by guest on june 9, 2013 … · q j med 1997; 90:61-73 cerebral vasculitis—recognition, diagnosis...

$: by guest on June 9, 2013 … · Q J Med 1997; 90:61-73 Cerebral vasculitis—recognition, diagnosis and management N.J. SCOLDING1, D.R.W JAYNE. 2, J.P. ZAJICEK P.A.R\ MEYER. 3, E.P$
Q J Med 1997; 90:61-73

Cerebral vasculitis—recognition, diagnosis and management

N.J. SCOLDING1, D.R.W. JAYNE2, J.P. ZAJICEK\ P.A.R. MEYER3, E.P. WRAIGHTand C M . LOCKWOOD 2

Departments of ^ Neurology,2Medicine,3Ophthalmology, and4Nuclear Medicine,Addenbrooke's Hospital, Hills Road, Cambridge, UK

Received 8 November 7 996

Summary

Cerebral vasculitis is a serious but uncommon condi-tion which presents considerable difficulties inrecognition, diagnosis and treatment. We studiedeight consecutive patients in whom this diagnosiswas made. Despite the great diversity of symptomsand signs, we noted three clinical patterns: (i) acuteor sub-acute encephalopathy, (ii) a picture withsome similarities to multiple sclerosis ('MS-plus'),and (iii) features of a rapidly progressive space-occupying lesion. The identification of these patternsmay help recognition of cerebral vasculitis. Thediagnostic value of four investigative procedures not

previously studied in cerebral vasculitis wasassessed: ophthalmological examination using low-dose fluorescein angiography with slit-lamp videomicroscopy of the anterior segment (abnormal in4/5 patients); spinal fluid oligoclonal band analysis(abnormal in 3/6 patients); anti-neutrophil cyto-plasm ic antibody assay (abnormal in 3/8 patients);and indium-labelled white-cell cerebral imaging(positive in only one patient). Treatment was withsteroid alone (n = 2) or steroid with cyclophospham-ide (n=6) . Seven patients responded clinically.

IntroductionThe vasculitides are a heterogenous group of dis-orders which share certain pathological features, inparticular intramural inflammation and leukocytocl-astic changes within the walls of blood vessels. Theymay be classified according to additional histologicalcharacteristics, including the size of vessel involvedand the presence of granulomata, or according tothe clinical context. Primary vasculitic disordersinclude Wegener's granulomatosis, microscopicpolyangiitis, and temporal arteritis; alternatively, vas-culitis can be secondary to collagen vascular orrheumatological disorders, malignancy (particularlymyeloproliferative disease), drugs, or infections suchas hepatitis B.

Involvement of the central or peripheral nervoussystem can occur in any of the systemic vasculit-ides.1'2 Additionally, primary or isolated vasculitis ofthe CNS or of the PNS is recognized, where little orno inflammation is apparent outside the nervoussystem.3"5 In both isolated and secondary CNS vascu-

litis, the neurological features arising from inflamma-tion and necrosis of the vasculature are similar, andrepresent principally the consequences of infarction.Cerebral vasculitis might therefore be considered asyndrome, of diverse aetiologies but a (largely)common cerebral pathological process. The mani-festations may be devastating and permanent, butCNS vasculitis remains under-recognized and notori-ously difficult to diagnose, further accentuating theclinical problem.

In recent years, there have been a number ofsignificant advances in the diagnostic approach tomultisystem vasculitis. Specifically, testing for anti-nuclear cytoplasmic antibodies, indium-labelledwhite-cell scanning, and ophthalmological exam-ination of ocular vessels have all been shown tooffer substantial diagnostic contributions.6""8 Novelimmunological therapies have also been described,and there has been a significant improvement in theprognosis of systemic vasculitic disorders.7'9"12 It is

Address correspondence to Dr N.J. Scolding, Department of Neurology, Addenbrooke 's Hospital, Hills Road, CambridgeCB2 2QQ

© Oxford University Press 1997

by guest on June 9, 2013http://qjm

ed.oxfordjournals.org/D

ownloaded from

http://qjmed.oxfordjournals.org/

62 N.J. Scolding et al.

therefore timely to re-examine vasculitic disease ofthe nervous system in the light of these recentadvances, and to investigate whether they presentnew opportunities for improving the diagnosis, treat-ment and outlook of these difficult and challengingdisorders.

As part of a structured clinical collaborationinvestigating vasculitis and the nervous system, wepresent an analysis of eight patients with the syn-drome of cerebral vasculitis. We have included aspectrum of aetiologies, ranging from primary CNSvasculitis to intracranial vasculitis secondary to rheum-atoid disease or lymphoma. This represents an attemptto address the practical diagnostic difficulties posedby the patient with suspected CNS vasculitis—wherea history of rheumatoid arthritis (for example) mightincrease suspicion of cerebral vasculitis, but also (withan associated history of immune suppressant treat-ment) raise other diagnostic possibilities, increasingthe importance of diagnostic accuracy. We haveattempted to analyse clinical patterns of disease, tofacilitate recognition of the disorder, and have assessedthe diagnostic value of four clinical or investigationalprocedures not previously assessed in cerebral vascul-itis: ophthalmological examination using low-dosefluorescein angiography with slit-lamp video micro-scopy of the anterior segment; spinal fluid oligoclonalband analysis; anti-neutrophil cytoplasmic antibodyassay; and indium-labelled white-cell cerebralimaging. We have also assessed the therapeuticimplications of the response of our patients to immunesuppressive treatments.

Methods

Patients

Eight patients in whom a diagnosis of cerebralvasculitis had been made were studied. Six werestudied prospectively from the time the diagnosiswas first considered, and their previous coursesretrospectively examined; the admission records ofthe remaining two were reviewed restrospectively.

Anti-neutrophil cytoplasmic antibody(ANCA) assay

All ANCA assays were performed in the same laborat-ory as previously described.13 Briefly, sera weretested by indirect immunofluorescence on alcohol-fixed normal human neutrophils, by acid-extractELISA, and by proteinase 3 and myeloperoxidaseELISAs.

Radioisotope-labelled white-blood-cell scan

Autologous leucocytes were separated by differentialcentrifugation and labelled using 16MBq 111indium

tropolonate as previously described.8'14 Whole-bodyscans were performed at 3 h and 20-24 h afterintravenous injection of labelled leucocytes, to obtainearly and late images; local images of the brain werealso obtained.

Cerebral perfusion scans by emissiontomography (SPECT)

These were performed after injection of 500 mBq"Tc-HMPAO in a quiet room with subdued lighting.

Medical ophthalmological assessment

Patients were examined using conventional slit-lampmicroscopy and fundus fluorescein angiography tostudy both anterior and posterior ocular circulations,supplemented by video slit-lamp microscopy record-ing of erythrocyte flow15 and low-dose anteriorsegment fluorescein angiography, both as previouslydescribed.16

ResultsIllustrative case histories

Patient 7, FT

This previously-well 28-year-old female becamedepressed one week after the uncomplicated deliveryof her second child. She developed headaches, earlymorning vomiting, and became drowsy. A CT brainscan in her local hospital showed a non-enhancingleft thalamic lesion with some space occupation(Figure 1a), felt most likely to be a glioma. She wastransferred to the regional neurosurgery unit whereno focal neurological signs were apparent on admis-sion. A CT-guided stereotactic biopsy was performed.She recovered well from surgery, but two days latersuffered a cardio-respiratory arrest; prompt resuscita-tion successfully re-established a normal cardiacoutput, but she did not recover consciousness andrequired continued mechanical ventilation. Sheremained deeply comatose, unresponsive to pain,with fixed gaze deviation to the right and bilateraldecorticate rigidity.

The biopsy showed clear necrotizing leukocytocl-astic vasculitis, with an infiltrate predominantly ofneutrophils and no granuloma formation (Figure 2).Her ESR was 60 (5 days after surgery), C-reactiveprotein (CRP) 49; anti-nuclear factor, rheumatoidfactor, hepatitis B, cryoglobulins, serum angiotensin-converting enzyme and ANCA assay were normal ornegative. Her CSF contained no cells, protein 0.3g/dl, normal glucose, and no evidence of oligoclonalimmunoglobulin bands. MRI (Figure 1b-e) showedmultiple T2-weighted high-signal-intensity lesions inthe periventricular areas, with larger areas in the left



ownloaded from


Cerebral vasculitis 63

Figure 1. Cerebral imaging, patient 1. a CT brain scan. A non-enhancing lesion is shown in the left thalamus, exhibitingslight mass effect. MRI (b-e) showed multiple T2-weighted high-signal-intensity lesions in the periventricular areas, withlarger areas in the left thalamus, the heads of both caudate nuclei, and the left occipital cortex. Some lesions showed highsignal on T1 -weighted images (d,e), suggesting a haemorrhagic element to focal ischaemic areas.

thalamus, the heads of both caudate nuclei, and theleft occipital cortex. Some lesions showed high signalon T1-weighted images, suggesting a haemorrhagicelement to focal ischaemic areas.

She was treated with intravenous cyclophospham-ide, 2 mg/kg/day, commencing 6 days after biopsy,and intravenous methylprednisolone, 500 mg dailyfor 5 days, then 60 mg prednisolone orally per day.She remained comatose for 9 days, but thereaftersteadily recovered, breathing independently 16 daysafter starting treatment, at which stage she had aright hemiparesis with bilateral extensor plantarresponses. Her improvement continued and she wasdischarged home 4 weeks later. When examined at6 months, she was asymptomatic, with no residualneurological signs on a tapering dose of cyclophos-phamide and prednisolone.

Patient 2, JM

This 61-year-old retired, previously-well female pre-sented with 2 weeks of increasing confusion and

forgetful ness, and suffered two generalized tonic-clonic convulsions. No abnormalities were apparenton general physical examination. She was disorient-ated, with poor and fluctuating registration, immedi-ate recall, and episodic memory. She exhibitedchoreiform movements of the limbs, predominantlydistal and symmetrical, together with facial grim-acing, but no other focal neurological signs.Psychiatric assessment suggested periodical hallu-cinations without delusions, paranoid features, occa-sional Ganserian responses, and crocydidmos.

Investigations included normal blood count andfilm, ESR, CRP, liver function, urea and electrolytesand thyroid function, and negative rheumatoid factor,anti-nuclear factor and thyroid antibody. A CT headscan showed only minimal atrophic changes, whileMR scanning was marred by gross movementartefact. CSF: normal protein and glucose, no organ-isms seen or cultured, 4 neutrophils and 2 lympho-cytes were present (per fil). ANCA serology waspositive at 22% on RIA, with competitive inhibition



ownloaded from


64 N.J. Scolding etal.

Figure 2. Cerebral biopsy, patient 1. Clear changes of necrotising leukocytoclastic vasculitis are present, with an intensecellular infiltrate composed predominantly of neutrophils.

revealing specific antibody. Indium-labelled white-cell scan showed patchy uptake in both cerebralhemispheres, more marked in the left (Figure 3), withadditional more marked uptake in the lungs. Cerebralperfusion SPECT scanning showed an isolated areaof reduced perfusion in the left parietal region.Bilateral carotid contrast arteriography showed vari-ation in arterial calibre on both sides, particularlyaffecting the distal vessels.

A diagnosis of isolated cerebral vasculitis wasmade and she was treated with oral prednisolone(60 mg daily). She made a good if slow recovery,ANCA serology becoming negative within 6 weeks,but was re-admitted 2 months later (on 30 mgprednisolone daily) with recurring confusion, albeitless severe. She was treated with 150mg daily oforal cyclophosphamide, and again recovered well.When seen 3 years later (6 months after changingcyclophosphamide to azathioprine) she was well,with no chorea, neurological signs or psychiatricabnormalities, although cognitive assessmentrevealed mild residual deficits of frontal and memoryfunctions.

Patient 3, MF

This previously-well female had episodes of right-sided twitching starting in 1988 when aged 44. From1990, she felt her memory steadily deteriorated. Inearly 1992 she developed nocturnal convulsions,and in March 1993, she had a short episode of statusepilepticus. CT head scan (normal in 1988) showeda low-density area in the left temporal area. In July

1993, she developed acute visual loss in her left eyeaccompanied by headache. Visual acuity wasreduced to 6/60, with an afferent pupillary defectand a pale optic disc. She had bilateral Babinskiresponses but no other neurological signs. ESR andtemporal artery biopsy were normal. Other serolo-gical tests, CRP, coagulation screen, cardiolipin anti-body, echocardiography, ACE, and ANCA tests werenormal. Her headache, but not her vision, improvedwith 80 mg prednisolone daily.

Four months later she was re-admitted in coma.Her consciousness level fluctuated over the next 3days but then improved and stabilized; at this stage,she was amnesic, with reduced registration. She hada pout reflex, an exaggerated jaw jerk, a right spasticparetic arm (MRC grade 4) with hyper-reflexia,bilateral flexor plantar responses and a non-specifically unsteady gait. Re-investigation revealedan ESR of 29 (repeated at 30), with a CRP of 68.CSF was acellular, but contained oligoclonalimmunoglobulin bands: identical serum bands werepresent. CT head scan now showed non-enhancinglow attenuation areas in the left frontal and bothoccipital lobes, and perfusion HMPAO SPECT scan-ning showed multiple focal ischaemic defects,affecting the left posterior parietal region, the leftfronto-parietal, and the right temporal areas(Figure 4). Bilateral carotid angiography was normal.An open cerebral biopsy of the left frontal lobe(abnormal on CT), which included meninges, whitematter and cortex, showed multiple small ischaemicareas, with nerve cell loss, myelin pallor, some



ownloaded from



Figure 3. Indium-labelled white-cell scan, patient 2.Patchy uptake is apparent in both cerebral hemispheres,more marked in the left. Additional pulmonary uptakewas also apparent and was found also in patient 7, whohad normal brain images.

macrophages but no vasculitic changes. Fluoresceinangiography, however, showed clear changes ofretinal vasculitis.

She was treated with intravenous cyclophospham-ide (four 750 mg doses over 6 weeks), followed byoral steroids. While symptomatically well thereafter,she had a bilateral inferior altitudinal hemianopiawhen examined 4 months later.

Clinical features

Details of all patients are summarized in Table 1. Asexpected, a wide variety of clinical features wereencountered. Three patients (3, 5 and 7) had relaps-ing and spontaneously remitting disease character-ized by optic neuropathy (3 episodes), brain stemevents (2), acute or sub-acute encephalopathic epis-odes (3), generalized or partial seizures (2 patients),and hemispheric stroke-like episodes (one patient,three events). Two such patients had additionalprogressive cognitive or neuropsychological dis-turbance.

The remaining five patients presented with acuteor subacute neurological encephalopathic illnesses;two had raised intracranial pressure. Three experi-enced seizures, two had accompanying focal neuro-logical signs, and one had chorea.

Three patients had an associated systemic disease:two had previously-established rheumatoid disease(without active arthritis at the time of neurologicalpresentation), and one had cerebral and systemicvasculitis in the context of lymphoma. The remainingfive patients had isolated CNS vasculitis (though onehad livedo reticulares).

Investigation and diagnosis

An elevated ESR or CRP accompanied neurologicalsymptoms in 5/8 patients; during two episodes (inone patient), dissociation was found, with a raisedCRP in the presence of a normal ESR. Conventionalautoimmune serology was abnormal in three, whileANCA serology was positive in two (patients 2 and8), with an additional false-positive result in onepatient with rheumatoid arthritis. Routine spinal-fluidexamination was abnormal in only 2/8, in one ofwhom (patient 2) the abnormality was very marginal(6 white cells). CSF oligoclonal bands were presentin 3/6 patients, in one (patient 7) showing a shiftingpattern during the protracted course.

Ophthalmological examination made an importantcontribution to the diagnosis of cerebral vasculitis(Table 2). Vasculitic changes were found in 4/5patients studied (patient 8 illustrated, Figure 5).Abnormalities revealed by dynamic video recordingvariably included marked slowing of flow, multifocalattenuation of arterioles, and erythrocyte aggregates.Fluorescein studies confirmed the variation in vesselcalibre, slowing of flow and red-cell aggregation,and also demonstrated areas of small-vessel infarc-tion, together with multifocal segments of intenseleakage from post-capillary and collecting venules.

CT scanning was abnormal in 7/8 patients, MRIin all four patients imaged. Cerebral perfusion SPECTwas abnormal in 2/3 patients, while labelled white-cell scanning indicated cerebral inflammation in 1/2patients—although in the patient with normal brainappearances, increased uptake in other sites wasfound and contributed usefully to diagnosis.

Cerebral biopsy was carried out in three patients;all were abnormal though only two showed unequi-vocal vasculitis, the third (patient 3), showed multiplesmall infarcts but no active vasculitis.

DiscussionWe have described eight patients with vasculitis ofthe central nervous system—in five confined to the



ownloaded from



B

B

Figure 4. Perfusion HMPAO SPECT scanning, patient 3. Multiple focal ischaemic defects are visible; these involved the leftposterior parietal region, the left fronto-parietal, and the right temporal areas.

brain, the remainder proving to have CNS vasculitiswith laboratory evidence or past history of systemicdisease. Of the latter three, two (patients 5 and 8)had established seropositive rheumatoid disease, ofwhich cerebral vasculitis is a rare but well-reportedcomplication.2'17 The third (patient 7) exhibited araised lymphocyte count at the onset of his illnesswhich eventually proved a consequence of alow-grade B-cell lymphoma; his illness clinicallyresembled lymphomatoid granulomatosis which notuncommonly transforms to lymphoma18, a recog-nized cause of systemic and cerebral vasculitis.

We elected to describe all eight patients togetherfor three reasons. Firstly, in all patients inflammatorydisease of the brain was overwhelmingly the prin-cipal cause of morbidity at the time of presentation,regardless of the underlying cause. Secondly, it seemslikely that all share similar cerebral processes.Thirdly, the investigation and management of allpatients was largely independent of any systemicprocess—which was either clinically quiescent atthe time of neurological presentation, or diagnosedonly in the course of investigation for suspectedcerebral vasculitis. As mentioned above, the clinicaldiagnostic problems presented by all eight patients

were similar, regardless of their underlying medicalbackground.

Cerebral vasculitis is unusual—in our unitaccounting for a maximum of 0.5% admissions, nomore than 3-4 patients per year (with a neurologicalcatchment area of approximately 2.4 million). It isdifficult to recognize, to diagnose, and to treat, butreports of successful therapies for other inflammatoryneurological diseases (j5-interferon and monoclonalantibodies for multiple sclerosis, plasmapheresis andintravenous immunoglobulin for inflammatory neuro-pathies and myopathies19"21 and for multi-systemvasculitis9'12) provide new hope for the treatment ofcerebral vasculitis, placing new emphasis on recogni-tion and treatment.

Our patients reflect the previously emphasizedwide variation in manifestations, course and severity,and the absence of a pathognomic or even typicalclinical picture.1'2'5'22 Focal and generalized seizures,stroke-like episodes, acute and sub-acute encephalo-pathies, brain-stem events, progressive cognitivechanges, chorea, optic and other cranial neuropathiesall were seen. Despite this variability, it is possible(and may be useful) to divide our patients into threebroad clinical groups.



ownloaded from


Tabl

e 1

Sum

mar

y of

the

clin

ical

fea

ture

s an

d in

vest

igat

ion

resu

lts i

n al

l ei

ght

patie

nts

Pat

ient

Clin

ical

fea

ture

sO

cula

rfin

ding

sIn

flam

mat

ory

Rou

tine

AN

CA

indi

ces

sero

logy

te

stin

g

ESR

C

RP

Spi

nal

fluid

N,

ogc

-ve

4n,2

l og

c -v

e

No

cells

ogc

+v

ese

rum

-l-v

e

N,

ogc-

ve

Imag

ing

CT

—th

alam

icle

sion

MR

I—m

ultip

lele

sion

sca

udat

e,th

alam

i,pe

riven

tric

ular

area

CT

—N

(sl

ight

prom

inen

ce o

fve

ntric

les)

wcs

—pa

tchy

cere

bral

upta

keSP

ECT—

sing

leis

chae

mic

are

aan

giog

raph

y—va

riatio

n of

vess

el c

alib

reC

T—

low

dens

ityte

mpo

ral

lobe

SP

EC

T—m

ultip

lepe

rfusi

onde

fect

san

giog

raph

y—no

rmal

CT— pe

riven

tric

ular

low

den

sitie

sS

PE

CT—

norm

al

His

topa

thol

ogy

Tha

lam

ic b

x:ne

crot

izin

gle

ukoc

ytoc

last

icva

scul

itis

ND

tem

pora

l ar

tery

bx.—

Nce

rebr

al b

x(m

enin

ges

corte

x &

whi

tem

atte

r):

mul

tiple

sm

all

infa

rcts

; no

visi

ble

vasc

uliti

sN

one

Dia

gnos

isba

sed

on

Tha

lam

ic b

iops

y

AN

CA

, w

hite

-ce

ll sc

an a

ndca

rotid

angi

ogra

phy

ESR

/CR

P; C

SFel

ectro

phor

esis

;SP

ECT;

cere

bral

biop

sy a

ndFF

A

Past

his

tory

,cu

rren

trh

eum

atoi

d &

card

iolip

inse

rolo

gy, E

SR&

CR

P, C

Tsc

an

8- si i" </> oc Si"

Pat

ient

1,

PT (F

)do

b 7.

4.66

Sub

acut

e is

olat

edC

NS

dis

ease

Pat

ient

2,

JM (F

)do

b 27

.4.4

0

Sub

acut

e is

olat

edC

NS

dis

ease

Pat

ient

3,

MF

(F)

dob

4.12

.44

Chr

onic

iso

late

dre

laps

ing

CN

Sdi

seas

e

Sub

acut

e po

st-p

artu

m

ND

head

ache

,vo

miti

ng,

drow

sine

ss;

card

iore

spira

tory

arre

st a

ndpe

rsis

tent

com

aaf

ter

biop

sy;

nosy

stem

ic f

eatu

res.

Sub

-acu

te c

onfu

sion

N

Dw

ith c

hore

iform

mov

emen

ts a

ndge

nera

lized

fits

6049

N

N

N

NN

+ v

e

5 ye

ar e

volu

tion,

star

ting

with

unila

tera

l tw

itchi

ngno

ctur

nal

fits.

Cog

nitiv

esy

mpt

oms.

Opt

icne

urop

athy

. A

cute

ence

phal

opat

hy.

FFA:

ret

inal

vasc

uliti

s30

68

N

Pat

ient

4,

VS

(F)

A

cute

psy

chos

is,

N78

90

dob

25.4

.35

Acu

te C

NS

dis

ease

in t

he c

onte

xt o

frh

eum

atoi

dar

thrit

is

follo

wed

by

conf

usio

nal

stat

e +

cog

nitiv

efa

ilure

. E

xten

sor

/.pl

anta

r. P

/Hrh

eum

atoi

d di

seas

e

RhF

+ v

e12

30lu

/ml

AC

A +

ve

(lgM

-68)

False

+

ve



ownloaded from


Tabl

e 1

(con

tinue

d)

Pat

ient

Pat

ient

5,

KR (

M)

dob

3.12

.46

Sub

acut

e re

laps

ing

CN

S d

isea

se,

lived

o re

ticul

ares

.

Pat

ient

6,

SP (M

)do

b 29

.8.5

8

Sub

acut

e is

olat

edC

NS

dis

ease

Clin

ical

fea

ture

s O

cula

rfin

ding

s

Acu

te j

aw,

toot

h &

S

ugge

sts

scal

p pa

in,

then

/.

vasc

uliti

svi

sual

los

s,fo

llow

ed b

y r.:

then

tra

nsie

nt /

.he

mis

pher

eep

isod

es.

Live

dore

ticul

ares

.C

ogni

tive

chan

ges

wit

h im

paire

dfr

onta

l fu

nctio

n.S

ubac

ute

head

ache

, N

Dvo

miti

ng,

papi

lloed

ema,

atax

ia

Infla

mm

ator

y R

outin

e A

NC

Ain

dice

s se

rolo

gy

test

ing

ESR

C

RP

14

<6

AC

A +

ve,

N

(lgM

-69)

1 <

6 N

N

Spi

nal

fluid

N 38L,

1N

prot

ein

0.3

pres

sure

—80

cm

wat

er

Imag

ing

CT— pe

riven

tricu

lar

low

den

sitie

s,pr

obab

lein

farc

ts,

gene

ralis

edat

roph

y

CT—

cere

bella

rm

ass

lesi

onw

ithhy

droc

epha

lus

MR

I—fo

cal

high

sign

al r

.ce

rebe

llar

hem

isph

ere

+up

per

pons

,m

id-b

rain

,ex

tern

alca

psul

e an

d r.

puta

men

,pe

riven

tricu

lar

area

s an

dce

ntru

mse

mio

vale

His

topa

thol

ogy

TA

—no

rmal

Live

r—m

ildci

rrho

ticch

ange

Ste

reot

actic

bx:

-gr

anul

omat

ous

vasc

uliti

s w

ithep

ithel

ioid

and

gian

t ce

lls,

lym

phoc

ytic

cuffi

ng,

reac

tive

chan

ges,

swol

len

astro

cyte

sK

veim

-ve

Dia

gnos

isba

sed

on

AC

A s

erol

ogy,

CT

sca

n an

doc

ular

fin

ding

s

CT,

MR

I an

dbi

opsy

Hig

h do

sest

eroi

ds a

ndcy

clos

porin

A.

Rel

apsi

ngco

urse

, di

edw

ithhy

droc

epha

lus

and

loca

lized

haem

orrh

agic

chan

ges

(on

CT)

^. •>- 8 o 5; <§'

2- r-



ownloaded from


Pat

ient

7,

JW (

M)

dob

11.1

0.34

Rel

apsi

ng

CN

S a

ndsy

stem

ic d

isea

se

Pat

ient

8,

LH (

M)

dob

4.9.

29

Sub

acut

e C

NS

dise

ase

in t

heco

ntex

t of

rheu

mat

oid

dise

ase

1986

: w

t lo

ss,

coug

h,he

adac

he;

pulm

onar

yin

filtr

ate,

rai

sed

Sug

gest

sly

mph

ocyt

e co

unt

arte

ritis

1992

: di

arrh

oea,

nigh

t sw

eats

, w

tlo

ss;

follo

wed

by

acut

e fe

brile

ence

phal

opat

hy19

93:

feve

r, n

ight

swea

ts,

abdo

. pa

inhe

adac

hes,

bra

inst

em e

piso

de19

94:

recu

rren

t br

ain

stem

epi

sode

with

/. V

I pa

lsy

Sub

acut

e flu

ctua

ting

Cle

ar v

ascu

litis

ence

phal

opat

hy,

chan

ges

head

ache

, pa

infu

lnu

mb

extr

emiti

es,

feve

r, n

ight

sw

eats

;dy

sart

hria

,pe

riphe

ral

sens

ory

loss

, tra

nsie

ntpu

ncta

te m

acul

arra

sh l

ower

leg

s.P

ast

treat

men

t fo

r R

Ain

clud

edC

ampa

th-1

H,

gold

,az

athi

oprin

e,cy

clos

porin

e

616 4

94 70

ND 27

39

191

23

9

N

NN

N

N

N

RhF

+2

39

, Tr

ue

+ve

,26

9 iU

/ml

ag n

otde

fined

ND

ogc

+ve

, 1

band

ser

umog

c +

ve,

seru

m+

ve

(diff

band

s)og

c +

ve,

seru

m-v

e

N o

cg

+v

ese

rum

+

ve

CT

—sm

all

infa

rct,

r.co

rona

irrad

iata

CT—

NM

RI—

mul

tiple

high

sig

nal

lesi

ons

WC

S—

incr

ease

dup

take

lun

gsan

d ili

ac f

ossa

,no

rmal

bra

inA

ngio

grap

hy—

norm

alM

RI—

high

sign

alpe

riven

tric

ular

area

, in

tern

alca

psul

e, s

ub-

cort

ical

are

a,po

ns,

/. gl

obus

palli

dus

CT—

exte

nsiv

elo

wat

tenu

atio

nce

rebr

al w

hite

mat

ter

MR

I—di

ffuse

high

sig

nal

inte

nsity

cent

rum

sem

i-ov

ale,

post

erio

rfr

onta

l lo

be,

exte

rnal

caps

ules

,pe

riven

tric

ular

area

s

lung

bx.

—in

flam

mat

ory

cells

rect

al b

x.—

arte

ritis

with

fibri

noid

necr

osis

Kve

im—

norm

albo

ne m

arro

w—

low

gra

deB

-cel

lly

mph

oma

Ren

al B

x—N

(tran

sien

tim

paire

d re

nal

func

tion)

Rec

tal

biop

sy,

MR

I, oc

ular

findi

ngs

Sys

tem

icsy

mpt

oms,

activ

erh

eum

atoi

dse

rolo

gy,

ESR

,C

RP

, A

NC

A,

ocul

arfin

ding

s.G

ood

resp

onse

to cycl

opho

spha

mid

e

? 8.>ral 2 In O C;

in"

dob,

dat

e of

birt

h; N

D,

not

done

; N

, no

rmal

; og

c, o

ligoc

lona

l ba

nd a

naly

sis;

/,

left;

r,

right

; w

cs,

whi

te-c

ell

scan

; bx

, bi

opsy

; FF

A, fu

ndus

flu

ores

cein

ang

iogr

aphy

;n,

neu

trop

hil;

I, ly

mph

ocyt

e; R

hF,

rheu

mat

oid

fact

or;

AC

A,

antic

ardi

olip

in a

ntib

ody.



ownloaded from


70

Table 2

Patient

3

4

5

7

8

Details of ocular

Visual activity

r1/60/ 1/60

r6/9/ 6/18

r6/5/6/5

r6/6/6/6

Not reliable

N.J. Scolding etal.

findings in the five cases fully examined

Examination

Anterior segment

Posterior segment

FFAConjunctival vesselsPosterior segments

Fields

Conjunctival vessels

Anterior and posteriorsegments

Conjunctival vesselsAnterior segmentFundus

FFAOcular examinationConjunctival fluorescein

videoangiogram

Results

IOP 31/30

Pale discs; arterial attenuation,venous beading, old venoussheathing r

No active vasculitisNo rbc aggregation/-focal arteriolar narrowing &

dry macular degenerationr-homonymous inferior

quadrantanopiaMultifocal venular dilation

and stasisNormal

NormalNormalr-cotton wool spot (nerve fibre

layer infarct)No other vascular abnormalityNormalMultifocal microvascular non-

perfusion and leakage

Ocular diagnosis

Steroid-induced ocularhypertension

Large vessel ischaemia inophthalmic artery territories

Previous clear retinal vasculitisNo vasculitis

/-posterior parietal infarct

Consistent with vasculitis

Compatible with systemicvasculitis

Characteristic of systemicvasculitis

(i) Atypical multiple sclerosis (more accurately'MS-plus'). Three patients (patients 3, 5 and 7)exhibited a spontaneously relapsing and remittingcourse, with neurological features which includedoptic neuropathy and brain stem episodes. Two hadCSF oligoclonal bands, one (of two scanned) hadmultifocal white-matter lesions on MRI; multiplesclerosis had been considered a possible diagnosisin all at some time. Each, however, had additionalfeatures less common in multiple sclerosis: seizures(1/3), severe headaches (2/3), encephalopathic epis-odes (2/3), or hemispheric stroke-like events (1/3).Systemic features were found in 2/3, including livedoreticulares, fever, or oligoarthropathy.

(ii) Intracranial mass lesion. This was found intwo patients on 'first pass' investigation (CT).

(Hi) Acute or sub-acute encephalopathy the pre-sentation in the remaining three patients.

The great majority of cases detailed in the literaturealso conform to these patterns, which we do notsuggest carry pathological or therapeutic implica-tions. Their occurrence might, however, lead to asuspicion of cerebral vasculitis, so they may improverecognition of this condition.

ANCA assays are now routinely used in theserological diagnosis of systemic vasculitis; theseantibodies may also have a pathogenic role.23 Anumber of case reports indicate the expected positiveANCA testing in patients with Wegener's granulom-atosis complicated by direct cerebral spread,24'25 but

we have been unable to identify any previousinvestigation of the role of ANCA serology in thediagnosis of cerebral vasculitis. We tested all eightpatients. Two had true-positive ANCA testing; in one(patient 2), the result was particularly useful sinceshe had isolated CNS disease and (otherwise) normalblood tests. We also performed ANCA-testing on oneCSF sample (patient 8, with a positive serum test).No antibody was detectable, concurring with ourexperience of CSF ANCA testing in cerebral diseasefrom ANCA-positive Wegener's granulomatosis(unpublished observations).

Spinal fluid was examined in all eight patients.Only one had a substantially elevated cell count(patient 6; 38 lymphocytes); Patient 2 had a verymarginally abnormal count (6 white cells). None hada raised total protein level. Previous published seriesreport abnormal results in 50-80% of cases,2'3'26"28

but comparisons are difficult, since the definition ofan abnormal result varies considerably (from > 1 , to> 4 cells per high-power field3).

There have been no published systematic studiesof CSF oligoclonal immunoglobulin band analysis incerebral vasculitis, although 2/4 cases of herpes-zoster-related cerebral vasculitis were positive,29 aswere '25-66%' of cases with various manifestationsof cerebral lupus.30 3/6 of our patients showedabnormalities. Patient 3 had different immunoglob-ulin bands in CSF and in serum, patient 8 hadidentical bands. In patient 7, a fluctuating pattern



ownloaded from



Figure 5. Anterior ocular vascular study, patient 8. Stillframes from slit-lamp video recording are shown.Abnormalities included multifocal attenuation of arterioles,and erythrocyte aggregates (a) Low-dose fluorescein stud-ies (b) confirmed the variation in vessel calibre, and alsodemonstrated areas of small-vessel infarction, togetherwith multifocal segments of intense leakage from post-capillary and collecting venules.

was found on serial testing over a 2-year period (seeTable 1). We suggest that oligoclonal band analysisis therefore worthwhile in suspected cerebral vascul-itis, an abnormal result (and perhaps particularly ofvariable pattern, or indicating intrathecal and sys-temic immunoglobulin synthesis) providing somediagnostic support, albeit without specificity.

MRI abnormalities were present in 4/5 of ourpatients. In the fifth, the images were not interpretabledue to movement. MRI has been suggested as asensitive but not specific screening test in cerebralvasculitis: Harris ef a/.31 reported significant abnor-malities in 9/9 cases of proven disease, and no false-negative scans. However, in an interesting correlativestudy, Greenan ef a/.32 found on careful regionalanalysis that 12/33 vascular territories with angio-graphic vasculitis exhibited no lesions on MR. Up to25% of cases may in fact have normal MRI scans,28'33

some additionally with normal angiography.34

The value of angiography is difficult to assess,many published series relying on this investigationfor diagnostic confirmation.26'28 Studies dependingon pathological examination indicate a false-negative

rate for angiography of 30-45%,3'27 although otherseries suggest angiography is diagnostically useful inonly 20-27% of cases.35'36 A 10% risk of transientneurological deficit is reported,37 with permanentdeficit in 1%. Only 1/4 of our patients had abnormalangiography, and although the value of this investi-gation may have been over-emphasized in the past,it clearly remains important to exclude atheromatousand other disease, and when positive, affords morespecificity than MRI.

We have found no reference to labelled-white-cell nuclear scanning in cerebral vasculitis, notwith-standing its increasingly recognized value in systemicvasculitides.8 We found abnormal cerebral accumu-lation of indium-labelled leucocytes in 1/2 patientstested, although the (clinically silent) increased pul-monary uptake in both was also diagnostically useful.Vasculitis in other organs is demonstrated indirectlyby leucocyte uptake in necrotic tissue, or ischaemicareas supplied by vasculitic vessels,14 whereas leuco-cyte infiltration of intracerebral vessels is unlikely tobe visualized directly. Nevertheless, further studiesmay be worthwhile. Functional cerebral imaging,including SPECT, was abnormal in 8/12 patients ina recent series,36 and in 1/2 patients we examined.Cerebral perfusion SPECT may well therefore be auseful if non-specific test in cerebral vasculitis,demonstrating focal ischaemia secondary to thevasculitic process.

Most helpfully, we found ophthalmological exam-ination with video microscopy and low-dose fluore-secin angiography of the anterior chamber to beextremely valuable in the assessment of our patients.Four of five patients examined had abnormal findingssuggestive of vasculitis. Studies of the ocular vas-culature have been shown to contribute to thediagnosis of multi-system vasculitis,6 and retinalvasculitis has been much studied in multiple sclerosis(where vasculitis generally refers to vascular leakageand occlusion without implying a histopathology ofleukocytoclastic vascular damage). We have, how-ever, been unable to find any reference in theliterature to the diagnostic contribution of theseophthalmological techniques to cerebral vasculitis.Previous studies have confirmed the clinicopatholog-ical correlation between fluorescein angiographicchanges and (localized ocular) vasculitis,38'39 andour findings suggest that ophthalmological assess-ment is a useful and important addition to theinvestigation of patients with suspected cerebralvasculitis.

Our study might be criticized for lacking patholo-gical data. This issue is, however, far from straightfor-ward. In two of the largest series surveying previouslypublished cases,3'27 biopsies had been performed inonly 29% (14/48) and 52% (37/71) of cases, andwere diagnostic only in 10 and 26 patients, respect-



ownloaded from



ively. Three of our eight patients underwent cerebralbiopsy: two in the course of urgent assessment foracute space-occupying lesions, biopsy revealingunsuspected vasculitis, the third undergoing 'blind'biopsy of cortex, white matter and meninges forsuspected vasculitis; in this patient, the findings werenon-specific and not diagnostic. Our results aretherefore numerically typical and representative ofpublished reports and series.

Cerebral biopsy is not a trivial procedure, carryinga significant risk of serious morbidity estimated at0.5-2%40, and in the majority of our patients wefound it impossible to justify. Patient 8 is illustrative:findings included strongly positive rheumatoid sero-logy, positive ANCA serology, marked elevations ofboth ESR and CRP, and unequivocal evidence ofsmall-vessel vasculitis in the anterior ocular circula-tion, with no other explanation for his subacuteencephalopathy after extensive investigation. We feltunable to build a compelling case for biopsy, particu-larly since, with our own experience of one non-diagnostic biopsy and published data consistentlyshowing the diagnostic yield of biopsy to be relativelylow at 70%,3'27 the decision to treat with cyclophos-phamide would not have been influenced by a non-diagnostic biopsy. A pronounced response to treat-ment in this patient provided further retrospectivesupport for the diagnosis.

The low incidence of vasculitic cerebral diseaserenders formal prospective therapeutic trialsextremely difficult: none has so far been reported.An informed approach to treatment therefore dependson the cumulative experience described in publishedretrospective and pooled series, and our patientsmay offer some useful lessons. Patients 1 and 6 hadvery similar features: biopsy-proven isolated cerebralvasculitis presenting with single cerebral mass lesionson CT but MRI evidence of multifocal disease, bothwith negative routine and ANCA serology. One wastreated with cyclophosphamide and steroids andmade a complete recovery with no further symptomssome 6 months later; the other received high-dosesteroids with cyclosporin and suffered a chronicrelapsing and ultimately fatal illness. Four otherpatients received cyclophosphamide. One had asingle further episode of optic neuropathy (whilst onoral steroid maintenance treatment alone, 3 monthsafter four pulses of intravenous cyclophosphamide)and remained well thereafter, and three others experi-enced a striking resolution of their illness. The fourthpatient, who had B-cell lymphoma, also died (ofpseudomonas pneumonia).

Two patients received steroids alone: one (patient4) with past rheumatoid disease, and one (patient5) whose initial presentation had suggested giantcell arteritis. Both responded well. Interestingly,Calabrese ef al. have defined a group of patients

with cerebral vasculitis who have more benigndisease which may, in fact, not require any treat-ment,41 a suggestion supported by others.26 In manycases of cerebral vasculitis, however, relapses mayoccur when steroids alone are used35 and we wouldadvocate early recourse to cyclophosphamide inpatients not responding rapidly to high dose intraven-ous steroids—as previously recommended in bothcerebral and systemic vasculitis.5'22'43 We have notas yet treated patients with cerebral vasculitis withany of the more experimental approaches, althoughthe promise of Campath-1 H humanized monoclonalantibody treatment in inflammatory demyelination20

and in systemic vasculitis12 may indicate significantpotential, cerebral vasculitis unresponsive to cyclo-phosphamide being well-described.26

References1. Moore PM, Calabrese LH. Neurologic manifestations of

systemic vasculitides. Semin Neurol 1994; 14:300-6.

2. Sigal LH. The neurologic presentation of vasculitic andrheumatologic syndromes. A review. Medicine 1987;66:157-80.

3. Calabrese LH, Mallek JA. Primary angiitis of the centralnervous system. Report of 8 new cases, review of theliterature, and proposal for diagnostic criteria. Medicine1988; 67:20-39.

4. Dyck PJ, Benstead TJ, Conn DL, et al. Nonsystemicvasculitic neuropathy. Brain 1987; 110:843-54.

5. Moore PM. Vasculitis of the central nervous system. SeminNeurol 1994; 14:307-12.

6. Charles SJ, Meyer PAR, Watson PC. Diagnosis andmanagement of systemic Wegener's granulomatosispresenting with anterior ocular inflammatory disease. BrJ Ophthalmol 1991; 75:201 -7.

7. Jennette JC, Falk RJ, Andrassy K, ef al. Nomenclature ofsystemic vasculitides: Proposal of an internationalconsensus conference. Arthritis Rheum 1994; 37:187-92.

8. Reuter H, Wraight EP, Qasim FJ, Lockwood CM.Management of systemic vasculitis: Contribution ofscintigraphic imaging to evaluation of disease activity andclassification. QJM 1995; 88:509-16.

9. Jayne DRW, Davies MJ, Fox CJV, et al. Treatment ofsystemic vasculitis with pooled intravenousimmunoglobulin. Lancet 1991; 337:1137-9.

10. Jennette JC, Falk RJ. Diagnostic classification ofantineutrophil cytoplasmic autoantibody- associatedvasculitides. Am J Kidney Dis 1991; 18:184-7.

11. Lai KN, Jayne DRW, Brownlee A, Lockwood CM. Thespecificity of anti-neutrophil cytoplasm autoantibodies insystemic vasculitis. Clin Exp Immunol 1990; 82:233-7.

12. Mathieson PW, Cobbold SP, Hale G, et al. Monoclonal-antibody therapy in systemic vasculitis. N EnglJ Med 1990;323:250-4.

13. Hagen EC, Andrassy K, Chernok E, ef al. The value ofindirect immunofluorescence and solid phase techniques forANCA detection. J Immunological Methods 1993;159:1-16.



ownloaded from



14. Fink AM, Miles KA, Wraight EP. lndium-111 labelledleucocyte uptake in aortitis. Clin Radiol 1994; 49:863-6.

15. Meyer PAR. Patterns of blood flow in episcleral vesselsstudied by low dose fluorescein videoangiography. Eye1988; 2:533-46.

16. Meyer PAR, Watson PC. Low dose fluorescein angiographyof the conjunctiva and episclera. BrJ Ophthalmol 1987;71:2-10.

17. Ramos M, Mandybur Tl. Cerebral vasculitis in rheumatoidarthritis. Arch Neurol 1975; 32:271-5.

18. Katzenstein A, Carrington CB, Liebow AA. Lymphomatoidgranulomatosis; A clinicopathological study of 152 cases.Cancer 1979; 43:360-73.

19. Jacobs LD, Cookfair DL, Rudick RA, etal. Intramuscularinterferon beta-1a for disease progression in relapsingmultiple sclerosis. Ann Neurol 1996; 39:285-94.

20. Moreau T, Thorpe J, Miller D, etal. Preliminary evidencefrom magnetic resonance imaging for reduction in diseaseactivity after lymphocyte depletion in multiple sclerosis.Lancet-\ 994; 344:298-301.

21. Thornton CA, Criggs RC. Plasma exchange and intravenousimmunoglobulin treatment of neuromuscular disease. AnnNeurol 1994; 35:260-8.

22. Calabrese LH, Duna GF. Evaluation and treatment of centralnervous system vasculitis. Curr Opin Rheumatol 1995;7:37-44.

23. Kallenberg CCM, Brouwer E, Weening JJ, Cohen TervaertJW. Anti-neutrophil cytoplasmic antibodies: Currentdiagnostic and pathophysiological potential. Kidney Int1994; 46:1-15.

24. Nishino H, Rubino FA, Parisi JE. The spectrum of neurologicinvolvement in Wegener's granulomatosis. Neurology 1993;43:1334-7.

25. Weinberger LM, Cohen ML, Remler BF, etal. IntracranialWegener's granulomatosis. Neurology 1993; 43:1831 -4 .

26. Abu Shakra M, Khraishi M, Grosman H, etal. Primaryangiitis of the CNS diagnosed by angiography. QJM 1994;87:351-8.

27. Hankey G. Isolated angiitis/angiopathy of the CNS.Prospective diagnostic and therapeutic experience.Cerebrovasc Dis 1991; 1:2-15.

28. StoneJH, Pomper MG, Roubenoff R, etal. Sensitivities ofnoninvasive tests for central nervous system vasculitis: Acomparison of lumbar puncture, computed tomography,

and magnetic resonance imaging. J Rheumatol 1994;21:1277-82.

29. Hilt DC, Buchholz D, Krumholz A, etal. Herpes zosterophthalmicus and delayed contralateral hemiparesis causedby cerebral angiitis: Diagnosis and managementapproaches. Ann Neurol 1983; 14:543-53.

30. Van Dam AP. Diagnosis and pathogenesis of CNS lupus.Rheumatol Int 1991; 11:1-11.

31. Harris KG, Tran DD, Sickels WJ, etal. Diagnosingintracranial vasculitis: The roles of MR and angiography.Am) Neuroradioh994; 15:317-30.

32. Greenan TJ, Grossman Rl, Goldberg HI. Cerebral vasculitis:MR imaging and angiographic correlation. Radiology 1992;182:65-72.

33. Alhalabi M, Moore PM. Serial angiography in isolatedangiitis of the central nervous system. Neurology 1994;44:1221-6.

34. Vanderzant C, Bromberg M, MacGuire A, McCuneJ. Isolated small-vessel angiitis of the central nervous system.Arch Neurol 1988; 45:683-7.

35. Koo EH, Massey EW. Granulomatous angiitis of the centralnervous system: Protean manifestations and response totreatment. J Neurol Neurosurg Psychiatry 1988;51:1126-33.

36. VollmerTL, GuarnacciaJ, Harrington W, etal. Idiopathicgranulomatous angiitis of the central nervous system:Diagnostic challenges. Arch Neurol 1993; 50:925-30.

37. Hellmann DB, Roubenoff R, Healy RA, Wang H. Centralnervous system angiography: Safety and predictors of apositive result in 125 consecutive patients evaluated forpossible vasculitis. J Rheumatol 1992; 19:568-72.

38. Ffytche TJ. Retinal vasculitis: a review of the clinical signs.Trans Ophthal Soc UK 1977; 97:457-61.

39. Stanford MR, Graham EM, Kasp E, etal. Retinal vasculitis:correlation of animal and human disease. Eye 1987;1:69-77.

40. Barza M, Pauker SG. The decision to biopsy, treat, or waitin suspected herpes encephalitis. Ann Intern Med 1980;92:641-9.

41. Calabrese LH, Gragg LA, Furlan AJ. Benign angiopathy: asubset of angiographically defined primary angiitis of thecentral nervous system. J Rheumatol 1993; 20:2046-50.

42. Lockwood CM. Approaches to specific immunotherapy forsystemic vasculitis. Semin Neurol 1994; 14:387-92.



ownloaded from


by guest on june 9, 2013 … · q j med 1997; 90:61-73 cerebral vasculitis—recognition, diagnosis...

Documents