The Spine Journal 6 (2006) 252S–267S

Surgical management of cervical myelopathy: indications andtechniques for laminectomy and fusion

Ricardo J. Komotar, MD, J. Mocco, MD, Michael G. Kaiser, MD*Department of Neurological Surgery, The Neurological Institute of New York, Columbia University Medical Center,

710 West 168th Street, Room 504, New York, NY 10032, USA

Abstract BACKGROUND: Cervical spondylotic myelopathy (CSM) is a commonly encountered surgicaldisease that may be approached through a variety of operative techniques. Operative goals in thetreatment of CSM include effective neural element decompression and maintaining spinal stabilityto avoid delayed deformity progression and neurologic compromise. Determining the most appro-priate operative approach requires careful consideration of the patient’s clinical presentation andradiographic imaging.PURPOSE: To review the indications and techniques for multilevel laminectomy and fusion in thetreatment of CSM.CONCLUSIONS: When indications permit, a multilevel laminectomy is an effective and safemethod of neural element decompression. Recognizing the potential for spinal instability is essen-tial to prevent neurologic compromise and intractable axial neck pain caused by deformity progres-sion. A variety of techniques have been described to supplement the posterior tension band afterlaminectomy; however, lateral mass fixation has evolved into the preferred stabilization technique.Although clinical success is well documented, a successful outcome is dependent on a comprehen-sive, individualized evaluation of each patient presenting with CSM. � 2006 Elsevier Inc. Allrights reserved.

Keywords: Cervical; Fusion; Indications; Laminectomy; Myelopathy; Techniques

Introduction

Cervical spondylotic myelopathy (CSM) is a commondiagnoses requiring surgical intervention among patientspresenting with disorders of the spine [1–8]. A variety ofwell-known pathological processes, both congenital and ac-quired, can lead to canal compromise and myelopathy;however, the presentation and history are difficult to predict[1,3,4,8]. Thus, the prognosis and management of thispatient population may be challenging. Although well-de-signed clinical outcome trials are lacking, the existing liter-ature suggests that operative intervention reliably arreststhe progression of myelopathy and may lead to functional

FDA device/drug status: approved for this indication (pedicle screws).

Nothing of value received from a commercial entity related to this

manuscript.

* Corresponding author. Department of Neurological Surgery, The

Neurological Institute of New York, Columbia University Medical Center,

710 West 168th Street, Room 504, New York, NY 10032. Tel.: (212) 305-

0378; fax: (212) 305-2026.

E-mail address: mgk7@columbia.edu (M.G. Kaiser)

1529-9430/06/$ – see front matter � 2006 Elsevier Inc. All rights reserved.

doi:10.1016/j.spinee.2006.04.029

improvement in the majority of patients [1,2,4–6,8–20].The success of any operative procedure is dependent ona comprehensive evaluation of the individual patient’s clin-ical and radiographic characteristics.

Both static and dynamic forces contribute to direct com-pression, distortion, and ischemia of the spinal cord, result-ing in injury that often extends beyond the limits of thecompressive pathology [21]. Degenerative changes com-promising the spinal canal are exaggerated by stretchingthe spinal cord across ventral pathology during flexionand in-folding of the ligamentum flavum with extension[22]. Repeated microtrauma contributes to a chronic pro-gressive course, while acute deterioration due to irrevers-ible cord injury may result from hyperextension. Patientssuffering from CSM typically manifest signs and symptomsincluding upper extremity weakness and paresthesias, lossof hand dexterity, gait instability, or bowel and bladder dys-function. Foraminal impingement will produce radicularcomplaints with pain and sensorimotor loss in a specificnerve root distribution.

The goals of operative intervention in the treatment ofcervical spondylotic myelopathy include the following:

253SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

(a) decompression of the spinal cord and nerve roots; (b)deformity prevention by maintaining or supplementing spi-nal stability; and (c) alleviating pain. Achieving these goalswill translate into improved clinical outcomes with stabili-zation or reversal of neurologic deficits, decreased pain,and maximal functional restoration.

A number of surgical strategies exist, including anterior,posterior, or circumferential approaches. Under most cir-cumstances one approach will produce optimal results.The surgical management of patients presenting withCSM requires a comprehensive and individualized approach.Designing the most effective surgical plan is dependent onnumerous factors, including the pathoanatomy of the patient,the patient’s neurologic presentation, medical comorbidity,assessment of procedure-specific risks, and the surgeon’s ex-perience and comfort level with specific procedures.

The multilevel cervical laminectomy has been proven tobe a safe and effective means to decompress the spinal ca-nal and nerve roots [6]. Although this procedure carries therisk of a late kyphotic deformity, improved neurologic out-come is possible under the appropriate conditions. Absenceof a fixed kyphotic deformity is mandatory when consider-ing a multilevel laminectomy. Disregarding this basicpremise will lead to suboptimal surgical results or worsen-ing neurologic deficits.

The presence or possibility of spinal instability must beanticipated in order to avoid a postoperative deformity thatcould lead to delayed neurologic deterioration. Selectingthe optimal posterior stabilization technique requires a thor-ough understanding of cervical biomechanics and the avail-able techniques of spinal reconstruction. Points of fixationto stabilize the cervical spine after resection of the laminaare limited to the lateral masses and pedicles. The choiceof fixation construct is dependent on the nature and extentof instability and the surgeon’s experience and comfortlevel [23]. Under the appropriate conditions, however,a multilevel laminectomy with or without an arthrodesisis a valuable and effective management strategy in the treat-ment of CSM.

Preoperative surgical planning

Deciding on the most appropriate surgical strategy is de-pendent on several factors, including neurologic presenta-tion, pathologic anatomy, and medical comorbidity. Thedecision to pursue surgery and its timing is influenced bya patient’s neurological presentation. A rapid neurologicdecline will require more urgent intervention whereas a sta-ble deficit may be approached on an elective basis. Prophy-lactic surgery for patients with stable deficits iscontroversial and requires careful consideration by boththe patient and surgeon. The decision is dependent ona careful assessment of the individual’s clinical and radio-graphic findings as well as the risk, whether operative orconservative, that the patient is willing to accept.

Once the decision to proceed with surgery is finalized,a comprehensive radiographic evaluation of the pathologicspine is absolutely necessary to determine the most appro-priate operative approach. Several important questions thatshould be considered when choosing the most appropriateoperative approach include:

(a) What is the alignment of the cervical spine? Is thespine lordotic, straight, or kyphotic? Posterior proce-dures are primarily indicated for lordotic, and possi-bly straight, spinal configurations.

(b) How rigid is the deformity? A single-stage posteriorapproach with stabilization remains an option even inthe presence of a kyphosis if postural reduction restoressagittal balance. Fixed kyphotic deformities are an ab-solute contraindication to the posterior approach.

(c) What is the nature of the pathologic process? Howmany spinal levels are involved? Does the patientdemonstrate congenital stenosis? Anterior interbodystrut grafting beyond two levels is associated withan increased failure rate. In such cases supplementalposterior stabilization is required or a single-stageposterior approach may be more appropriate.

(d) Does a static or mobile subluxation exist? Severe sub-luxation or any increase visualized on dynamic imag-ing would eliminate the possibility of a stand-alonelaminectomy and require inclusion of a stabilizationand fusion.

(e) What is the patient’s baseline medical status? Olderpatients typically are more prone to complications re-lated to swallowing dysfunction with anterior ap-proaches [7]. Severe osteoporosis will increase thechance of interbody graft subsidence. Under bothcircumstances a posterior approach may be moreappropriate.

(f) Does the patient complain of significant axial neckpain? The presence of axial neck pain, if attributedto motion across a spondylotic segment, may supportthe addition of an arthrodesis.

Advantages of the posterior cervical approach

A major advantage of the posterior approach in the man-agement of CSM is the familiarity with the surgical proce-dure. A multilevel laminectomy is a commonly performedprocedure, considered technically easier, with shorter oper-ative times and fewer perioperative complications whencompared with an equivalent anterior procedure [6]. Ante-rior exposure may prove exceedingly difficult for obese pa-tients or patients with short thick necks. There is less risk ofunintentional durotomy and cerebrospinal fluid leak witha posterior approach, particularly in patients with ossifica-tion of the posterior longitudinal ligament where the ventraldura is attenuated or absent. The risk of swallowing dys-function or recurrent laryngeal nerve palsy is eliminated

254S R.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

with the posterior approach. Finally, posterior fusion tech-niques eliminate the complications associated withintervertebral strut grafting, such as subsidence or dislodge-ment, particularly in the osteoporotic patient.

Disadvantages of the posterior cervical approach

The indirect mechanism of neural element decompres-sion is one limitation of a multilevel laminectomy in thetreatment of CSM. In the presence of ventral pathology, op-erative success is dependent on the dorsal translation of theneural elements. The posterior approach is therefore ideallysuited for patients demonstrating a minimal degree of lor-dosis. The presence of a straight or kyphotic alignment, es-pecially if associated with significant ventral pathology,will limit dorsal migration and cause the spinal cord andnerve roots to drape across ventral pathology, leading tofurther neurologic compromise [24].

The posterior approach also tends to be associated withincreased postoperative pain and morbidity owing to the de-nervation and devascularization of the paraspinal muscles.Unsightly cosmetic defects resulting from atrophy of theparaspinal muscles may be cause for concern in certain pa-tients. Release of the posterior tension band increases therisk of postoperative kyphotic deformity. The potential forthis complication is influenced by numerous factors, includ-ing the underlying pathologic process, the number of spinallevels involved, the extent of the decompression, and thepatient’s age and underlying medical condition (eg, presenceof osteoporosis) [6]. Although the theoretical risk of spinalcord injury may be lower with posterior approaches, thereare more specific neurologic deficits that are associated witha posterior decompression, in particular dysfunction of theC5 nerve root producing dissociated motor loss [25].

Flexible deformities or subluxation may be exacerbatedwith a stand-alone multilevel laminectomy, and thereforethe inclusion of a posterior stabilization construct must beconsidered. The potential for open deformity reduction islimited with the more common posterior cervical fixationtechniques because of the limited purchase of lateral massscrews. Although the use of cervical pedicle screws mayprovide a biomechanical advantage, the insertion of thesedevices is technically demanding with the potential for cat-astrophic neurovascular complications.

Indications for multilevel cervical laminectomy

For patients presenting with CSM, poor prognostic indi-cators and, therefore, absolute indications for surgery are:(1) progression of neurologic signs and symptoms; (2) pres-ence of myelopathy for 6 months or longer; or (3) severespinal cord compression, defined by a compression ratio ap-proaching 0.4 or transverse spinal cord area of 40 squaremillimeters or less [1,2,4,6,9,12,18,26,27]. Under these

circumstances, conservatively treated patients nearly al-ways experience neurological deterioration [19], with sur-gical intervention being the most reliable method toprevent disease progression.

Factors that influence the surgical decision-making pro-cess include the extent of disease, location of compressivepathology, spinal alignment, and the presence of congenitalcanal stenosis (Table 1). For cases with anterior compres-sion limited to one or two levels, fixed kyphotic deformity,and no significant developmental narrowing of the canal, ananterior or circumferential decompression and stabilizationis favored [12,14,15,18,26,28]. In contrast, patients withcompression extending more than two levels, developmen-tal narrowing of the canal, lordotic alignment, and primaryposterior compressive pathology are candidates for the mul-tilevel laminectomy (Fig. 1) [9,16,20,24,29–32].

Among these factors, appropriate assessment of the cer-vical alignment is the most important factor when deter-mining if a multilevel laminectomy is appropriate. Benzeldefines an ‘‘effective’’ cervical lordosis as the configurationwhere no dorsal component of C3 to C7 crosses a line fromthe posterior caudal corner of C2 to an identical point onC7. Associated with this line is a zone of uncertainty, wherea surgeon’s bias and experience determine if the global con-figuration is consistent with a lordosis or kyphosis. Othersconsider a configuration that falls into this ‘‘gray’’ zoneas a ‘‘straightened’’ spine [33]. The decision to performa multilevel laminectomy in the presence of a straight spinerequires consideration of additional factors, such as the de-gree of canal encroachment. Dorsal migration of the spinalcord away from ventral pathology may be insufficient witha straight cervical spine. Limits of ventral encroachmenthave been suggested; however, no definitive value exists.Hamanishi and Tanaka considered a minimum lordosis of10 degrees required for adequate dorsal migration of thespinal cord [34]. Yamazaki and coworkers observed contin-ued ventral compression after posterior decompressionwhen the lordosis was less than 10 degrees and the extentof ventral canal encroachment exceeded 7 mm in patientspresenting with ossification of the posterior longitudinal lig-ament [35]. Under these questionable circumstances, the de-cision often rests on the surgeon’s clinical experience andprocedural bias as there are no defined standards of treatment.After the decision has been made to perform a multilevel

Table 1

Indications for anterior and posterior approaches for

cervical spondylotic myelopathy

Anterior approach Posterior approach

Primarily anterior compression Primarily posterior compression

1-2 level of disease $3 levels of disease

Kyphotic deformity Normal or lordotic deformity

Absence of congential canal

stenosis

Presence of congenital canal stenosis

Associated radiculopathy No radiculopathy

Absence of OPLL Presence of OPLL

255SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

laminectomy, the surgeon must then decide whether the addi-tion of a posterior cervical fusion is indicated.

Indications for posterior cervical fusion

The primary surgical goals when performing a posteriorcervical stabilization and fusion include restoring stability,maintaining alignment, providing stability until fusion hasmatured, and alleviation of pain. These constructs providestability by reinforcing the posterior tension band that iscompromised by the pathologic process or surgical decom-pression. Determining the presence and extent of instabilityrests on careful assessment of both static and dynamic im-aging. There have been numerous models and definitionscreated in an attempt to identify and quantify spinal insta-bility [36,37]. Although these models provide a foundationfor operative decision making, each has limitations andcannot be universally applied to all clinical situations. Froma practical standpoint, the presence of instability must bedetermined on an individual basis. Findings on radiographsthat are suggestive of instability include: subluxation ofmore than 3.5 mm on static lateral radiographs; more than11 degrees of angulation between adjacent segments; and

Fig. 1. Sagittal T2-weighted magnetic resonance imaging demonstrating

spinal cord compression and characteristics considered appropriate for

a multilevel laminectomy approach including a lordotic alignment, the

presence of posterior compressive pathology, and multilevel degenerative

disease.

more than 4 mm of subluxation on dynamic views [17]. Un-der these circumstances a posterior decompression withoutstabilization is likely to lead to a progressive deformity thatwill require intervention at a future date.

Although not an absolute indication for supplementalposterior stabilization, the presence of a straight cervicalspine should alert the surgeon to an increased potentialfor a postlaminectomy kyphosis. The extent of decompres-sion may prove more significant under these circumstances,and has been correlated to postoperative spinal instability,deformity progression, accelerated spondylotic changes,and constriction of the dura mater by formation of extradur-al scar tissue [6,8,11,16,20,24,27,32,38]. In vitro studieshave demonstrated significant mobility and potential forpostoperative instability with the resection of 50% of thefacet complex [39,40]. Age has also been correlated tothe development of a postlaminectomy kyphosis, requiringthe surgeon to consider a stabilization when performinga multilevel decompression in a younger patient [41–43].

Contraindications to the posterior approach

Specific to the posterior approach, the presence of a fixedkyphotic deformity is an absolute contraindication whenconsidering a multilevel laminectomy, even with the inclu-sion of a posterior cervical arthrodesis. These patients oftenrequire a circumferential approach to adequately decom-press the neural elements, stabilize the spine, and optimizesagittal balance. A similar approach may be considered forpatients with severe osteoporosis, who require 360 degreestabilization to prevent construct failure. Chronic injuryto the spinal cord, as demonstrated on preoperative mag-netic resonance imaging, is also considered a contraindica-tion. Imaging characteristics consistent with intramedullarycystic necrosis, syrinx formation, or myelomalacia indicatepermanent cord injury that is unlikely to improve withoperative intervention [44].

Relative contraindications, not specific to the posteriorapproach, include a history of significant medical comor-bidity, such as older age (O70 years), diabetes, coronarydisease, obstructive pulmonary disease, peripheral vasculardisease, stroke, and social history of tobacco use or alcohol-ism [45–47]. Under these circumstances the risks of ope-rative intervention often do outweigh the benefits,particularly if there is no chance of neurologic recovery.

Surgical technique

Patient positioning/operative setup

An awake, fiberoptic intubation is preferred for symp-tomatic patients with severe cord compression. During theinduction of general anesthesia the natural protective mech-anisms resisting neck motion are inhibited, allowing unre-stricted cervical manipulation that could result in spinal

256S R.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

cord injury. Accurate blood pressure monitoring is requiredto avoid hypotension that could result in spinal cord ische-mia or infarction. Before patient positioning, baseline elec-trophysiological monitoring, including somatosensoryevoked potentials, motor evoked potentials, and free run-ning electromyography, is obtained, particularly for pa-tients with severe cord compression. Although the valueof electrophysiological monitoring is debatable, it can pro-vide useful information during reversible maneuvers, suchas patient positioning [48–50]. Monitoring, however,should not be regarded as an insurance policy for poor sur-gical technique.

The patient is positioned prone with the arms tucked atthe sides and appropriate padding to prevent pressure neu-ropathies. A Mayfield head holder or tongs with traction areused to secure the head (Fig. 2). A neutral position isfavored because prolonged periods in either a flexed orextended posture may not be tolerated, especially in thepresence of severe spinal cord compression. A second trac-tion line is set up to extend the neck and maximize lordosisonce neurologic decompression is achieved. After final po-sitioning, repeat electrophysiological monitoring is per-formed. If a change is identified, factors that may affectpotentials such as anesthetics or alterations in blood

pressure should be verified and corrected. Neck positionshould be rechecked and returned to a more neutral posi-tion. Intraoperative fluoroscopic imaging is useful to verifycervical alignment after final positioning and confirm spinallevel during the operative procedure. Also alignment maybe rechecked during the procedure to optimize sagittal bal-ance before any stabilization is performed.

Multilevel laminectomy

Once the patient is positioned and skin localization con-firmed, the spine is exposed through a midline incision.Maintaining a midline approach in the avascular fascialplane will help decrease blood loss and minimize postoper-ative pain. The paraspinal muscles are elevated in a subper-iosteal fashion using monopolar cautery. Dissection of thefacet joints is completed if an arthrodesis is intended. Careis taken not to disrupt the facet capsule of uninvolved adja-cent segments or if only a laminectomy is intended (Fig. 3).Localization can be performed with anatomic landmarks,such as the prominent C2 spinous process; however, confir-mation with intraoperative imaging is recommended.

Resection of the lamina en bloc is the favored approach.If required, keyhole foraminotomies are performed before

Fig. 2. Patient placed in a prone position for multilevel laminectomy and arthrodesis with Gardner Wells tongs and traction. Initially a neutral alignment is

preferred; however, a second traction line, with an upward directed vector, will enhance lordosis once the decompression is completed by placing the neck in

extension.

257SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

Fig. 3. The paraspinal muscle and soft-tissue dissection necessary to expose the posterior cervical spine are depicted in a schematic and corresponding intra-

operative photograph. If no fusion is intended, violation of the facet capsule should be avoided.

resection of the lamina so that the lamina can act as a pro-tective barrier during drilling of the foramen. The en blocresection tends to be less time-consuming and avoids theinsertion of any instrument into the central canal beforethe decompression. Using a high-speed drill equipped witha matchstick or small round burr, troughs are drilled bilat-erally at the lamina–facet junction. The drill is held perpen-dicular to the dorsal surface of the lamina to provide theshortest route to the canal and avoid drilling of the lateralmass (Fig. 4). Drilling initially passes through the dorsalcortex followed by the soft cancellous core, and finally thedense ventral cortex of the lamina. Bleeding from the medialwall of the trough is easily controlled with wax. Althougha diamond bit may be used, this tends to slow the processand generate more heat. Regardless of the drill bit used, con-tinuous irrigation is recommended to avoid thermal injury.The rostral aspect of the lamina tends to dive deep and re-quires more extensive drilling to completely disarticulatethe lamina. Entry into the epidural space is confirmed withgentle palpation with either the drill bit or nerve hook. Theventral cortex may also be resected by inserting a small foot-plated rongeur into the trough.

Once the bony troughs are completed, the spinous pro-cesses at the rostral and caudal extremes of the decompres-sion are elevated with clamps and the ligamentousattachments of the ligamentum flavum within the troughs

are stripped with rongeurs. Epidural bleeding encounteredduring this maneuver is easily controlled with bipolar cau-tery and gelfoam. Symmetric elevation of the lamina is re-quired to avoid levering the bone into the spinal canal(Fig. 5). Once all ligamentous attachments are resected,the entire complex is removed. The underlying dura is in-spected and epidural bleeding controlled. If no fusion is in-tended, a hemovac drain is placed to prevent postoperativehematoma and a multilayer closure is performed. The he-movac drain is typically removed within 24 hours of the op-eration. Patients may be placed in a soft cervical collar forcomfort measure for several weeks.

Posterior cervical stabilization and fusion

The earliest attempts at stabilizing the subaxial cervicalspine involved the placement of autologous bone along theposterior elements; however, this approach lacked immedi-ate stability and required prolonged periods of immobiliza-tion. Wiring techniques were first described in 1891 byHadra [51], and since that time a variety of posterior wiringtechniques have been developed that have demonstrated ac-ceptable fusion potential with relatively low complicationrates [52–56]. Postlaminectomy wiring techniques requiredthe wire to be passed between adjacent facets or tied to ei-ther bone grafts or metallic rods (Fig. 6) [57,58]. Although

258S R.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

Fig. 4. The en bloc laminectomy approach avoids placement of instruments into the central canal before the decompression. Bilateral troughs are drilled at

the lamina–facet junction with the drill bit oriented perpendicular to the dorsal surface of the lamina to avoid violation of the lateral mass.

facet wiring techniques have largely been replaced by morerigid, newer generation constructs, wiring continues to bea viable option.

Over the last several decades the emergence of lateralmass fixation has become the stabilization technique ofchoice [59–63]. Lateral mass fixation has proven to be a safeand biomechanically superior technique when comparedwith wiring procedures. Gill et al. demonstrated the in-creased rigidity and fusion potential with lateral mass fixa-tion compared with posterior wiring techniques [64]. Theintrinsic strength of the lateral mass screw provides immedi-ate stabilization, allowing early mobilization and possiblyeliminating the need for external bracing. The placementof cervical pedicle screws has also been described, demon-strating superior fixation when compared with lateral massand anterior plating techniques [65]. Despite the biomechan-ical rationale for inserting cervical pedicle screws, this tech-nique is not routinely practiced because of the technicaldemands and the potential neurovascular complications [66].

Lateral mass fixation

Fixation to the lateral masses has evolved from earlygeneration screw-plate constructs to more versatile multi-axial screw-rod constructs. Plating systems were generallyunyielding because of the predetermined position of thescrew hole within the plate. These constructs could notadapt well to variations in the patient’s anatomy, instead

the patient’s anatomy was forced to conform to the con-struct. Because of the rigidity of the plate in the coronalplane, all screws had to be positioned in line. In addition,points of fixation were sacrificed if the plate hole did notmatch the entry site within the lateral mass (Fig. 7). Dueto the dynamic interface between the screw and plate, thepotential to maintain reduction was limited and required bi-cortical screw purchase to obtain maximal stability [67].Surface area for fusion formation was compromised be-cause the plate was positioned flush against the dorsal wallof the lateral mass. Finally, revision surgery required re-moval of all the screws in order to disengage the plate.These deficiencies have largely been overcome with theevolution of multi-axial screw rod constructs [62].

Various lateral mass screw insertion techniques havebeen described in the literature (Fig. 8). The Roy-Camilletechnique uses the midpoint of the lateral mass as the inser-tion site and directs the screw without any rostral or caudalangulation and 10 degrees laterally [68]. Jeanneret and Ma-gerl describe a starting point 1 to 2 mm medial and superiorto the midpoint of the lateral mass with the trajectory aimed30 degrees superior and 15 to 25 degrees lateral [69]. Anand colleagues recommended a starting point 1 mm medialto the lateral mass center and angled 15 degrees cranial and30 degrees lateral [70]. Finally, Anderson et al. modifiedthe Magerl technique with a starting point 1 mm medialto the lateral mass center and aimed 20 to 30 degrees cra-nially and 10 to 20 degrees laterally [71].

259SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

Several studies have investigated the biomechanicalcharacteristics and anatomic relationships of the variouslateral mass screw techniques [72–75]. Montesano and co-workers demonstrated a 40% greater pullout strength whenscrews were placed using the Magerl compared with the

Fig. 5. The posterior elements, including the lamina and spinous pro-

cesses, are resected en bloc making sure the rostral and caudal ends are

simultaneously elevated to avoid leveraging one end into the spinal canal.

Curettes or rongeurs are inserted into the troughs to release any ligamen-

tous tissue tethering the lamina.

Roy-Camille technique [74]. From a clinical perspective,however, there has been no difference in the complicationrate between the different techniques [73]. Bicortical pur-chase of the lateral mass is associated with greater compli-cations without a clear biomechanical advantage [75,76].Ultimately measurement of exact angles for screw place-ment is impractical, and a trajectory with a bias in the me-dial to lateral direction with a rostral angulation startingclose to the lateral mass midpoint will create an appropriatescrew trajectory (Fig. 9). The surgeon needs to developa clear understanding of the anatomy, perform an appropri-ate exposure, and study preoperative imaging in order toobtain the optimal screw insertion.

Exposure for lateral mass fixation requires soft-tissuedissection to be extended until the far lateral margin ofthe facet is identified. To avoid compromise of uninvolvedsegments, care must be taken to maintain the facet cap-sule at the rostral and caudal extremes of the intended fu-sion. Lateral mass landmarks must be clearly identified forappropriate placement of screws. The medial border is de-fined by the valley at the lamina–lateral mass junction. Therostral and caudal boundaries are defined by the adjacent facetjoints; and the lateral boundary by the exposed lateral edge.Although the relationship of the vertebral artery and nerveroot to the subaxial lateral masses is generally consistent, ap-propriate screw placement requires a detailed understandingof these anatomical relationships. The vertebral artery lies an-terior to the lamina–lateral mass junction whereas the nerveroot passes anterolaterally, deep to the caudal superior facet.

Once the soft-tissue exposure is completed, entry sitesfor the lateral screws are marked by scoring the dorsalcortical wall with a high-speed drill. This site is typicallyin close proximity to the midpoint of the lateral mass. Thedrill bit is placed into the entry site, perpendicular to thedorsal cortex. Once the bit engages the bone, the trajec-tory is altered aiming from the infero-medial quadrant to-ward the supero-lateral quadrant, away from the nerveroot and vertebral artery. Resection of the spinous

Fig. 6. Wiring techniques used to stabilize the cervical spine after a multilevel laminectomy include passage of the wire between adjacent facets, securing

individual wires to a structural graft, or to a rigid rod, such as the Hartshill rectangle.

260S R.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

Fig. 7. This anterior-posterior radiograph demonstrates how plate con-

structs have a limited ability to conform to the surgical anatomy. The right

sided screw hole second from the top is positioned over the facet joint,

eliminating the possibility for screw placement.

processes may be necessary to obtain the optimal trajec-tory (Fig. 10). Laminectomies are not performed until af-ter the holes are drilled to protect the dura and neuralelements. Once completed, the screw holes may be tappedand filled with bone wax to prevent excessive bleeding.The lamina are then resected with the en bloc technique(Fig. 11). If a fusion is performed, all bone resected issaved as autograft. Lateral mass screws are inserted intothe predrilled holes aiming in a supero-lateral direction(Fig. 12). Malleable rods are cut and contoured, and lock-ing mechanisms are engaged. Rods should be contouredso that significant force is not required to seat the rodwithin the screw. The posterolateral spine, including thefacet joints, is decorticated and the grafting material is im-pacted across the spine and into the facet joints (Fig. 13).Closure is performed in the same manner as for anisolated laminectomy.

Cervical pedicle screws

Specific indications for cervical pedicle screws havenot been defined. With the exception of C7, the placementof subaxial cervical pedicle screws is not routinely per-formed due to the potential for catastrophic neurovascularinjury, morphologic characteristics and variation of thecervical pedicles [77], the technical difficulty involved with

Fig. 8. The entry site for the screw with the Roy-Camille method is the midpoint of the lateral mass, perpendicular to the dorsal cortex in the sagittal plane.

The screw is directed laterally by 10�. Using the Magerl technique the screw entry site is located slightly medial and cranial to the midpoint of the lateral

mass. The screw trajectory is parallel to the facet joint in the sagittal plane and directed 25� laterally in the transverse plane. The entry site for the Anderson

technique is located approximately 1 mm medial to the lateral mass midpoint with a rostral angulation of 30–40� and a lateral angle of 10�. Finally, the An

technique uses an entry site similar to the Anderson technique but only 15� of rostral angulation and a lateral angulation of 30�. These trajectories are in-

tended for placement of screws into the lateral masses of C3 to C6. (Reprinted with permission from Barrey et al. [72] and Xu et al. [75])

261SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

Fig. 9. The drawing demonstrates the general trajectory required for effective and safe lateral mass screw placement. A lateral trajectory directs the screw

away from the transverse foramen and vertebral artery, while the rostral angle avoids the nerve root traversing deep to the superior facet of the caudal spinal

segment. Bone volume is sufficient with this trajectory to enhance screw purchase.

insertion, and the lack of clinical data supporting the superi-ority of cervical pedicle screws over lateral mass screws. De-spite the technical difficulties, placement of cervical pediclescrews may serve as a pragmatic solution under unusual cir-cumstances where lateral mass fixation cannot be achieved.In animal and cadaveric biomechanical studies, the

placement of cervical pedicle screws has demonstrated supe-rior stability, fixation, and reduction potential when com-pared with lateral mass fixation [65,78].

In vitro studies have demonstrated an unacceptable po-tential for injury when depending only on anatomic topog-raphy for the placement of cervical pedicle screws [79,80].

Fig. 10. Lamina are left intact during drilling of the lateral masses to act as a protective barrier. The prominent spinous processes can obstruct the trajectory

for appropriate lateral mass drilling; therefore resection will allow the optimal screw trajectory. Once drilling of the lateral masses is completed, the lamina

are resected.

262S R.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

The inability for accurate insertion based on visualized an-atomic landmarks is primarily due to the inherent morpho-logic variability of the cervical pedicle trajectories and theminimal margin of error allowed for appropriate screwplacement (Fig. 14). Increased accuracy has been observedwith direct exposure of the pedicle through a laminoforami-notomy and the addition of computer-assisted navigationaltechniques [79–81]. Even with a laminoforaminotomy,however, in vitro studies have demonstrated a 39.6%incidence of critical violation, with possible nerve root orvertebral artery injury [66].

Abumi et al. have published extensively on the place-ment of cervical pedicle screws [66,82–86]. They chosean entry site just lateral to the midpoint of the lateral massin close proximity to the posterior edge of the superior ar-ticular surface. The cortical bone is drilled away to exposethe dorsal cancellous bone of the pedicle. Trajectorythrough the pedicle is created with a hand-held probe witha lateral to medial inclination ranging from 25 to 45 de-grees. Cannulation of the pedicle is performed with real-time fluoroscopic imaging to supplement tactile feedback.Using this technique, Abumi and colleagues reported ratesof pedicle violation ranging from 5.3% to 6.9%; however,no injury to the vertebral artery was observed. Using thesame technique, Ludwig et al. [97] reported a 12% rate

Fig. 11. An intraoperative photograph showing the thecal sac after the

lamina has been resected and the lateral masses drilled. Bleeding in the

epidural space is controlled with bipolar cautery and hemostatic agents,

such as gelfoam. Bone wax or gelfoam are inserted into the screw holes

to prevent continuous bleeding during the laminectomies.

of critical injury to the vertebral artery, nerve root, or spinalcord in a cadaveric model.

Based on the available literature, it seems that successfulplacement of cervical pedicle screws not only requiresa comprehensive understanding of the three-dimensionalpedicle anatomy and the surrounding neurovascular tissues,but also requires supplementation with direct visualizationof the pedicle or utilization of advanced navigational tech-niques. Despite these maneuvers, considerable risks remainand more conventional methods of posterior cervical stabi-lization should strongly be considered before the placementof these implants.

Complications

Iatrogenic deformity, as a result of a previous laminec-tomy, is considered one of the more common causes ofa progressive cervical kyphosis. Because of the increasedflexibility of the cervical spine, stability is more depen-dent on the integrity of the posterior muscles and liga-ments. The posterior exposure causes denervation andatrophy of the posterior cervical musculature and compro-mises the ligamentous capsule of the facet joints[42,87,88]. The incidence of kyphosis after dorsal cervi-cal procedures has been reported to be as high as 21%[88]. As many as 53% of patients under 18 years ofage who undergo a multilevel laminectomy will developa progressive kyphosis [41]. Younger patients are at anincreased risk as a result of incomplete ossification ofthe vertebral bodies and reduced resistance to the incre-mental compressive forces that develop with compromiseof the posterior tension band [42,43]. Additional risk fac-tors include compromised preoperative sagittal balance,extent of posterior element disruption, and the numberand location of lamina resected [42,89–91]. Cadavericstudies have demonstrated a significant increase in cervi-cal mobility after resection of more than 50% of the facetcomplex [39,40].

Neurologic complications can be divided into cord androot injuries. The incidence of spinal cord injury rangesfrom 0% to 3%, whereas injury to an individual nerve roothas been reported as high as 15% [47,92]. Technical error,introduction of instruments under the central lamina, aswell as ischemic injury due to hypoperfusion or aggres-sive distraction can lead to irreversible spinal cord injury.Nerve root injury may result from direct manipulation buthas also been attributed to the dorsal translation of the spi-nal cord after midline decompression [47]. The resultingloss of motor function in the absence of sensory losshas been termed dissociated motor loss. Radicular com-plaints are more commonly described in the C5 and C6nerve root distributions. Symptom onset typically occurswithin hours to days after surgery, with specific painand weakness. Spontaneous recovery is the general rule;however, deficits may persists for months. Isolated reportshave been published that demonstrate neurologic recovery

263SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

Fig. 12. Once the decompression is completed, the lateral mass screws are inserted. The screws are inserted with a rostral and lateral angulation with respect

to the lateral mass. The computed tomographic image demonstrates appropriately placed screws and their relationship to the transverse foramen.

with aggressive posterior decompression after onset ofsymptoms [25].

Complications associated with lateral mass fixation re-sult from impingement of the traversing nerve root or pen-etration of the vertebral artery. Potential for nerve rootinjury is increased when attempting bicortical purchase.Cadaveric studies have investigated the potential for inap-propriate placement with the various fixation techniques[75,76]. Seybold and colleagues using a modified Magerltechnique, demonstrated a 17.4% incidence of nerve rootimpingement and 5.8% incidence of potential vertebralartery injury when placing bicortical screws [76]. No suchinjuries were recorded with unicortical screws, and nosignificant difference in pullout strength was observed be-tween unicortical and bicortical screws. Xu and coworkersobserved a 95%, 90%, and 60% incidence of nerve rootviolation with the Magerl, Anderson, and An techniques re-spectively [75]. In this study a fixed screw length of 20 mmwas used to ensure bicortical purchase. The Magerl tech-nique commonly violated the dorsal ramus, whereas theAn technique violated the ventral ramus.

The risk of injury is not only related to the technique butalso to the morphology of the lateral masses. Barrey et al.observed an increased incidence of poor screw position atC3–C4 with the Magerl technique and at C5–C6 with theRoy-Camille technique [72]. This variation was attributedto the change in the height/thickness ratio as one descends

down the spine, with more caudal lateral masses becominglonger and thinner. These authors concluded that variationsin the technique of lateral mass screw placement should beconsidered to achieve optimal purchase and reduce inci-dence of nerve injury. Despite the potential for neurovascu-lar injury, most studies have demonstrated that fixation tothe lateral masses is a safe and effective means of posteriorcervical stabilization [93].

Insertion of cervical pedicle screws is associated withrisks to the vertebral artery, nerve root, and spinal cord. Be-cause of the technical difficulty, small size, and morpholog-ical variation in the cervical pedicle, the risk is consideredgreater than observed with lateral mass fixation. The greatestpotential for injury is to the vertebral artery injury owing tothe steep lateral to medial inclination required for appropriatescrew placement. Abumi and Kaneda observed a 6.7% inci-dence of pedicle penetration; however, only 4% of thesescrews produced a clinically significant radiculopathy [66].In vitro studies have demonstrated rates of critical pedicle vi-olation as high as 65.5% depending on the insertion tech-nique [79,81]. A greater degree of accuracy was achievedwhen direct visualization of anatomic landmarks was supple-mented with computer-assisted stereotactic navigation tech-niques. Pedicle screw placement is generally reserved forlevels with larger diameter pedicles, such as C7, when cross-ing the cervicothoracic junction, or when fixation to the lat-eral masses is not possible.

264S R.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

Fig. 13. The final stabilization construct is depicted in the schematic and intraoperative photograph. The lateral masses and facet joints are decorticated and

the bone graft impacted along these surfaces. To enhance stability, the newer generation screw-rod constructs allow insertion of cross-links to create a rect-

angular construct. Placing bone graft along the exposed dural surfaces, especially along a decompressed nerve root, should be avoided to prevent the pos-

sibility of recurrent stenosis.

Outcomes

Under the appropriate conditions, a multilevel laminec-tomy with or without arthrodesis is an effective

management strategy when treating cervical myelopathy.Most clinical series are retrospective in nature and describean individual surgeon’s experience. Laminectomy asa stand-alone procedure has demonstrated comparable

Fig. 14. The drawings demonstrate the trajectories required for appropriate cervical pedicle screw placement in both the sagittal and axial planes. A steep lateral to

medial inclination and the significant variability of the pedicle trajectory make the insertion of cervical pedicle screws technically demanding. Potential for neurovas-

cular injury exists because of the proximity of the vertebral artery, nerve root, and spinal cord. (Reprinted with permission from Ladd et al. [98] and Abumi et al. [99])

265SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

immediate postoperative results to anterior procedures andlaminoplasty [18]. Other series report a significant inci-dence of delayed deterioration, with rates as high as 40%[10,18,94].

Herkowitz, in 1988, compared results of patients under-going anterior decompression and strut grafting, lamino-plasty, and laminectomy without fusion [15]. Only 66%of patients after laminectomy reported good to excellentoutcomes, whereas the anterior group and laminoplastygroup reported good to excellent outcomes in 92% and86% respectively. Twenty-five percent of the patients inthe laminectomy group developed a postlaminectomykyphosis within 2 years of the operation.

Outcomes after multilevel laminectomy and arthrodesisfor cervical degenerative disease have been favorable[71,95,96]. Long-term improved neurologic status andmaintenance of cervical alignment have been demonstrated[3]. Retrospective studies have compared the complexity,safety, and clinical results of laminectomy versus corpec-tomy in patients with multilevel cervical spondylosis[6,9,11,15,18,26,28]. Whereas anterior decompression andfusion procedures at one or two motion segments have pre-dictable results, procedures involving three or more levelsare associated with increased morbidity. Outcome parame-ters have included operative time, blood loss, length of stay,and rate of complications. These investigations have con-cluded that posterior decompression by laminectomy formultilevel stenosis (O2 levels) requires shorter operativetime and results in fewer complications than anterior cervi-cal decompression with similar clinical results. Factors thathave been shown to correlate with poor outcome after lam-inectomy for CSM include advanced age (O70 years at thetime of the first surgery), severe original myelopathy, andrecent trauma [7,20,32,38]. In short, laminectomy, with orwithout fusion, may be successfully employed to addressmultilevel cervical pathology in carefully selected patients.

Cervical laminectomy with posterior fusion is valuablein cases of CSM when preoperative dynamic imaging dem-onstrates instability. Several retrospective studies have fo-cused on this topic. One review by Huang et al. involving32 patients with multilevel CSM treated by laminectomyand lateral mass plate fusion confirmed the clinical utilityof this procedure [38]. The etiology of disease in thesepatients was either cervical spondylosis or ossification ofthe posterior longitudinal ligament, the mean postoperativefollow-up was over 15 months, and in all cases myelopathywas graded using the Nurick system. The authors noted themean Nurick score improved from 2.6 (range 1–4) to 1.8(range 0–3) postoperatively (p!.0001). More specifically,22 patients (71%) had an improvement in Nurick gradeof at least one point and nine showed no improvement.Of note, no patients had neurologic deterioration. Severemyelopathy, advanced age, and myelomalacia on preopera-tive magnetic resonance imaging had no prognostic valuefor improvement of myelopathy. Overall, this study demon-strated that laminectomy and fusion may result in excellent

outcomes for carefully selected patients with multilevelcervical myelopathy.

Conclusions

Under the appropriate conditions, a multilevel cervicallaminectomy with or without a supplemental fusion is aneffective management strategy for the treatment of CSM.Careful preoperative assessment of the pathologic anatomy,clinical presentation, and medical comorbidity is essentialto achieve operative success. Patients requiring surgerywith evidence of a reducible deformity or static lordosisare ideal candidates for a posterior decompression. The po-tential or presence of instability must be recognized andprevented with a supplemental stabilization technique. Lat-eral mass fixation is the stabilization procedure of choice,demonstrating superior biomechanical properties, technicalsimplicity, and low complication rate. These operative tech-niques are essential to the surgeon’s armamentarium whentreating patients with CSM.

References

[1] Edwards CC 2nd, Riew KD, Anderson PA, Hilibrand AS,

Vaccaro AF. Cervical myelopathy: current diagnostic and treatment

strategies. Spine J 2003;3:68–81.

[2] Emery SE. Cervical spondylotic myelopathy: diagnosis and treat-

ment. J Am Acad Orthop Surg 2001;9:376–88.

[3] Kumar VG, Rea GL, Mervis LJ, McGregor JM. Cervical spondylotic

myelopathy: functional and radiographic long-term outcome after

laminectomy and posterior fusion. Neurosurgery 1999;44:771–7; dis-

cussion 777–8.

[4] McCormack BM, Weinstein PR. Cervical spondylosis. An update.

West J Med 1996;165(1–2):43–51.

[5] McCormick WE, Steinmetz MP, Benzel EC. Cervical spondylotic

myelopathy: make the difficult diagnosis, then refer for surgery.

Cleve Clin J Med 2003;70:899–904.

[6] Mehdorn HM, Fritsch MJ, Stiller RU. Treatment options and results

in cervical myelopathy. Acta Neurochir Suppl 2005;93:177–82.

[7] Smith-Hammond CA, New KC, Pietrobon R, Curtis DJ,

Scharver CH, Turner DA. Prospective analysis of incidence and risk

factors of dysphagia in spine surgery patients: comparison of anterior

cervical, posterior cervical, and lumbar procedures. Spine 2004;29:

1441–6.

[8] Wiberg J. Effects of surgery on cervical spondylotic myelopathy.

Acta Neurochir (Wien) 1986;81(3–4):113–7.

[9] Wong AS, Massicotte EM, Fehlings MG. Surgical treatment of cervi-

cal myeloradiculopathy associated with movement disorders: indica-

tions, technique, and clinical outcome. J Spinal Disord Tech

2005;18(Suppl):S107–14.

[10] Snow RB, Weiner H. Cervical laminectomy and foraminotomy as

surgical treatment of cervical spondylosis: a follow-up study with

analysis of failures. J Spinal Disord 1993;6:245–50; discussion

250–1.

[11] Sevki K, Mehmet T, Ufuk T, Azmi H, Mercan S, Erkal B. Results of

surgical treatment for degenerative cervical myelopathy: anterior cer-

vical corpectomy and stabilization. Spine 2004;29:2493–500.

[12] Ozgen S, Naderi S, Ozek MM, Pamir MN. A retrospective review of

cervical corpectomy: indications, complications and outcome. Acta

Neurochir (Wien) 2004;146:1099–105; discussion 1105.

[13] Nakamura K, Kurokawa T, Hoshino Y, Saita K, Takeshita K,

Kawaguchi H. Conservative treatment for cervical spondylotic

266S R.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

myelopathy: achievement and sustainability of a level of ‘‘no disabil-

ity’’. J Spinal Disord 1998;11:175–9.

[14] Huang JJ, Niu CC, Chen LH, Lai PL, Fu TS, Chen WJ. Anterior cer-

vical spinal surgery for multilevel cervical myelopathy. Chang Gung

Med J 2004;27:531–41.

[15] Herkowitz HN. A comparison of anterior cervical fusion, cervical

laminectomy, and cervical laminoplasty for the surgical management

of multiple level spondylotic radiculopathy. Spine 1988;13:774–80.

[16] Epstein NE. Laminectomy for cervical myelopathy. Spinal Cord

2003;41:317–27.

[17] Epstein N, Epstein JA. Treatment of cervical myelopathy: Part A.

Laminectomy. In: The cervical spine. TCSR Society, editor. Philadel-

phia: Lippincott Williams & Wilkins, 2005:1043–56.

[18] Ebersold MJ, Pare MC, Quast LM. Surgical treatment for cervical

spondylitic myelopathy. J Neurosurg 1995;82:745–51.

[19] Clarke E, Robinson PK. Cervical myelopathy: a complication of cer-

vical spondylosis. Brain 1956;79:483–510.

[20] An HS, Ahn NU. Posterior decompressive procedures for the cervical

spine. Instr Course Lect 2003;52:471–7.

[21] Ono K, Ikata T, Yamada H, et al. Cervical myelopathy secondary to

multiple spondylotic protrusions. Spine 1977;2:218-21.

[22] Yuan Q, Dougherty L, Margulies SS. In vivo human cervical spinal

cord deformation and displacement in flexion. Spine 1998;23:

1677–83.

[23] Sawin P, Sonntag VKH. Techniques of posterior subaxial cervical fu-

sion. Operative Techniques in Neurosurgery 1998;1:72–83.

[24] Epstein N. Posterior approaches in the management of cervical spon-

dylosis and ossification of the posterior longitudinal ligament. Surg

Neurol 2002;58(3–4):194–207; discussion 207–8.

[25] O’Toole JE, Olson TJ, Kaiser MG. Surgical management of dissoci-

ated motor loss following complex cervical spine reconstruction.

Spine 2004;29:E56–60.

[26] Geck MJ, Eismont FJ. Surgical options for the treatment of cervical

spondylotic myelopathy. Orthop Clin North Am 2002;33:329–48.

[27] Hohmann D, Liebig K. [Surgical therapy of spondylogenic cervical

myelopathy: indications and techniques]. Orthopade 1996;25:

558–66.

[28] Casella E, Chiappetta F, Dell’Aquila G, Fiume D, Massari A,

Scarda G. [Surgical treatment of cervical spondylogenic myelopathy

(surgical indications and long-term results)]. Riv Neurobiol 1979;25:

435–48.

[29] Benzel EC, Lancon J, Kesterson L, Hadden T. Cervical laminectomy

and dentate ligament section for cervical spondylotic myelopathy.

J Spinal Disord 1991;4:286–95.

[30] Epstein NE. Laminectomy with posterior wiring and fusion for cervi-

cal ossification of the posterior longitudinal ligament, spondylosis,

ossification of the yellow ligament, stenosis, and instability: a study

of 5 patients. J Spinal Disord 1999;12:461–6.

[31] Heller JG, Edwards CC 2nd, Murakami H, Rodts GE. Laminoplasty

versus laminectomy and fusion for multilevel cervical myelopathy:

an independent matched cohort analysis. Spine 2001;26:1330–6.

[32] Yonenobu K, Fuji T, Ono K, Okada K, Yamamoto T, Harada N.

Choice of surgical treatment for multisegmental cervical spondylotic

myelopathy. Spine 1985;10:710–6.

[33] Benzel EC. Degenerative and inflammatory diseases of the spine. In:

Benzel EC, editor. Biomechanics of spine stabilization. Chicago,

Illinois: American Association of Neurological Surgeons, 2001:

45–60.

[34] Hamanishi C, Tanaka S. Bilateral multilevel laminectomy with or

without posterolateral fusion for cervical spondylotic myelopathy: re-

lationship to type of onset and time until operation. J Neurosurg

1996;85:447–51.

[35] Yamazaki A, Homma T, Uchiyama S, Katsumi Y, Okumura H. Mor-

phologic limitations of posterior decompression by midsagittal split-

ting method for myelopathy caused by ossification of the posterior

longitudinal ligament in the cervical spine. Spine 1999;24:32–4.

[36] Denis F. Spinal instability as defined by the three-column spine

concept in acute spinal trauma. Clin Orthop Relat Res 1984;189:

65–76.

[37] White AA 3rd, Johnson RM, Panjabi MM, Southwick WO. Biome-

chanical analysis of clinical stability in the cervical spine. Clin Or-

thop Relat Res 1975;109:85–96.

[38] Huang RC, Girardi FP, Poynton AR, Cammisa FP Jr. Treatment of

multilevel cervical spondylotic myeloradiculopathy with posterior

decompression and fusion with lateral mass plate fixation and local

bone graft. J Spinal Disord Tech 2003;16:123–9.

[39] Raynor RB, Pugh J, Shapiro I. Cervical facetectomy and its effect on

spine strength. J Neurosurg 1985;63:278–82.

[40] Zdeblick TA, Zou D, Warden KE, McCabe R, Kunz D, Vanderby R.

Cervical stability after foraminotomy: a biomechanical in vitro anal-

ysis. J Bone Joint Surg Am 1992;74:22–7.

[41] Bell DF, Walker JL, O’Connor G, Tibshirani R. Spinal deformity af-

ter multiple-level cervical laminectomy in children. Spine 1994;19:

406–11.

[42] Deutsch H, Haid RW, Rodts GE, Mummaneni PV. Postlaminectomy

cervical deformity. Neurosurg Focus 2003;15:E5.

[43] Yasuoka S, Peterson HA, Laws ER Jr, MacCarty CS. Pathogenesis

and prophylaxis of postlaminectomy deformity of the spine after mul-

tiple level laminectomy: difference between children and adults. Neu-

rosurgery 1981;9:145–52.

[44] Al-Mefty O, Harkey LH, Middleton TH, Smith RR, Fox JL. Myelo-

pathic cervical spondylotic lesions demonstrated by magnetic reso-

nance imaging. J Neurosurg 1988;68:217–22.

[45] Epstein NE. Circumferential surgery for the management of cervical

ossification of the posterior longitudinal ligament. J Spinal Disord

1998;11:200–7.

[46] Epstein NE. The value of anterior cervical plating in preventing ver-

tebral fracture and graft extrusion after multilevel anterior cervical

corpectomy with posterior wiring and fusion: indications, results,

and complications. J Spinal Disord 2000;13:9–15.

[47] Saunders RL, Pikus HJ, Ball P. Four-level cervical corpectomy. Spine

1998;23:2455–61.

[48] Uematsu S, Tolo V. Electrophysiological recordings during spinal

surgery. Stereotact Funct Neurosurg 1989;52(2–4):145–56.

[49] Kitagawa H, Itoh T, Takano H, et al. Motor evoked potential monitor-

ing during upper cervical spine surgery. Spine 1989;14:1078–83.

[50] Epstein NE, Danto J, Nardi D. Evaluation of intraoperative somato-

sensory-evoked potential monitoring during 100 cervical operations.

Spine 1993;18:737–47.

[51] Hadra B. Wiring the spinous process in Pott’s disease. Trans Am Or-

thop Assoc 1891;4:206–10.

[52] Brodsky AE, Khalil MA, Sassard WR, Newman BP. Repair of symp-

tomatic pseudoarthrosis of anterior cervical fusion. Posterior versus

anterior repair. Spine 1992;17:1137–43.

[53] Lovely TJ, Carl A. Posterior cervical spine fusion with tension-band

wiring. J Neurosurg 1995;83:631–5.

[54] Maurer PK, Ellenbogen RG, Ecklund J, Simonds GR, van Dam B,

Ondra SL. Cervical spondylotic myelopathy: treatment with posterior

decompression and Luque rectangle bone fusion. Neurosurgery

1991;28:680–3; discussion 683–4.

[55] Miyazaki K, Tada K, Matsuda Y, Okuno M, Yasuda T, Murakami H.

Posterior extensive simultaneous multisegment decompression with

posterolateral fusion for cervical myelopathy with cervical instabi-

lity and kyphotic and/or S-shaped deformities. Spine 1989;14:

1160–70.

[56] Weiland DJ, McAfee PC. Posterior cervical fusion with triple-wire

strut graft technique: one hundred consecutive patients. J Spinal Dis-

ord 1991;4:15–21.

[57] Garfin SR, Moore MR, Marshall LF. A modified technique for cervi-

cal facet fusions. Clin Orthop Relat Res 1988;230:149–53.

[58] Murphy MJ, Daniaux H, Southwick WO. Posterior cervical fusion

with rigid internal fixation. Orthop Clin North Am 1986;17:55–65.

267SR.J. Komotar et al. / The Spine Journal 6 (2006) 252S–267S

[59] Alexander E Jr. Posterior fusions of the cervical spine. Clin Neuro-

surg 1981;28:273–96.

[60] Cooper PR. Posterior stabilization of the cervical spine. Clin Neuro-

surg 1993;40:286–320.

[61] Domenella G, Berlanda P, Bassi G. Posterior-approach osteosynthesis

of the lower cervical spine by the R. Roy-Camille technique. (Indica-

tions and first results). Ital J Orthop Traumatol 1982;8:235–44.

[62] Horgan MA, Kellogg JX, Chesnut RM. Posterior cervical arthrodesis

and stabilization: an early report using a novel lateral mass screw and

rod technique. Neurosurgery 1999;44:1267–71; discussion 1271–2.

[63] Schwartz MS, Anderson GJ, Horgan MA, Kellogg JX,

McMenomey SO, Delashaw JB Jr. Quantification of increased expo-

sure resulting from orbital rim and orbitozygomatic osteotomy via the

frontotemporal transsylvian approach. J Neurosurg 1999;91:1020–6.

[64] Gill K, Paschal S, Corin J, Ashman R, Bucholz RW. Posterior plating

of the cervical spine: a biomechanical comparison of different poste-

rior fusion techniques. Spine 1988;13:813–6.

[65] Kotani Y, Cunningham BW, Abumi K, McAfee PC. Biomechanical

analysis of cervical stabilization systems: an assessment of transpe-

dicular screw fixation in the cervical spine. Spine 1994;19:2529–39.

[66] Abumi K, Kaneda K. Pedicle screw fixation for nontraumatic lesions

of the cervical spine. Spine 1997;22:1853–63.

[67] Heller JG, Estes BT, Zaouali M, Diop A. Biomechanical study of

screws in the lateral masses: variables affecting pull-out resistance.

J Bone Joint Surg Am 1996;78:1315–21.

[68] Roy-Camille R, Saillant G, Mazel C. Internal fixation of the unstable

cervical spine by posterior osteosynthesis with plates and screws. In:

The cervical spine. TCSR Society, editor. New York: Lippincott,

1989:390–403.

[69] Jeanneret B, Magerl F, Ward EH, Ward JC. Posterior stabilization of

the cervical spine with hook plates. Spine 1991;16(3 Suppl):S56–63.

[70] An HS, Gordin R, Renner K. Anatomic considerations for plate-screw

fixation of the cervical spine. Spine 1991;16(10 Suppl):S548–51.

[71] Anderson PA, Henley MB, Grady MS, Montesano PX, Winn HR.

Posterior cervical arthrodesis with AO reconstruction plates and bone

graft. Spine 1991;16(3 Suppl):S72–9.

[72] Barrey C, Mertens P, Jund J, Cotton F, Perrin G. Quantitative ana-

tomic evaluation of cervical lateral mass fixation with a comparison

of the Roy-Camille and the Magerl screw techniques. Spine 2005;30:

E140–7.

[73] Heller JG, Carlson GD, Abitbol JJ, Garfin SR. Anatomic comparison

of the Roy-Camille and Magerl techniques for screw placement in the

lower cervical spine. Spine 1991;16(10 Suppl):S552–7.

[74] Montesano PX, Jauch E, Jonsson H Jr. Anatomic and biomechanical

study of posterior cervical spine plate arthrodesis: an evaluation of

two different techniques of screw placement. J Spinal Disord

1992;5:301–5.

[75] Xu R, Haman SP, Ebraheim NA, Yeasting RA. The anatomic relation

of lateral mass screws to the spinal nerves: a comparison of the Ma-

gerl, Anderson, and An techniques. Spine 1999;24:2057–61.

[76] Seybold EA, Baker JA, Criscitiello AA, Ordway NR, Park CK,

Connolly PJ. Characteristics of unicortical and bicortical lateral mass

screws in the cervical spine. Spine 1999;24:2397–403.

[77] Panjabi MM, Duranceau J, Goel V, Oxland T, Takata K. Cervical hu-

man vertebrae: quantitative three-dimensional anatomy of the middle

and lower regions. Spine 1991;16:861–9.

[78] Jones EL, Heller JG, Silcox DH, Hutton WC. Cervical pedicle screws

versus lateral mass screws: anatomic feasibility and biomechanical

comparison. Spine 1997;22:977–82.

[79] Ludwig SC, Kramer DL, Balderston RA, Vaccaro AR, Foley KF,

Albert TJ. Placement of pedicle screws in the human cadaveric

cervical spine: comparative accuracy of three techniques. Spine

2000;25:1655–67.

[80] Ludwig SC, Kramer DL, Vaccaro AR, Albert TJ. Transpedicle

screw fixation of the cervical spine. Clin Orthop Relat Res 1999;359:

77–88.

[81] Ludwig SC, Kowalski JM, Edwards CC 2nd, Heller JG. Cervical ped-

icle screws: comparative accuracy of two insertion techniques. Spine

2000;25:2675–81.

[82] Abumi K, Itoh H, Taneichi H, Kaneda K. Transpedicular screw fixa-

tion for traumatic lesions of the middle and lower cervical spine: de-

scription of the techniques and preliminary report. J Spinal Disord

1994;7:19–28.

[83] Abumi K, Kaneda K, Shono Y, Fujiya M. One-stage posterior decom-

pression and reconstruction of the cervical spine by using pedicle

screw fixation systems. J Neurosurg 1999;90(1 Suppl):19–26.

[84] Abumi K, Shono Y, Ito M, Taneichi H, Kotani Y, Kaneda K. Compli-

cations of pedicle screw fixation in reconstructive surgery of the cer-

vical spine. Spine 2000;25:962–9.

[85] Abumi K, Shono Y, Kotani Y, Kaneda K. Indirect posterior reduction

and fusion of the traumatic herniated disc by using a cervical pedicle

screw system. J Neurosurg 2000;92(1 Suppl):30–7.

[86] Abumi K, Shono Y, Taneichi H, Ito M, Kaneda K. Correction of cer-

vical kyphosis using pedicle screw fixation systems. Spine 1999;24:

2389–96.

[87] Schlenk RP, Stewart T, Benzel EC. The biomechanics of iatrogenic

spinal destabilization and implant failure. Neurosurg Focus

2003;15:E2.

[88] Steinmetz MP, Kager CD, Benzel EC. Ventral correction of postsur-

gical cervical kyphosis. J Neurosurg 2003;98(1 Suppl):1–7.

[89] Guigui P, Benoist M, Deburge A. Spinal deformity and instability

after multilevel cervical laminectomy for spondylotic myelopathy.

Spine 1998;23:440–7.

[90] Herman JM, Sonntag VK. Cervical corpectomy and plate fixation for

postlaminectomy kyphosis. J Neurosurg 1994;80:963–70.

[91] Katsumi Y, Honma T, Nakamura T. Analysis of cervical instability

resulting from laminectomies for removal of spinal cord tumor. Spine

1989;14:1171–6.

[92] Yonenobu K, Hosono N, Iwasaki M, Asano M, Ono K. Neurologic

complications of surgery for cervical compression myelopathy. Spine

1991;16:1277–82.

[93] Wellman BJ, Follett KA, Traynelis VC. Complications of posterior

articular mass plate fixation of the subaxial cervical spine in 43 con-

secutive patients. Spine 1998;23:193–200.

[94] Kato Y, Iwasaki M, Fuji T, Yonenobu K, Ochi T. Long-term follow-

up results of laminectomy for cervical myelopathy caused by ossifi-

cation of the posterior longitudinal ligament. J Neurosurg 1998;89:

217–23.

[95] An HS, Coppes MA. Posterior cervical fixation for fracture and

degenerative disc disease. Clin Orthop Relat Res 1997;335:101–

11.

[96] Fehlings MG, Cooper PR, Errico TJ. Posterior plates in the manage-

ment of cervical instability: long-term results in 44 patients. J Neuro-

surg 1994;81:341–9.

[97] Ludwig SC, Kowalski JM, Edwards CC 2nd, Heller JG. Cervical ped-

icle screws: comparative accuracy of two insertion techniques. Spine

2000;25:2675–81.

[98] Ladd JE, Heller JG, Silcox HD, Hutton WC. Cervical pedicle screws

versus lateral mass screws: Anatomic feasibility and biomechanical

comparison. Spine 1997;22:977–82.

[99] Abumi K, Kaneda K. Pedicle screw fixation for nontraumatic lesions

of the cervical spine. Spine 1997;22:1853–63.