3tournal ofNeurology, Neurosurgery, and Psychiatry 1995;58:607-613

Midcervical central cord syndrome: numb andclumsy hands due to midline cervical discprotrusion at the C3-4 intervertebral level

Masashi Nakajima, Keizo Hirayama

AbstractEight patients with midline cervical discprotrusion at the C3-4 intervertebrallevel showed unusual clinical signs:numbness in the finger tips and palms,clumsiness of the hands, and a tighteningsensation at the midthoracic level. Theproprioceptive and cutaneous sensoryafferents essential for motor control ofthe upper limbs were preferentiallyinvolved, tactile discrimination of pas-

sively given stimuli being spared.Somatosensory evoked potentials subse-quent to median nerve stimulationshowed conduction failure through thefasciculus cuneatus, as evidenced byabsent or delayed and attenuatedmedullary and scalp potentials. Thepotential originating in the lower cervicalcord (N13a) had a low amplitude, indica-tive of the caudal extension of the lesion.On the basis ofthe functional anatomy ofthe intraspinal pathways, especially of thedorsal columns, it is concluded that involve-ment of the central cord at the C3-4intervertebral level and its caudal exten-sion is responsible for the syndrome.

(7 Neurol Neurosurg Psychiatry 1995;58:607-6 13)

Keywords: numb and clumsy hands; C3-4 disc;central cord syndrome

"Numb and clumsy hands" has beendescribed as a sign of high cervical spinal cordcompression,'-5 but whether selective involve-ment of lower cervical segments may becaused by a higher lesion is not clear. Weinvestigated the transverse and longitudinalextent of cord lesions clinically and electro-physiologically in eight patients who had numband clumsy hands due to midline cervical discprotrusion at C3-4.

Department ofNeurology, School ofMedicine, ChibaUniversity, 1-8-1Inohana Chuo-ku,Chiba 260, JapanM NakajimaK HirayamaCorrespondence to:Dr Masashi Nakajima,Department of Neurology,School of Medicine, ChibaUniversity, 1-8-1 InohanaChuo-ku, Chiba 260, Japan.Received 31 May 1994and in revised form20 October 1994.Accepted 14 December 1994

Patients and methodsPATIENTSTable 1 gives clinical findings for the eightpatients tested. The case records of five havebeen published elsewhere in preliminaryform.6 Numbness of both hands, especially ofthe finger tips and palms, was an early clinicalmanifestation. A similar sensation was notedin the soles of five of the patients. Clumsinessof the fingers and hands restricted daily activi-ties, especially in vision deprived situations.They could not fasten a collar button or pick

up a corn in a pocket. In the later stage of thecondition, seven patients noticed tight, bandlike sensations of deep pain at the midthoraciclevel that were circumferential in three andfrontal in four. Neck pain that occasionallyradiated to the back was experienced by threepatients. Five complained of unsteady gait,which was severe enough that patient 2required assistance; the others did not.

Homer's syndrome was present in six ofthe patients and two had acrocyanosis.Sensory and motor impairment that essen-tially was symmetric, predominated in theupper limbs. Pin prick and thermal sensationswere mildly diminished below the C3 derma-tome. Hypoalgesia occasionally was limited tothe lower cervical segments. Impairment ofcrude touch and vibration predominated inthe hands, but sometimes occurred up to theshoulders. Atrophy of the shoulder girdle andhands was noted in patient 4. Details of thehand dysfunctions are given in the results.

In the lower limbs, vibration sense waspreferentially impaired. Diminished pain sen-sation, weakness, and ataxia were present invarious combinations. These long tract signswere mild when present, except for the ataxiaof patient 2. Bladder disturbances wereabsent, except for patient 3, who had frequenturination. Muscle stretch reflexes were exag-gerated below the cervical segments in five ofthe patients, but were normal in patient 1 anddiminished in patient 2. Patient 6 had exag-gerated finger jerks but normal reflexes in thelower limbs.

All the patients showed a midline cervicaldisc protrusion at C3-4 with or withoutspondylotic changes on myelography or MRI.Spinal cord compression was anteroposteriorat this level (fig 1A left and 1B), being massivein five and moderate in three of the patients.At the lower cervical level, five patients hadmild spondylotic cord compression, the otherthree having no pathological findings.Electromyography showed chronic denerva-tion of the C5 innervated muscles in fivepatients, two of whom had changes thatextended to the lower cervical segments.Patient 7 had neurogenic changes in the lowerC8 rather than in C5.

Anterior cervical discectomy and fusion atC3-4 was performed for four of the patients.The outcome was excellent for two, paraes-thesia remaining only at the finger tips, andthey became free from clumsiness of thehands. Patient 2, who had a history of necktrauma and had the most severe compression,could undertake independent daily activities

607

on March 2, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.5.607 on 1 M

ay 1995. Dow

nloaded from

Nakajima, Hirayama

Table 1 Clinical and laboratory findings

History Signs in the upper limbsPatient Signs in the lower limbsAgelSex Onset Evolution Hypoalgesia Others and bladder

1 Numbness of finger Clumsiness of hands, tight band at C3-C8 Diminished vibration in Abolished vibration53 y/F tips T5, for 4 months hands

2 Numbness of Clumsiness of hands, gait C3-C8 Diminished touch, pressure, Diminished pain, vibration,58 y/M palms and soles 2 unsteadiness, tightness in abdomen and vibration below touch, and joint movement

months after neck at T10, then at T8, for 1 month elbows, mild weakness sense below knees, Rombergtrauma sign, ataxic gait

3 Nuchal pain and its Numbness and clumsiness of hands, C4-C8 Diminished touch and Diminished pain below cervical52 y/M electrical radiation gait unsteadiness, tightness of vibration below shoulders, level, abolished vibration

to back chest at T5, for 6 months mild weakness below knees, weakness andspasticity, Babinski sign,frequent micturition

4 Numbness of finger Dull aching around neck, clumsiness C4-C6 Diminished touch and Diminished vibration, mild48 y/F tips of hands, weakness of right arm, vibration in hands, spasticity

tightness of chest at T6, for 32 acrocyanosis, weaknessmonths and atrophy

5 Numbness and Acrocyanosis (right), numbness below C6-C8 Diminished touch and Diminished vibration and touch47 y/M clumsiness of knees, gait unsteadiness, tight band vibration in hands; mild in feet, mild spasticity,

hands at T8-9, for 3 months weakness and spasticity, Babinski signacrocyanosis

6 Dullness of neck Numbness and clumsiness of hands, C4-C8 Diminished touch and Diminished pain and vibration59 y/F and shoulders tight band at T6, numbness of vibration in hands, mild below cervical level

soles, for 8 months weakness

7 Numbness of palms Clumsiness of hands, gait C6-C8 Diminished vibration in Diminished pain below cervical55 y/M and soles unsteadiness, tightness of chest hands level, mild spasticity

at T6, for 36 months

8 Gait unsteadiness Numbness of palms and soles, C7-C8 Diminished touch in hands Ataxic gait, Romberg sign,76 y/F clumsiness of hands, for 12 months diminished vibration in feet,

Babinski sign

after surgery but considerable neurologicaldeficits remained. Generalised hyporeflexiawas converted to hyper-reflexia after decom-pression. For patient 4, the outcome forlaminectomy from C3 to C7 with foraminec-tomy at C5-6 on the right was fair. The tight,

Figure 1 MRI andmetrizamide CT ofpatient2. (A) MRIs before (left)and after (right)discectomy, showing thelevel of cord compression.(B) Preoperativemetrizamide CT at theC3-4 level shows massivecompression of the cervicalcord by a protruding softdisc.

band like sensation at the midthoracic leveldisappeared in all the patients who underwentsurgery.

EXAMINATION OF PROPRIOCEPTIVE ANDDISCRIMINATIVE HAND FUNCTIONS

Proprioception in the hands was examinedwith respect to the sense of passive and activemovements and position.7-9 The sense ofpassive movement was evaluated by theperception of the direction of movementgiven passively to the fingers. The sense ofactive movement was examined by thefinger-to-nose test with the eyes open. Thesense of position was evaluated with the eyesclosed using the finger-to-nose test and thewandering of outstretched fingers (pseudo-athetosis).

Tests of tactile discrimination included two

point discrimination and graphaesthesia"° inthe palms and finger tips. Stereoaesthesiallwas used to assess both proprioceptive andtactile discrimination in the hands.'2 In thestereoaesthesia test, the ability to identify theshapes, consistencies, and textures of smallobjects, and dexterity in manipulating an

object placed on the palm were evaluated.

SOMATOSENSORY EVOKED POTENTIALS

Somatosensory evoked potentials (SEPs) wererecorded after median nerve stimulation at thewrist (median SEPs) with a connected earsreference.'3 In addition, bipolar recordings ofshort latency SEPs'4 were made. Recordingelectrodes on the scalp at C3 and C4 (theinternational 10-20 system), the seventhcervical spine (Cv7), and Erb's point were ref-erenced to connected ears (A +). Antero-

608

on March 2, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.5.607 on 1 M

ay 1995. Dow

nloaded from

Midcervical central cord syndrome: numb and clumsy hands due to midline cervical disc protrusion at the C3-4 intervertebral level

Homer's C3-C4 discl Electrodiagnostic Surgery and outcomeReflexes syndrome Spondylosis Additional radiologicalfindings findings PostsurgicalfindingsNormal No Massive/No None Normal EMG Anterior, excellent.

Tingling of finger tips

Decreased in No Massive/No Venous angioma on the lumbar EMG not done; Anterior, fair.four limbs ampulla normal H reflexes Exaggerated reflexes, numbness and

clumsiness, diminished pain from C5 to C8

Exaggerated Yes (left) Massive/Yes C2-3 vertebral fusion; spondylotic Neurogenic EMG Anterior, good.below C5 narrow canal at C5-6 in C5 Diminished touch and vibration in hands

without hypoalgesia, residual leg dysfunction

Decreased in Yes Massive/Yes Narrow cervical canal; multiple Neurogenic EMG Posterior, fair.upper, and (bilateral) discs at C4-5 and C5-6 from C5 to C7 Numbness and clumsiness,increased in hypoalgesia in C6, diminished vibration inlower limbs hands and legs, weakness and atrophy in right

arm

Exaggerated Yes (left) Moderate/Yes Atlantoaxial dislocation without Neurogenic EMG Anterior, excellent.below C5 cord compression in C5 Tingling of finger tips

Exaggerated Yes (left) Massive/Yes Mild canal stenosis from C4-5 to Neurogenic EMG Not donefinger jerks C6-7 from C5 to TI

Exaggerated Yes (right) Moderate/Yes Narrow cervical canal; dynamic Neurogenic EMG Not donebelow C5 dislocation at C5-6 in C8

Exaggerated Yes (right) Moderate/Yes Narrow cervical canal Neurogenic EMG Not donebelow C7 in C5

posterior bipolar leads connected Cv7 to theanterior neck just above the thyroid cartilage.Longitudinal bipolar leads connected Cv2 toa point 3 cm rostral to the inion. In the bipolarrecordings, negativities at the posterior orcaudal electrodes were shown upward. Forthe SEPs subsequent to posterior tibial nervestimulation at the ankle (tibial SEPs), therecording electrodes were placed on the scalpat Fz, Cz, and Pz together with a connectedears reference, and on the 12th thoracic spine(T12) with a reference on the iliac crest con-tralateral to the stimulated limb.'5 Silver-silverchloride disc electrodes were attached to theskin, and the impedance maintained at lessthan 3000 ohms. Square wave pulses of0-2 ms duration were delivered at the rate offour per second to a unilateral nerve trunk.The stimulus intensity was adjusted to pro-duce a minimal twitch. Potentials were ampli-fied and filtered between 10 and 1500 Hz(3dB down). One thousand to 2000 epochswere averaged for an analysis time of 30 mswith a 2 ms delay between the shock and thesweep onset.

Somatosensory evoked potentials of agematched normal controls were obtained from32 median nerves of 22 subjects (13 womenand nine men) who had a mean age of 52-4(range 38 to 74) years, and 17 tibial nerves of16 subjects (11 women and five men) whohad a mean age of 55 9 (range 38 to 81) years.Figure 2 shows median SEPs of a normal sub-ject. The first small diphasic negative-positivewave with a positive peak (P10) and subse-quent large negative wave (N13a) were identi-fied in the anteroposterior bipolar leads. Alarge negative potential (N13b), identified in

the longitudinal bipolar leads, had a peaklatency similar to N13/P13 in SEPs with aconnected ears reference. The absolute laten-cies of the median and tibial SEPs and theinterpeak latency between lumbar N20 andscalp P37 in the tibial SEPs increased linearlyin proportion to the subject's height.

ResultsPROPRIOCEPTIVE AND DISCRIMINATIVE HANDFUNCTIONS (TABLE 2)Sense of position was invariably impaired.The exploring hand wandered about andrepeatedly missed the nose when the eyes

2 12 22 32ms

C3-A+

C4-A+

Cv7-A+

Cv2-INCv7-AN

Erb-A+1:35 sV

Figure 2 Median SEPs ofa normal subject. PIO, Nl3a,and N13b in the bipolar leads, and N13/P13 and Nl9 inthe earlobe reference leads are shown by arrow heads.Vertical dashed lines represent the latencies ofN13a andNl3b in thisfigure and in figures 3 and 4.

609

on March 2, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.5.607 on 1 M

ay 1995. Dow

nloaded from

Nakajima, Hirayama

Table 2 Hand dysfunction

Position Discriminative functions

FNT with Pseudo- Passive Graphaes- Stereo- Dexterity ofPatientlagelsex eyes closed athetosis movement Two point thesia aesthesia manipulation1/53/F 2* Yes* 0 1* 0 2* 2*2/58/M 2 Yes 2* 2 1* 2 23/52/M 2* Yes* 1* 1 1 * 2* 2*4/48/F 2 Yes 2* 2 1* 2 25/47/M 1* No 0 1 0 1* 2*6/59/F 2 Yes 1 2 1 2 27/55/M 1 No 0 1 0 1 18/76/F 1 No 0 1 0 1 1

*Signs profoundly diminished after surgery; FNT = finger-to-nose test; 0 = normal; 1 = mildly impaired; 2 = severely impaired orlost.

were closed. Five patients who had severe tar-get pointing errors also had pseudoathetosis.Perception of passive movement and targetpointing with the eyes open were less affected.In the discriminative sensory functions, loss ofstereoaesthesia was remarkable by contrastwith the relative preservation of two point andtactile directional discrimination. Two pointdiscrimination was possible at distances of 5to 10 mm on the finger tip and 20 to 30 mmon the palm in the mildly impaired patients.

Recognition of numerical symbols outlinedon the finger tips or palms was possible, occa-sional errors being made by the mildlyimpaired patients.As well as the inability to identify small

objects, dexterity of manipulation wasimpaired. The patients had difficulty in bring-ing an object into a position between theindex finger and thumb, the manner by whichnormal subjects identify such objects.'6 Theyclenched and opened the fist, and pressed theobject to the palm using the fingers, but withno rolling of the thumb and index fingers.They sometimes lost the object but continuedtheir exploring movements, perception of thegiven tactile stimuli being preserved.Awkwardness of manipulation was an essen-tial feature of their hand clumsiness.

SOMATOSENSORY EVOKED POTENTIALSIn the median SEPs, the absolute latencies ofErb's potential and P10 in the anteroposteriorbipolar lead (figs 3 and 4) adjusted to160-0 cm were always within the normalmeans plus 3 SD. Table 3 shows interpeak

Figure 3 Median SEPsin patient 7. N13b andsubsequent components aredelayed. All thecomponents except Erb 'spotential and the firstdiphasic wave in Cv7-A +and Cv7-AN leads havelow amplitudes.

2

2 12 22 32 msC3-A+

C4-A+

Cv2-1N

PloCv7-AN 0.25 sV

2 12 22 32i

C3-A+

C4-A+ N19

Ni3b

Cv2-IN

Cv7-AN NISA

ms

Figure 4 Median SEPs in patient 2 before (top) andafter (bottom) surgery. Pl0 and Nl3a of low amplitudesare present with no subsequent activities (top). Afterdiscectomy, small, delayed N13b and N19 are present andN13a remains low in amplitude (bottom). In Cv7-ANleads, thefirst diphasic wave is larger than subsequentN13a.

12 22 32 ms

C3-A+

C4-A+

Cv7-A+

Cv2-INCv7-AN

latencies of N13a and Nl3b from P10, and ofN19 from N13b. In the preoperative record-ings, although P10 was always identified,N13a, N13b, and N13/P13 were occasionallyabsent. In patient 2, N13a was present butwithout identifiable N13b and scalp potentials(fig 4 top). In the other recordings showingboth N13a and N13b, the prolonged P10-N13b interpeak latency was remarkable incomparison with the normal P10-N13ainterpeak latency. N19 evoked in the scalpelectrodes contralateral to the stimulus sidewas attenuated (figs 3 and 4 bottom) orErb-A+

610

on March 2, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.5.607 on 1 M

ay 1995. Dow

nloaded from

Midcervical central cord syndrome: numb and clumsy hands due to midline cervical disc protrusion at the C3-4 intervertebral level

Table 3 Interpeak latencies ofsomatosensory evoked potentials

Median SEPs (ms) Tibial SEPs (ms)

P10-N13a P10-N13b N13b-N19 N20-P37*PatientlHeight Left Right Left Right Left Right Left Right

1/152-5 2-64 2-82 7-44 7-14 5-52 6-06 16-0(2-82) (3 00) (4 08) (4-14) (5 70) (6 00)

2/159-0 3-18 25-0(3-38) (3 66) (6 60) (5 94) (624) (6-36) (21 1)

3/168-2 - 3-66 - 7-02 0 6-66(3-18) (2 58) (6-42) (5 64) (6 36) (5 70)

4/148-8 - - - - - -(2 70) (2-58) (3-96) (7 80) (5 79) (5-89) (19-3)

5/160-0 3-00 2-52 5-58 - 6-55 -(3-00) (2 70) (3 84) (4 02) (6-72) (6-18) (18-0) (18-2)

6/154-0 2-28 2-22 6-00 5-46 - 6-18 19.17/166-6 3 00 2-76 4-98 5-46 6-36 5 94 20-6 20 28/147-0 2-76 6-24 5-94 21-9

Normal values:Mean (SD) 2-64 (0 35) 3 03 (0 43) 5-94 (0-47) 17-1 (1-27)Upper limitst 3-69 4-32 7-35 20-9

*N20-P37 IPL is adjusted to 160-0 cm; its normal value is at that height; tNormal mean + 3SD. Values for SEPs after surgery are given in parentheses. - = Thelater components (N13a, N13b, or N19) were not evoked.

Figure 5 Preoperative(closed circles) andpostoperative (open circles)P10-N13a amplitudevalues for median SEPs.Data on left stimulationare arranged on the left ofeach patient's number.Data on the zero line showthat no N13a was evoked.The stippled area indicatesthe normal limits within 2SDfrom the mean.

4 0>

o 3-0-------Z::r2 2.0---z

° 1 0-CD10

-

1 2 3 4 5

Patients

absent (fig 4 top), but when 1normal interpeak latency from cPostoperatively, N13a, N13b, a

evoked in all the nerves tested (In each patient who had pioperative P10-N13b interpeakinterpeak latency was shortenedmal values in patients 1 and 5.The peak to peak amplitude

P10 (P10-N13a amplitude) wa

the preoperative and postoperexcept for a significant postopeiin patient 5 (fig 5). When the altude of P10-Nl3a was low, tratio of P10-Nl3a to the first dwas low as well (figs 3 and 4)SEPs, the lumbar potential (N;P37 were evoked. The N20-1latency was normal or mildly prthe nerves tested (table 3).

DiscussionThe eight patients with chroniccompression at the C3-4 interhad stereotypical clinical n

Dissociated impairment of thlower limbs, of the propriocepcriminative senses in the upper I

sense of vibration and other fwlower limbs were pronounced.the palms and finger tips, and c

the soles, was an early symptom. A tight, bandlike sensation at the midthoracic level was anotable complaint.Hand dysfunction without significant

involvement of the lower limbs is an outstand-ing feature of cervical spondylotic myelo-

* pathies. The SEPs showed more prominentabnormalities after median stimulation thanafter tibial stimulation. The cervical N13 inthe median SEPs has at least two separate

-- generators,'4 one of which originates low in6 7 8 the cervical cord.'7 18 On the basis of results of

bipolar recordings, Kaji and Sumner'4 dividedcervical N13 into N13a of lower cervical cordorigin and N13b of medullary origin. Theirtechnique showed that the P10-N13a

present had a interpeak latencies were normal and the P10-:ervical N13b. N13b interpeak latencies greatly prolongedand N19 were with attenuated or absent scalp potentials infig 4 bottom). our patients. The interpeak latencies of N13brolonged pre- were reduced after decompressive surgery andlatencies, the accompanied regenerated scalp potentials.and had nor- These findings, which suggest conduction

failure through the fasciculus cuneatus, areof N13a from compatible with a cord lesion level at theIs low in both C3-4 disc, which is of the C4 or C5 cord seg--ative periods, ment.19-21 By contrast, conduction through therative increase fasciculus gracilis is relatively preserved, asbsolute ampli- shown by the normal or mildly increasedthe amplitude N20-P37 interpeak latencies.Liphasic wave'7 Clumsiness of the hands is an early mani-. In the tibial festation of, and a cardinal sign in, the syn-20) and scalp drome. In proprioceptive modalities, theP37 interpeak sense of position in the upper limbs-mostrolonged in all evident in the hands-was affected.

Impairment of proprioception is responsiblefor the clumsiness shown by the uncoordi-nated handling of a small object with the eyesclosed in the test of stereoaesthesia. Stereo-

cervical cord aesthesia itself may be caused by disturbancesvertebral level of active touch and proprioception."2 2223 Bynanifestations. contrast, tactile discrimination of passivelyLe upper and given stimuli, graphaesthesia in particular, is)tive and dis- relatively preserved. A similar dissociationlimbs, and the between stereoaesthesia and graphaesthesia innctions in the high cervical cord disorders has beenNumbness in described by Boshes and Padberg."ccasionally in Functional lamination of the dorsal

611

.

on March 2, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.5.607 on 1 M

ay 1995. Dow

nloaded from

Nakajima, Hirayama

columns has been shown in animal experi-ments24 and human studies,25 as reviewed byRoss.26 These studies suggest that the discrim-ination of given tactile stimuli is anatomicallyrelated to the outermost layer of the dorsalcolumn, vibration sense to the innermost layerclose to the commisure, and proprioception tothe area between them. In our patients, sen-sory dissociation in the hands denoted lesionlocalisation in the inner aspect of the fasciculicuneati. According to the topographicalanatomy of the human dorsal column estab-lished by Smith and Deacon,27 the fasciculuscuneatus occupies most of the ventral aspectof the column adjacent to the commisurefrom the second to sixth cervical cord seg-ment. The proposed lesion in the inner aspectof the dorsal column may affect the innermostpart of the fasciculi graciles that is responsiblefor vibration, but spare the middle partresponsible for proprioception within thefasciculi.

In conclusion, our patients' various dissoci-ated forms of impairment are explained by alesion to the inner aspect of the dorsal columnat the C3-4 intervertebral level. Homer's syn-drome in six patients indicates involvement ofthe deeper lateral funiculus, where the sympa-thetic tract descends.28Numbness at the distal ends of the upper

limbs, sometimes associated with numbnessof the lower limbs, is the earliest manifestationof illness. Tingling in the finger tips remainedin two patients who had excellent surgicaloutcomes. In general, the initial symptom has alocalising value; therefore, the probable site ofthe lesion responsible for the paraesthesiae isin the cervical cord at C3-4. This peripheralneuropathy type sensory presentation is wellrecognised but ill defined in cervical spondy-lotic myelopathy. The band like tightness withthe character of deep pain at the thoracic cordlevel was a later complaint, which disappearedafter surgery. Schvarcz29 stimulated the baseof the dorsal columns of conscious patients atthe cervicomedullary junction and obtained abasic response pattern that consisted ofparaesthesiae starting at the soles of both feet.In a stimulation study, bilateral burning trun-cal sensations were sometimes describedwhen "the central canal region" was stimu-lated.30 Involvement of the dorsal column orafferent propriospinal fibres explains thesecaudal sensory phenomena. The somatotopicorganisation of the dorsal column at a highercervical cord level3' accounts for the paraes-thesiae of the extremities. The propriospinalsystem within the central grey matter or in theperigriseal white matter that connects the dif-ferent segments of the body32 33 probably con-tributes to truncal tightness.

Whether there is profound and enduringimpairment of motor acts after dorsal cordo-tomy is a matter of controversy.'234 In ourpatients dysfunctions within the grey mattercontributed to some manifestations. Dimin-ished pain and temperature sense, neurogenicchanges in electromyography extending to thelower cervical segments, and decreased ampli-tudes of the lower cervical P10-NI 3a indicate

the involvement of the cord caudal to C3-4.These findings for patients 1, 2, and 5, whoshowed no lower spondylotic changes, arestrong evidence of lower grey matter involve-ment due to higher cord compression.Dysfunctions of the dorsal grey matter mayalso contribute to hand clumsiness. There arepropriospinal neurons in cats that function inforelimb target reaching that are located in thecervical cord at C3 and C4, called "C3-C4propriospinal neurons".35 The propriospinalsystem is considered to receive impulses fromboth cutaneous and group I muscle afferents,and their deafferentation results in hyperme-tria in forelimb target reaching.36 37 In primateexperiments, dorsal column lesion extendingto the neighbouring grey matter or the peri-canal region at the C3-4 level was shown toimpair motor performance.38 39

Central cord involvement due to compres-sive myelopathy4' has been explained as beingcaused by border zone arterial insufficiency,4'or by direct compressive and shearing forces.42Caudal involvement of the grey matter causedby a higher compression has been explainedby venous stagnation,43 or dynamic propaga-tion of the compressive force.44 Our conclu-sion, based on clinicoelectrophysiologicalevidence of the transverse and longitudinalextents of the lesion, is that involvement ofthe central cord and its caudal extension,which is caused by high cervical compressionat the C3-4 intervertebral level, is responsiblefor the syndrome.

We thank Dr Seisuke Watanabe for his advice on the SEPtechniques, and Dr Mitsuru Kawamura for the referral ofpatient 2. We also deeply appreciate the suggestions made byDr Ryuji Kaji, Department of Neurology, Faculty ofMedicine, Kyoto University.

1 Daniels LE. Compression of cervical portion of the spinalcord. Neurology 1956;6:344-9.

2 Good DC, Couch JM, Wacaser L. "Numb, clumsy hands"and high cervical spondylosis. Surg Neurol 1984;22:285-91.

3 Savettieri G, Castiglione MG, D'Arpa A, Aiello S, FierroB. Loss of joint position sense with sparing of vibrationsense in two cases of compressive cervical myelopathy.Acta Neurol (Napoli) 1984;6:11-5.

4 England JD, Hsu CY, Vera CL, Barone BM, Rawe SE.Spondylotic high cervical compression presenting withhand complaints. Surg Neurol 1986;25:299-303.

5 Voskuhl RR, Hinton RC. Sensory impairment in thehands secondary to spondylotic compression of the cervi-cal spinal cord. Arch Neurol 1990;47:309-1 1.

6 Nakajima M, Hirayama K. High cervical disc herniationpresenting with deep sensory disturbances of the upperlimbs. Shinkei Kenkyu no Shinpo (Japan) 1993;37:235-44. (In Japanese.)

7 McCloskey DI. Kinesthetic sensibility. Physiol Rev1978;58:763-820.

8 Gandevia SC. Neurophysiological mechanisms underlyingproprioceptive sensations. In: Struppler A, Weindl A,eds. Clinical aspects of sensory motor integration. Berlin:Springer-Verlag, 1987:14-24.

9 Adams RD, Victor M. Principles of neurology. 5th ed. NewYork: McGraw-Hill, 1993:138-9.

10 Bender MB, Stacy C, Cohen J. Agraphesthesia: a disorderof directional cutaneous kinesthesia or disorientation incutaneous space. Neurol Sci 1982;53:531-55.

11 Boshes B, Padberg F. Studies on the cervical spinal cord inman: sensory pattems after interruption of the posteriorcolumns. Neurology 1953;3:90-101.

12 Vierck CJ, Jr. The spinal lemniscal pathways. In: DavidoffRA, ed. Handbook of the spinal cord. Vols 2 and 3. NewYork: Marcel Decker, 1984:673-749.

13 Yamada T, Kimura J, Nitz DM. Short latency somato-sensory evoked potentials following median nervestimulation in man. Electroencephalogr Clin Neurophysiol1980;48:367-76.

14 Kaji R, Sumner AJ. Bipolar recording of short-latencysomatosensory evoked potentials after median nervestimulation. Neurology 1987;37:410-8.

15 Chiappa KH. Short-latency somatosensory evoked poten-

612

on March 2, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.5.607 on 1 M

ay 1995. Dow

nloaded from

Midcervical central cord syndrome: numb and clumsy hands due to midline cervical disc protrusion at the C3-4 intervertebral level

tials: Methodology. In: Chiappa KH, ed. Evoked poten-tial in clinical medicine. New York: Raven Press, 1990:308-70.

16 Pause M, Kunesch E, Binkofski F, Freund HJ.Sensorimotor disturbances in patients with lesions of theparietal cortex. Brain 1989;112:1599-625.

17 Mauguiere F, Restuccia D. Inadequacy of the foreheadreference montage for detecting abnormalities of thespinal N13 SEP in cervical cord lesions. Electro-encephalogr Clin Neurophysiol 199 1;79:448-56.

18 Desmedt JE, Cheron G. Prevertebral (oesophageal)recording of subcortical somatosensory evoked poten-tials in man: the spinal P13 component and the dualnature of the spinal generators. Electroencephalogr ClinNeurophysiol 1981;52:257-75.

19 Payne EE, Spillane JD. The cervical spine: an anatomicopathological study of 70 specimens (using a special tech-nique) with particular reference to the problem of cervicalspondylosis. Brain 1957;80:571-96.

20 Mair WGP, Druckman R. The pathology of spinal cordlesions and their relation to the clinical features in pro-trusion of cervical intervertebral discs: a report of fourcases. Brain 1953;76:70-91.

21 Wilkinson M. The morbid anatomy of cervical spondylosisand myelopathy. Brain 1960;83:589-617.

22 Gibson Jn. Observations on active touch. Psychol Rev1962;69:477-9 1.

23 Brinkman J, Porter R. Movement performance and affer-ent projections to the sensorimotor cortex in monkeyswith dorsal column lesions. In: Gordon G, ed. Activetouch. Oxford: Pergamon Press, 1978:119-37.

24 Uddenberg N. Differential localization in dorsal funiculus offibres originating from different receptors. Exp Brain Res1968;4:367-76.

25 Schneider RC. Syringomyelia. In: Schneider RC, KhanEA, Crosby EC, Taren JA, eds. Correlative neurosurgery.3rd ed. Springfield: Charles C Thomas, 1982:1022.

26 Ross RT. Dissociated loss of vibration, joint position anddiscriminatory tactile senses in disease of spinal cord andbrain. Can JNeurol Sci 1991;18:312-20.

27 Smith MC, Deacon P. Topographical anatomy of the pos-terior columns of the spinal cord in man. Brain 1984;10:671-98.

28 Nathan PW, Smith MC. The location of descending fibresto sympathetic neurons supplying the eye and sudomo-tor neurons supplying the head and neck. Y NeurolNeurosurg Psychiatry 1986;49: 187-94.

29 Schvarcz JR. Stereotactic extralemniscal myelotomy.J Neurol Neurosurg Psychiatry 1976;39:53-7.

30 Schvarcz JR. Functional exploration of the spino-medullary junction. Acta Neurochirurgica (suppl) 1977;24:179-85.

31 Whitsel BL, Petrucelli LM, Sapiro G, Ha H. Fiber sortingin the fasciculus gracilis of squirrel monkeys. Exp Neurol1970;29:227-42.

32 Nathan PW, Smith MC. Fasciculi proprii of the spinalcord in man. Brain 1959;82:610-68.

33 Cook AW, Nathan PW, Smith MC. Sensory consequencesof commisural myelotomy: A challenge to traditionalanatomical concepts. Brain 1984;107:547-68.

34 Davidoff RA. The dorsal columns. Neurology 1989;39:1377-85.

35 Illert M, Lundberg A, Tanaka R. Integration in thedescending motor pathways controlling the forelimb inthe cat: 3. convergence on propriospinal neurones trans-mitting disynaptic excitation from the corticospinal tractand other descending tracts. Exp Brain Res 1977;29:323-46.

36 Alstermark B, Gorska T, Johannisson T, Lundberg A.Hypermetria in forelimb target-reaching after interrup-tion of the inhibitory pathway from forelimb afferents toC3-C4 propriospinal neurones. Neurosci Res 1986;3:457-61.

37 Kitazawa S, Urushihara Y, Hongo T. Forelimb skin deaf-ferentation impaired limb movements in the cat.Neurosci Res (suppl) 1991;16:104.

38 Ferraro A, Barrerra SE. Effects of experimental lesions ofthe posterior columns in macacus rhesus monkeys. Brain1934;57:307-32.

39 Giiman S, Denny-Brown D. Disorders of movement andbehaviour following dorsal column lesions. Brain 1966;89:397-418.

40 Crandall PH, Batzdorf U. Cervical spondylotic myelopa-thy. J Neurosurg 1966;25:57-68.

41 Turnbull IM, Brieg L, Hassler C. Blood supply of cervicalspinal cord in man: a microangiographic cadaver study.J Neurosurg 1966;24:951-65.

42 Panjabi M, White A. Biomechanics of nonacute cervicalspinal cord trauma. Spine 1988;13:838-42.

43 Taylor AR, Byrnes DP. Foramen magnum and highcervical cord compression. Brain 1974;97:473-80.

44 Hashizume Y, Iijima S. Pencil-shaped softening of thespinal cord: pathologic study in 12 autopsy cases. ActaNeuropathol (Berl) 1983;61:219-24.

613

on March 2, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.5.607 on 1 M

ay 1995. Dow

nloaded from