Diagnosis management to radiculomyelitis

Journal ofNeurology, Neurosurgery, and Psychiatry, 1979, 42, 12-18

Diagnosis and management of tuberculous paraplegiawith special reference to tuberculous radiculomyelitisDAVID FREILICH AND MICHAEL SWASH

From the Department of Neurology, The London Hospital, London

S U M M ARY Paraplegia occurred in eight of 17 patients with central nervous system tuberculosis.In six of these paraplegia was the presenting feature. Paraplegia may complicate tuberculousmeningitis, or vertebral tuberculosis, but it may also occur, as in three of our cases, as a primarylocalised spinal tuberculous radiculomyelitis. These cases are presented in relation to theconcept that paraplegia complicating these forms of tuberculosis is caused by radiculomyelitis.

Until recently tuberculous infection of the centralnervous system appeared to be of decreasing im-portance in the United Kingdom, although it hasremained a common clinical problem elsewhere,particularly in Asia (Dastur et al., 1968; Tandonand Pathak, 1973; Wadia, 1973a). In thesecountries tuberculous paraplegias caused by Pott'sdisease, extradural and intradural spinal tuber-culomas, spinal radiculomyelitis, or other chronicforms of tuberculous meningitis are relativelycommon (Dastur and Wadia, 1969; Tandon andPathak, 1973; Wadia, 1973b). Kocen and Parsons(1970) and McNicol et al. (1971) have describedsome features of the changing pattern of tuber-culosis of the nervous system in Britain, particu-larly among non-European immigrants. In thispaper we shall emphasise the importance of tuber-culous radiculomyelitis (arachnoiditis), hithertoan almost unrecognised cause of paraplegia inBritain, and compare our recent experience withthat at The London Hospital 50 years ago.

Subjects

The case records of all patients with tuberculousinfection of the nervous system admitted to TheLondon Hospital in the years 1972 to 1977 werereviewed. Seventeen patients with tuberculosis ofthe central nervous system were admitted in thissix year period. Their ages ranged from 18 to 50years (mean 46). Thirteen (76%) were male.

Address for correspondence and reprint requests: Dr M. Swash,Department of Neurology, The London Hospital, Whitechapel,London El I BB.Aczepted 9 July 1978

Eleven (64%) were immigrants; seven of the lattercame from the Indian subcontinent. Two patientsdied. The clinical presentation of these 17 patientsvaried. Eight (47%) presented with the syndromeof tuberculous meningitis and six (36%) withparaplegia caused by spinal tuberculosis. The re-maining three patients had cerebral tuberculomas.Eight patients (47%) developed paraplegia.

PARAPLEGIA COMPLICATING TUBERCULOUSMENINGITISTwo of the eight patients with tuberculous menin-gitis developed a paraplegia during the course ofthe illness.

Case IThis 34 year old Asian woman had migrated tothe United Kingdom from India a year beforeadmission to hospital. In August 1977 she was ad-mitted to another hospital with tuberculous menin-gitis. For two weeks she had complained ofanorexia, vomiting, headache, and remittent fever.On examination she was drowsy and confused,with marked stiffness, but there were no othersigns. The CSF contained 160 lymphocytes/mm3,1.8 g protein/I, and 1.9 mmol glucose/l. Acid-fastbacilli were not isolated from the CSF, but theywere later cultured from the urine. The Mantouxtest was postive at 1: 10 000. Treatment was begunwith rifampicin, isoniazid, and pyridoxine. A weeklater bilateral lateral rectus palsies were noted. Atthis time the CSF protein was 3.6 g/l. She wastreated with prednisolone (30 mg daily) for fourweeks. She improved, and was discharged inOctober 1977, still taking antituberculous drugs.

In November 1977 she complained of pain in12

Protected by copyright.

on April 3, 2022 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.42.1.12 on 1 January 1979. Dow

nloaded from

http://jnnp.bmj.com/

Diagnosis and management of tuberculous paraplegia

her back, radiating round her waist on both sides.Two weeks later she noted bilateral sciatic pain,and difficulty in walking. In the next few days herlegs became weaker, and she was readmittedscarcely able to walk. Examination revealed aflaccid, areflexic paraparesis with slight distalsensory impairment to pinprick and touch in thelegs below the knees. The right plantar responsewas extensor; the left was equivocal. The armswere normal, and there was no bladder disturb-ance. There was no spinal tenderness or deformity.The CSF contained 103 lymphocytes/mm3 and13.5 g protein/l but the CSF glucose was normaland acid-fast bacilli were not isolated from thespecimen. Plain radiographs of the spine and amyelogram were normal. Antituberculous chemo-therapy was continued, and prednisone (30 mgdaily) was begun. During the next month therewas rapid improvement in strength, gait, and sen-sory disturbance. The tendon reflexes and theCSF abnormalities gradually resolved.Comment Despite apparently successful man-agement of her tuberculous meningitis this patientdeveloped a flaccid, areflexic paraplegia, with onlyslight sensory disturbance, but with localised andradicular pain suggestive of a lesion at T10 root.These signs were preceded by an increase in theCSF protein. There was no radiographic evidenceof vertebral tuberculosis, and the myelogram wasnormal. The clinical features were consistent withradiculomyelitis, and there was a rapid response totreatment with prednisone, while antituberculousdrugs were continued.

Case 2This 58 year old Asian man, who had enteredBritain from Bangladesh 15 years previously, wastransferred to The London Hospital for manage-ment of tuberculous meningitis. Eight monthsearlier he had complained of low back pain, butradiographs of the lumbar spine showed osteo-arthrosis without other changes. For three monthshe had also complained of loss of appetite, and ofincreasingly severe right sided sciatica. It wasthought that he had a lumbosacral disc prolapsebut there was no improvement after a period ofbed rest. He had been admitted to anotherhospital five weeks previously because of headacheand intermittent fever. Examination was normal,but repeat radiographs of the lumbar spine showederosion of the lower margin of the body of thesecond lumbar vertebra. The chest radiograph wasnormal. The CSF was clear and colourless, con-taining 0.4 g protein/l and 2.2 mmol glucose/l, butno cells. His fever continued. Two weeks later,neck stiffness was noted, and a second lumbar

puncture showed yellow, viscous CSF containing20 lymphocytes/mm3, 9.5 g protein/l, and 4.4mmol glucose/l. Rifampicin, streptomycin, andisoniazid were begun, but he graduallydeteriorated.On admission to The London Hospital he was

drowsy, mumbling incoherently. The right pupilwas larger than the left, but both reacted to light.The fundi were normal. There was a flaccid are-flexic paraparesis, more marked on the right. Bothplantar responses were extensor. The bladder waspalpable. Sensation to pinprick was impaired inboth legs but no upper level could be defined. Thearms were normal.

Prednisolone (20 mg daily) was started. A fewdays later left sided focal seizures developed. Alumbar puncture revealed yellow fluid under lowpressure with spinal block. The CSF contained 16 gprotein/l, 4.83 mmol glucose/l, and 4 leucocytes/mm3. Cerebral angiography showed narrowing ofthe right internal carotid artery at the base of thebrain, irregular narrowing of the right middlecerebral artery, and slow perfusion of its branches.These appearances were considered consistent witharteritis.The subsequent clinical course was complicated

by hyponatraemia and seizures and, because ofdoubt about the diagnosis of tuberculous menin-gitis, a meningeal biopsy was performed through aright parietal burrhole. A granulomatous meningealinfiltrate was found in the biopsy but no acid-fastbacilli were seen. A ventricular reservoir was im-planted and 10 mg streptomycin was given into theventricle daily for 10 days. He improved, but ayear later he remained disabled by dementia andby weakness of the legs. The tendon reflexes hadreturned, the plantar responses were flexor, andsensory loss was no longer apparent. During thistime the lesion of the second lumbar vertebrabecame calcified. Acid-fast bacilli were neverisolated from CSF, sputum, or urine.Comment In this patient the bony lesion, situ-ated adjacent to the intervertebral disc, wascharacteristic of tuberculous infection and this wasprobably the site of invasion of the subarachnoidspace, although acid-fast bacilli were never iso-lated. Sciatic pain early in the course of the illnesswas caused by nerve root involvement by tuber-culoma or localised arachnoiditis, but the para-paresis which developed later resulted from cordinvolvement at a level much higher than that ofthe bony lesion. This was probably caused byvasculitis similar to that seen at angiography andin the cortical meningeal biopsy. The developmentof spinal block was an indication of impendingparaplegia (Brooks et al., 1954).

13



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


David Freilich and Michael Swash

PARAPLEGIA COMPLICATING VERTEBRALTUBERCULOSISThe typical bony changes of vertebral tuberculosismay also be complicated by a paraplegia of mixedradicular and cord origin, but this is usually pre-ceded by a phase of chronic spinal tuberculousmeningitis.

Case 3This 29 year old Asian nurse, born in St Lucia,had lived in England for eight years. In 1970 shehad been treated for sarcoidosis, manifested byarthralgia, granulomatous uveitis, a persistent drycough, pulmonary hilar lymphadenopathy, hyper-calcaemia, and a raised ESR. During this illnessthe Mantoux test was negative, and the Kveimtest positive. Her symptoms and signs graduallyresolved with steroid therapy.

In February 1974 she noted persistent low backpain with left sided sciatica. There was no im-provement after a period of bed rest and, in June1974, a radiograph of the lumbar spine revealedcollapse of the body of the fifth lumbar vertebra.The ESR was 75 mm/hr. During the next threeweeks she developed weakness and numbness firstof the left leg, and then of the right leg. A diag-nosis of metastatic carcinoma was considered butno primary neoplasm could be found. TheESR rose to 108 mm/hr, and the Mantoux testbecame positive at 1: 10000. A chest radiographshowed slight residual hilar lymphadenopathy.A lumbar puncture revealed turbid CSF con-taining 1000 leucocytes/mm3 (mostly neutrophils),5 g protein/I, and 2.67 mmol glucose/l. Veryscanty acid-fast bacilli were seen in smears ofthis specimen, and cultures were later reportedpositive.The patient was transferred to The London

Hospital. On examination at that time there wasmarked weakness of both legs, especially of hipflexion and of all movements at the ankles. Shecould not sit up unaided. Tone was reduced andboth ankle jerks were absent. In the right leg therewas sensory impairment in the L5, S1, and S2dermatomes, and in the left leg in part of the L5dermatome. The lower lumbar spine was tenderand painful. Straight leg raising was limited bypain to about 200. There was retention of urine,with incontinence. Treatment with rifampicin,ethambutol, and isoniazid was started and sixweeks later her legs were stronger although distalweakness, especially of eversion and dorsiflexionat the ankles, remained prominent. The sensorydisturbance improved, and bladder function re-turned. Some weeks later a spinal fusion was per-formed, and a sequestrum was removed from the

body of the fifth lumbar vertebra. During the nextfew months there was an excellent functionalrecovery.Comment In most reported cases of tuberculosisafter sarcoidosis, both phases of the illness haveaffected the same organ, usually the lung (Scadding,1967), but in our patient there was no clinical evi-dence of pulmonary tuberculosis. The subsequentvertebral tuberculous infection was only recognisedwhen lumbar puncture showed the features oflocalised spinal tuberculous meningitis. The radio-logical feature of bony collapse was misleading,since tuberculosis of the spine more commonlypresents with erosion of the margin of a vertebralbody, as in our case 2. Conservative managementwith antituberculous drugs, without myelography,followed by spinal fusion resulted in an excellentfunctional recovery.

Case 4A 38 year old Asian man presented with bilateralsciatic pain. Radiographs of the spine showed onlyminor degenerative changes, and the chest filmwas normal. His pain continued, and he developedweakness of eversion of the left foot, with sensoryimpairment on the lateral aspect of the foot. Theleft ankle jerk was absent, and both plantar re-sponses were flexor. There was slight tenderness inthe lower lumbar region. A lumbar myelogramrevealed almost complete extradural obstruction ofthe column of contrast medium at the level ofthe body of the fifth lumbar vertebra. Tomogramsof this region showed a small destructive lesion ofthe body of the fifth lumbar vertebra. The adjacentintervertebral disc spaces were normal. The CSFcontained 0.5 g protein/l and 4 lymphocytes/mm3.The ESR was 55 mm/hr. It was consideredthat the most likely diagnosis was metastaticneoplasm.At laminectomy a mass, arising from the body

of the fifth lumbar vertebra, was found to containa creamy fluid, from which acid-fast bacilli weresubsequently cultured. Biopsy specimens showedgranulomatous inflammatory changes, with giantcells. Postoperatively he was treated with anti-tuberculous chemotherapy, and a slow but fullrecovery ensued.Comment Vertebral tuberculosis with extraduralcompression of the cauda equina by tuberculousgranuloma was not suspected before operation.The CSF protein was normal, and the minimallyraised cell count in the CSF did not arouse sus-picion. Although the tuberculous granuloma wasopened and partly excised without a preliminaryperiod of treatment with antituberculous drugs,meningitis did not follow the operation.

14



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



INTRADURAL SPINAL TUBERCULOMA

Extradural tuberculomas usually arise as a com-

lication of tuberculosis of the spine (see cases 3and 4) although a few cases have been reported inwhich no bony involvement or extradural com-

ponent was discovered (Tandon and Pathak, 1973).Intradural spinal tuberculomas, however, are un-

common. Dastur (1972) found only six examplesin a series of 266 tuberculomas of the central ner-

vous system. We have observed one patient inwhom paraplegia caused by cord compression fromintradural tuberculoma seems to have developedas a complication of tuberculous meningitis.

Case 5

This 24 year old Asian man had come to Britainfrom Tanzania in 1963. In 1970 he developed pul-monary tuberculosis, complicated by tuberculousmeningitis. Acid-fast bacilli were isolated fromboth sputum and CSF. During treatment withstreptomycin, isoniazid, and para-amino salicylicacid he suddenly developed moderate flaccid weak-ness of both legs, with reduced tendon reflexes, butwith extensor plantar responses. There was no painor sensory loss. The CSF protein at this time was

25 g/l. This weakness recovered during a period ofabout two weeks without change in management.Two years later, while taking antituberculous

chemotherapy, he presented with a three monthhistory of increasing weakness and numbness ofhis legs, with urgency and frequency of micturi-tion. For two weeks he had noticed persistent painin the middle of his back and for a week he hadbeen unable to stand unaided. On examinationthere was a dense flaccid paraplegia with extensorplantar responses. The tendon reflexes were absent.There was dense sensory impairment to all mo-

dalities below a level at T12 dermatome. Radio-graphs of the spine were normal. The chestradiograph showed pleural thickening. but no

evidence of active pulmonary tuberculosis. A lum-bar myelogram showed complete obstruction tothe flow of contrast medium at the level of thebody of the eleventh thoracic vertebra. At laminec-tomy a dorsally situated subarachnoid tuberculomawas excised. This tuberculoma consisted of numer-

ous small tuberculoid granulomata which extendedover several segments of the cord. The cord was

slightly indented by this lesion. His legs remainedweak and spastic but bladder control and distal sen-

sation returned.Comment In this patient a zone of intraduralgranuloma and arachnoiditis extended over severalspinal segments. In most previous reports this dis-order has presented as a progressive paraplegiawithout a history of meningitis (Tandon and

Pathak, 1973). In some of these cases localisedsolid tuberculoma compressed the cord (Arseniand Samitcar, 1960; Jenkins and Hill, 1963; Par-sons and Pallis, 1965), but in others a more d-ffusegranulomatous mass surrounded the cord and wasassociated with localised arachnoiditis (Bucy andOberhill, 1950). Our case seemed to combinefeatures of both these descriptions.

TUBERCULOUS RADICULOMYELITIS (ARACHNOIDITIS)Tuberculous radiculomyelitis occurs typically asthe first manifestation of tuberculosis of the centralnervous system (Wadia, 1973b). This occurred inthree of our patients.

Case 6This 18 year old Asian had entered Britain fromhis native Bangladesh three weeks before his ad-mission to The London Hospital. For six weekshe had observed weakness in his arms and legs,and for two weeks he had been unable to walkunaided. During this time he had developed diffi-culty in micturition with occasional incontinence.He also complained of sharp pains in his arms andlegs, especially in a sciatic distribution. On exam-ination there was a flaccid quadriparesis, affectingthe legs more than the arms. The tendon reflexeswere absent in the legs, but the plantar responseswere flexor. There was no sensory loss. Radio-graphs of the chest and spine, and a myelogramwere normal. The Mantoux test was negative. TheCSF contained 0.47 g protein/I and 46 lympho-cytes/mms but a second specimen taken a weeklater contained 2.6 g protein/l, 480 leucocytes(mainly lymphocytes)/mm:, and 2.1 mmol glucose/1. The blood glucose was 5.0 mmol/l. During thisshort period of observation he deteriorated so thathe was unable to stand without assistance. Treat-ment was, therefore, started with rifampicin,ethambutol, isoniazid, and pyridoxine. He graduallyimproved, and was discharged 10 weeks laterwalking unaided.Two months later there was no clinical abnor-

mality, although the CSF contained 0.9 g protein/land 10 lymphocytes/mm:'. The CSF glucose atthis time was 5.0 mmol/l. Acid-fast bacilli werenever isolated from CSF or urine. Antituberculouschemotherapy was continued for a further periodof six months, and two years later he was well andin fulltime employment.

Case 7This patient, an Asian aged 48 years, who hadmoved to Britain from India several years pre-viously, presented with a five week history of pro-gressive paraplegia, impaired sensation in the legs.

15



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



pains in buttocks and thighs, and impotence.Urinary incontinence developed a few days beforeadmission.On examination there was a flaccid paraparesis,

with sensory impairment to light touch below avague level at the mid thigh bilaterally. Positionsense was impaired in the toes, but painful stimuliwere relatively well perceived. The ankle jerkscould scarcely be elicited but the knee jerks werebrisk and both plantar responses were extensor.The bladder was palpable and required catheteris-ation. The Mantoux test was positive at 1: 1000. Aradiograph of the chest revealed a calcified hilargland, but films of the spine and a myelogram werewithin normal limits. The CSF contained 2 g pro-tein/l, 150 leucocytes (mainly lymphocytes)/mm3,and 2.5 mmol glucose/l. He was treated withrifampicin, ethambutol, and isoniazid, and duringthe next few weeks strength in his legs slowly im-proved, and voluntary micturition returned. How-ever, his legs remained weak and he continued tocomplain of impotence, urinary frequency, andurgency. Acid-fast bacilli were never isolated fromthe CSF or urine.

Case 8This 61 year old Englishman, who had lived allhis life in London, was transferred from anotherhospital for investigation of weakness of both legs.Six months previously he had noted low back pain,radiating into the left leg in a sciatic distribution,and followed a few weeks later by weakness andnumbness of the left foot. There had been noprevious injury. He gradually became aware ofincreasing weakness, affecting both legs. On exam-ination there was a severe flaccid paraparesis. Theleft ankle jerk was absent but the other tendonreflexes in the legs were increased. There was sen-sory loss in the left L5 dermatome. The spine wasnormal. Straight leg raising was painless to 700 inboth legs. The chest radiograph was normal. Plainfilms of the spine revealed spondylotic changes inthe cervical and lumbar regions with narrowing ofthe L5/S1 intervertebral space, but there was nobony erosion at any level. A metrizimide myelo-gram showed a small filling defect at the left L5/SI intervertebral disc, and thickening of all thelower lumbar and sacral nerve roots. The CSFcontained more than 2 g protein/l, 163 leucocytes(mainly lymphocytes)/mm3, and 2.5 mmol glucose/1. The ESR was 45 mm/hr.Carcinomatous meningitis was considered the

most likely diagnosis but cytological examinationof the CSF and sputum failed to show malignantcells. An exploratory laminectomy was performed.The roots of the cauda equina were thickened and

reddened. A biopsy of the meninges revealed agranulomatous inflammatory response, withoutgiant cells or acid-fast bacilli. Cultures of thebiopsy specimen, and of the CSF failed to growacid-fast or other bacilli, or fungi. The Mantouxtest was positive at 1 :1000. A liver biopsy wasnormal and the Kveim test negative. Serologicaltests for syphilis were negative.

After the laminectomy he continued to deterior-ate, so that he became unable to stand, and had tobe catheterised. Sensory impairment developedbelow a level at about T1O dermatome. Anti-tuberculous chemotherapy consisting of rifampicin,streptomycin, and isoniazid was started but therewas no improvement. Four weeks later a righthemiplegia, without visual field defect or aphasia,developed. Steroid therapy was added withouteffect on the paraparesis, but the hemiplegia gradu-ally improved.Comment In case 8 the diagnosis remained un-certain. It was assumed that the granulomatousmyeloradiculitis was caused by tuberculous infec-tion, since fungal infection, syphilis, carcinoma,and sarcoidosis had been excluded. There was noevidence of systemic vasculitis, of lumbosacral discprotrusion, or of Pott's disease. Despite treatmentwith antituberculous drugs and steroids there wasno improvement.

TUBERCULOSIS OF THE CENTRAL NERVOUS SYSTEM50 YEARS AGOWe have examined the microfilmed case recordsof all patients with tuberculosis of the centralnervous system admitted to The London Hospitalin the years 1920 and 1921. There were 66 patients.The majority were children (mean age 8.5 years),and nearly all were of English or European origin.Thirty-six were female.

All these patients suffered from tuberculousmeningitis and all died; 94% survived for less thantwo weeks after admission. In 52 (79%) the illnesslasted less than two weeks before diagnosis, but inthe remainder there was a nonspecific prodromalillness lasting one to three months, consistingprincipally of malaise and weight loss. Sixty-twopatients presented with the clinical syndrome ofuncomplicated meningitis. In four patients therewere additional features. One of these developedfocal epilepsy, two had hemiplegia, and one para-plegia. The CSF was examined in 51 of the 66patients. The CSF protein was invariably raised(but it could not be measured accurately at thistime), and the cell count ranged from 8 to 381leucocytes/mm3 (mean 85/mm3). Acid-fast bacilliwere found in smears of the CSF sediment in 15cases (29%).

16



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



Postmortem examination was performed in 42cases. In all of them there was tuberculous menin-gitis with pulmonary and miliary tuberculosis.There was hydrocephalus in one case. Solitarycerebral tuberculomas were found in six cases;one of these patients had focal epilepsy, and theother had a hemiplegia. The other patient with ahemiplegia had tuberculous arteritis of the middlecerebral artery.

Discussion

The clinical distinction between the various formsof spinal tuberculosis with paraplegia is difficultto maintain in practice (Tandon and Pathak, 1973).In recognising the similar clinical and pathologicalfeatures of these patients, Wadia and Dastur (1969)and Wadia (1973b) have suggested that the desig-nation tuberculous radiculomyelitis be used as ageneric term to include cases previously categorisedas arachnoiditis (Ransome and Monteiro, 1947),intradural spinal tuberculoma (Jenkins and Hill,1963; Parsons and Pallis, 1965), or granuloma(Bucy and Oberhill, 1950; Arseni and Samitcar,1960), or as spinal cord complications of tuber-culous meningitis (Brooks et al., 1954; Dibble andCascino, 1956; Bawa and Wahi, 1961; Kocen andParsons, 1970). These patients' tuberculous para-plegias differ, however, both in origin and course,from those caused by primary tuberculous radi-culomyelitis. Wadia (1973b) has taken account ofthis difference and has suggested the followingclassification: (1) primary tuberculous spinal radi-culomyelitis (cases 6, 7, and 8); (2) radiculomyelitissecondary to tuberculous basal meningitis (cases 1and 5); (3) radiculomyelitis secondary to vertebraltuberculosis (cases 2, 3, and 4).

This concept allows recognition of the almostinvariable combination of root and cord signs inpatients with lumbar infection, accounting for thelocal and radicular pain, and the flaccid, areflexicparaparesis with distal sensory impairment and ex-tensor plantar responses. The observation thatspinal block often occurs before the developmentof paraplegia (Wilson, 1940; Brooks et al., 1954)both in patients with classical tuberculous menin-gitis (cases 1 and 2) and in patients with vertebraltuberculosis (cases 3 and 4) is thus easilyunderstood.Primary tuberculous radiculomyelitis accounted

for 20 of the 38 cases of tuberculous radiculomy-elitis reported by Wadia and Dastur (1969), but ithas not been recognised in this country. Indeed,Elkington (1936, 1951) entirely discounted tuber-culosis as a cause of arachnoiditis. Our cases 6, 7,and 8 were typical examples of the disorder. In

case 8, however, the tuberculous origin of the radi-culomyelitis could not be established, even aftermeningeal biopsy. Syphilis, pyogenic, or fungal in-fection, carcinomatous meningitis, and a variety ofnoninfective agents may cause this syndrome, anda similar localised spinal arachnoiditis may occurrarely after myelography, or after other intra-thecal infections, particularly in the presence ofother spinal disorders. (Shaw et al., 1978). Nonethe-less, Wadia and Dastur (1969) failed to find acause in 10 of their 70 cases. Primary tuberculousradiculomyelitis probably arises as a localisedarachnoiditis from a primary focus in the meninges(Rich, 1951). The CSF changes are thus typical ofan indolent meningitis; indeed, the CSF glucosemay be normal (Wadia and Dastur, 1969; Wadia,1973b). Myelography usually reveals multiple,irregular filling defects, cyst formation, or spinalblock (Wadia, 1973b; Shaw et al., 1978). Treatmentwith antituberculous drugs may need to be pro-longed, and some patients show little improvement(Wadia, 1973b). However, two (cases 6 and 7) ofour three cases of primary tuberculous radiculo-myelitis made an excellent functional recovery.Four (cases 1, 3, 4, and 5) of our five patients withradiculomyelitis secondary to vertebral tuber-culosis or to tuberculous meningitis also recoveredsubstantially. The role of steroid therapy is un-certain, but in case 1 there seemed to be a rapidimprovement when steroids were started.

Arachnoiditis (radiculomyelitis), usually due totuberculosis (Wadia, 1973b), is a common cause ofparaplegia in India (Ramamurthi, 1961; Mani etal., 1969) but similar spinal complications of tuber-culosis have, until recently, been unusual inBritain, as shown by the experience at The LondonHospital 50 years ago, when tuberculous meningitiswas a uniformly fatal disease. However, spinaltuberculomas are recorded in the older Englishliterature (Hunter, 1902; Horsley, 1909), andWilson (1940) recognised that transverse myelitiscould occur both as a presenting feature, and as acomplication of tuberculous meningitis. Ourcurrent experience of these disorders reflects aninterplay of several factors, including differencesin host immunity because of improved nutritionalstatus, survival from acute tuberculous infectionresulting from effective antituberculous treatment,and the arrival of Asian immigrants who seempeculiarly liable to develop subacute or chronicforms of the disease.

References

Arseni, C., and Samitcar, D. C. (1960). Intraspinaltuberculous granuloma. Brain, 83, 285-292.

17



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from



Bawa, Y. S., and Wahi, P. L. (1961). Spinal tuber-culous meningitis. Journal of the Indian AledicalAssociation, 37, 449-452.

Brooks, W. D. W., Fletcher, A. P., and Wilson, R. R.(1954). Spinal cord complications of tuberculousmeningitis: a clinical and pathological study.Quarterly Journal of Medicine, NS, 23, 275-290.

Bucy, P. C., and Oberhill, H. R. (1950). Intraduralspinal granulomas. Journal of Neurosurgery, 7, 1-12.

Dastur, H. M. (1972). A tuberculoma review with somepersonal experiences. II. Spinal cord and its cover-ings. Neurology (India), 20, 127-131.

Dastur, D. K., Lalitha, V. S., and Prabhakar, V.(1968). Pathological analysis of intracranial space-occupying lesions in 1000 cases including children.Part 1: age, sex and pattern; and the tuberculomas.Journal of the Neurological Sciences, 6, 575-593.

Dastur, D. K., and Wadia, N. H. (1969). Spinal menin-gitides with radiculomyelopathy. Part 2: pathologyand pathogenesis. Journal of the NeurologicalSciences, 8, 261-297.

Dibble, J. B., and Cascino, J. (1956). Tuberculoma ofthe spinal cord. Journal of the American MedicalAssociation, 162, 461-462.

Elkington, J. St. C. (1936). Meningitis circumscriptaspinalis. Brain, 59, 181-203.

Elkington, J. St. C. (1951). Arachnoiditis. In ModernTrends in Neurology, vol. 1. Edited by A. Feiling.Butterworth: London.

Horsley, V. (1909). Chronic spinal meningitis: itsdifferential diagnosis and surgical treatment. BritishMedical Journal, 1, 513-517.

Hunter, W. K. (1902). A case of tubercular tumourof the spinal cord in a child two years old. Brain,25, 226-239.

Jenkins, R. B., and Hill, C. (1963). Intradural spinaltuberculoma with genito-urinary symptoms. Archivesof Neurology (Chicago), 8, 539-543.

Kocen, R. S., and Parsons, M. (1970). Neurologicalcomplications of tuberculosis: some unusual mani-festations. Quarterly Journal of Medicine, NS, 39,17-30.

Mani, K. S., Mani, A. J., and Montgomery, R. D.(1969). A spastic paraplegia syndrome in SouthIndia. Journal of the NeurologiCal Sciences, 9, 179-199.

McNicol, M. W., Mikhail, J. R., and Sutherland, I.(1971). Tuberculosis in Brent. Postgraduate MedicalJournal, 47, 691-693.

Parsons, M., and Pallis, C. A. (1965). Intradural spinaltuberculomas. Neurology (Minneapolis), 15, 1018-1022.

Ramamurthi, B. (1961). Spinal arachnoiditis. IndianJournal of Medical Sciences, 15, 776-781.

Ransome, G. A., and Monteiro, E. S. (1947). A rareform of tuberculou3 meningitis. British MedicalJournal, 1, 413-414.

Rich, A. R. (1951). The Pathogenesis of Tuberculosis.Second edition. Oxford University Press: London.

Scadding, J. (1967). Sarcoidosis. Eyre and Spottis-woode: London.

Shaw, M. D. M., Russell, J. A., and Grossart, K. W.(1978). The changing pattern of spinal arachnoiditis.Journal of Neurology, Neurosurgery, and Psychiatry,41, 97-107.

Tandon, P. N., and Pathak, S. N. (1973). Tuberculosisof the central nervous system. In Tropical Neur-ology, pp. 37-62. Edited by J. D. Spillane. OxfordUniversity Press: London.

Wadia, N. H. (1973a). An introduction to neurologyin India. In Tropical Neurology, pp. 25-36. Editedby J. D. Spillane. Oxford University Press: London.

Wadia, N. H. (1973b). Radiculomyelopathy associatedwith spinal meningitides (arachnoiditis) with specialreference to the spinal tuberculous variety. InTropical Neurology, pp. 63-72. Edited by J. D.Spillane. Oxford University Press: London.

Wadia, N. H., and Dastur, D. K. (1969). Spinalmeningitides with radiculo-myelopathy. Part 1:clinical and radiological features. Journal of theNeurological Sciences, 8, 239-260.

Wilson, S. A. K. (1940). Neurology. Edward Arnold:London.

18



http://jnnp.bmj.com

/J N

eurol Neurosurg P


nloaded from


Documents

Diagnosis management to radiculomyelitis