sign in sign up
<<
Home , Meningitis, Stroke

Journal ofNeurology, , and Psychiatry 1994;57:111-115 ill J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.1.111 on 1 January 1994. Downloaded from CLINICOPATHOLOGICAL CASE CONFERENCE

Progressive paraparesis, lymphocytic , and in a 37-year-old man

J Bamford, J Ironside, H McNaughton, C P Warlow

Case presentation Further investigations which were found to The patient is a right-handed male, born in be normal included: barium meal, upper gas- 195 1, and is married. He is an electronics trointestinal endoscopy, CT, MRI of solderer. the brain and cervical cord, sural nerve In January 1989 this man began to feel biopsy, aspiration and trephine, tired and lethargic and noticed weight loss serum amylase, antinuclear factor, rheuma- despite a reasonable appetite. He also toid factor, , anticardi- described night sweats, possibly in association olipin antibodies, and complement levels, with a temperature. By March he was begin- B12 and , blood porphyrins, Treponema ning to notice weakness in the right leg, and pallidum haemagglutination assay (TPHA), then the left leg. He developed tingling and HIV antibodies, CD4/CD8 ratio, antibodies numbness in the left foot, which spread up to against toxoplasmosis, mycoplasma, borrelia, the knee, and with aching in the left leg. He Epstein-Barr , herpes zoster, herpes began to notice in the right testicle and simplex, cytomegalovirus, measles, and right leg; the latter gradually became numb. virus titres, and an ultrasound scan of By May he was hardly able to walk and he the abdomen including liver, spleen, and developed . His health was kidneys. Nerve conduction was normal in the otherwise good. He did not smoke and drank arms but in the legs, the left sural sensory about three pints (1 7 litres) of a day. He action potential was absent and the right was previously had had multiple exostoses (famil- small. There was no conduction in the per- ial) removed from both hands. There was no oneal nerves. Int. the legs, the muscles other family history of illness. appeared to be slightly denervated on EMG. On examination in June 1989 he was thin A muscle biopsy was non-specifically abnor- and was noted at times to have mild pyrexia. mal, however, with no definite evidence of He had lost about 25 kg in weight over the denervation. There was selective type II fibre previous six months. General examination atrophy. http://jnnp.bmj.com/ was otherwise, and remained, normal. On By the end ofJune 1989 the numbness and neurological examination, higher functions, weakness had extended to his abdomen and and both arms were normal. he was still occasionally pyrexial (up to 38°C) He was areflexic in the legs and plantar with night sweats. He was started on cortico- reflexes were bilaterally extensor. Tone was and improved somewhat neuro- normal. Both legs were globally weak, the logically but developed pressure sores. right one more than the left. Pain sensation Azathioprine was added and the steroids were was reduced in both feet on September 26, 2021 by guest. Protected copyright. Department of and along the inner gradually reduced in October. Ten days after , St James's aspect of the right leg. Joint-position the steroids were stopped, he awoke to find University Hospital, was lost in both feet. At this stage a urinary that he was weak in the left arm and face, and Leeds, UK catheter was inserted and he was confined to had slurred J Bamford speech. On examination that day a . there was and an upper motor neu- Western General Hospital, Edinburgh, The following initial investigations were ron weakness of the left side of his face. He UK found to be normal: full blood count, calcium, was weak in the left arm and the left arm Department of phosphate, electrophoresis, alkaline reflexes were brisker than the right. In the Pathology phosphatase, thyroid function, , liver legs he had no movement at all, absent J Ironside function (except y-glutamyltransferase = 69 reflexes, and plantar reflexes were flexor. Department of units/l, normal <42), chest radiograph, ECG, There was a sensory level at T4 and there was Clinical blood cultures, and a full myelogram. The no joint-position sensation in the feet. A CT H McNaughton CSF was xanthochromic, however, and pro- scan on the next day showed a low-density C P Warlow tein levels ranged from 3-1 to 7-0 g/l. The area lateral to the right lateral ventricle. The Correspondence to: immunoglobin G level and index were nor- CSF protein was still raised at 2-3 and Dr H McNaughton, g/l Department of Clinical mal. There were 189 /ml. The there were 38 lymphocytes/1l. All CSF cul- Neurosciences, Western level of glucose in the CSF was 2-6 mmol/l tures were again negative and the blood tests General Hospital, Crewe Road, Edinburgh, compared with 6-2 mmol/l in the blood. were unremarkable. The haemoglobulin was EH4 2XU, UK Cultures for , fungi, and 11 2 g/l and the erythrocyte sedimentation Received 31 August 1993 were negative on several occasions; India ink rate was 24 mm in the first hour. The and in revised form pre- 12 October 1993. stains were also negative. There were no vious tests were repeated and were again nor- Accepted 18 October 1993 malignant cells. mal. Brain biopsy was declined by the patient. 112 Bamford, Ironside, McNaughton, Warlow J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.1.111 on 1 January 1994. Downloaded from Figure 1 Non-contrast on axial brain CTscan early since the history is of a global weak- showing haemorrhagic ness in the legs. This, rather than the focal infarct in the region ofthe weakness usually associated with cauda left basal ganglia. equina , suggests cord involvement, albeit that the signs (areflexia and flexor plan- tars), point to a problem. Additionally, one notes that proprioception is much more affected than spinothalamic modalities. In vasculitic conditions affecting peripheral nerves, the larger myelinated fibres may be picked out preferentially. The normal myelograms reasonably exclude a mass and I note the absence of thickened or infil- trated nerve roots which might have been a source of a tissue diagnosis, although the can be normal even with extensive lymphomatous infiltration of nerve roots. Finally, the testicular pain might be radicular in origin rather than due to primary testicular pathology. So we come to the CSF: there is a very high protein content and the immediate assumption would be that the xanthochromia relates to this, although, in view of subse- quent events, the possibility of a haemor- rhagic pathology needs to be remembered. There is a response and a reduc- tion in the CSF glucose compared with the blood of more than 50%. The of these abnormalities is extensive, Over the next several months he developed but in the clinical context, tuberculous infected pressure sores, occasional chest meningitis (TBM) must be high on the list. , and urinary infections. He had Hospitalised patients with TBM have CSF intermittent pyrexia. In May and June 1990 white cell (predominantly lymphocyte) counts the were being reduced when, of up to 1000/ml with high protein and low on 17 June, he suddenly became aphasic with glucose levels. The direct smear is positive for weakness of the right arm, possibly with a acid- bacilli in 10-85% of cases, depend- right homonymous hemianopia. A plain CT ing on technique, with positive cultures in scan (fig 1) showed a haemorrhagic infarct in 45-90%.l The was normal the left basal ganglia. His was so in this case but up to 50% of patients with badly impaired that a nasogastric tube was TBM have no evidence of tuberculosis else- inserted. An EEG showed widespread slow- where.' So far, polymerase chain reaction ing bilaterally. The corticosteroids were again techniques for rapid diagnosis of TBM have http://jnnp.bmj.com/ increased and there may have been some been disappointing. The clinician here was improvement. He was still completely para- faced with the problem of an active CSF lysed in his legs, catheterised, and confined to compatible with TBM, in a man whose neu- a wheelchair. He was eventually placed in rological condition was deteriorating rapidly. long-term care, and gradually declined. He Furthermore, the use of steroids was being died in January 1992. As far as we know there contemplated. There are those clinicians, and

were no further, overt neurological events, I think I may be one of them, who would on September 26, 2021 by guest. Protected copyright. except that, weeks before his , he had start treatment with antituberculosis in an episode of and weakness in his this situation while awaiting the results of cul- right arm. A postmortem examination was tures. On the other hand, the normal, con- performed. trast-enhanced CT scan, in particular the absence of basal meningeal enhancement or , would argue against TBM. Discussion was obviously looked for DR JOHN BAMFORD: with India ink stains on multiple CSF speci- This man's initial symptoms are non-specific, mens, although I would have preferred to but give hints of an inflammatory, infective, have had the result of the cryptococcal anti- or neoplastic aetiology. The neurological gen test.2 deterioration was rapid with the initial symp- clearly merits consideration. toms of right leg weakness in March, and yet Matthews' review3 remains the most useful by the same summer he was paraplegic and here. There are neurological symptoms in 5% catheterised. The presenting symptoms sug- of patients with sarcoid although gest a cauda equina syndrome although there pathologists find evidence of CNS involve- were some unusual features, not the least ment more frequently. Neurosarcoid may being the fact that the motor deficit preceded occur without demonstrable involvement the sensory one. I wonder whether there was elsewhere, and may present in ways that central and peripheral disease in the neuraxis could explain this patient's symptoms. A Progressive paraparesis, lymphocytic meningitis, and stroke in a 37-year-old man 113

cauda equina syndrome is a recognised pre- event in the other hemisphere, and I would J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.1.111 on 1 January 1994. Downloaded from sentation of neurosarcoid, although less com- have requested an echocardiogram, although mon than isolated cranial nerve palsies and a this may be negative in marantic . granulomatous leptomeningitis. Furthermore, For what it is worth, there is no hyperdensity it is capable of giving a mixture of central and in the middle cerebral on the non- peripheral disease. Most of the CSF findings enhanced CT scan, which would have of the case were compatible with neurosar- suggested the presence of . coid, although one notes that the protein level A CNS is looking more likely was originally higher than those quoted by now. Of the primary vasculitides, polyarteritis Matthews. nodosa has a predilection for the peripheral Although uncommon in the UK, brucel- , and a CSF reaction of this losis should be mentioned. There are six order would be very unlikely. Wegener's types of the small, non-sporing, non-motile, granulomatosis is a possible candidate Gram-negative coccobacillus. All but one of although the lack of chest symptoms counts these are recognised human and against it. Sarcoid remains a distinct possibil- host specificity is not exclusive. Acute brucel- ity and , although unlikely, losis rarely presents to neurologists but they needs to be considered. Although a lympho- should be aware of the distinction between cytic response of this magnitude can be seen chronic and neurobrucellosis. The in the CSF of patients with Behcet's disease, hallmark is the active CSF. There may be an the protein is probably too high, and he acute , but also a chronic would not fulfil the usual diagnostic criteria. meningitis, with or without a vasculitic or Isolated CNS angiitis usually presents with granulomatous component, or both.4 In the and . Secondary vasculitis present case, the three months of lassitude may be seen with lymphoma, histiocytosis, and weight loss at the onset would be com- and hairy cell leukaemia, but none of the patible with brucellosis, as would the subse- other investigations support these diagnoses. quent neurological symptoms and CSF Similarly, there is a lack of support for an findings. -related vasculitis, such as that fol- A comment should be made about the lowing herpes zoster. . There are asymmetrical The second cerebral event was a haemor- sural potentials of a degree that would be rhagic infarct which was larger than the abnormal for a man of this age, and thus sup- infarct on the right, although probably still port a multifocal disorder rather than a dif- within the striate territory. Miller Fisher" sug- fuse polyneuropathy. We are not told which gested that such haemorrhagic change was muscle was biopsied, and therefore one can- the hallmark of cardiogenic , with not make too much of the apparent discrep- the impacting and causing ischaemia ancy between the EMG and the muscle of the distal capillaries followed by fragmen- biopsy with respect to denervation. I would tation and the rupture of those same capillar- also have done a muscle biopsy, although the ies when re-exposed to systemic blood findings turned out to be rather non-specific: pressure. More recently, CT studies have in arteritic conditions, muscle biopsy has suggested that it may be the size of the been said to have a yield of 35-40%, and infarct, rather than the aetiology, that influ- granulomas are sometimes found.5 ences this change, and, in at least one case, http://jnnp.bmj.com/ At this stage I would have liked to have haemorrhagic change has been demonstrated seen the results of an antineurotrophil cyto- distal to an occluded artery. Anecdotally, vas- plasmic antibody (ANCA) test, and may well cultic infarcts are more likely to undergo have asked for ENT and ophthalmic opin- haemorrhagic change. In the absence of a car- ions, thinking of Wegener's granulomatosis diac lesion, or evidence of degenerative vas- and uveitis respectively. It is clear that lym- cular disease, the presence of in

phoma was, quite correctly, repeatedly con- both hemispheres points towards a vasculitic on September 26, 2021 by guest. Protected copyright. sidered as a diagnostic possibility, but all process. reasonable tests were negative. I think the diagnosis rests between sar- Following treatment with steroids, he had coidosis and isolated angiitis of the CNS, and a stroke-like episode involving the left arm I would favour the latter. Although it usually and face, with dysarthria. The CT scan shows presents with cerebral symptoms, such as a low-density area just above the internal cap- change in mentation, headache, and hemi- sule on the right, which would be quite com- , 20% present with paraparesis, , patible with an infarct. The area is too large and weight loss. There are published cases to be in the territory of a single perforating presenting with a fluctuating , and artery, despite the fact that the clinical details one case of a cauda equina syndrome.7 As far we are given would be compatible with a par- as the CSF is concerned, a number of cases tial lacunar syndrome (although one cannot presenting with a myelopathy had high CSF be sure that adequate testing of non-domi- and a lymphocyte response similar to nant cortical function was performed). Thus this case. Intracerebral and subarachnoid one is thinking of disorders that might involve haemorrhage occurs (perhaps reflected in the multiple striate branches; a vasculitis related xanthochromic CSF). The condition is vari- to a basal meningitis or an embolus to the ably responsive and the overall clinical proximal for example. pattern of this case would be compatible with Marantic endocarditis needs to be consid- the diagnosis. It is the condition par excellence ered, especially in view of the subsequent where a leptomeningeal biopsy is indicated. 114 Bamford, Ironside, McNaughton, Warlow

Figure 2 The cut surfaces to J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.1.111 on 1 January 1994. Downloaded from ofthe cerebrum show a secondary infection. Cachexia was noted recent haematoma in the and there were widespread pressure sores. On rightparietal lobe, with old gross examination of the brain, multiple small infarcts in the head of the foci of right caudate nucleus haemorrhage within the subarachnoid (which appears shrunken) space were noted. The left occipital lobe and in the globus pallidus, showed a greyish discoloration with multiple , internal capsule, areas of recent and on the haemorrhagic suggestive left. A smaller area of infarction. On cross section, a small area of haemorrhagic infarction is established infarction was noted in the mid- noted in the left parietal and at several regions in the cerebellar lobe. hemispheres. In the cerebral hemispheres, haematomas were present within the parietal lobes bilaterally, and established areas of infarction were noted in the basal ganglia on the left side and in the head of the caudate nucleus on the right (fig 2). Multiple DR JAMES IRONSIDE: smaller areas of haemorrhagic infarction were The postmortem examination was performed noted throughout the cerebrum. The spinal by a general pathologist in a different town. cord was markedly abnormal, being shrunken The brain and were sent here for and discolored and reduced to a ribbon-like detailed examination. strip in the mid-thoracic region. On cross sec- The major findings were of bi- tion, there was extensive tissue destruction at lateral suppurative bronchopneumonia with this site, and multiple smaller areas of haem- hilar lymphadenopathy thought to be orrhage and congestion were noted elsewhere around blood vessels in the cord. Histological A examination revealed an inflammatory infil- trate around blood vessels in the and parenchyma, with necrosis of small arter- ies and occlusion of some vessels (fig 3A). The veins were well preserved and spared by the inflammatory process. The inflammatory cell infiltrate was polyclonal and there were no organisms including fungi. No malignant t,*:. K ,t . w.f cells were present. In the brain itself there were collections of cells around blood vessels f.I consisting of epithelioid macrophages. These were by definition granulomas, and contained * occasional giant cells (fig 3B). In the spinal f*~.4',r ; Z 'S , 4. 7; cord, the anterior spinal artery was involved by the inflammatory process with granulo- matous change. Some of the vessels in the cord were occluded and the cord itself was very damaged. In cross section the mid- http://jnnp.bmj.com/ thoracic region was demyelinated with tissue destruction. A section of the cauda equina showed inflammatory change around small vessels and a small .*. 4* i* w ;. * r granuloma. s.. There was an acute bronchopneumonia in .-.. .. s * * **u 4' . ;It.. the . Lymph nodes from the hilum v* 4'>;, * 's showed large, pink, epithelioid macrophages on September 26, 2021 by guest. Protected copyright. Oq:* with no necrosis or caseation although fibro- 5409. sis was present. Giant cells were present but stains for tuberculosis were negative and 44 there were no inclusions characteristic of sar- coid. In summary, there was haemorrhage and infarction throughout the CNS, with a granu- v~~~~ ~ ~~~~~~~~3*.; lomatous angiitis. The significance of the lymph node changes in the hilum are uncer- -~~~~~~~~4 tain but they are not diagnostic of sarcoid. Other organs were extensively sampled and *~~~~: v*.' .'* ...... :.'~4:.: ~~~~~~a* there was no evidence of granulomas else- where including the testes.

Figure 3(A) In the leptomeninges there is a dense perivascular inflammatory infiltrate Conclusions around small and arterioles (right). These cells comprise lymphocytes, plasma Dr J7ohn Bamford's diagnosis: isolated angiitis cells and epithelioid macrophages. One ofthe small arteries showsfibrinoid necrosis of the CNS. with occlusion of the lumen. The veins in this region (left) appear relatively spared. (B) Granulomatous with multinucleate giant cells is present in the Pathological diagnosis: Granulomatous angi- around a cerebral arteriole. No evidence ofnecrosis or caseation is itis of the CNS; granulomatous lymphadenitis noted. of hilum. Progressive paraparesis, lymphocytic meningitis, and stroke in a 37-year-old man 115 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.57.1.111 on 1 January 1994. Downloaded from 1 Molavi A, LeFrock JL. . Med Clin Comment North Am 1985;69:315-31. Whether the granulomatous changes in the 2 Sabetta JR, Andriole VT. Cryptococcal infection of the hilar lymph nodes were an incidental finding, . Med Clin North Am 1985; 69:333-44. or connected with the neurological process 3 Matthews WB. . In: Vinken PJ, Bruyn remains unclear. Granulomatous angiitis of GW, eds. Handbook of clinical neurology. Amsterdam: Elsevier/North Holland; 1979;38:521-42. the CNS is probably a non-specific pathologi- 4 Bharucha NE, Bhabha SK, Bharucha EP. Infections of cal reaction to a number of disease the nervous system. In: Bradley WG, Daroff RB, Fenichel GM, Marsden CD, eds. Neurology in clinical processes.7 The differentiation between gran- practice. Boston: Butterworth-Heinemann 1991; ulomatous angiitis of the CNS and neurosar- 1079-80. 5 Loughlin M. Muscle biopsy: a laboratory investigation. coidosis, for example, may be a matter of Oxford: Butterworth-Heinemann 1993;143. definition rather than representing two sepa- 6 Fisher CM, Adams RD. Observations on brain embolism with special reference to the mechanisms of hemor- rate disease processes.8 rhagic infarction. Neuropathol Exp Neurol 1951;10: 92-4. 7 Hankey GJ. Isolated angiitis/ of the central nervous system. Cerebrovasc Dis 1991;1:2-15. This CPC was presented at the Edinburgh Advanced 8 Uhrich H. Neurosarcoidosis or granulomatous angiitis. A Neurology course, April 1 1993. problem of definition. Mt Sinai _r Med 1977;44:718-25. http://jnnp.bmj.com/ on September 26, 2021 by guest. Protected copyright.