Thomas

S. Dma,

Primary versus

MD

Central Toxoplasmosis

The imaging studies of 16 patients with acquired immunodeficiency syndrome (AIDS) and proved primary central nervous system (CNS) lymphoma were reviewed. All studies included computed tomography (CT); six also included magnetic resonance (MR) imaging. A periventricular lesion was seen in 50% of patients. At least one such lesion cxhibited subependymal spread or ventricular encasement in 38%. One-third of lesions in three of five patients who underwent nonenhanced CT were hyperattenuating. Five lesions were at least in part hypointense on T2-weighted MR images. The specificity of these findings was evaluated with a similar review of the imaging studies in 28 patients with AIDS and proved toxoplasmosis. Only 3% of lesions were periventricular. None exhibited subependyntal spread or encasement. None were hyperattenuating on nonenhanced CT scans. Similar findings in other CNS lesions in AIDS patients could not be found in the literature. A focal enhandng mass with subependymal spread on CT or MR Images and hyperattenuation at nonenhanced CT were the most reliable features in distinguishing between primary CNS lyinphoma and toxoplasmosis in AIDS patients. Index terms: Acquired immunodeficiency syndrome (AIDS) #{149} Brain neoplasms, CT,

10.1211, 10.34 #{149} Brain neoplasms, MR studies, 10.1214, 10.34 #{149} Lymphoma, CT, 10.1211, 10.34. Lymphoma, MR studies, 10.1214, 10.34 #{149} Toxoplasmosis, 10.2074 Radiology

1991; 179:823-828

Nervous

T

System in AIDS’

Lyinphoina

of central nervous system (CNS) lymphoma in patients with the acquired immunodeficiency syndrome (AIDS) has inHE incidence

creased.

Malignant

lymphoma

is sec-

ond in frequency only to toxoplasmoMs among CNS mass lesions in AIDS patients (1). However, reports of the computed tomographic (CT) appearance of primary CNS lymphoma in AIDS patients vary; some authors describe focal lymphoma mass lesions as being similar to those in patients without AIDS, while others describe differences (2,3). Also, while the CT appearance is generally described as an enhanced mass or multiple masses exhibiting diffuse or

ring tion

enhancement

with

a predilec-

for the corpus callosum, basal ganglia, and periventricular areas, the reported frequency of such findings varies (4-14). This variance is at least in part due to the relatively small number of cases of lymphoma, even in reviews of large numbers of CNS lesions in patients with AIDS. The spectrum of CT appearances often includes normal CT scans as well as nonspecific

findings

of atrophy

and white matter area of low aftenuation in a number of proved cases, leaving a smaller number of focal lesions from which to tabulate findings. On CT scans, toxoplasmosis may also appear as solitary or multiple, ring- or nodular-enhancing masses (3,9,15-17). Delayed double-dose contrast material-enhanced CT increases the detection rate and conspicuity of lesions, and magnetic resonance (MR) imaging further increases sensitivity in lesion detection (1 1,13,18). Imaging studies, however, remain nonspecific. It is therefore

3 From the Department of Radiology, George Washington University Medical Center, 901 23rd St, NW, Washington, DC 20037. Received October 9, 1990; revision requested November 26; revision received February 13, 1991; accepted February 25. Address reprint requests to the author. 0 RSNA, 1991

generally accepted that primary CNS lymphoma is indistinguishable from toxoplasmosis (5,12,19). We have encountered several cases of focal lymphoma mass lesions that were subependymal in location and surrounded and constricted (encased) a portion of the ventricle. A subependymal location of enhancing focal lymphoma mass lesions has been described in both nonimmunocompromised and AIDS patients, but without documentation of the relative prevalence

of this

finding

or differ-

entiation lesions

from other periventricular (8,11). The finding of ventricular encasement has been illustrated without comment as to the frequency or potential significance of the finding (1 1,13,20). Ventricular encasement has been described on CT scans in two of 32 patients by Goldstein et al(14). The purpose of our study was to review our experience with the neuroimaging of primary CNS non-Hodgkin lymphoma and toxoplasmosis in the AIDS population and to review reported cases to further define their imaging characteristics. To be specific, we wished to determine the preyalence of subependymal spread and ventricular encasement in lymphoma and the prevalence, if any, of this appearance in toxoplasmosis. PATIENTS

AND

METHODS

The neuroimaging studies and medical records of patients with a confirmed diagnosis of AIDS were reviewed. Of 59 patients with diagnoses of AIDS and lymphoma, 41 had systemic non-Hodgkin lymphoma with or without CNS involvement. Thirty-one of these patients had extranodal

lymphoma

(including

four

with

CNS involvement); in 10, the full extent of disease was not defined. There was one

Abbreviations: ficiency

syndrome,

AIDS CNS

-

acquired central

immunodenervous sys-

tem.

823

case

of Hodgkin

maining

lymphoma.

17 patients

sions,

systemic

who

In the

re-

brain

le-

had

non-CNS

lymphoma

Table

1

Lesion

was

Number,

Presentation,

excluded on clinical grounds. The history and physical examination were supported by

negative

men,

CT

and/or

studies

of the

pelvis

appropriate

and

negative

lymph

node

biopsies,

and

considered

chest,

bone

Lesion presentation Solitary

with

Lesions

Although

primary

CNS

was excluded in which specimen was suggesbut insufficient for deThe lymphoma study

group

included

therefore

imaging

studies

and

or autopsy-proved lymphoma. There

woman years).

16 patients biopsy(n 15)

(n were

aged 23-46 The 15 men

12) or bisexual likely acquired transfusion.

years were

with and/

(mean, 36.6 homosexual

lesions

each

were

of the

plasmosis sponse

identical,

=

to a clinical

anti-Toxoplasma tients (in one

complete,

of the

sone, resulting but no radiographic

in

tients tested Toxoplasma

=

(n

studies fluid for preparations

bacteria

titers,

The

of 40 patients for the proved

nosis

of toxoplasmosis

To be included also

imaging clinical nosis

fined lesion cific

=

studies

presentation; was based

on

imaging response.

patients, sone in

with the five. Since

gree or intensity be due to steroid

to marked mass was

medication

effect. limited

12);

axial

nique

able.

Systems,

[30%])

included

sections.

underwent

technique,

contiguous

Several

immediate

but

were

a

receiving

10-mm-

earlier

studies

)

11

=

scanning

a

tech-

of satisfactory

quality

for

MR imaging was performed (Signa; GE Medical The initial imaging sequence

with a 1.5Systems). included

T imager

Ti-weighted 1.0-mm skip, msec/echo

time

coronal, images

skip, 2,000/35-70). administration

dimeglumine Wayne,

images 600/20

(5.0 mm [repetition

msec])

followed

by

if indicated) long-rep(5.0 mm thick, 2.5After the of 0.1 mmol/kg

NJ),

intravegad-

(Magnevist; Ti-weighted

Bercoronal, was

to 6-8 mm to afford full brain All sequences employed a 256 x 192 matrix and one signal averaged. Lesions were characterized as to location, number, and size, as well as CT at-

le-

and

MR white

pattern

signal matter,

intensity and the

relative presence

to

of enhancement.

rate and enhancea valid

21 were hoand three

At the

Number

and

time

of presentation,

many less

than

contrast

(n

lesions

21

diameat least one larger the greatest cm, and in one3 cm. group, the 1 cm

in

material-enhanced

CT scan demonstrated sions in 17 (61%) and

multiple lesolitary lesions in 1 1 (39%). Five of the 1 1 patients with solitary lesions had undergone single-dose contrast enhancement at outside institutions. The number of lesions per patient ranged from 1 to 12 (mean, 3.3). The greatest number of lesions (n = 47 [52%J) were less than 1 cm in size. Thirty-three lesions (36%) were 1-3 cm in diameter, and 1 1 (12%) were greater than 3 cm in diameter. At least one lesion 1-3 cm in diameter was present in 64% of patients, and 36% of patients had at least one lesion greater than 3 cm in greatest diameter.

Contrast-enhanced CT was performed in 15 of the 16 lymphoma patients. Sixteen (76%) of 21 small ( 1 cm

double-dose

sagittal thick,

treatdiag-

obtained

as a moderate size, medical

medical

underwent presumptive

proved with 4), or clinical response (n

Lesion

pa-

serum ancillary

in this

therapeutic

low-up sion

(n

All

examination.

the

toxoplasmosis

12), autopsy

24

(%)

pattern < 1 cm

Milwaukee).

this

India ink fluid, anti-

and

fulfilled

had

of cerebrospinal

were

patients

(b) imaging

negative of other

who or

viewed. (a)

also

9800

single-dose

dexametha-

cytologic

ords ment

na: (n

had

studies

(36) 11(12)

of toxoplasmosis patients (n performed at outside institutions

to

clinical improvement improvement.

A host culture

and

14 pawas in-

directly

and fungi, of cerebrospinal

imaging

47(52)

38 (53) 12(17)

CT was performed

Toxolack of re-

14 patients

with

10)

titers. included

body

went

treatment

21(30)

1-3 cm >3cm Enhancement Lesions

Excluding

or GE

in

at least history

the

another

Eleven

undergone

in

3.3

of

of appropriate

therapy patient,

and

biopsy).

trial

1-12

39*