA, Peterson D, Hayward RD: Useof an adjustable, transportable, radiolucent spinal immobiliza-tion device in the comprehensive management of cervical spineinstability: Technical note. J Neurosurg 85:1177–1180, 1996.
765. Tuli S, Tator CH, Fehlings MG, Mackay M: Occipital condylefractures. Neurosurgery 41:368–377, 1997.
766. Tulyapronchote R, Selhorst JB, Malkoff MD, Gomez CR: De-layed sequelae of vertebral artery dissection and occult cervicalfractures. Neurology 44:1397–1399, 1994.
767. Turgut M, Akpinar G, Akalan N, Ozcan OE: Spinal injuries inthe pediatric age group: A review of 82 cases of spinal cord andvertebral column injuries. Eur Spine J 5:148–152, 1996.
768. Turnbull IM: Microvasculature of the human spinal cord.J Neurosurg 35:141–147, 1971.
769. Turnbull IM: Blood supply of the spinal cord: Normal andpathological considerations. Clin Neurosurg 20:56–84, 1973.
770. Turnbull IM, Brieg A, Hassler O: Blood supply of cervicalspinal cord in man: A microangiographic cadaver study.J Neurosurg 24:951–965, 1966.
771. Urculo E, Arrazola M, Arrazola M Jr, Riu I, Moyua A: Delayedglossopharyngeal and vagus nerve paralysis following occipitalcondyle fracture: Case report. J Neurosurg 84:522–525, 1996.
Bibliography S197
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
772. Vaccaro AR, An HS, Lin S, Sun S, Balderston RA, Cotler JM:Noncontiguous injuries of the spine. J Spinal Disord 5:320–329,1992.
773. Vaccaro AR, Falatyn SP, Flanders AE, Balderston RA, NorthrupBE, Cotler JM: Magnetic resonance evaluation of the interver-tebral disc, spinal ligaments, and spinal cord before and afterclosed traction reduction of cervical spine dislocations. Spine24:1210–1217, 1999.
774. Vaccaro AR, Klein GR, Flanders AE, Albert TJ, Balderston RA,Cotler JM: Long-term evaluation of vertebral artery injuriesfollowing cervical spine trauma using magnetic resonance an-giography. Spine 23:789–795, 1998.
775. Valaskatzis EP, Hammer AJ: Fracture of the occipital condyle:A case report. S Afr Med J 77:47–48, 1990.
776. Vale FL, Burns J, Jackson AB, Hadley MN: Combined medicaland surgical treatment after acute spinal cord injury: Results ofa prospective pilot study to assess the merits of aggressivemedical resuscitation and blood pressure management.J Neurosurg 87:239–246, 1997.
777. Verbiest H: Anterolateral operations for fractures and disloca-tions in the middle and lower parts of the cervical spine: Reportof a series of forty-seven cases. J Bone Joint Surg Am 51A:1489–1530, 1969.
778. Verheggen R, Jansen J: Hangman’s fracture: Arguments in fa-vor of surgical therapy for type II and III according to Edwardsand Levine. Surg Neurol 49:253–262, 1998.
779. Verska JM, Anderson PA: Os odontoideum: A case report ofone identical twin. Spine 22:706–709, 1997.
780. Viccellio P, Simon H, Pressman BD, Shah MN, Mower WR,Hoffman JR: A prospective multicenter study of cervical spineinjury in children. Pediatrics 108:E20, 2001.
781. Vital JM, Gille O, Senegas J, Pointillart V: Reduction techniquefor uni- and biarticular dislocations of the lower cervical spine.Spine 23:949–955, 1998.
782. Vogel LC: Unique management needs of pediatric spinal cordinjury patients: Etiology and pathophysiology. J Spinal CordMed 20:10–13, 1997.
783. Waddell JP, Reardon GP: Atlantoaxial arthrodesis to treatodontoid fractures. Can J Surg 26:255–260, 1983.
784. Wagner FC Jr, Johnson RM: Cervical bracing after trauma. MedInstrum 16:287–288, 1982.
785. Wallace MC, Tator CH: Successful improvement of blood pres-sure, cardiac output, and spinal cord blood flow after experi-mental spinal cord injury. Neurosurgery 20:710–715, 1987.
786. Walsh M, Grant T, Mickey S: Lung function compromised byspinal immobilization. Ann Emerg Med 19:615–616, 1990.
787. Walters BC: Clinical practice parameter development in neuro-surgery, in Bean JR (ed): Neurosurgery in Transition: The Socio-economic Transformation of Neurological Surgery. Baltimore, Wil-liams & Wilkins, 1998, pp 99–111.
788. Walton G: A new method of reducing dislocation of cervicalvertebrae. J Nerv Ment Disord 20:609, 1893.
789. Walton R, DeSalvo JF, Ernst AA, Shahane A: Padded vs unpad-ded spine board for cervical spine immobilization. Acad EmergMed 2:725–728, 1995.
790. Wang GJ, Mabie KN, Whitehill R, Stamp WG: The nonsurgicalmanagement of odontoid fractures in adults. Spine 9:229–230,1984.
791. Wang J, Vokshoor A, Kim S, Elton S, Kosnik E, Bartkowski H:Pediatric atlantoaxial instability: Management with screw fixa-tion. Pediatr Neurosurg 30:70–78, 1999.
792. Wani MA, Tandon PN, Banerji AK, Bhatia R: Collet-Sicardsyndrome resulting from closed head injury: Case report.J Trauma 31:1437–1439, 1991.
793. Waring WP, Karunas RS: Acute spinal cord injuries and theincidence of clinically occurring thromboembolic disease. Para-plegia 29:8–16, 1991.
794. Washtenaw/Livingston County Medical Control Authority:Spinal Injury Assessment and Immobilization: EMS Protocols. AnnArbor, Washtenaw/Livingston County Medical Control Au-thority, 1997.
795. Wasserberg J, Bartlett RJV: Occipital condyle fractures diag-nosed by high-definition CT and coronal reconstructions. Neu-roradiology 37:370–373, 1995.
796. Watanabe M, Toyama Y, Fujimura Y: Atlantoaxial instability inos odontoideum with myelopathy. Spine 21:1435–1439, 1996.
797. Waters RL, Adkins RH, Yakura JS: Definition of complete spi-nal cord injury. Paraplegia 29:573–581, 1991.
798. Waters RL, Adkins RH, Yakura JS, Vigil D: Prediction of am-bulatory performance based on motor scores derived fromstandards of the American Spinal Injury Association. ArchPhys Med Rehabil 75:756–760, 1994.
799. Waters RL, Meyer PR Jr, Adkins RH, Felton D: Emergency,acute, and surgical management of spine trauma. Arch PhysMed Rehabil 80:1383–1390, 1999.
800. Watridge CB, Orrison WW, Arnold H, Woods GA: Lateralatlantooccipital dislocation: Case report. Neurosurgery 17:345–347, 1985.
801. Watson N: Venous thrombosis and pulmonary embolism inspinal cord injury. Paraplegia 6:113–121, 1968.
802. Watson N: Anti-coagulant therapy in the prevention of venousthrombosis and pulmonary embolism in the spinal cord injury.Paraplegia 16:265–269, 1978.
802a. Weingarden SI, Weingarden DS, Belen J: Fever and thrombo-embolic disease in acute spinal cord injury. Paraplegia 26:35–42, 1988.
803. Weinstein PR, Karpman RR, Gall EP, Pitt M: Spinal cord injury,spinal fracture, and spinal stenosis in ankylosing spondylitis.J Neurosurg 57:609–616, 1982.
804. Weller SJ, Rossitch E Jr, Malek AM: Detection of vertebralartery injury after cervical spine trauma using magnetic reso-nance angiography. J Trauma 46:660–666, 1999.
805. Wells JD, Nicosia S: Scoring acute spinal cord injury: A study ofthe utility and limitations of five different grading systems.J Spinal Cord Med 18:33–41, 1995.
806. Wessels LS: Fracture of the occipital condyle: A report of 3cases. S Afr Med J 28:155–156, 1990.
807. White AA, Panjabi MM: Clinical Biomechanics of the Spine. Phil-adelphia, J.B. Lippincott, 1990, ed 2.
808. White P, Seymour R, Powell N: MRI assessment of the pre-vertebral soft tissues in acute cervical spine trauma. Br J Radiol72:818–823, 1999.
809. Wholey MH, Bruwer AJ, Baker HL: The lateral roentgenogramof the neck (with comments on the atlanto-odontoid-basionrelationship). Radiology 71:350–356, 1958.
810. Wilberger JE: Immobilization and traction, in Tator CH, BenzelEC (eds): Contemporary Management of Spinal Cord Injuries: FromImpact to Rehabilitation (Neurosurgical Topics). Park Ridge,AANS, 2000, ed 2, pp 91–98.
811. Williams TG: Hangman’s fracture. J Bone Joint Surg Br 57B:82–88, 1975.
812. Willis BK, Greiner F, Orrison WW, Benzel EC: The incidence ofvertebral artery injury after midcervical spine fracture or sub-luxation. Neurosurgery 34:435–442, 1994.
S198 Guidelines for Management of Acute Cervical Spinal Injuries
Neurosurgery, Vol. 50, No. 3, March 2002 Supplement
813. Wilson JT, Rogers FB, Wald SL, Shackford SR, Ricci MA: Pro-phylactic vena cava filter insertion in patients with traumaticspinal cord injury: Preliminary results. Neurosurgery 35:234–239, 1994.
814. Winemiller MH, Stolp-Smith KA, Silverstein MD, TherneauTM: Prevention of venous thromboembolism in patients withspinal cord injury: Effects of sequential pneumatic compressionand heparin. J Spinal Cord Med 22:182–191, 1999.
815. Wing PC, Nance P, Connell DG, Gagnon F: Risk of avascularnecrosis following short term megadose methylprednisolonetreatment. Spinal Cord 36:633–636, 1998.
816. Wolf A, Levi L, Mirvis S, Ragheb J, Huhn S, Rigamonti D,Robinson WL: Operative management of bilateral facet dislo-cation. J Neurosurg 75:883–890, 1991.
817. Wood-Jones F: The ideal lesion produced by judicial hanging.Lancet 1:53, 1913.
818. Woodring JH, Lee C: Limitations of cervical radiography in theevaluation of acute cervical trauma. J Trauma 34:32–39, 1993.
819. Woodring JH, Lee C, Duncan V: Transverse process fractures ofthe cervical vertebrae: Are they insignificant? J Trauma 34:797–802, 1993.
820. Woodring JH, Selke AC Jr, Duff DE: Traumatic atlantooccipitaldislocation with survival. AJR Am J Roentgenol 137:21–24,1981.
821. Worsing RA Jr: Principles of prehospital care of musculoskel-etal injuries. Emerg Med Clin North Am 2:205–217, 1984.
822. Xiong XH, Bean A, Anthony A, Inglis G, Walton D: Manipula-tion for cervical spinal dislocation under general anaesthesia:Serial review for 4 years. Spinal Cord 36:21–24, 1998.
823. Yamaguchi N, Ikeda K, Ishise J, Yamashita J: Traumatic atlanto-occipital dislocation with long-term survival. Neurol Med Chir(Tokyo) 36:36–39, 1996.
824. Yamashita Y, Takahashi M, Sakamoto Y, Kojima R: Atlantoaxialsubluxation: Radiography and magnetic resonance imagingcorrelated to myelopathy. Acta Radiol 30:135–140, 1989.
825. Yarkony GM, Roth E, Lovell L, Heinemann A, Katz RT, Wu Y:Rehabilitation outcomes in complete C5 quadriplegia. Am JPhys Med Rehabil 67:73–76, 1988.
826. Yashon D, Tyson G, Vise WM: Rapid closed reduction of cer-vical fracture dislocations. Surg Neurol 4:513–514, 1975.
827. Yavuz N, Tezyurek M, Akyuz M: A comparison of two func-tional tests in quadriplegia: The Quadriplegia Index of Functionand the Functional Independence Measure. Spinal Cord 36:832–837, 1998.
828. Yelnik A, Dizien O, Bussel B, Schouman-Claeys E, Frija G,Pannier S, Held JP: Systematic lower limb phlebography inacute spinal cord injury in 147 patients. Paraplegia 29:253–260,1991.
829. Young W, Bracken MB: The Second National Acute Spinal CordInjury Study. J Neurotrauma 9[Suppl 1]:S397–S405, 1992.
830. Young B, Ott L, Rapp RP, Norton J: The patient with criticalneurological disease. Crit Care Clin 3:217–233, 1987.
831. Young B, Ott L, Twyman D, Norton J, Rapp R, Tibbs P, HaackD, Brivins B, Dempsey R: The effect of nutritional support onoutcome from severe head injury. J Neurosurg 67:668–676,1987.
832. Young WF, Rosenwasser RH, Getch C, Jallo J: Diagnosis andmanagement of occipital condyle fractures. Neurosurgery 34:257–261, 1994.
833. Zach GA, Seiler W, Dollfus P: Treatment results of spinal cordinjuries in the Swiss Paraplegic Centre of Basel. Paraplegia14:58–65, 1976.
834. Zampella EJ, Duvall ER, Langford KH: Computed tomographyand magnetic resonance imaging in traumatic locked-in syn-drome. Neurosurgery 22:591–593, 1988.
835. Zavanone M, Guerra P, Rampini P, Crotti F, Vaccari U: Trau-matic fractures of the craniovertebral junction: Management of23 cases. J Neurosurg Sci 35:17–22, 1991.
836. Zeidman SM, Ling GS, Ducker TB, Ellenbogen RG: Clinicalapplications of pharmacologic therapies for spinal cord injury.J Spinal Disord 9:367–380, 1996.
837. Zigler JE, Waters RL, Nelson RW, Capen DA, Perry J: Occipito-cervico-thoracic spine fusion in a patient with occipito-cervicaldislocation and survival. Spine 11:645–646, 1986.
838. Zimmerman E, Grant J, Vise WM, Yashon D, Hunt WE: Treat-ment of Jefferson fracture with a halo apparatus: Report of twocases. J Neurosurg 44:372–375, 1976.
Early representations of spinal vertebrae. Courtesy, Rare Book Room, NorrisMedical Library, Keck School of Medicine, University of Southern California,Los Angeles, California.
Bibliography S199