Letters to the Editor

otherwise normal. Motor nerve conductionvelocity (MNCV) in the right ulnar nervewas slowed at 47 m/s and 48 m/s in the fore-arm and around the elbow respectively, butseverely reduced to 10m/s in the upper armwhere there was conduction block. Theevoked compound muscle action potential(CMAP) at the wrist, below elbow, aboveelbow, and in the upper arm was 4-3, 3 3,2-4, and 1-4 mV respectively. MNCV wasmoderately reduced, to 37 m/s in the leftulnar nerve in the forearm, but to 1 6m/saround the elbow where there was conduc-tion block. The amplitude of the CMAP fellfrom 3 6 mV at the wrist to 2-5 mV belowand 1-2 mV above the elbow. The distalmotor latency (DML) was normal in bothnerves; F waves were absent. MNCV in theright peroneal and tibial nerves was border-line at 40 and 42 m/s without block and withnormal DML values. F wave latency in thetibial nerve was increased at 60-64 ms; noperoneal nerve F responses were obtained.The right median sensory nerve actionpotential (SAP) was of normal amplitude(22 pV) and velocity (63 m/s). The left suralSAP was 555,V with a velocity of 46m/s; itwas absent on the right.The P100 visual evoked potential laten-

cies were increased bilaterally (117 ms onright, 112 ms on left). Cranial MRI showedlesions of high T2 signal in cerebral whitematter (figure), many abutting the walls of

the lateral ventricles, especially their poste-rior portions. Lesions were also present inthe corpus striatum, pons, and middle cere-bellar peduncles.The electrophysiological results indicated

a patchy demyelinating disorder with con-duction block affecting the peripheral nervesconsistent with CIDP with concomitantCNS involvement. Clinical examination alsoshowed CNS abnormalities. The cranialMRI showed clear evidence of multifocaldemyelinating pathology. The patient wastreated with prednisolone with resolution ofher symptoms apart from persistent paraes-thesiae distally in the limbs and objectiveevidence of improvement on examination.

This patient is of interest in relation to theonset of relapsing symptoms in early child-hood. It is uncertain whether the two previ-ous episodes represented peripheral orcentral nerve disease. Cranial nerve lesionsare known to occur with either in CIDPwith central involvement. } The rarity of mul-tiple sclerosis in Afrid populations suggeststhat the central demyelinating lesions in thistype of case may reflect a different form ofdemyelination than occurs in multiple scle-rosis. Experimental allergic neuritis has beenconsidered to provide a model for theGuillain-Barre syndrome and its chronicvariant for CIDP.5 Possibly the chronic vari-ant of experimental allergic encephal-omyelitis" provides an equivalent for the

central demyelination seen in the presentcase and other instances of CIDP with cen-tral involvement. The question as to whetherchronic relapsing experimental allergicencephalomyelitis is a valid model for multi-ple sclerosis continues to be controversial.

P K THOMASA VALENTINE

B D YOULRoyal Free Hospital and School ofMedicine,

London, UK

Correspondence to: Professor P K Thomas,Department of Clinical Neurosciences, Royal FreeHospital School of Medicine, Rowland Hill Street,London NW3 2PF, UK.

1 Thomas PK, Walker RWH, Rudge P, et al.Chronic demyelinating peripheral neuropa-thy associated with multifocal central ner-vous system demyelination. Brain 198711O:53-76.

2 Mendell JR, Kolkin S, Kissel JT, et al.Evidence for central nervous system demyeli-nation in chronic inflammatory demyelinat-ing polyradiculoneuropathy. Neurology 1987;37:1291-9.

3 Rubin M, Karpati G, Carpenter S. Combinedcentral and peripheral myelinopathy.Neurology 1987;37: 1287-90.

4 Waddy HM, Misra VP, King RHM, et al.Focal cranial nerve involvement in chronicinflammatory demyelinating polyneuropathy:clinical and MRI evidence of peripheral andcentral lesions. J Neurol 1989;236:400-1 0.

5 Pollard JD, King RHM, Thomas PK. Re-current experimental allergic neuritis. Anelectron microscope study. J Neurol Sci1975;24:365-83.

6 Raine CS, Snyder DH, Valsamis MP, StoneSH. Chronic experimental allergicencephalomyelitis in inbred guinea-pigs-anultrastructural study. Lab Invest 1974;31:369-76

Cranial MRI showing lesions of high T2 signal in cerebral white matter as detailed in the text.

Neurothekeoma of the cauda equina

Neurothekeoma or nerve sheath myxoma isa rare benign tumour of peripheral nervoustissue. Histogenesis from Schwann cells' orundifferentiated nerve sheath precursorcells2 has been suggested. Intrathecal loca-tion is rare.An 84 year old woman presented in May

1994 with severe low back pain and painradiating into both legs. She had fallen on tothe floor at home three weeks previously.The pain was exacerbated by coughing,sneezing, and straining. She had increasingdifficulty walking, requiring the aid of oneperson. She had difficulty initiating micturi-tion and there was occasional faecal soiling.Clinical examination disclosed a flexed pos-ture. She was unable to stand erect due topain. There was pain on femoral stretchingand on straight leg raising bilaterally.Tendon reflexes were absent in the lowerlimbs. There was sensory loss below theright knee and in the left foot. Saddle sensa-tion and anal tone were normal. Power test-ing was not possible due to pain.

Blood biochemistry and haematologywere normal. Magnetic resonance imagingdisclosed a 3 x 1-7 x 1 2 cm intrathecalmass displacing the lower conus anteriorlyand to the left. The tumour had high signalintensity on T2 weighted images (figure, a).After gadodiamide there was pronouncedenhancement of the tumour mass with asmall central non-enhancing component(figure, b). No adjacent bone or extraduralabnormality was identified.

At operation a thoracolumbar laminec-tomy was made and the thecal sac wasopened longitudinally. The tumour seemedexophytic from the right side of the caudalconus and displaced the roots of the caudaequina. The major exophytic part of the

530

on July 2, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.61.5.530 on 1 Novem

ber 1996. Dow

nloaded from

Letters to the Editor

tumour was removed with an ultrasonicaspirator. A thin layer of tumour remainedattached to the conus. A good plane of dis-section between the tumour and the surfaceof the conus could not be safely achieved.

Postoperatively, she made a slow butsteady recovery. She was relieved of lowback and radicular pain. At follow up fourmonths after surgery, there was an area ofnumbness in the L4 dermatome on theright. She had returned home, where shelooked after herself.The tumour specimen was an irregular,

soft, greyish portion of tissue 2-2 x 1-5 x0 7 cm in size. Microscopically it displayed amultinodular structure, in which pale,sparsely cellular islands were separated byconnective tissue septa or bands of morecompactly cellular neoplastic tissue (figure,c). The pale nodules had a myxomatousappearance composed of stellate or elon-gated cells in a strongly alcianophilic,mucoid matrix. The tumour cells in the

intervening areas were mostly spindleshaped and had a more bulky eosinophiliccytoplasm. Both components stained posi-tively for S-100 protein (figure, d).

Electron microscopy showed that thetumour cells in the myxomatous nodulesand those in the internodular areas hadelongated cytoplasmic processes, mostlycovered by a continuous external lamina.Wide spaced collagen fibres (Luse bodies)were also identified.

Neurothekeomas most often occur in theskin of the face and arms of young adults.They arise from small cutaneous nervebranches and not from the major peripheralnerves. They are benign lesions which arecured by excision. Rare recurrences havebeen reported after incomplete resection.Cutaneous neurothekeomas have been clas-sified into cellular and myxomatous sub-types, the second arising at an older age.3

Neurothekeomas have rarely been foundintrathecally. There is one report of the

tumour arising from a single nerve root mthe lower cauda equina.4 In that case thetumour was completely excised. In the pre-sent case complete excision was not possibledue to adherence to the conus and multiplenerve roots. The early postoperative coursehas, however, been satisfactory with relief ofpain and restoration of ambulation.The MRI appearance is similar to that of

other intrathecal extramedullary tumours.Ependymoma, neurofibroma, and menin-gioma were the major differential diagnoses.

Connective tissue myxomas have occa-sionally been described as arising fromparaspinal muscle5 and from the vicinity of afacet joint.6 An intramedullary location hasalso been described.7 The compact cellularschwannomatous component, the positivestaining for S-100 protein, and the ultra-structural appearances clearly distinguish thepresent tumour from connective tissue myx-oma.

GEORGE F KAARSAAD H BASHIRJAMES M N'DOW

Department ofNeurosurgeryPHILIP V BEST

Department ofPathologyLESLEY N GOMERSALL

Department ofRadiology,Aberdeen Royal Hospitals NHS Trust,

Aberdeen AB9 2ZB, UK

bV...

at, y J

%. X

-i ...4

4..POST GA0ODIA1341UDEr

i

,.l

far~~~~~~~~~~~~~~~~~~~

'- ..

#

AN»

;* * . ..

$ jr 441.8$.. ,,

p,,O.'4~~~~~~~~~~~~~~~~~~0

(a) Sagittal T2 weightedMR image (TR = 4500 ms, TE = 130 ms) showing tumour distorting theregion of the conus (arrow). (b) Postgadodiamide injection Tl weightedMR image (TR = 450 ms,TE = 15 ms), showing tumour (black arrow) displacing conus (white arrow) anteriorly andlaterally. (c) Photomicrograph (haematoxylin and eosin x 10) showing pale myxomatous nodulesseparated by narrow septa. (d) Photomicrograph ( x 40) of the tumour stained with theimmunoperoxidase method to demonstrate S-100 protein. Note the dark (positive) staining in nucleiand cytoplasmic strands.

Correspondence to: Mr G F Kaar, Ward 40,Aberdeen Royal Infirmary, Foresterhill, AberdeenAB9 2ZB, UK.

1 Webb JN. The histogenesis of nerve sheathmyxoma: report of a case with electronmicroscopy. jPathol 1979;127:35-7.

2 Angervall L, Kindblom LG, Haglid K. Dermalnerve sheath myxoma. A light and electronmicroscopic, histochemical, and immunohis-tochemical study. Cancer 1984;53: 1752-9.

3 Barnhill RL, Mihm MC Jr. Cellular neu-rothekeoma. A distinctive variant of neu-rothekeoma mimicking melanocytictumours. Am J Surg Pathol 1990;14: 113-5.

4 Paulus W, Jellinger K, Pemeczky G. Intra-spinal neurothekeoma (nerve sheath myx-oma). A report of two cases. Am J ClinPathol 1991;95:511-6.

5 Tahmouresie A, Farmer PM, Stokes N.Paraspinal myxoma with spinal cord com-pression. Case report. J7 Neurosurg 1981;54:542-4.

6 Bell WO, Gill A, Babiak T, et al. Epiduralmyxoma causing compression of the caudaequina. Case report. Neurosurgery 1983;12:325-6.

7 Pasaoglu A, Patiroglu TE, Orhon C, YildizhanA. Cervical spinal intramedullary myxoma inchildhood. Case report. J Neurosurg1988;69:772-4.

Voluntary facial palsy with a pontinelesion

Emotional voluntary dissociation is notuncommon in patients with central facialpalsies.'-3 The neuroanatomical basis of thisdissociation is not well understood. A recentcase report in this J3ournal suggested that apontine stroke could result in unilateral vol-untary facial palsy.4 We have recently hadthe opportunity to study a very similar syn-drome.A teacher aged 57 was referred three days

after the onset of progressive neurologicalsymptoms and signs that had included, inorder, unsteadiness, dysarthria, dysphagia,and weakness of the right arm, leg, and face.On admission, the patient was awake,dysarthric but not aphasic, and had normalvision and oculomotor function. He com-plained of impaired swallowing and rightfacial weakness. Neurological examinationdisclosed a right central facial palsy for vol-

531

on July 2, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.61.5.530 on 1 Novem

ber 1996. Dow

nloaded from