74ournal of Neurology, Neurosurgery, and Psychiatry 1995;58:374-378

CLINICOPATHOLOGICAL CONFERENCE

Papilloedema and lymphocytic meningitis in a 68year old man

Chris Allen, Jeanne Bell, Harry McNaughton, Charles Warlow

Department ofNeurology,Addenbrooke'sHospital, CambridgeC AllenDepartment ofNeuropathology,Western GeneralHospital, EdinburghJ BellDepartment ofClinicalNeurosciences,Western GeneralHospital, EdinburghH McNaughtonC WarlowCorrespondence to:Dr H McNaughton,Department of Medicine,Hutt Hospital, Private Bag31967, High Street, LowerHutt, New Zealand.Received 15 August 1994.Accepted 2 September 1994

Case presentationThis 68 year old man presented in lateFebruary 1992 with a one month history ofintermittent blurring of vision, lasting sec-onds, in the right eye. He also had mild inter-mittent occipital headache, mostly at night,with some nausea in the morning. He hadnoticed some hearing loss in the left ear withtinnitus for about a year, and his wife thoughthe was a little unsteady on his feet. He was adiabetic controlled on diet, and there hadbeen some recent weight loss. In earlier yearshe had been a heavy drinker, but no longer.He had never smoked. He was known to bein atrial fibrillation and had previously had amyocardial infarction. Many years ago therehad been a "shadow on the lung", possiblytuberculosis.On examination he was in atrial fibrillation

but there were no other general medicalabnormalities; his chest and abdomen werenormal; there was no lymphadenopathy.There was bilateral papilloedema much moreon the right than the left but he had normalvisual acuities. There was mild sensorineuraldeafness, more on the left than the right.There were some slight cerebellar signs onthe left, his reflexes were brisk but symmetric,

and the plantar responses were flexor.Computed tomography (fig 1) was

reported as normal, although there was con-cern about aqueduct stenosis. He was poly-cythaemic (haemoglobin 180 g/l and packedcell volume 54%). The other blood indiceswere normal, and the erythrocyte sedimenta-tion rate was 2 mm/h. Numerous other inves-tigations were done at that point and over thenext few weeks and were normal. Theseincluded urea, electrolytes, liver function,proteins and protein electrophoresis, calcium,alkaline phosphatase, B12, folate, serum iron,coagulation screen, antinuclear factor,rheumatoid factor, anticardiolipin antibodies,lupus anticoagulant, creatine kinase, calcium,and chest radiograph. His ECG showed atrialfibrillation.

Because of the possibility of aqueductstenosis a right parietal access device wasinserted and the intracranial pressure peakedat 30-40 cm H2O. The protein concentrationin the CSF was normal (0 3 g/l), but therewere 46 white blood cells/,ul (90% lympho-cytes). The CSF glucose concentration was7-1 mmol/l with a blood glucose concentra-tion of 12A4 mmol/l. No organisms were seenon Gram stain, no acid fast bacilli were seen,

Figure 1 Contrast enhanced axial CT scans of the head showing dilated lateral ventricles with normal basal cisternswithout midline shift.

374

-M.M-.k..

x:w;:. -1.

x .:. ....

on June 1, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.58.3.374 on 1 March 1995. D

ownloaded from

Papilloedema and lymphocytic meningitis in a 68 year old man

and nothing was grown either then or eventu-ally in the way of bacteria, fungi, or mycobac-teria from this specimen. Oligoclonal bandswere absent.No further investigations were done at this

time and it was decided to insert a ventricu-loperitoneal shunt on 9 March to control theraised CSF pressure. By 16 March the pres-sure was less than zero, but there were still 95white blood cells/,ul in the ventricular CSF(60% lymphocytes and 40% neutrophils).The protein concentration was raised at1 84 g/l and Gram positive cocci were seen. Ashunt infection was diagnosed and he wastreated with intrathecal vancomycin and oralflucloxacillin. Some scanty coagulase negativestaphylococci were grown from the CSF. Thepatient's symptoms were not changing much,but by 21 March his ventricular CSF whitecount was normal, even though the proteinconcentration was still raised at 1 38 g/l. Hewas discharged on 27 March.He was readmitted on 1 April complaining

of headache, although on this occasion itoccurred when lying flat. The ventricularCSF was under negative pressure but it wasstill lymphocytic, with a white cell count of60 white blood cells/,ul, and a raised proteinof 2-67 g/l. The shunt was thought to beinfected, and it was changed on 14 April. Hecontinued to lose weight, although the papil-loedema was resolving. The ventricular CSFwhite blood cell count went down to 3 cells/,ulby 21 April, there was no growth, and he wasdischarged home on the 29 April.He was readmitted to hospital on 12 May

complaining of increasing unsteadiness on hisfeet, tiredness, nausea, and sometimes vomit-ing. The papilloedema was much less obviousbut nothing else was found. He was again dis-charged home.He was readmitted on 20 May complaining

of confusion and increasing ataxia of gait,weight loss, and anorexia. On this occasionthe ventricular CSF contained 51 white bloodcells/,ul, with a glucose concentration of

6-6 mmol/l, and a blood glucose concentra-tion of 9 7 mmol/l; the protein concentrationwas still raised at 1-76 g/l. No organisms wereseen in or grown from the CSF. At aroundthis time CSF cytology was requested and nomalignant cells were found, although thespecimen was contaminated with blood. Itwas thought that he had a low grade shuntinfection and he was treated with antibiotics.On 17 June the ventricular CSF was muchthe same, and CT (fig 2) showed bilateral butsmall subdural collections with possibly someenhancement of the basal cisterns, suggestinga meningitic process. On 8 July he had a bar-ium meal, which showed some ulceration, butendoscopy was negative. On 14 July the shuntwas revised. By 21 July he was deterioratingand was thought to have bilateral lower motorneuron facial weakness and bilateral ptosis.He had a period of left ventricular failureprobably due to fast atrial fibrillation, and aurine infection, which was treated with gen-tamicin. His white blood cell count went upfor the first time to 14-7 x 109/1, his haemo-globin was a little lower at 149 g/l, and hiserythrocyte sedimentation rate was 4 mm/h.On 29 July his alkaline phosphatase was229 IU/1.On 1 August a new registrar rotated to the

neurology ward and performed a diagnosticexamination-a procedure not requiring anyinstruments-which was well within hiscapabilities. The patient died four days later.

DR CHRIS ALLENWe have a man who presented at the age of68 with a one month history of blurring ofvision. It lasted seconds and was in the righteye. This sort of intermittent blurring ofvision does not sound like amaurosis fugax, itis not a sudden blackout of vision, and it doesnot seem to be optic neuritis (a less suddensubacute white-out of vision), but suggeststhe transient visual obscurations of someonewith raised intracranial pressure, or at leastpapilloedema. The mild intermittent occipital

_ .

-

__-

b ..

-i. :*:....... ig ...

w* . } n. wY^- i . g * *^* . . . .. . i. ...... . 0.X ^, s 0* ;. uF: . i :o: ::

- w XO: -. .' X* .. t.g.C y .. S:.X §e e

t9 .. eO..t:R;. ..:: & i; o

...., ..j.y, j. bj b K ,,. ...j ,.t ,l. Ng.....":: :C p °'^X y; .' .':'ao::-:

::.^....:....: ,: a..>.}--o=B- :: .e* .i:, s6, .S,: -. . ;se. . :. . .4w: ,{ !: r: 11k- .t :. . ._wfs :: YS: xo-;w ......... F - nBfffl 3 SO gi$!:: . :2wpX P.if-.3 .}gif a"tPS Wj8,,l, >X< >,<*. s q*ie=7g S.: b3ES M @SSe

X Gl.. se. ; , jF Xi' R' :S i 3.^P3:'*. :: ..: *>SSCe..:i...:_.tS.. st:y .:S :N :.:^:9§C'^..XC: §|°w

.a°s^-jk'' jOiC3R "° i -:.4P -, e ...fteVSEe-S : -o J4M*. ::es.zR.O O. 6 . : : .Q:e* ......... tu.<: . : ::. zan o..:. . :. ::: :.:>b. juj:.'sEg ! . . . .¢:. .; .:* - ' ^- i jd .... vfa . 9 ^ .^::9::-.}: -. A* -e*t .... . M .: : .., : : . ,W=-x.; t: e, - § & ,a >^* . X q&.: :u' ^'B;.' g

'.I

Figure 2 Contrast enhanced axial CT scans of the head showing ventricular drainage devices, less definition than before of the basal cisterns, and a

peripheral hyperdense region in the left cerebral hemisphere.

U

.*1

375

on June 1, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.58.3.374 on 1 March 1995. D

ownloaded from

Allen, Bell, McNaughton, Warlow

Figure 3 Section of the prostate gland showing infiltration of subepithelial stroma bylymphoma cells that surround the vessels and penetrate to the immediate subepithelialtissue. The urethral epithelium is shown to the right of the figure. Haematoxylin andeosin. x 100.

I.

f ;

--ii,

7c .o i.

Figure 4 Section of the pons showing infiltration of the subarachnoid space (centre) withlymphoma cells, and infiltration of the Vth cranial nerve (right) with tumour.

Haematoxylin and eosin x 40.

eio*-I

Figure 5 Section ofprostatic lymphoma stained immunocytochemically with an antibodto the B cell marker, L26. The background prostatic stroma is negative. Pleomorphism ofthe neoplastic cells is seen x 200.

headache, mostly at night with nausea in themorning, is suspicious and raises the possibil-ity that he has a mass lesion causing papil-loedema or hydrocephalus, or some othercause of raised intracranial pressure such asone of the syndromes which I like to callpseudotumour cerebri, itself sometimescaused by cerebral venous sinus thrombosis.The hearing loss is most likely to be inciden-tal presbycusis because it has been present forabout a year, with tinnitus. Diabetes suggestsodd vascular complications or local infectionssuch as mucormycosis. Weight loss and thefact that he died within six months of onsetmakes me think of malignancy, chronic sys-temic infection, or inflammation. The rest ofthe medical background raises other possibili-ties: cirrhosis or hepatoma in a previouslyheavy drinker, atrial myxoma or other heartdisease with atrial fibrillation, and tuberculo-sis, sarcoidosis, or aspergillosis with the oldlung shadowing.

Examination showed atrial fibrillation andno other general medical abnormalities. Hischest and abdomen were normal and therewas no lymphadenopathy. I was taught as amedical student, by the surgeons of course,that a rectal examination was part of theabdominal examination so I hope that a rectalexamination was performed here. There wasbilateral papilloedema and normal visualacuities. The normal visual acuities more orless rule out the other causes of swollen look-ing discs. The history and examinationstrongly suggest raised intracranial pressure.Raised intracranial pressure without focalsigns makes one think of hydrocephalus or alesion in a silent part of one hemisphere. Hecould have had raised intracranial pressurefrom a pseudotumour syndrome due tomeningeal disease or cerebral venous sinusthrombosis. Papilloedema with normal CSFpressure can occur in the presence of a serumparaprotein. For example, a recent patient inCambridge presented with an odd vasculiticdisease of the brain demonstrated at biopsyand associated with a serum paraprotein, andone of the features she had on admission waspapilloedema with normal intracranial pres-sure at lumbar puncture. Other possibilitiesare leukaemic infiltration or sarcoidosis.Posterior fossa structural or meningeal dis-ease could explain the ataxia and cerebellarsigns. The hearing loss could be due tomeningeal disease as well.

P. The contrast enhanced CT (fig 1) showsslightly dilated ventricles but the sylvian fis-sures and the cisterns around the brainstemare visible. I cannot see any mass lesions in

4 the posterior fossa and what I can see of hisfourth ventricle is central and not dilated.The cerebral sulci are visible and there is no

4 midline shift, so this man had mild hydro-cephalus with a normal fourth ventricle. I do

4 not think that the hydrocephalus caused hispapilloedema.The only abnormal tests were a raised

haemoglobin, packed cell volume, and ECGevidence of atrial fibrillation. Not available,perhaps by simple omission, were measure-

376

on June 1, 2020 by guest. Protected by copyright.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.58.3.374 on 1 March 1995. D

ownloaded from

Papilloedema and lymphocytic meningitis in a 68year old man

ments of acid phosphatase or prostate specificantigen. The polycythaemia, in a non-smokerwithout obvious lung disease (apart from oldtuberculous shadowing), suggests the possi-bility of cerebellar hemangioblastoma whichwe would have liked to have seen in the pos-terior fossa on CT or a renal tumour, particu-larly adenocarcinoma. Other renal disease,even mild hydronephrosis, can cause poly-cythaemia. Twenty per cent of one series ofpolycythaemia induced by tumours were dueto hepatomas, with testicular tumours lesscommon.' One of the investigations this mandid not have was an ultrasound of hisabdomen. To clarify the diagnosis, I thinkthat he should probably have had an MRangiogram, to make sure there was no cere-bral venous sinus thrombosis, and an ultra-sound or CT of his abdomen. He should havehad some CSF cytology from his lumbar CSFand tests for acid phosphatase and prostatespecific antigen.

I think that he must have been under theneurosurgeons to begin with because theyobviously disagreed with me about the aque-duct stenosis or perhaps were nervous aboutobtaining CSF, which would be the next nec-essary investigation. So he had a right parietalaccess device inserted and his intracranialpressure was raised. The CSF protein was 0-3g/l. This CSF was presumably taken from hisventricle. Ventricular CSF should have a pro-tein concentration about half that of lumbarCSF so this is abnormal.2 He has a low gradelymphocytic meningitis with a normal CSFglucose. Although CSF cultures were nega-tive, no blood cultures were done. Listeria, inparticular, is found more readily in the bloodthan in the CSF.3 Nevertheless, I think that itis unlikely that infection explains the CSFfindings and this includes Lyme disease.He continued to have a headache. This

could be due to carcinomatous meningitis,which commonly presents with a headacheand is the commonest single symptom in aseries from the Sloane Kettering Institute ofmalignant meningitis,4 and he was unsteadyon his feet, which is another common symp-tom of malignant meningitis. If he did havemalignant meningitis, I think that he wasmost likely, based on the information wehave, to have had an abdominal malignancy,gastric or other adenocarcinoma or lym-phoma. Alternatively, a paraneoplastic syn-drome with cerebellar atrophy could causeunsteadiness and an abnormal CSF.Papilloedema from raised intracranial pres-sure would be unusual in this situationalthough papilloedema without raised pres-sure is described in the paraneoplastic syn-drome with optic neuritis.5

Eventually CSF was sent for cytology. TheCSF was probably obtained from the ventri-cle but should preferably come from bothventricular and lumbar sites. There are wellrecorded series in which malignant cells havebeen consistently absent from one of the twosites, ventricular or lumbar, and present inthe other.6 In malignant meningitis the CSFis positive for malignant cells in about 50%

on the first sample, and in the SloaneKettering Institute series eventually 85% werepositive.4 On 17 June the ventricular CSF wasmuch the same and CT was again performed.

This CT (fig 2) is different from the previ-ous one. We can no longer make out the cis-terns clearly and his third ventricle is notvisible although he has had more than oneshunt. There is loss of definition of the cen-tral structures and it looks to me as thoughhis brain is generally rather swollen. The CTreport comments on possible meningealenhancement, which is always difficult tojudge. He has small subdural collections. Iam not sure what the superficial cord-likehigh density in the left hemisphere is. Thiscould be the "cord sign" of cerebral venoussinus thrombosis although I have never seenit.7 I thought that it might be enhancingmeningeal infiltration. The barium meal andendoscopy help to rule out gastric cancer. By21 July he was becoming worse and wasthought to have bilateral lower motor neuronfacial weakness and bilateral ptosis. Suddenlyone wonders about a myasthenic syndromebut I think that in the current contextmeningeal infiltration is much more likely.Now we come to the nice bit at the end of

every clinicopathological conference whenthey half tell you something to try and teaseyou. On 1 August a new registrar rotates onto the neurology ward and performs "a diag-nostic examination, not a procedure requiringany instruments, which was well within hiscapabilities". The patient died four days later.What on earth could he have done?To get the final diagnosis right it is neces-

sary to guess what the registrar did. So, didhe do a rectal examination and find a malig-nant prostate, or a rectal carcinoma? Or didhe examine the abdomen, which we are toldwas previously normal? Had the patientStauffer's syndrome, a syndrome of paraneo-plastic enlargement of the liver withoutmetastasis associated with renal adenocarci-noma?8 Or did the registrar find a greatcraggy lump of a hepatoma? Or did thepatient have melanomatosis and a metastasisfrom this? Did the new registrar go even fur-ther and examine the testicles? Did he find aperitoneal mass from the spread of hismeningeal melanomatosis via his shunt intohis peritoneum? This is unlikely. Did he openthe patient's mouth and look at his genitalsagain and find ulcers suggesting Behqet's dis-ease? Did he look at the skin and find lupuspernio, suggesting sarcoidosis, or amelanoma? Or did he find a lymph nodeenlarged by lymphoma?

In summary, I think that this patient hadmalignant meningitis from an adenocarci-noma. My best guess is a prostatic carcinomabut it may be another intra-abdominal adeno-carcinoma. Forced to provide a differentialdiagnosis, I would suggest a paraneoplasticencephalopathy with one of the previouslymentioned cancers or a cerebral venous sinusthrombosis with one of those cancers. I thinkthat if he had cerebral venous sinus thrombo-sis it is unlikely that he would have died from

377 on June 1, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.3.374 on 1 M

arch 1995. Dow

nloaded from

Allen, Bell, McNaughton, Warlow

this without one of the complications, such asa venous infarction or a major stroke-likeevent.

DR JEANNE BELLWe received two CSF specimens for cytologyduring this man's illness. The initial specimenwas negative for malignant cells. He had afurther CSF sample analysed a few daysbefore he died, which showed malignant cellsthat were thought to be adenocarcinomacells. It was, however, negative for epithelialmarkers. We were thinking of further teststhat we might do on additional specimenswhen he died.

At necropsy his brain was unremarkable.He had a bronchopneumonia and no sign ofany tumour in his lungs. He had heavy lungsand a very large heart with an old myocardialinfarction. There was a 2 cm white nodule inthe cortex of the left kidney. The prostate wasgreatly enlarged by malignant tumour, whichwas also adherent to the base of the bladderand posteriorly to the rectum. I checked thespare cytospins of the CSF that we had had afew days before death to look for prostaticacid phosphatase. This was negative.Nevertheless the presumptive diagnosis wasadenocarcinoma of the prostate metastatic tothe subarachnoid space.

Examination of the fixed brain showedsome atheroma in his vessels, the tracks of theshunt and access devices, a small aqueduct,and mildly swollen brain parenchyma. Theventricles were of normal size. Microscopicexamination of the prostate showed a non-Hodgkin's B cell malignant lymphoma (fig 3).The nodule noted in the kidney was a lym-phoma with lymphoma cells also in the peri-adrenal fat. Pneumonia was confirmed in thelung and the brain showed wide subarachnoidspread of lymphoma involving the cranialnerve roots (fig 4). He had terminal hypoxicchanges in his brain.

Immunocytochemical staining showed pos-itivity for L26 (a B cell marker) and negativityfor MT1 (a T cell marker), so this was a highgrade non-follicular B cell non-Hodgkin'slymphoma. The subarachnoid space wasfilled with tumour cells that had spread intothe parenchyma of the brain with a relatedglial response. Nearly all the roots examinedwere infiltrated diffusely by tumour.

I need to comment on the differential diag-nosis of the CSF cytology. Looking back onthis case I think that I should have placedmore reliance on our immunocytochemicaltests. The CSF that came to us with malig-nant cells was assumed to show carcinomacells and when the cells were negative for theepithelial markers it was not checked for lym-phoma with T and B cell markers. If the cellsare clearly malignant cytologically and theyturn out to be B cells this is a strong indicator

of B cell lymphoma in the CSF pathway.Nevertheless, CSF specimens are often lim-ited in volume and we can only make agraded list of what we would choose to dowith the specimen. Carcinoma is morecommon than lymphoma and in the presenceof malignant cells we would certainly go forepithelial markers first.

ConclusionsDr Chris Allen's diagnosis: Prostatic adeno-carcinoma with malignant meningitis.Pathological diagnosis: Prostatic lymphomametastatic to the subarachnoid space andnerve roots and metastatic to the kidney andperiadrenal fat.

CommentLymphoma affecting the prostate gland,either as primary extranodal disease or as sec-ondary spread, is very rare. By 1985 only 95cases had been reported. Of the 13 patientsstudied by Bostwick and Mann (collectedfrom 1935 to 1983), seven had primary extra-nodal disease.9 In none of the 13 patients hadthis spread to the CNS. The subarachnoidspace is the CNS compartment most com-monly involved by non-Hodgkin's lym-phoma. In 60% of cases with subarachnoidspread of the non-Hodgkin's lymphoma thereis focal parenchymal infiltration and morethan 80% show cranial nerve root invasion."'The investigation performed by the new

registrar was a rectal examination, whichshowed a hard craggy prostate gland.

This clinicopathological conference was presented at theEdinburgh advanced clinical neurology course, April 1994.We thank Judi Clarke for her help with transcription of theoriginal.

1 Hammond D, Winnick S. Paraneoplastic erythrocytosisand ectopic erythropoietins. Ann NY Acad Sci 1974;230:219-27.

2 Merritt HH, Fremont-Smith F. The cerebrospinal fluid.Philadelphia: WB Saunders, 1938.

3 Schuchat A, Broome CV. Infections caused by listeriamonocytogenes. In: Isselbacher KJ, Braunwald E,Wilson JD, et al, eds. Harrison's principles of internalmedicine. 13th ed. New York: McGraw-Hill, 1994;63 1-33.

4 Wasserstrom WR, Glass P, Posner JB. Diagnosis andtreatment of leptomeningeal metastases from solidtumors: experience with 90 patients. Cancer 1982;49:759-72.

5 Bunn PA, Ridgway EC. Paraneoplastic syndromes. In:Devita VT, Hellman S, Rosenberg SA, eds. Cancer:principles and practice of oncology 4th ed. Philadelphia: JBLippincott, 1993;2042.

6 Murray JJ, Greco FA, Wolff SN, Hainsworth JD.Neoplastic meningitis: marked variations of cere-brospinal fluid composition in the absence of extraduralblock. AmJ7Med 1983;75:289-94.

7 Rao KC, Knipp HC, Wagner EJ. Computed tomographicfindings in cerebral sinus and venous thrombosis.Radiology 1981;140:391-8.

8 Stauffer MH. Nephrogenic hepatosplenomegaly. Gastro-enterology 1961;40:694.

9 Bostwick DG, Mann RB. Malignant lymphomas involvingthe prostate: a study of 13 cases. Cancer 1985;56:2932-38.

10 Law IP, Dick FR, Blom J, Bergevin PR. Involvementof the central nervous system in non-Hodgkin'slymphoma. Cancer 1975;36:225-31.

378 on June 1, 2020 by guest. P

rotected by copyright.http://jnnp.bm

j.com/

J Neurol N

eurosurg Psychiatry: first published as 10.1136/jnnp.58.3.374 on 1 M

arch 1995. Dow

nloaded from