Neuroradiology Harris
S. Chrysikopoulos,
John
R. Hesselink,
#{149} Gary
MD
MD
#{149} Clayton
Encephalitis Immunodeficiency MR Imaging Clinical and
A. Press, MD A. Wiley, MD
tion,
13.203,
Brain,
MR
ter, 13.253 ma, 10.349
Radiology
Brain,
CT,
13.2054. studies,
13.2065, 13.1214
#{149} Encephalitis, #{149} Viruses,
1990;
13.1211
#{149} Brain,
13.2066, #{149} Brain,
13.253 13.2065,
MD
#{149}
by Human Virus: CT and Manifestations with Pathologic Correlation’
N
EUROLOGIC
infec-
findings
white #{149} Kaposi
travenous
occurs
may
be related
to AIDS-
associated neoplasms tic infections, a more
on opportuniscommon abnon-
mality
subconticab
is progressive
mentia, thy
termed
(1,2).
AIDS
This
is
encephalitis.
AND
METHODS
Subjects Review of neuropathologic records over a 6-year period (1982-1987) at our institution revealed 140 autopsies on patients
with
AIDS.
HIV
alone
or together
with other infectious agents or lymphoma was identified in the brains of 39 of scans,
patients. MR
Premortem
images,
for the
cranial
or both
evaluation
had
CT
been
ob-
symptoms
sarco-
tients and constitute the focus of this meport. The study population included 23 men and one woman with a mean age of 37 years (range, 27-67 years). Risk factors included homosexuality in 20 patients, in-
13.2066
175:185-191
signs
in
24 of these
patient
remaining
of a patient
with
AIDS was diagnosed to the criteria of the
Control
School
of
vember C
RSNA,
22,
Medicine,
1989; 1990
UCSD
accepted
of Radiology of Pathology Medical
January
and
the Magnetic Resonance (M.R.G., C.A.W.), University
Center,
225
9. 1990.
Address
Dickinson
reprint
St.
San
pa-
requests
Institute (H.S.C., GAP., of California, San Diego Diego,
CA
to GAP.
92103.
Received
been
the
In every
clinically Centers
wife
case,
according for Disease
(12,13).
Fourteen aminations
patients underwent and 22 underwent
MR cxCT studies.
The hospital and outpatient records, in addition to autopsy data, were available with
these
the imaging
cal Systems,
data
were
findings
All MR studies 1 .5-T superconducting
were
correlated
of the CNS.
performed with a magnet (GE Mcdi-
Milwaukee).
In all patients,
proton densityand T2-weighted images in the axial plane were acquired with the following parameters: 2,000/25, 70 (two excitations) or 3,000/30, 80 (one excitation) (repetition time [TR] msec/echo
time [TE] msec). In six patients, the initial sequence was repeated in the coronal plane and/or a Ti-weighted sequence (600/25, two excitations) was performed in the sagittal plane. Section thickness was 5 mm, with an intersection gap of 2.5 mm in all examinations. The field of view was 20 or 24 cm, with an acquisition and display matrix size of 256 X 256. In the 22 patients who underwent CT, axial, 10-mm-thick, contiguous sections from the fonamen magnum to the vertex
were
obtained
nous
administration
(150
mL
before
and after
intrave-
of contrast
of Conray-60
material
[Mallinckrodt
Mcd-
ical, St Louis]) in all patients. Three of the authors reviewed metrospectively all of the MR and CT studies
findings were divided into four categonies: (a) Atrophy-Dilatation of the yentniculan system out of proportion to the sulci and dominantly
cisterns central
enlargement
ficiency
was interpreted atrophy (there
of the
Abbreviations: From the Department J.R.H.) and the Department I
had
AIDS.
in
in one. The
without knowledge of distribution of the pathologic abnormalities. The radiologic
of neunologic
mat-
and
homosexuality
one
for all patients;
Although
many reports illustrate findings of AIDS-associated tumors and infections at computed tomography (CT) or magnetic resonance (MR) imaging (3-9), few depict the imaging manifestations of HIV encephalitis (10,1 1). In this study, we describe the cranial appearance of 24 patients with histopathologically confirmed HIV encephabitis at MR imaging and CT and correlate imaging results with clinicab and pathologic data. PATIENTS
use and
transfusion
Neuroradiology
encephabopathy
(HIV)-HIV
drug
two, and blood
de-
encephabopa-
thought to be due to infection within the central nervous system (CNS) by human immunodeficicncy virus
tamed
13.2074
dysfunction
frequently in patients with acquired immunodcficiency syndrome (AIDS). Although focal neurobogic
these terms:
R. Grafe,
Caused
To determine the computed tomographic (CT) and magnetic resonance (MR) imaging manifestations of central nervous system (CNS) infection by the human immunodeficiency virus (HIV), the authors analyzed the results of imaging of the CNS in 24 patients with HIV encephalitis confirmed at autopsy. Careful pathologic correlation demonstrated that neither CT nor MR imaging enabled detection of microglial nodules with multinucleated giant cells, the hallmark of HIV encephalitis seen in all 24 affected patients. The most common abnormality observed on images of the CNS was atrophy, demonstrated in 18 patients. Demyelination and vacuolation of white matter tracts accompanying severe HIV infection caused hypoattenuation on CT scans and hyperintensity on T2-weighted MR images. These lesions had no mass effect. MR imaging was more sensitive than CT in the detection of lesions caused by HIV or other superimposed infectious agents. Although it is often difficult to attribute any radiologic appearance to a single etiologic agent in patients with acquired immunodeficiency syndrome, the combination of atrophy and symmetric, periventricular or diffuse white matter disease suggests HIV encephalitis. Index
#{149} Marjorie
AIDS
syndrome.
= cerebrospinal munodeficiency
ed giant
cell,
PML
virus,
horns
acquired
CMV
CSF No-
temporal
as prewas no
to
immunode-
cytomegalovirus.
fluid, HIV MNGC
progressive
=
human multinucleat-
im-
multifocal
leukoencephalopathy.
185
Figure
1.
Axial,
noncontrast
CT images
from 33-year-old man with severe CNS HIV encephalitis only, obtained 3.5 months before death, demonstrate a greater degree of ventricular
dilatation
ing (arrows), phy
greater
suggest sulci
(V) than sulcal with central peripheral atrophy.
widenatro-
compatible than
hydrocephalus); exceeding
prominence
that
of
of the ventricles
was
interpreted as predominantly peripheral atrophy. The location of parenchymal volume loss (supra- and/or infratentonial) was recorded also. (b) White matter disease-Abnormal WM density or signal intensity without mass effect. (c) Gray matter disease-Abnormal gray matter density or signal intensity without mass effect. (d) Mass lesions-Lesions associated with edema, compression of adjacent structures, or both. Neuropathology Every brain was sectioned in the coronab plane at i-cm intervals. Routine regions (frontal, temporal, parietal, and occipital cortex; basal ganglia; thalamus; pineal gland; cerebellum; midbrain; and medulla), in addition to grossly abnormal regions, were sampled and paraffin em-
bedded
for histologic
several
cases,
examination.
additional
a.
b.
In
grossly
unre-
markable but radiologically abnormal mcgions were sampled for histologic study. All sections were stained with hematoxybin and eosin and examined for histopathologic features. Sections of temporal cortex cortex,
(including hippocampus), panietal basal ganglia, and thalamus were
stained for the HIV envelope glycoprotein 4i with methods described previously (14). Furthermore, sections for histopathologically abnormal regions were stained
with
(a) HIV
4i;
hematoxybin
for which a polycbonal sciences, Wamnington, (c) herpes
simplex antiserum
lycbonal
and
for
eosin
(CMV), antiserum (PobyPa) was used (15);
(b) cytomegabovirus
virus, (Dako,
for which Carpentenia,
a po-
was used; (d) Toxoplasma organisms, for which a primary cheetah antiserum was used; and (e) papovavirus (which Calif)
causes
progressive
mubtifocab
leukoence-
phalopathy
[PML]), for which in situ hybridization with a biotinylatcd probe to the JC virus was used (16). HIV encephalitis was graded mild, moderate, or severe according to the number of infected cells per focus and the
number
of foci
per
slide.
The
pres-
ence of rare HIV-positive cells (fewer than one cell per 10-power field) with or without microglial nodules was designated mild infection. The finding of occasional HIV-positive cells (more than one cell per ten 10-power fields but fewer than one cell per 10-power field) with microgliab
nodules
with
or without
multinu-
cleated giant cells (MNGCs) was graded moderate. The presence of frequent HIVpositive cells (at least one cell per 10power field) with frequent microglial nodules and with MNGCs was graded severe. 186
Radiology
#{149}
a. Figure cephalitis
b. 2.
Axial, only,
matter lucency (b) compatible
noncontrast obtained 4.5
(L) without with
CT images from 30-year-old man months prior to death, demonstrate
mass
HIV-induced
effect
within
demyelination
RESULTS Clinical
and
Laboratory
Data
The diagnosis of AIDS was establishcd from 2 months to 3 years prior to death in each patient by means of well-established clinical criteria (12,13); moreover, the serum of every patient who presented with symptoms after 1984 tested positive on the enzyme-linked immunosorbent assay and Western blot test for HIV antibodies. The most common initial manifestation of AIDS was Pneumocystis carmu pneumonia (13 patients [54%]),
the forceps seen
with severe CNS diffuse, symmetric,
minor
(a) and
corona
HIV
enwhite
madiata
at autopsy.
followed by opportunistic infections (eg, candidiasis, CNS toxoplasmosis, or disseminated infection from Mycobacterium organisms, Cryptococcus organisms, herpes simplex virus, or CMV) (seven patients [29%]), Kaposi
sarcoma
(three
lymphoma Neumobogic
(one
patients
[12%]),
patient symptoms
[4%]). and
were
the
in six
patients
(25%):
seizures
in one,
four, sis
presenting
in one.
ic dysfunction population
cranial
The
spectrum
occurring included
ncuropathics,
features
and signs
of AIDS
dementia
and
in
hcmipamcof neurobog-
in our peripheral
study and
myclitis,
ataxia, April
1990
b. Figure
3.
Images
sity-weighted out mass effect
trum
of 27-year-old
(b, c) images within the
semiovale
(arrows,
obtained pons and
(patient
1 month prior middle cerebellar
in 15 Patients Correlation Focal
Studied
1
Not done White
3 4
ease; focal (temporal lobe and pons) NFL NFL*
5
Not
6
Deep
matter
dis-
done
7
NFL
8
White
severe
CNS
to death demonstrate peduncles (arrows,
3 Months
matter
or Less
Supratentorial Atrophy
CT
2
disease, tamen)
1) with
Lesions
MR Imaging
gray
c.
14 in Table
HIV encephalitis
symmetric a), forceps
regions major,
before
Death:
CT/MR Imaging Interval
HIV Encephalitis
NFL
Severe
NA
Mild
NFL*
None
3 d
Severe
Not done NFL
Mild Mild
NA
Mild
1 mo
Moderate
NFL
Mild
NA
Moderate
Moderate
NA
Severe
Moderate
3 wk
Moderate
Moderate
5 d
Severe
Not
done
focal (puNFL*
matter
White
dis-
matter
dis-
9 10
ease; diffuse, confluent (supra- and infratentonial) Not done Not done
ease,* focal (occipital lobe, right cerebellum) Normal Normal
None None
NA NA
Severe Severe
11
NFL*
NFL
Mild
1.5 mo
Severe
12
NFL
NFL*
None
10 d
Mild
13
White matter disease*; multifocal
NFL
Mild
1 wk
Moderate
NFL*
Mild
1 mo
Severe
NFL
Moderate
NA
Mild
14
(frontal lobes) White matter disease; diffuse (su-
pra- and infraten-
15 Note-NFL * Initial
tonal)
and
matter
disease
gray
(un-
ifocal, putamen) Not done =
only.
Axial
of abnormally forceps minor,
no focal
examination
bowel
and
lesions,
NA
bladder
not applicable.
=
incontinence.
(75%) suffered encephabopathy
(somnolence, decreased attention span, slowed speech, flat affect, paired memory, disorientation, difficulty in word finding) for 175
proton
den-
months) revealed (a) normal results in 1 1 patients (seven of whom had severe HIV encephalitis), (b) a mild increase in CSF total protein levels without pleocytosis in eight patients with moderate or severe HIV encephalitis (with on without supcmimposed CMV, PML, or toxoplasmosis), and (c) a marked increase in CSF total protein levels (greater than twice the upper limit of normal in our babonatory) or pleocytosis in three patients with HIV and superimposed active CNS toxoplasmosis, cryptococcosis, or lymphoma (proved with cultures
or biopsy
performed.
Eighteen patients from a progressive
Volume
(a) and
increased signal intensity withbasal ganglia (arrows, b), and cen-
sy in most patients. However, three patients with severe HIV cnccphalitis had no dementia, and one with seveme HIV encephalitis and PML had only mild dementia. Measurements of total protein and glucose bevels as well as cellular analysis of 39 cenebrospinal fluid (CSF) samples from 22 patients (obtained 1 week to 24 months prior to death; the mean pmemontcm interval was 6
#{149} Number
1
imand 3-12
months
prior
to death.
Most
patients
in this group were unable to care for themselves for varying periods of time prior to death. Theme was good correlation between the severity of dementia and severity of HIV encephalitis at autop-
samples).
Routine
and viral (not including tunes of all specimens growth.
Pathologic and
T2-weighted
c).
Table 1 HIV-induced Lesions Radiologic-Pathologic
Patient
man
HIV) revealed
bacterial cubno
Data
The brains of our 24 study patients demonstrated diverse pathologic findings consistent with previously described
subacute
encephalitis
20). The
distinct
histopathobogic
pcarance reliable
of HIV encephalitis differentiation from Radiology
(i7-
appermits other vi#{149} 187
nab encephalitides. CNS inflammation affected predominantly the white matter and deep nuclear structunes; infiltration of the parcnchyma and penivascular spaces by bymphocytes and macnophages was seen. Reactive astrocytosis was demonstrated frequently. In severe infections, MNGCs were identified within the mononuclear infiltrate. Multiple small nodules composed of microgha, astrocytes, and occasional macrophagcs (microglial nodules) were often present, most commonly in the
cortex
and
basal
ganglia
and
infre-
quently in the white matter. Focal aneas of demyclination on diffuse white matter vacuolation were also seen occasionally. HIV encephalitis was disclosed in seven patients (29%) without other superimposed infectious on neoplastic processes. Coexisting disease was found in the remaining 17 patients (71%): CMV infection in nine (38%), CMV infection and toxoplasmosis in three (12%), PML in two (8%), lymphoma in two (8%), and CMV infection and cryptococcosis in one (4%). The severity of HIV encephalitis was graded mild in six patients (25%), moderate in five patients (21%), and severe in 13 patients (54%). The vast majority of HIV lesions detected by means of histologic cxamination were not seen at gross inspection of the brain. No focal mass lesions could be attributed to HIV infection alone.
a.
b.
d.
C.
Figure
4.
alitis,
Images
CMV
weighted
Imaging
Studies
Twenty-two patients were exammed with CT; 12 underwent multiple examinations. MR imaging studies were performed in 14 patients, four of whom underwent serial scanning. The interval between the final pncmortem MR imaging study and autopsy was less than 3 months in eight patients, 3-6 months in four patients, and 12-18 months in two patients. The interval between the final premortem CT study and autopsy was less than 3 months in 13 patients, 3-6 months in four patients, and 7-15 months in five patients. In 15 paticnts, the final premortem imaging study was performed 3 months or less before death. Their MR. CT, and pathologic findings attributable to HIV arc presented in Table 1 . Because of the relatively slow progmession of subacute HIV encephalitis, we believe that a 3-month window accommodates accurate madiobogic and pathologic correlation. A detailed description of imaging studies 188
Radiology
#{149}
from
38-year-old
encephalitis,
and
(b) images
man
(patient
toxoplasmosis.
demonstrate
Axial
toxoplasmosis
2 in
Table
1) with
T2-weighted
abscesses
severe
(a) and
with
surrounding
CNS
proton
HIV
enceph-
density-
edema
and
mass effect in both basal ganglia (arrow, a, b). Lower 12-weighted sections (c, d) show two additional lesions caused by a combination of HIV (predominantly) and CMV in the medial temporal lobe (arrow, c) and the right middle cerebellar peduncle (arrows, d). The temporal lobe lesion had a hypenmntense center, isointense rim, and surrounding hypenintense edema on 12- and proton density-weighted images. This lesion was indistinguishable from the toxoplasmosis seen also in this patient. The infratentorial lesion was smaller, poorly defined, and hyperintense; minimal effacement of the adjacent fourth ventricle is seen (d).
in all
24 patients
follows.
HIV encephalitis-Seven patients had encephalitis caused exclusively by HIV. Cerebral atrophy (moderate on severe) was seen in five patients (central atrophy greater than peniph-
cral
atrophy
in four
trab
atrophy
equivalent
atrophy in one crab, symmetric
patients
and
ccn-
to peripheral
patient) white
(Fig matter
1). Bilatdisease
was seen in three patients: diffuse white matter disease in two patients and third
pemivcntmicubam patient (Figs
had normal CT of death despite
encephalitis sions
in any
were
patient
disease in the 2, 3). Two patients
scans within 3 weeks severe, diffuse HIV
at autopsy. seen
on
CT
in this
No mass or MR
group.
be-
images
HIV and CMV encephalitis-Of nine patients with HIV and CMV encephalitis, eight had mild or modemate CNS atrophy (central atrophy predominant in five, peripheral atmophy predominant in two, and central atrophy equivalent to peripheral atmophy in one); two patients had a singbc small high-attenuation mass be-
sion
caused
by a hemorrhagic
infarc-
tion. These infanctions were attributed to CMV because pathobogic examination disclosed a focal accumubation of cytomegalic cells. Two patients had mubtifocab white matter infamcts, also attributed to CMV, and a single focus of deep gray matter disease attributed to HIV. One patient with HIV and CMV encephalitis April
1990
on gray matter disease in two patients, multiple foci of white matter disease in four patients, and diffuse white matter disease in one patient.
b. Figure 5. Images from 44-year-old man with severe HIV encephalitis and PML. Coronal T2-weighted (a) and Ti-weighted (b) images obtained 5.5 months prior to death demonstrate asymmetric white matter disease without mass effect within the region of the parietooccipital junction (arrows). At pathologic examination, these lesions were shown to have been caused by HIV and PML. Dilated occipital horns (v) are seen also.
also in this patient smaller mass lesion
Table 2 Distribution of HIV Lesions in Imaging Studies in All 24 Patients No.
White matter disease Diffuse (confluent,
supra-
bi4
Peniventricular
(lateral
ventricle)
1
Focal and/or multifocal (supra- and/or infratentonial) Deep gray matter disease Putamen Note-Only nantly due
of
and/or
infratentonial)
4
lymphoma.-There
lesions exclusively or predomito HIV are included in this table.
with
simultaneous
by HIV
alitis.-Thmcc
patients
had
this prior enceph-
simulta-
neous CNS infection with HIV, CMV, and Toxoplasma organisms. Mild or moderate atrophy was seen in two patients (greater centrally in one patient and peripherally in the other). Mass lesions attributed to toxoplasmosis (Fig 4) were seen in all. One patient had diffuse white matter disease, attributed to HIV, seen at MR imaging. In one patient, two mass lesions due to a combination of HIV and CMV were noted in the right tempoma! lobe and night middle cerebellar peduncle. The temporal lobe lesion had an isointense periphery and hypenintense center (compared with white matter on proton densityand T2-weightcd images) and induced
surrounding
high-signal-intensity
edema; this lesion abbe from Toxoplasma
Volume
175
#{149} Number
was
indistinguish-
abscesses 1
density-
coinfection (Fig HIV encephalitis
2
had a normal CT scan; however, scan was obtained 10 months death. HIV, CMV, and Toxoplasma
proton
weighted images. HIV encephalitis two patients with and PML demonstrated (moderate, central peripheral atrophy and severe atrophy dominance in the both had bilateral white matter disease
Patients
Lesions
lateral,
on both
seen
to
was phy ease lesions two
mild
and
(Fig 4). The other, was hypemintense
and
T2-
and PML.-Thc HIV encephalitis CNS atrophy atrophy equal to in one patient with central pmeother). In addition, but asymmetric in regions of
5). and primary CNS were two patients CNS involvement
lymphoma.
Lymphoma
associated with mild central atnoand unifocab white matter disin one patient and multiple mass in the other patient. In these patients, HIV encephalitis was
without
any
radiobogic
manifes-
tations. HIV, CMV, and Cryptococcus encephalitis.-Mild cortical atrophy was the only abnormality identified in the one patient with this multipleagent-induced encephalitis. Table 2 summarizes the distnibution of lesions exclusively on predominantly due to HIV and identified on CT scans, MR images, on both in all 24 patients. Comparison of MR imaging and CT.Both MR imaging and CT were performed in 12 patients. MR imaging was more sensitive for the detection of CNS abnormalities in eight pa-
tients:
MR
imaging
enabled
detection
of additional lesions caused by HIV in four patients, toxoplasmosis in three patients, and lymphoma in one patient. The additional HIV lesions detected with MR imaging appeared as solitary foci of white matter and!
MR and CT were equably sensitive in the remaining four patients: one patient with diffuse asymmetric white matter disease and three patients with atrophy only. When compared with Ti-weighted and proton density-weighted images, the T2-weighted images were superior for detecting lesions, delineating their margins, and separating a central mass from surrounding edema (Fig 4).
HIV
Encephalitis: Imaging Pathologic Correlation
and
Several abnormalities revealed by imaging of the CNS were confidently attributed to HIV infection at pathologic examination. In some patients, the abnormalities were undoubtedly accentuated by coexistent infections: Atrophy.-Supratcntonial atrophy occurred in 18 patients. In 12 patients, central atrophy revealed by CT, MR imaging, on both exceeded peripheral volume loss and conrebated well with the neuropathobogic distribution of the HIV lesions. In the autopsy material, HIV affected predominantly the deep white matten with lesser involvement of the subcortical white matter and relative sparing of the cerebral cortex. Infratentonial atrophy was seen after imaging studies in 12 patients in conjunction with atrophy of the cerebral hemispheres. Serial studies (CT, MR imaging, or both) oven an interval of V/2-ii months in 16 patients mcvealed progression of atrophy in 10 of them. White matter disease.-Extcnsivc, nearly symmetric, deep white matter disease associated with severe HIV encephalitis was seen by means of CT
on MR imaging with tonal seen
moderate atrophy. as regions
in four
of 1 1 patients
or marked supratenThese lesions were of bow attenuation
on
CT scans or as regions of high signal intensity on proton densityor T2weighted MR images relative to normal panenchyma. HIV also caused smaller (less than 1 cm) white matter or deep gray matter (basal ganglia) lesions evident only as focal regions of hypenintensity on T2-weighted images. Posterior fossa white matter disease induced by HIV (seen in two patients) was patchy on was merged with abnormally hypcnintcnsc fiber tracts descending from the cerebral cortex. Isolated infmatcntonial white Radiology
#{149} 189
matter disease was not seen. Abnormalities detected with imaging of the CNS correlated well with pathologic findings. Nevertheless, many HIV lesions seen at pathologic examination were not detected with MR imaging and CT. For example, cortical microglial nodules seen at autopsy in eight patients remained undisclosed by imaging of the CNS. In addition, although cerebral white matter and/or deep gray matter HIV lesions were present at autopsy in all 24 patients in our study, such lesions were demonstrated by imaging of the CNS in only eight patients. In only two of eight patients with brain stem or cerebellan HIV lesions at pathobog-
ic examination
did
MR
imaging
on
CT show an abnormality in the comesponding location. Even when positive, both MR imaging and CT were grossly inadequate in showing the true extent of parcnchymal disease in all instances.
DISCUSSION The neurologic AIDS arc protean cal series report
39% (21,22).
complications and common; an incidence
In a significant
of cliniof 31%-
propon-
tion (10%) of AIDS patients, ncunologic abnormalities arc the presenting complaint (22). Recently, much attention has been directed to the AIDS-dementia complex that afflicts 36%-75% of patients
during
the
course
of their
illness
(1,2,21-24). The results of extensive research suggest a direct role of HIV in the pathogenesis of this prognessive encephalitis: (a) The virus has been recovered from CSF, the brain, spinal cord, retina, and peripheral nerve of patients with HI V-related neurobogic syndromes (25-27); (b) HIV DNA and RNA sequences and HIV proteins have been identified in the CSF of such patients by antigen-capture techniques (14,2830); (c) HIV-specific immunoglobulin and antigen have also been detected in CSF (25,31); and (d) HIV within the CNS has been identified at momphobogic examination of patients with AIDS-related cncephabopathy (14,28-30,32). Studies of the neuropathologic pcct of AIDS have shown a wide
spectrum
of abnormalities
as-
in more
than 75% of autopsies; approximately 33% of the cases demonstrate cvidence of viral encephalitis. Prior to the advent of immunocytochemical and nucleic acid probes, it was impossible to define the specific viral agent responsible for the CNS in190
Radiology
#{149}
flammation.
These
techniques HIV and
new
laboratory
often demonstrate CMV in regions
cephalitis.
Microglial
both of viral en-
nodules
and
pemivascuban MNGCs accompanying gliosis of deep white and gray matter are most commonly associated with HIV infection alone or in combination with CMV. Several reports mdicate that the amount of virus (HIV or CMV) present in the CNS does not correlate with the severity and extent of the inflammatory changes (14). In our series, MR imaging and CT depicted reliably the CNS volume
loss
that
may
accompany
HIV
infec-
tion. The predominantly central vobume loss shown in most affected patients correlates well with the pmeponderance of deep white and gray matter HIV lesions seen at pathologic examination. Nevertheless, no come-
lation
was
found
between
the
degree
of volume boss and the severity of HIV infection in our series; indeed, in several patients with severe HIV
encephalitis at pathologic examination, no atrophy was demonstrated by imaging studies performed within 3 months of death. In addition, there was no correlation between the severity of dementia and the severity of atrophy or white matter disease. Both imaging techniques were insensitive for the detection of microgliab nodules induced by HIV or CMV. Cortical involvement by either virus was not seen on CT scans or MR images of any patient. Only the white matter regions most severely involved with HIV infection were detected with radiobogic studies. In our study, MR imaging enabled iden-
tification lesions
of more infratentonial than reported previously
HIV (10).
The use of a middle-field-strength (0.5-T) magnet and 10-mm-thick
the
microglial
The
long
interval
between the MR imaging
in four
of our
nodule (7-18
of our study. Imaging this group of patients
This
observation
of antigen
for
fied with methods).
each
organism
of the high polymicrobiab
sions
where
multiple
in
that
in the
it is very attnifound etiobogic extends to of CNS be-
agents,
evaluation
AIDS-rebated
particu-
mass
is more lesions
are
matter
disease,
mass lesions, involvement
MR
sensitive than CT. If detected by means of HIV infection alone presence of diffuse
imaging studies, is unlikely. The white
of suspected
encephabopathy,
imaging
in addition
to
suggests superimposed with HIV. Asymmetric
white matter disease predominating in the pamietooccipitab regions can occur with either PML alone or PML combined with HIV infection. The diagnosis of HIV encephalitis can be
suggested
by the
combination
trab atrophy and tnicubam or diffuse ease. U
rum the
symmetric white
was San
Primary
donated Diego
by
Kurt
of cen-
penivenmatter dis-
cheetah
antise-
Benirschke,
MD,
and
Zoo.
References 1.
Navia AIDS tunes.
2.
Navia
BA,
Jordan
as the
3.
BD,
BA,
Price
RW.
presenting Neuro! WM,
1987;
findings. MJD,
ma!
5.
Post Central quired spective
contrast 1985;
MJD,
fea-
acquired
immucomplex
manifestation
of
virus
infection.
44:65-69.
Brant-Zawadzki
M.
Acquired
syndrome: Radiology
neuronadio1983;
Kunsunoglu
149:485-
SJ. Hensley
Chan JC, Moskowitz Cranial CT in acquired syndrome: spectrum AJNR
The
dementia
or sole
immunodeficiency
4.
The
immunodeficiency
Arch
491. Post
RW.
I. Clinical 19:517-524.
syndrome
Kelly logic
Price
dementia complex. Ann Neurol 1986;
nodeficiency
fumthcm
the insensitivity of MR imaging and CT in the evaluation of HIV cncephabitis, especially during its early phase. Although a large number (7i%) of
of nco-
larby HIV and CMV, are found in individuab foci and even in individual cells (33). In conclusion, our data indicate
months)
of the CNS was nommab.
frequency and/on
plastic conditions in AIDS, often difficult to confidently bute radiobogic abnormalities in such patients to a single agent. The same difficulty the pathologic evaluation
human
CT on death
(identi-
immunocytochemicab
Because coexistent
is a limitation
emphasizes
the relative contribution of process to any given lesion.
Such quantitative assessment was carried out by correlating the severity of encephalitis with the amount
(10).
final prcmomtem examination and
patients
grade each
Acknowledgments: 5cc-
tions for some of the examinations in the previous study (10) may account for this discrepancy. The regions of focal on diffuse white matter abnommality detected occasionally by means of imaging studies most often reflects secondary demyclination and gliosis rather than the primary CNS
lesion,
our patients had additional infectious agents or neopbasms that were superimposed on a background of HIV encephabitis, careful pathologic assessment allowed us to separate and
GT,
LB.
Hoffman TA. immunodeficiency of diseases and opti-
enhancement
techniques.
6:743-754.
Sheldon
JJ,
Hensley
CT,
et a!.
nervous system disease in acimmunodeficiency syndrome: procorrelation using CT, MR imag-
April
1990
ing,
and
pathologic
studies.
Radiology
virus
6.
Levy
RM, Rosenbloom
Neuroradiologic
7.
8.
in AIDS: a review of 200 cases. AJNR 1986; 7:833-839. Gilleux MH, Steiner RE, Young JR. MR imaging in progressive multifocal leukoencephalopathy. AJNR 1986; 7:1033-
17.
1035.
18.
Sze
G, Brat-Zawadzki
Ramsey tions
1 1.
12.
of AIDS:
CT
MR
findings.
AJR
19.
RM, et al. HIV
encephali-
tis and meningitis. Olsen WL, Longo
AJR 1988; 151:373-380. FM, Mills CM, Norman
D.
disease
White
matter
ings at MR imaging. 169:445-448. Centers for Disease the case definition
in AIDS:
Radiology
20.
find-
for
Revision immuno-
national
MMWR
21. of
report-
1984;
34:373-
22.
syndrome.
1987; 36(suppl
1):3-l5. Wiley CA, Schrier RD. Nelson JA, Lampert PW, Oldstone MB. Cellular localization of human immunodeficiency virus the
brains
munodeficiency
n’s in AIDS
virus
and
encephalitis.
24. im-
W,
Burge
PC,
Snider
Levy
WD,
Neurol
1986;
Simpson
DM,
#{149} Number
1
de la Monte Hirsch MS. cephalomyelitis to HTLV-III 37:562-569.
27.
Pomerantz
RJ,
SM,
Infection
in
CE, Posner
JB.
RM,
Bredesen
DE,
Ann
Neurol
1988;
28.
GM,
quired
immune
29.
30.
31.
ML.
syndrome:
J Clin Pathol
virus
type I. N
Harper
infection
ME,
Hahn
BH,
in brains
et a!.
of children
with AIDS encephalopathy. 1985; 277:177-181. Gabuzda DH, Ho DD, de Ia Monte SM, Hirsch MS. Rota TA, Sobe! RA. Immunohistochemical identification of HTLV-III antigen in brains of patients with AIDS. Ann Neurol 1986; 20:289-295. Pumarola-Sune 1, Navia BA, Corolon-
C, Cho ES, Price RW.
in the brains of patients dementia complex. Ann 21:490-496. Resnick L, diManzo-Veronese synthesis tients
Intra
32.
HIV antigen
with the AIDS Neurol 1987; F, Schup-
blood-brain-barrier
of HTLV-III-specific with
neurologic
IgG symptoms
ed with AIDS or AIDS-related EngI J Med 1985; 313:1498-1504.
of
a
de La Monte retina by hu-
adults
Cardo
CD, in ac-
DR.
of the
Science
33.
Am
Shaw
and
23(suppl):527-
SL, Petito CL, Urmacher JB. Subacute encephalitis
Kuritzkes
1987; 317:1643-1647.
bach J, et a!.
533. Nielsen Posner
study.
J Med
HTLV-III
Neurologi-
Rosenblum
deficiency
et a!.
Engl
5, Gold
immune deof 50 patients.
SM, Ho DD, Schooley RI, Richardson EP. Subacute enof AIDS and its relation infection. Neurology 1987;
man immunodeficiency
19:525-535.
Nielsen
in paassociat-
complex.
N
Gyorky F, Melnick JL, Gyorky P. Human immunodeficiency virus in brain biopsies of patients with AIDS and progressive encephalopathy. J Infect Dis 1987; 155:870876. Nelson JA, Reynolds-Kohler C, Oldsone
MBA, Wiley brain
cells
gy 1988;
CA. in patients
HIV and CMV coinfect with
AIDS.
Viro!o-
165:286-290.
1984;
82:678-682. 25.
Ho
DD,
lation id and
TR,
Schooley
of HTLV-III
Rota
from
neural syndromes
tissues
RI.
et a!.
cerebrospinal
of patients related
immunodeficiency syndrome. Med 1985; 313:1493-1497.
175
26.
Pro-
Neurological manifestations of the acquired immunodeficiency syndrome (AIDS): experience at UCSF and review the literature. J Neurosurg 1985; 62:475495. Price RW, Sidtis J, Rosenblum M. AIDS dementia complex: some current ques-
rologic
Volume
et a!.
leukoencephalopathy
Ann
postmortem
J Pathol
C, Nelson J. virus and JC
progressive
patients with acquired immune deficiency syndrome (AIDS). Acta Neuropathol (Berl) 1985; 68:333-339. Petito CK, Navia BA, Cho E-S, Jordan BD, George DC, Price RW. Vacuolar myelopathy pathologically resembling subacute combined degeneration in patients with the acquired immunodeficiency syndrome. N EngI J Med 1985; 312:874-879. Anders KH, Guerra WF, Tomiyasu U, Vinters HV. The neuropathology of AIDS. Am J Pathol 1986; 124:537-558. Navia BA, Cho E-S, Petito CK, Price RW. The AIDS dementia complex. II. Neuro-
tions.
cytomegalovi-
Am
1988; 133:73-81. Wiley CA, Grafe M, Kennedy Human immunodeficiency
23.
of acquired
immune deficiency syndrome patients. Proc Natl Acad Sci 1986; 83:7089-7093. Wiley CA, Nelson JA. Role of human
with
cal complications of acquired ficiency syndrome: analysis Ann Neurol 1983; 14:403-418.
Control. Revision of case definition for ac-
immunodeficiency
within
P, Lang
diffuse
JWM, Metroka
ing: United States. 374. Centers for Disease
infection
Kleihues
pathology.
1988;
Control. of acquired
syndrome
MMWR
16.
D,
of 22:42-53.
CNS complica-
Post MJD, Tate LG, Quencer CT, MR and pathology in
quired
15.
GK.
and
CDC surveillance
14.
Norman
patients
leukoencephalopathy. Acta (Berl) 1988; 76:338-346.
gressive
151:449-454.
deficiency
13.
M,
The neuroradiology Roentgenol 1987;
RG, Geremia
1988;
10.
LV.
findings
Newton TH. AIDS. Semin
9.
5, Perret
in AIDS
multifocal Neuropathol
1986; 158:141-148.
flu-
with
to the
Isoneu-
acquired
N EngI
Radiology
191
#{149}