We present 2 cases of caudaequina tumours thatcaused delay and confu-

sion in diagnosis because initially thelesions were not localized to any par-ticular level. Neurosurgeons are awareof the mobility of spinal tumours;however, only 7 cases of mobile in-traspinal tumours have been docu-mented.1–5 Redundancy of the nerveroots of the cauda equina enables thetumours to migrate through severalvertebral levels, causing only transientparoxysmal symptoms from involve-ment of lumbar and sacral roots at var-

ious times. Objective signs may de-velop only months later as the tumourenlarges enough to become impactedat one level. Until then, the fleeting,bizarre symptoms may be consideredas non-neurogenic.Magnetic resonance imaging (MRI)

is desirable and reliable, but even in in-stitutions where it is available, myelog-raphy and CT scanning are often usedfor initial screening. When correctlyperformed, these neuroradiologic pro-cedures are sufficient to localize a masslesion within the lumbar spinal canal.Our 2 patients were seen before the in-

troduction of MRI. In mobile tu-mours, discrepancy between the clini-cal and neuroradiologic findings andthe surgical location of the tumour mayprevail. Therefore, awareness of mobil-ity is important for early diagnosticconfirmation when dealing with caudaequina tumours.

CASE REPORTS

Case 1

A 65-year-old man had paroxysmsof backache and perianal spasms for a

Case ReportÉtude de cas

MOBILE TUMOURS IN THE LUMBAR SPINAL CANAL:A DIAGNOSTIC PROBLEM

George Varughese, MB, MSc, FRCSC; Rida Mazagri, MD*

From the Division of Neurosurgery, Department of Surgery, University of Saskatchewan, Saskatoon, Sask.

*Resident in neurosurgery

Accepted for publication June 25, 1996

Correspondence to: Dr. George Varughese, Division of Neurosurgery, Royal University Hospital, 103 Hospital Dr., Saskatoon SK S7N 0X0

© 1997 Canadian Medical Association (text and abstract/résumé)

In two cases of mobile tumours in the lumbar spinal canal there was difficulty and delay in clinical and radi-ologic diagnosis because the early symptoms did not correspond to any particular dermatome. Myelogra-phy and computed tomography (CT) are the initial diagnostic procedures used in most institutions, evenwhere magnetic resonance imaging (MRI) is available. The purpose of these 2 case reports is to remindclinicians that it is possible for certain tumours attached to the roots in the lumbar spinal canal to migrate,because the roots tend to be redundant or lax. Multilevel search is essential in neuroradiologic studies forearly diagnostic confirmation of mobile tumours.

Dans deux cas de tumeur mobile dans la partie lombaire du canal rachidien, le diagnostic clinique et radi-ologique a été difficile et retardé parce que les premiers symptômes ne correspondaient pas à un dermatomeen particulier. La myélographie et la scanographie sont les premières interventions de diagnostic utiliséesdans la plupart des établissements, même ceux qui disposent d’un service d’imagerie par résonnance mag-nétique (IRM). Ces deux rapports de cas visent à rappeler aux cliniciens qu’il est possible que certainestumeurs fixées aux racines de la partie du canal rachidien se déplacent, parce que les racines ont tendance àêtre redondantes ou lâches. Il est essentiel de procéder à une analyse à niveaux multiples lors de l’examenneurologique pour confirmer un premier diagnostic de tumeur mobile.

14799 February/97 CJS /Page 59

CJS, Vol. 40, No. 1, February 1997 59

year. He also had pain and paresthesiasin 1 leg or the other, without any con-sistency. His transient symptoms wors-ened over 6 months. There were nobladder or bowel symptoms. No mus-cle weakness or sensory changes werenoted by the physician, who admittedthe patient for surgical management ofspondylolisthesis, which was diagnosedby plain x-rays at the L3–4 level. Themyelogram, surprisingly, was sugges-tive of an intradural tumour at L5(Figs. 1 [left] and 2 [left]). However,CT scanning a few hours later at the L4to S1 levels failed to show the expectedabnormality. The radiologist did notpursue the search for a tumour at anyother level. Surgery for the spondylolis-thesis was cancelled. The neurosurgeon

was consulted when the right knee jerkwas noted to be absent, and a secondmyelogram (1 week after the first) lo-cated the tumour at L4 (Figs. 1 [right]and 2 [right]). CT scanning a fewhours later revealed that the tumourhad migrated further up to L3, abovethe L3–4 spondylolisthesis. Afterlaminectomy at L3, L4 and L5, an in-tradural schwannoma attached to aroot of the cauda equina, still movingup and down, was removed.The patient made a smooth recovery

and had no further paroxysmal radicu-lar symptoms over a 3-year follow-up.

Case 2

A 78-year-old, previously healthy,

man suffered pain in the right lumbararea like that from an electric shock,with occasional radiation to the ante-rior and lateral thigh. Sometimes heexperienced pain or paresthesias radi-ating to the calf or even to the coccyxon the right side. These spells weretransient and paroxysmal for severalmonths and then became constantduring the 3 weeks before admission.No objective neurologic signs werenoted on admission. The clinical diag-nosis was spinal stenosis, because theradiologist reported degenerativechanges at multiple lumbar levels,without spondylolysis or spondylolis-thesis. MRI was not available. Routinemyelography, surprisingly, showed anovoid intradural tumour just below

VARUGHESE AND MAZAGRI

14799 February/97 CJS /Page 60

60 JCC, Vol. 40, No 1, février 1997

FIG. 1. Case 1. Lateral view of myelograms 1 week apart, locating the lesion at L5 (left) and at L4 (right), still below the L3–4 spondylolisthesis.

the L2–3 disc space. The subsequentCT scan demonstrated the lesionabove the L2–3 level. After multiplelumbar laminectomies (L1, L2 andL3), before the dura was opened, thetumour mass could be seen movingup and down between L1 and L3 witheach respiration. A schwannoma froman intradural root was removed. Thepatient’s symptoms resolved immedi-ately and he remained well over a 4-year follow-up.

DISCUSSION

With the introduction of nonionicwater-soluble contrast myelography,the lumbar spinal canal can be visual-ized much better, so a mass lesion can

be demonstrated without great diffi-culty. These 2 cases suggest a migra-tory tendency of a cauda equina masslesion, especially if the symptoms per-taining to different lumbar or sacralroots are transient and paroxysmal.These are not the usual clinical presen-tations in spondylolisthesis (case 1) orspinal stenosis (case 2). The myelo-gram and CT scan are needed to makethe correct diagnosis if MRI is notavailable. In case 1, even when themyelogram showed a mass lesion, theCT scan failed to confirm the diagno-sis because the radiologist limited thescanning to the myelographically ab-normal levels only. With an awarenessof the migratory tendency for some ofthese tumours, we were able to

demonstrate the lesion on CT withless difficulty in case 2, as the wholelumber spinal canal was scanned. Mo-bile spinal tumours may be rare anddifficult to diagnose preoperatively.The movement of the tumour massobserved in the neuroradiologic stud-ies and during surgery before openingthe dura is caused by the laxity of thenerve root to which the tumour is at-tached. An ependymoma of the filumterminale has been reported to moveover a distance of 7 cm.5 In lumbarspinal stenosis, redundancy of thecauda equina nerve root has been re-ported.6,7 Such redundancy allows mo-bility of an attached tumour. Thelumbar spinal canal allows mobility forthe cauda equina and for any mass le-

MOBILE TUMOURS IN THE SPINAL CANAL

14799 February/97 CJS /Page 61

CJS, Vol. 40, No. 1, February 1997 61

FIG. 2. Case 1. Two myelograms 1 week apart showing the lesion at the L5 (left) and L4 (right) vertebral levels.

VARUGHESE AND MAZAGRI

14799 February/97 CJS /Page 62

62 JCC, Vol. 40, No 1, février 1997

sion arising from it, as the nerve rootscan float in the spinal fluid belowL1–2, where the cord ends in mostadults.In summary, there may be various

reasons for the migratory tendency ofsome lumbar spinal tumours. Exces-sive mobility of the tumour is respon-sible for some bizzare, transient parox-ysmal radicular symptoms with verylittle objective findings. Such mobiletumours, although rare, can bedemonstrated by multilevel neuroradi-ologic investigations. Awareness of thepossibility that tumours of the caudaequina may migrate is most importantfor the early diagnostic confirmationof the lesions and for prompt surgicalmanagement.

References

1. Hollin SA, Drapkin AJ, Wancier J,Huang YP. Mobile schwannoma ofthe cauda equina. J Neurosurg 1978;48:135-7.

2. Orunesu AP, Viale ES, Turtas S, Zi-rattu G. Mobile neurinoma of thecauda equina. Acta Neurochir (Wien)1982;60:115-7.

3. Tavy DL, Kuiters RR, Koster PA,Hekster RE. Elusive tumor of thecauda equina. J Neurosurg 1987;66:131-3.

4. Toyohiko I, Yoshinobu I, Minoru A,Masafumi N, Hiroshi A. Mobileschwannoma of the cauda equina di-agnosed by magnetic resonance imag-ing. Neurosurgery 1989;25:968-71.

5. Wortzman G, Botterell EH. A mobileependymoma of the filum terminale.J Neurosurg 1963;20:164-6.

6. Ehni G, Moiel R, Bragg T. The re-dundant or knotted nerve root: a clueto spondylotic cauda equina radicu-lopathy. J Neurosurg 1970;32:252-4.

7. Varughese G. Extradural extrusion ofroots of the cauda equina. Surg Neu-rol 1976;5:161-3.