chronic lymphocytic inflammation with pontine perivascular - brain

BRAINA JOURNAL OF NEUROLOGY

Chronic lymphocytic inflammation withpontine perivascular enhancementresponsive to steroids (CLIPPERS)Sean J. Pittock,1,2 Jan Debruyne,3 Karl N. Krecke,4 Caterina Giannini,2 Jelle van den Ameele,3

Veerle De Herdt,3 Andrew McKeon,1,2 Robert D. Fealey,1 Brian G. Weinshenker,1

Allen J. Aksamit,1 Bruce R. Krueger,1 Elizabeth A. Shuster5 and B. Mark Keegan1

1 Department of Neurology, Mayo Clinic College of Medicine, Rochester, MN 55905, USA

2 Department of Laboratory Medicine and Pathology, Mayo Clinic College of Medicine, Rochester, MN 55905, USA

3 Department of Neurology, Ghent University Hospital, Ghent, B.9000, Belgium

4 Department of Radiology, Mayo Clinic College of Medicine, Rochester, MN 55905, USA

5 Department of Neurology, Mayo Clinic College of Medicine, Jacksonville, FL 32224, USA

Correspondence to: Dr B. Mark Keegan,

Department of Neurology,

Mayo Clinic College of Medicine,

200 First St. SW Rochester,

MN 55905, USA

E-mail: [email protected]

The classification and pathological mechanisms of many central nervous system inflammatory diseases remain uncertain. In this

article we report eight patients with a clinically and radiologically distinct pontine-predominant encephalomyelitis we have

named ‘chronic lymphocytic inflammation with pontine perivascular enhancement responsive to steroids’ (CLIPPERS). The pa-

tients were assessed clinically, radiologically and pathologically at Mayo Clinic, USA and Ghent University Hospital, Belgium

from 1999 to 2009. Median follow-up duration from clinical onset was 22 months (range 7–144 months). Patients underwent

extensive laboratory (serum and cerebrospinal fluid), radiological and pathological testing (conjunctival, transbronchial and brain

biopsies) to search for causes of an inflammatory central nervous system disorder. All eight patients (five female, three male)

presented with episodic diplopia or facial paresthesias with subsequent brainstem and occasionally myelopathic symptoms and

had a favourable initial response to high dose glucocorticosteroids. All patients had symmetric curvilinear gadolinium enhance-

ment peppering the pons and extending variably into the medulla, brachium pontis, cerebellum, midbrain and occasionally

spinal cord. Radiological improvement accompanied clinical response to glucocorticosteroids. Patients routinely worsened fol-

lowing glucocorticosteroid taper and required chronic glucocorticosteroid or other immunosuppressive therapy. Neuropathology

of biopsy material from four patients demonstrated white matter perivascular, predominantly T lymphocytic, infiltrate without

granulomas, infection, lymphoma or vasculitis. Chronic lymphocytic inflammation with pontine perivascular enhancement re-

sponsive to steroids is a definable, chronic inflammatory central nervous system disorder amenable to immunosuppressive

treatment. The T cell predominant inflammatory pathology in affected central nervous system lesions and the clinical and

radiological response to immunosuppressive therapies is consistent with an immune-mediated process.

Keywords: brain stem; neuroinflammation; encephalitis

Abbreviations: CLIPPERS = chronic lymphocytic inflammation with pontine perivascular enhancement responsive to steroids

doi:10.1093/brain/awq164 Brain 2010: 133; 2626–2634 | 2626

Received February 15, 2010. Revised March 17, 2010. Accepted May 14, 2010. Advance Access publication July 17, 2010

� The Author (2010). Published by Oxford University Press on behalf of the Guarantors of Brain. All rights reserved.

For Permissions, please email: [email protected]

Dow

nloaded from https://academ

ic.oup.com/brain/article/133/9/2626/352369 by guest on 13 February 2022

IntroductionThe immunopathogeneses of most inflammatory CNS disorders,

including those with and without accompanying demyelination,

remain poorly understood (Kalman, 2008). Advances in clinical,

radiological, serological and pathological evaluation have facili-

tated disease classification and differentiation of distinctive disease

entities (Kalman, 2008). Examples include: (i) revised diagnostic

criteria for multiple sclerosis (Polman et al., 2005; based on clinical

and radiological data and supportive paraclinical investigations

importantly with ‘no better explanation’); (ii) novel autoantibody

markers of autoimmune optic neuritis and myelitis [aquaporin-

4-immunoglobulin G (IgG) (Lennon et al., 2005; Wingerchuk

et al., 2007); collapsin response mediator protein (CRMP-5-IgG)

(Cross et al., 2003; Keegan et al., 2008)] and encephalitis [voltage

gated potassium channel (Tan et al., 2008) Ma (Dalmau et al.,

2004) and NMDA receptor antibodies (Dalmau et al., 2008)] and

(iii) immunopathological characterization of neuromyelitis optica

[i.e. loss of aquaporin-4 immunoreactivity, vasculocentric distribu-

tion of immunoglobulin deposition and complement activation

(Roemer et al., 2007)]. We report clinical, radiological and patho-

logical features of a treatable, brainstem-predominant, clinical and

radiological entity, which we term chronic lymphocytic inflamma-

tion with pontine perivascular enhancement responsive to steroids

(CLIPPERS).

Patients and methodsThe study was approved by Mayo Clinic Institutional Review Board.

The 8 patients were assessed clinically, radiologically and pathologically

at Mayo Clinic (n = 6) (Rochester, MN and Jacksonville, FL) and

at Ghent University Hospital (n = 2) (Ghent, Belgium) from 1999 to

2009. Patients underwent extensive laboratory (serum and cerebro-

spinal fluid), radiological and pathological testing (conjunctival,

transbronchial and brain biopsies) to search for causes of an inflam-

matory brainstem-predominant CNS disorder.

Autoimmune serological evaluationFive of six Mayo Clinic patients had serum (n = 4) or CSF (n = 2) avail-

able for neural autoantibody testing including: cation channel antibo-

dies [voltage-gated calcium channels (P/Q-type and N-type),

voltage-gated potassium channels and nicotinic acetylcholine receptors

(muscle-type and ganglionic-type)]; skeletal muscle striational antibo-

dies; anti-neuronal nuclear autoantibodies types 1, 2 and 3; Purkinje-

cell cytoplasmic autoantibodies types 1, 2 and Tr; anti-glial/neuronal

nuclear antibody type 1; collapsin response-mediator protein-5 IgG;

amphiphysin IgG and glutamic acid decarboxylase and thyroid auto-

antibodies. Testing for potential novel or unclassified neural-specific

autoantibodies was also performed blinded in a routine clinical labora-

tory setting using a composite 4 mm frozen section of mouse cerebel-

lum and brainstem, gut and kidney (standardized to detect

paraneoplastic autoimmunity) (Pittock et al., 2006). To minimize inter-

ference by non-neural-specific antibodies, sera were preabsorbed (at

1 : 240 dilution) with bovine liver powder. Fluorescein-conjugated goat

anti-human IgG (Southern Biotechnology; Birmingham, AL) detected

bound IgG.

ImagingBrain MRI was performed in 1.5 T MRI scanners, and images from

T1 (pre and post administration of gadolinium), T2, fluid attenuation

inversion recovery and proton density sequences were acquired.

A median of five MRIs (range 3–12) per patient were available for re-

view, spanning a median interval of six months (range 1–104 months)

and including pre and post therapy images. Vascular neuroimaging

was available for review in four patients, specifically magnetic reson-

ance angiography (n = 2) and conventional cerebral angiography

(n = 2). Spinal cord MRI exams with gadolinium were available for

five patients. All patients underwent chest CT imaging to address

the possibility of sarcoidosis, and three patients also had CT examin-

ation of abdomen and pelvis.

Pathological methodsThree patients underwent brain biopsy at Mayo Clinic (Patients 1,

2 and 5) and one at Ghent University Hospital (Patient 6). Tissue

slides were made from formalin fixed paraffin embedded brain speci-

mens and stained with haematoxylin and eosin, luxol fast blue coun-

terstained with periodic acid-Schiff and silver impregnations (modified

Bielschowsky). Immunohistochemical stains obtained included: glial fi-

brillary acid protein for astrocytes (clone GA5, 1/1,000; BioGenex),

CD68 (clone KP-1, 1/200 dilution; Dako) for microglia/macrophages,

CD3, CD1a, CD20, CD117 and placental alkaline phosphatase.

Results

Clinical characteristics

Symptoms

The clinical characteristics of the eight patients (median age at

onset 45.5 years; range 16–86 years; five females and three

males) are shown in the Table 1. All patients had subacute gait

ataxia and diplopia. Seven patients also developed dysarthria (pri-

marily ataxic but with some spastic features). An altered sensation

or tingling of the face or scalp was reported by five of eight pa-

tients. Other symptoms included non-specific dizziness, nausea,

dysgeusia, pseudobulbar affect (pathological crying or laughter),

tinnitus, tremor, nystagmus, paraparesis, sensory loss and spasti-

city. None had significant systemic symptoms nor weight loss,

fever, meningism, lymphadenopathy, polydipsia/polyuria or other

symptoms of hypothalamic dysfunction, uveitis or oral and/or

genital ulcers.

Neuroimaging

Brain MRI in all eight patients revealed a characteristic pattern of

punctate and curvilinear enhancement peppering the pons (Fig. 1)

and extending variably into the medulla, brachium pontis and

mid-brain. A subtler radiating pattern of similar enhancing lesions

extended into the basal ganglia and inferior cerebellar white

matter in three patients, corpus callosum in one patient and

spinal cord in three patients (Fig. 2). Lesions were typically less

numerous and smaller as distance from the pons increased. The

largest single punctate lesion was 9 mm in greatest dimension, but

most lesions measured between 1 and 3 mm. One patient pre-

sented with asymmetric enhancement in the left pons and right

CLIPPERS Brain 2010: 133; 2626–2634 | 2627

Dow



Tab

le1

Cli

nic

alfe

ature

sof

eight

pat

ients

wit

hC

LIPPER

S

Pat

ient

Sex

Age

atonse

t(y

ears

)

Dura

tion

of

foll

ow

-up

from

onse

t(m

onth

s)

Sym

pto

ms

atonse

tSy

mpto

ms

duri

ng

illn

ess

evolu

tion

Loca

tion

of

per

ivas

cula

rco

ntr

ast

enhan

cem

ent

on

MR

I

CSF

abnorm

alit

ies

Tim

efr

om

onse

tto

bra

inbio

psy

(month

s)

Tim

efr

om

onse

tto

glu

coco

rtic

ost

eroid

ther

apy

(month

s)

Init

ial

dose

,ro

ute

and

resp

onse

toglu

coco

rtic

ost

eroid

s

Mai

nte

nan

cetr

eatm

ent

and

foll

ow

-up

1F

78

20

Ata

xia,

dip

lopia

Ata

xic

dys

arth

ria,

dip

lopia

,al

tere

dfa

cial

sensa

tion,

action

trem

or,

pse

udobulb

araf

fect

(pat

holo

gic

alcr

ying),

pal

atal

myo

clonus

Sym

met

ric

inm

edulla

,pons

(bra

chiu

mpontis)

,m

idbra

inan

dbas

algan

glia

10

white

blo

od

cells

,pro

tein"

,N

SE"

912

Pre

dnis

one

80

mg

dai

ly:

radio

logic

alim

pro

ve-

men

tw

ithout

mar

ked

clin

ical

impro

vem

ent

Ora

lpre

dnis

one,

reso

lution

of

gad

olin

ium

enhan

cing

lesi

ons

but

per

sist

ent

impai

rmen

t

2M

37

144

Epis

odic

faci

altinglin

gA

taxi

a,sp

astic

atax

icdys

arth

ria,

dip

lopia

,pse

udobulb

araf

fect

(pat

holo

gic

alla

ughte

r),

nau

sea,

diz

zines

s,neu

rogen

icbla

dder

Sym

met

ric

inm

edulla

,pons,

infe

rior

and

super

ior

cere

bel

lar

ped

uncl

es.

Upper

cerv

ical

spin

alco

rdin

volv

emen

t

Pro

tein"

,O

CB"

36

41

gIV

MP

dai

lyfo

r5

day

s:dra

mat

icim

pro

vem

ent

Initia

llyuse

dpuls

eIV

MP

and

pre

dnis

one

taper

for

firs

tfe

wye

ars

then

com

bin

edora

lpre

dnis

-one

and

met

hotr

exat

e.Pat

ient

now

stab

leon

met

hotr

exat

em

onoth

erap

y3

M47

7Ep

isodic

horizo

nta

ldip

lopia

Ata

xia,

atax

icdys

arth

ria,

dip

lopia

,al

tere

dsc

alp

and

limb

sensa

tion,

dys

phag

ia,

spel

ls(las

ting

seco

nds)

of

spee

char

rest

and

leg

stiffe

nin

g

Sym

met

ric

par

amid

line

mid

bra

in,

pons,

cere

bel

lar

ped

uncl

es

Pro

tein"

Bio

psy

not

done

51

gIV

MP

dai

lyfo

r5

day

s:m

oder

ate

impro

vem

ent

Ora

lpre

dnis

one

4F

86

25

Horizo

nta

ldip

lopia

Ata

xia,

atax

icdys

arth

ria,

dip

lopia

,al

tere

dfa

cial

sensa

tion,

diz

zines

s,in

term

itte

nt

lighth

eaded

nes

s

Sym

met

ric

mid

bra

in(lef

tce

rebra

lped

uncl

e),

pons

and

exte

ndin

gm

inim

ally

into

the

mid

dle

cere

bel

lar

ped

uncl

es

OC

B"

,N

SE"

Bio

psy

not

done

71

gIV

MP

dai

lyfo

r5

day

s:dra

mat

icim

pro

vem

ent

1g

IVM

Pev

ery

10-1

4day

s:die

d16

month

sla

ter

from

pulm

onar

yem

bolu

s

5M

70

22

Horizo

nta

ldip

lopia

Ata

xia,

atax

icdys

arth

ria,

alte

red

faci

alan

dlim

bse

nsa

tion,

nau

sea,

fatigue,

tinnitus,

tongue

wea

knes

s

Sym

met

ric

dors

alm

edulla

,pons

and

mid

bra

in

Pro

tein"

20

21

1g

IVM

Pdai

lyfo

r5

day

s:m

oder

ate

impro

vem

ent

Ora

lpre

dnis

one

60

mg

for

thre

em

onth

sta

per

ing

by

10

mg

ever

ytw

ow

eeks

6F

41

14

Dip

lopia

Ata

xia,

atax

icdys

arth

ria,

diz

zines

s.C

ereb

ella

rped

uncl

esright

gre

ater

than

left

,pons

Initia

l"O

CB

then

OC

Bnorm

aliz

edon

repea

tC

SF

13

1IV

MP;

reso

lution

within

2w

eeks

Initia

ted

mitoxa

ntr

one

inad

ditio

nto

glu

coco

rtic

ost

eroid

s.

7F

16

43

Dip

lopia

Ata

xia,

par

apar

esis

,dip

lopia

,ve

rtic

alupbea

tnys

tagm

us

Sym

met

ric

pons,

cere

bel

lum

,m

idbra

in.

Cer

vica

lan

dth

ora

cic

spin

alco

rdin

volv

emen

t

Norm

alBio

psy

not

done

11

gIV

MP

dai

lyfo

r5

day

sw

ith

ora

lpre

dnis

one

taper

for

2w

eeks

;co

mple

tere

solu

tion

of

bra

inst

emsy

mpto

ms

Wors

ened

follo

win

gglu

coco

rtic

ost

eroid

taper

.A

zath

ioprine

added

late

rto

glu

coco

rtic

ost

eroid

sdid

not

contr

olre

curr

ence

8F

44

7H

orizo

nta

ldip

lopia

Ata

xia,

dys

arth

ria,

dys

geu

sia,

par

apar

esis

,hyp

erac

usi

s

Sym

met

ric

pons,

cere

bel

lum

,m

idbra

in.

Cer

vica

lan

dth

ora

cic

spin

alco

rdin

volv

emen

t

Pro

tein"

Bio

psy

not

done

31

gIV

MP

dai

lyfo

r5

day

sw

ith

ora

lpre

dnis

one

taper

for

2w

eeks

;dra

mat

icim

pro

vem

ent

Wors

ened

follo

win

gpre

dnis

one

taper

s.R

einitia

ted

pre

dnis

one.

NSE

=neu

ron

spec

ific

enola

se;

OC

B-C

SF=

uniq

ue

olig

ocl

onal

ban

ds;

IVM

P=

intr

aven

ous

met

hyl

pre

dnis

olo

ne;

M=

mal

e;F

=fe

mal

e;"

=el

evat

ed.

2628 | Brain 2010: 133; 2626–2634 S. J. Pittock et al.

Dow



mesencephalon that progressed over 6 months to become more

extensive and symmetric (Fig. 3). Cervical and thoracic spinal cord

MRI performed with and without contrast was normal in two

patients, and in three patients showed foci of punctate gadolinium

enhancement within the spinal cord (Fig. 2). There were no other

remote supratentorial parenchymal white or grey matter enhan-

cing lesions, no hypothalamic-pituitary involvement and no bony,

meningeal or pial lesions. There was patchy non-specific increased

T2 signal in the corresponding regions of gadolinium enhance-

ment. None of the lesions had significant mass effect.

Intracranial magnetic resonance angiography performed in two

patients and conventional angiography of neck and intracranial

vessels in two other patients were normal without angiographic

evidence of vasculitis.

Differential diagnosesDiagnoses considered included: neurosarcoidosis, CNS lymphoma,

lymphomatoid granulomatosis, CNS vasculitis, Bickerstaff brain-

stem encephalitis, paraneoplastic disease, chronic perivascular

infectious process (tuberculosis, neurosyphilis, Whipple’s disease,

parasitic infection), glioma, CNS demyelinating disease, CNS

Behcet’s and histiocytosis including Langerhans cell histiocytosis

and Erdheim-Chester disease.

Laboratory investigationsExtensive systemic work-up included: complete blood count; renal

and liver function (all values normal in eight patients tested); sedi-

mentation rate (normal in four tested); angiotensin converting

enzyme (one of seven patients tested had an elevated value of

74 U/l; normal range 7–46 U/l); anti-nuclear antibody [one of six

tested was seropositive (4.8 U; normal51 U; double stranded DNA

antibody was negative)]; antibodies against extractable nuclear

antigens [negative in three tested; Sjogren syndrome B (SSB) min-

imally elevated in one (value 1.1 U; normal 51.0 U who was

anti-nuclear antibody negative)]; and viral, fungal and syphilis serol-

ogies (negative in eight tested). Radiological investigations negative

in all patients were: CT chest (all eight patients), CT abdomen and

pelvis (three patients), whole body PET (two patients), testicular

ultrasound (two patients) and mammogram (two patients).

No neural specific IgGs were identified in the five patients

(sera n = 4, CSF n = 2) tested. Serum antibodies against the

aquaporin-4 water channel (neuromyelitis optica immunoglobulin

G; NMO-IgG) were assessed and negative in 4 Mayo Clinic

patients and one patient whose CSF was tested. One Belgian

patient’s serum (Patient 7) tested was NMO-IgG negative.

Pathological investigations other than brain biopsy included:

conjunctival biopsy in three patients and transbronchial lung

biopsy in 1, all of which yielded normal tissue.

Electrophysiological assessments done and normal were visual

evoked response (n = 2), brainstem auditory evoked response

(n = 2) and somatosensory evoked potentials (n = 1).

Cerebrospinal fluid analysesCSF abnormalities were found in four of eight patients

tested: mildly elevated protein in four (47–65 mg dl-1; normal

reference range 15–45 mg dl-1) and pleocytosis in 1 (10 white

blood cells ul�1, 98% lymphocytes). Elevated unique CSF oligoclo-

nal bands were found in three of six tested, including one patient

whose electrophoretic pattern reverted to normal. CSF cytology

Figure 1 Characteristic pontine-predominant MRI characteristics of CLIPPERS syndrome. MRI brain with gadolinium showing foci of

pontine enhancement with a curvilinear pattern highly suggestive of a perivascular distribution in all eight patients.

CLIPPERS Brain 2010: 133; 2626–2634 | 2629

Dow



with flow cytometry was negative for malignant cells in all eight

patients. Fungal cultures (7) and Tropheryma whippelii polymer-

ase chain reaction (2) were also negative.

NeuropathologySites chosen for brain biopsies in four of eight patients were the

pons in one and cerebellum in three. The biopsy findings were

similar in each case and revealed a marked lymphocytic infiltrate

in the white matter with perivascular predominance, but also a

more diffuse parenchymal inflammatory infiltrate (Fig. 4). The

lymphocytic infiltrate was composed predominantly of CD3 react-

ive T lymphocytes and some CD20 positive B lymphocytes. CD68

positive histiocytes and activated microglia were present in mod-

erate number. Myelin was intact. Special stains for fungi

and mycobacteria (Grocott methenamine silver and auramine

rhodamine) were negative. Characteristic findings of sarcoidosis,

histiocytosis, lymphoma, lymphomatoid granulomatosis, multiple

sclerosis or other diseases were not found.

TreatmentAll patients were treated initially with glucocorticosteroids [intra-

venous methylprednisolone (1 g daily for 5 days) in seven and oral

prednisone (80 mg daily) in one]. The median time from symptom

onset to glucocorticosteroid treatment initiation was 6.75 months

(range 1–21 months). All seven patients treated with intravenous

methylprednisolone showed improvements in gait, diplopia, head

and limb sensation and dysarthria between Days 2 and 4 of

treatment. Patient 1 treated with 80 mg oral prednisone did not

experience clinical improvement. The number and size of perivas-

cular enhancing lesions was reduced on post-treatment MRI

exams in all patients (Fig. 5). Lesion distribution continued to be

concentrated in the pons. Long-term therapeutic response varied,

ranging from excellent (Patient 2) to incomplete with ongoing

inflammatory disease despite immunosuppressive medications

(Patient 6).

All treated patients required maintenance immunosuppression to

sustain clinical improvements (Table 1). In 6 cases in which with-

drawal or reduction of glucocorticosteroids was attempted, clinical

relapse ensued.

Illustrative case history—Patient 2Two years prior to evaluation at Mayo Clinic, the patient noticed

episodic facial tingling but was well otherwise. Six months prior,

he developed painless, horizontal, binocular diplopia and subse-

quently gait ataxia, worsening over 2 months. Brain MRI

showed brainstem gadolinium enhancing abnormalities centred

within the pons. Serological investigations were normal and CSF

demonstrated elevated unique oligoclonal bands. He was treated

for five consecutive days with 1000 mg intravenous methylpredni-

solone and his symptoms improved. Further investigations were

unrevealing. He remained in clinical remission without ongoing

treatment for 1 year, after which he experienced worsening gait

ataxia, dysarthria and recurrent horizontal diplopia. Repeat brain

MRI demonstrated enlarging areas of gadolinium enhancing le-

sions within the brainstem (Fig. 5A). A pontine biopsy was per-

formed while on no therapy (Fig. 4D–F). He was retreated with

five consecutive days of 1000 mg intravenous methylprednisolone

and he again had marked symptomatic improvement. He then had

three more similar episodes of symptomatic worsening within the

following 8 months that all reliably improved with intravenous

corticosteroids, only to relapse following their discontinuation.

Chronic therapy with oral prednisone 60 mg every other day

was therefore initiated and he experienced steady improvement

Figure 2 Representative MRI findings outside of primary

brainstem abnormalities. Similar foci of gadolinium

enhancement seen in upper thoracic cord (A), upper cervical

cord (B), mid-lower thoracic cord (C) and basal ganglia (D).


Dow



in diplopia, dysarthria, gait ataxia and dysphagia. Brain MRI

showed significant improvement in the gadolinium enhancing le-

sions (Fig. 5B).

Oral hydroxychloroquine (400 mg daily) was introduced and

prednisone was weaned over months because of glucocorticoster-

oid side effects. Six months later, brain MRI showed radiological

progression of the inflammatory lesions (Fig. 5C) although clinic-

ally he remained stable. Following reinitiation of oral prednisone

(60 mg every other day), a follow-up brain MRI scan showed

improvement, the hydroxychloroquine was discontinued and oral

methotrexate (7.5 mg once weekly) was started and gradually

increased to 20 mg weekly.

After 1.5 years of clinical and radiological stability on metho-

trexate and prednisone, alternate day treatment with oral prednis-

one was reduced by 5 mg each month and at last follow-up he

remained on oral methotrexate (10 mg weekly) monotherapy

without radiological recurrence (Fig. 5D). He has developed

neither symptoms or radiological evidence of other CNS diseases

such as multiple sclerosis nor a systemic disease despite follow-up

of 12 years.

DiscussionWe suggest that the constellation of clinical, radiological and

pathological findings reported in these eight patients represents

a definable, treatable inflammatory CNS, brainstem-predominant

syndrome (CLIPPERS). Clinical and radiological findings point to a

pontine-centric disorder with variable involvement of adjacent

structures. All patients experienced cranial sensory abnormalities,

diplopia, ataxia and dysarthria, and had a signature MRI punctate

pattern of patchy gadolinium enhancement ‘peppering’ the pons,

brainstem, cerebellum and spinal cord. Neuropathology in four of

eight patients revealed a prominent perivascular lymphocytic infil-

trate. Immunosuppressive therapy was successful in improving

clinical symptoms and radiological disease; however, early gluco-

corticosteroid taper was accompanied by recurrence of both symp-

toms and radiological progression, and therefore prolonged

therapy appears to be necessary for the majority of cases.

Other disorders likely to present with these findings were care-

fully excluded. Neurosarcoidosis (Spencer et al., 2005; Brown

et al., 2008), CNS lymphoma (Batchelor and Loeffler, 2006) and

CNS vasculitis (Salvarani et al., 2007; Miller et al., 2009) were the

three alternative diagnoses most strongly favoured in the initial

evaluation. Although not characteristic, each of these entities

could show a similar pattern of perivascular or infiltrative gadolin-

ium enhancement on MRI. Pathological review of lung, conjunc-

tival or brain biopsy tissues failed to reveal non-necrotizing

granulomas with giant cells typical of sarcoidosis. CSF cytology,

systemic imaging, haematologic investigations and neuropathology

revealed no evidence of lymphoma. Vasculitis was not found on

conventional or magnetic resonance angiography, and pathology

showed no partial or transmural destructive vascular inflammation

involving leptomeningeal or parenchymal vessels (Miller et al.,

2009) in biopsy specimens.

Figure 3 Progression of lesions over 6 months in Patient 3. MRI brain (T1 weighted post gadolinium) at presentation shows asymmetric

enhancement in the left pons and right mesencephalon (A, top panels), which progressed over 6 months (B, bottom panels) to

become more extensive and symmetric.

CLIPPERS Brain 2010: 133; 2626–2634 | 2631

Dow



Figure 4 Neuropathology from Patients 1, 2, and 5. Marked perivascular and parenchymal lymphocytic infiltrates in the white matter

(haematoxylin and eosin stain), for Patients 1 (A�100, B�200), 2 (D�40, E� 200), and 5 (G�100). The lymphocytic infiltrate

was composed predominantly of CD3 positive (T) lymphocytes in all cases (Patient 1, C� 400; Patient 2, F� 400; Patient 5, H� 200);

similar pathology was demonstrated in Patient 6 (results not shown).


Dow



Other diagnoses considered but excluded were: histiocytosis X

(Prayer et al., 2004; Langerhans cells were absent and neuroima-

ging was inconsistent), paraneoplastic disease (Ellison et al., 2008;

systemic workup for malignancy, paraneoplastic autoantibody

markers were negative and neuropathology lacked characteristic

microglial nodules and neuronophagia), demyelinating disease

(Lucchinetti et al., 2008; no evidence of demyelination on path-

ology, nor were there typical MRI findings). The clinical course,

negative microbiology workup, improvement with immunosup-

pressants and absence of well-formed microglial nodules, necrotiz-

ing granulomata or stainable microorganisms argue strongly

against infectious processes.

Bickerstaff brainstem encephalitis deserves particular attention,

as it is also a brainstem-predominant inflammatory disease. This

diagnosis was excluded in these patients by the observed neuro-

pathology and radiological features and lack of peripheral nerve

involvement (Odaka et al., 2003; Ito et al., 2008). Diagnostic

features of Bickerstaff brainstem encephalitis include drowsiness

or coma, progressive external ophthalmoplegia, ataxia and corti-

cospinal tract signs; the vast majority of cases occur following

viral illness and have a monophasic course and a good prognosis.

MRI abnormalities occur in a minority (10–31%) of patients with

Bickerstaff brainstem encephalitis and, when present, typically

appear as a homogenous, non-gadolinium enhancing lesion

easily differentiated from the pepper-like gadolinium enhancement

seen as a hallmark MRI feature in our patients (Chataway et al.,

2001; Mondejar et al., 2002). Furthermore, peripheral polyneur-

opathy and serum anti-ganglioside GQ1b IgG antibodies are

present in the majority of patients with Bickerstaff brainstem

encephalitis, and thus it is now considered to be in the spectrum

of Miller Fisher Guillain Barre syndrome with coexistent CNS

involvement.

Although only eight subjects were identified, we suspect that

this disorder may be under-recognized, considering the relatively

short duration over which these patients were identified in three

academic centres. Despite a characteristic clinical and radiological

syndrome, other, as yet undefined, aetiologies could present in a

similar fashion and perhaps were not excluded in the four patients

lacking pathological examination. Those without pathological

confirmation seemed to have similar imaging, outcome and

radiological response to therapy as those in whom this was

accomplished. Although the median duration of follow-up was

short (median 20 months), it was reassuring that those patients

with the longest follow-up experienced neither dissemination of

the disease in the CNS nor any underlying systemic illness nor

evidence of idiopathic inflammatory demyelinating disease such

as multiple sclerosis despite follow-up as long as 12 years.

Despite the eloquence of the primarily involved CNS region

(brainstem, cerebellar peduncles), pathological examination was

accomplished in four cases and showed identical pathology that

corresponded to the neuroimaging abnormalities noted in all the

patients. It remains possible, given the sample size limitations

Figure 5 Marked reduction in extent of gadolinium enhancement following immunotherapy in Patient 2. Axial T1 MRI brain with

gadolinium. (A) At time of relapse, off of all immunosuppressive therapy and prior to brain biopsy. (B) Near complete resolution of

gadolinium enhancing lesions following chronic oral prednisone. (C) Recurrent gadolinium enhancing lesions following discontinuation

of prednisone. (D) Improvement and stability for over 1.5 years on methotrexate monotherapy.

CLIPPERS Brain 2010: 133; 2626–2634 | 2633

Dow



required to prevent neurological injury, other pathology may have

been missed. However, the biopsies were of substantial quality

and size for pathological interpretation and there were no cases

where the biopsy suggested an alternative pathology, even around

the margins of the submitted sample. Extensive evaluations were

done in all patients and no alternative diagnoses were found.

The clinical course and neuroimaging features were similar

between those with and those without tissue pathology, although

an alternative pathological process could not be excluded as

causing this identical syndrome in patients where pathology was

not documented.

The pathogenesis of CLIPPERS is unknown. The presence

of a perivascular and parenchymal inflammatory cell infiltrate

in affected CNS tissue combined with the clinical response to

immunosuppressive therapies suggests an autoimmune or other

inflammatory mediated pathogenesis. If organ-specific auto-

immunity is its basis, then the location of the inflammatory infil-

trate suggests that the target autoantigen is likely to be located

in perivascular regions. A limited form of multiple sclerosis is

excluded given the lack of development of more typical attacks

and MRI lesions and the neuropathological findings.

We recommend full evaluation with both non-invasive and min-

imally invasive procedures to attempt to make a clear diagnosis if

similar cases are encountered. The patients who underwent biop-

sies did not experience adverse effects of the neurosurgical pro-

cedure. Given the location of the prominent abnormalities within

the brainstem, however, caution needs to be taken prior to con-

sideration of this in similar cases. Biopsy should be considered in

patients when alternative diagnoses remain likely and experienced

neurosurgical colleagues recommend that it may be performed

safely. Conversely, if alternative diagnoses are rigorously excluded,

the clinical and radiological features may be sufficiently distinctive

that CLIPPERS could be diagnosed and treated without

pathological examination.

Conflict of interestDr. Keegan is compensated as a Section Editor for Neurology� and

as Chief Editor for eMedicine. The authors report no other conflict

of interests.

ReferencesBatchelor T, Loeffler J. Primary CNS lymphoma. J Clin Oncol 2006; 24:

1281–8.

Brown E, Gray F. Bacterial infections. In: Love S, Louis DN, editors.Greenfield’s Neuropathology 8th edn. London: Edward Arnold

Publishers Ltd; 2008. p. 1391–445.

Chataway S, Larner A, Kapoor R. Anti-GQ1b antibody status, magnetic

resonance imaging, and the nosology of Bickerstaff’s brainstemencephalitis. Eur J Neurol 2001; 8: 355–7.

Cross S, Salomao D, Parisi J, Kryzer T, Bradley E, Mines J, et al.

Paraneoplastic autoimmune optic neuritis with retinitis defined by

CRMP-5-IgG. Ann Neurol 2003; 54: 38–50.

Dalmau J, Gleichman A, Hughes E, Rossi J, Peng X, Lai M, et al. Anti-

NMDA-receptor encephalitis: case series and analysis of the effects of

antibodies. Lancet Neurol 2008; 7: 1091–8.

Dalmau J, Graus F, Villarego A, Posner J, Blumenthal D, Thiessen B, et al.

Clinical analysis of anti-Ma2-associated encephalitis. Brain 2004; 127:

1831–44.

Ellison DW, Perry A, Rosenblum M, Asa S, Reid R, Louis DN. Tumours:

non-neuroepithelial tumours and secondary effects. Greenfield’s

Neuropathology 8th edn. London: Edward Arnold Publishers Ltd;

2008. p. 2002–2146.

Ito M, Kuwabara S, Odaka M, Misawa S, Koga M, Hirata K, et al.

Bickerstaff’s brainstem encephalitis and Fisher syndrome form a

continuous spectrum. J Neurol 2008; 255: 674–82.

Kalman B. Disorders of the central and peripheral nervous systems

related to known or assumed system-immune abnormalities. In:

Kalman B, editor. Neuroimmunology in clinical practice. Oxford,

UK: Blackwell Publishing, Ltd, 2008. p. 179–259.

Keegan BM, Pittock S, Lennon V. Autoimmune myelopathy associated

with collapsin response-mediator protein-5 immunoglobulin G.

Ann Neurol 2008; 63: 531–4.

Lennon V, Kryzer T, Pittock S, Verkman A, Hinson S. IgG marker of

optic-spinal multiple sclerosis binds to the aquaporin-4 water channel.

J Exp Med 2005; 202: 473–7.

Lucchinetti C, Gavrilova R, Metz I, Parisi J, Scheithauer B, Weigand S,

et al. Clinical and radiographic spectrum of pathologically confirmed

tumefactive multiple sclerosis. Brain 2008; 131: 1759–75.

Miller D, Salvarani C, Hunder G, Brown R, Parisi J, Chirstianson T, et al.

Biopsy findings in primary angiitis of the central nervous system.

Am J Surg Pathol 2009; 33: 35–43.

Mondejar R, Santos J, Villalba E. MRI findings in a remitting-relapsing

case of Bickerstaff encelphalitis. Neuroradiology 2002; 44: 411–4.

Odaka M, Yuki N, Yamada M, Koga M, Takemi T, Hirata K, et al.

Bickerstaff’s brainstem encephalitis: clinical features of 62 cases and

a subgroup associated with Guillain-Barre syndrome. Brain 2003; 126:

2279–90.

Pittock S, Yoshikawa H, Ahlskog J, Tisch S, Benarroch E, Kryzer T, et al.

Glutamic acid decarboxylase autoimmunity with brainstem, extrapyr-

amidal, and spinal cord dysfunction. Mayo Clin Proc 2006; 81:

1207–14.

Polman C, Reingold S, Edan G, Filippi M, Hartung H, Kappos L, et al.

Diagnostic criteria for multiple sclerosis: 2005 revisions to the

"McDonald Criteria". Ann Neurol 2005; 58: 840–6.

Prayer D, Grois N, Prosch H, Gadner H, Barkovich A. MR imaging pre-

sentation of intracranial disease associated with Langerhans cell histio-

cytosis. AJNR Am J Neuroradiol 2004; 25: 880–91.

Roemer S, Parisi J, Lennon V, Benarroch E, Lassmann H, Bruck W, et al.

Pattern-specific loss of aquaporin-4 immunoreactivity distinguishes

neuromyelitis optica from multiple sclerosis. Brain 2007; 130:

1194–205.

Salvarani C, Brown R, Calamia K, Christianson T, Weigand S, Miller D,

et al. Primary central nervous system vasculitis: analysis of 101 patients.

Ann Neurol 2007; 62: 442–51.

Spencer T, Campellone J, Maldonado I, Huang N, Usmani Q,

Reginato A. Clinical and magnetic resonance imaging manifestations

of neurosarcoidosis. Semin Arthritis Rheum 2005; 34: 649–61.

Tan K, Lennon V, Klein C, Boeve B, Pittock S. Clinical spectrum of

voltage-gated potassium channel autoimmunity. Neurology 2008;

70: 1883–90.

Wingerchuk D, Lennon V, Lucchinetti C, Pittock S, Weinshenker B.

The spectrum of neuromyelitis optica. Lancet Neurol 2007; 6:

805–815.


Dow



chronic lymphocytic inflammation with pontine perivascular - brain

Documents