Download as pdf or txt
Download as pdf or txt
You are on page 1of 9

Clinical Radiology (1996) 51, 842-850

The Spectrum of MRI Findings in CNS Cryptococcosis in AIDS


K. A. MISZKIEL*, M. A. HALL-CRAGGS*, R. F. MILLERt, B. E. KENDALL*, I. D. WILKINSON*,
M. N. PALEY~ and M. J. G. HARRISON~
*MRI Unit, Imaging Department at UCL Hospitals (NHS) Trust, /'Division of Pathology and Infectious Diseases and
~Department of Neurology, UCL Medical School, The Middlesex Hospital, London UK

We retrospectively reviewed the cranial MRI appearances of 25 patients with AIDS and
microbiologically proven central nervous system (CNS) cryptococcosis. Four patients had a
normal scan. Ten patients had dilated perivascular Virchow-Robin spaces that were hyper-
intense on T2-weighted images. Nine of these patients developed progressive cryptococeomas,
eight in the basal ganglia and one in the cerebral white matter. The cryptococcomas displayed
high signal on T2-weighted and intermediate to low signal on Tl-weighted images. None
enhanced after dimeglumine gadopentetate. No abnormal dural or leptomeningeal enhance-
ment was detected in any patient. One patient developed an acquired arachnoid cyst during
treatment of CNS cryptococcosis which was thought to represent a focal collection of
organisms and mucoid material within the subarachnoid space. In addition either cerebral
atrophy and/or background white matter hyperintensity on T2-weighted images was present in
19/25 patients. In two patients the neuropathological findings at autopsy correlated well with
the imaging abnormalities. In conclusion, this spectrum of MRI appearances in CNS
cryptococcosis reflects the pathological mechanism of invasion by the fungus, but a normal
scan or one with features of CNS HIV infection such as atrophy or white matter hyperintensity
does not exclude the diagnosis. Miszkiel, K.A., Hall-Craggs, M.A., Miller, R.F., Kendall,
B.E., Wilkinson, I.D., Paley, M.N. & Harrison, M.J.G. (1996). Clinical Radiology, 51,842-
850. The Spectrum of MRI Findings in CNS Cryptococcosis in AIDS

Accepted for Publication 23 April 1996

Cryptococcus neoformans is a ubiquitous saprophytic CNS cryptococcosis and compared them to computed
fungus isolated from soil contaminated by bird excreta tomography [6-8]. In this study we have reviewed the
which may become particularly pathogenic in immuno- MRI findings in a cohort of twenty-five patients with
compromised patients. Although most human infection AIDS and microbiologically confirmed CNS cryptococ-
is probably secondary to inhalation, pulmonary infec- cosis and offer a pathological explanation for the ima-
tion with cryptococcosis is often asymptomatic or serf- ging findings.
limited with the fungus having a particular predilection
for haematogenous spread to the central nervous system
(CNS).
C. neoformans is the third most common pathogen to PATIENTS AND METHODS
cause CNS infection in the acquired immune deficiency
syndrome (AIDS) after the human immunodeficiency We retrospectively reviewed the cranial MRI appear-
virus (HIV) itself and Toxoplasma gondii [1,2]. Approxi- ances of twenty-five patients, twenty-three male homo-
mately 5% of patients with AIDS will develop CNS sexuals (twenty-one Caucasians and two Afrocaribbeans)
cryptococcosis although it rarely develops in patients and two female heterosexuals (of Afrocaribbean origin)
with CD4+ lymphocyte counts of greater than aged 24-56 years (median = 36 years), with a median
200 x 106/1 [3]. Most cases of C. neoformans CNS CD4+ lymphocyte count of 60 (normal range = 350-
infection present as a subacute or chronic illness with 2200 x 106/1) and with microbiologically confirmed CNS
altered mental state, behavioural changes, headache and C. neoformans infection, who were referred to a specialist
fever. Focal neurological signs are unusual, occurring in HIV/AIDS inpatient facility at UCL Hospitals between
only 15% of patients [4]. In contrast to CNS cryptococ- August 1991 and October 1995. We recorded the results
cosis in non-HIV immunosuppressed patientS, in AIDS of CSF culture and staining, CSF and serum crypto-
it is frequently part of disseminated infection [5]. The coccal antigen latex agglutination (CRAG) titres and
diagnosis is made on the basis of a series of microbiolo- whether C. neoformans was cultured from blood. Two
gical investigations including cerebrospinal fluid (CSF) patients had a post-mortem examination. At the time
culture and positive identification of the fungus with of diagnosis MRI studies had been performed on a
India ink or mucicarmine staining, raised cryptococcal 1.5T scanner (Magnetom 63SP, Siemens Medical Sys-
antigen latex agglutination titres in CSF and blood and tems, Erlangen, Germany) with a circularly polarized
positive blood cultures. transmit and receive head coil. A dual spin-echo imaging
Several studies have described the magnetic resonance sequence (TR = 3500ms, TE = 20/90ms, 192 x 256
imaging (MRI) findings in small cohorts of patients with acquisition matrix, FOV 230 mm, 5 mm slice thickness,
NE x 1) was acquired with 30 contiguous slices orien-
Correspondence to: Dr K. A. Miszkiel, MRI Unit, The Middlesex tated transversely along the plane of the temporal
Hospital, Mortimer Street, London W l N 8AA, UK. lobes. In all patients Tl-weighted spin-echo sequences
© 1996 The Royal College of Radiologists.
MRI 1N CNS CRYPTOCOCCOSIS I N AIDS 843

Table 1 - Results o f microbiological investigations in 25 patients with C. neoformans meningitis

Patient number CSF India ink stain C S F culture CSF CRAG titre Blood CRAG titre Blood culture

1^ + + 1:2000 1:131000 +
2A + + 1:1000 1:512 N/P
3 + + 1:2000 1:8000 -
4 + + 1:64 1:2000 N/P
5 - + - 1:2 N/P
6 + + 1:4000 1:256000 +
7 + + 1:4000 1:512 -
8 + + + * 1:8000 -
9 - - * + *** -
10 + + 1:32 1:4000 N/P
11 - + 1:4 1:16 N/P
12 + + 1:256 1:4000 +
13 + + 1:256 1:256 N/P
14 + + 1:16000 1:256000 +
15 + + 1:4000 1:128000 ÷
16 + + 1:1000 1:4000 +
17 + + 1:4000 1:4000 N/P
18 + + 1:512 1:2048 +
19 + + 1:128000 1:8000 -
20 + + 1:4000 1:4000 +
21 + + 1:512 1:4096 -
22 + + 1:512 1:16000 N/P
23 + + 1:128 1:128 +
24 - - 1:512 1:2000 +
25 - + 1:4 - -

C ~ A G = C r y p t o c o c c a l latex a g g l u t i n a t i o n titre; + = positive result, - = n e g a t i v e result, N / P = test n o t p e r f o r m e d . *Insufficient s p e c i m e n for titre;


**titre n o t d e t e r m i n e d ; ^ a u t o p s y p e r f o r m e d .

(TR = 640ms, TE = 14ms 192 x 256 acquisition matrix meningism and photophobia (n = 4), dizziness and
with 3/4 rectangular 230FOV, NE x 2) were also visual symptoms (n = 3 for both), diarrhoea and confu-
acquired coronally with 5 mm slices interspaced by a s i o n - ( n = 2 for both). Results of the neurological
1 mm gap, before and after 0.1 mmol/kg of intravenous examination at the time of presentation showed that
dimeglumine gadopentetate (Gd-DTPA, Magnevist, eight of 25 patients had new neurological signs asso-
Schering Health Care Ltd). ciated with CNS cryptococcosis including hypertonia/
The cranial M R I was reviewed as a batch, retrospec- hyperreflexia (n = 5), meningism/photophobia (n = 5),
tively and independently by two radiologists (MAHC altered mental state, extensor plantars or oculomotor
and BEK) who were experienced in the neuroradiology signs (n = 2 in each category) and hemiparesis (n = 1).
of HIV infection. Both radiologists were aware of the Four patients had a previously documented abnormality
diagnosis of CNS C. neoformans infection in the patient and 13 patients had no neurological signs. All patients
cohort. In each case consensus was achieved. In all were treated with appropriate antifungal therapy.
patients the following features were recorded: (1) abnor- Twenty-three patients survived the immediate crypto-
mal dilated Virchow-Robin (V-R) spaces which were coccal infection; one patient died from progressive CNS
defined as small punctate rounded lesions greater than C. neoformans infection and one patient died 3 months
2ram but less than 3 m m diameter [9,10], which were later; both these patients had an autopsy.
hyperintense on T2-weighted images and of low signal on
the Tl-weighted images, (2) cryptococcomas; these were
defined as lesions greater than 3 mm in size which were Microbiological Findings
hyperintense on T2-weighted images and of low or
intermediate signal on Tl-weighted images [6,11], (3) Results of the microbiological investigations are
shown in Table 1. In three patients with positive
abnormal dural or leptomeningeal enhancement, (4)
features of HIV or viral encephalopathy as indicated by blood cultures, C. neoformans was also cultured from
sputum.
cerebral atrophy and/or any patchy or diffuse background
white matter hyperintensity on the T2-weighted images
[12,13], and (5) other mass lesions or pathology. For the
purpose of this study the basal ganglia referred to include MRI Findings
the lentiform and caudate nuclei and the thalami. Cranial M R I was performed during the course of CNS
cryptococcosis in all patients. The median time interval
between lumbar puncture and initial MRI was 1 day
RESULTS (range = 0-22 days). Eighteen patients had follow-up
scans; the median number of scans performed in each
Clinical Presentation patient was 2 (range = 1-5). The median time interval at
The most common presenting symptoms in our patient which the first follow-up scan was performed was 36 days
cohort were headache (n = 23), nausea and vomiting (range = 3-510 days). Twenty-two patients were given
(n = 13) and fevers/sweats (n = 10). Less frequently intravenous Gd-DTPA, either at the time of the initial
occurring symptoms were malaise/fatigue ( n = 4 ) , scan (n = 18) or during follow-up (n = 4).
© 1996 The Royal College of Radiologists, Clinical Radiology, 51, 842 850.
844 CLINICAL RADIOLOGY

Four patients had a normal M R I at the time of were characteristically of high signal on T2-weighted and
diagnosis of CNS cryptococcosis. of intermediate to low signal on Tl-weighted sequences
The abnormalities detected in the remaining scans with a stippled appearance early on in the course of the
were as follows. illness. With disease progression the lesions became more
confluent, some better defined, with some lesions exert-
Dilated Virchow Robin Spaces/Cryptococcomas ing mild mass effect, but none had surrounding vaso-
genic oedema or haemorrhage. In three patients the basal
Multiple dilated V - R spaces were seen on the initial ganglia lesions extended superiorly to involve the corona
M R I in eight patients. These were present in the basal radiata. One patient had a different pattern of disease
ganglia (n = 8) (Fig. 1), brainstem (n = 3) (Fig. 2a and b) with multiple poorly defined high signal cryptococcomas
or the cerebral white matter (n = 3). All eight patients present on the T2-weighted images in the cerebral
with dilated V - R spaces in the basal ganglia had white matter, some in a periventricular distribution with
involvement of the lentiform nuclei, six had involvement
of the heads of the caudate nuclei and two patients had
thatamic involvement. Two patients with normal V - R
spaces on the initial M R I developed dilated V - R spaces
during the course of the illness detected on follow-up
scans; in one patient they were detected in the basal
ganglia at 8 weeks and in the other patient in the cerebral
white matter and basal ganglia at 5 weeks' follow-up.
Thus, overall ten patients developed dilated V - R spaces
as a result of CNS cryptococcosis, all with basal ganglia
involvement.
Cryptococcomas were detected in seven patients on
the initial MRI; six of these had cryptococcomas in the
basal ganglia with additional periventricular cryptococ-
comas around the lateral ventricles in three patients. The
distribution of the basal ganglia cryptococcomas was
similar to the distribution of dilated V - R spaces; six
patients had involvement of the lentiform nuclei, five
patients caudate head involvement and two patients had
cryptococcomas in the thalami. A further two patients
who did not have cryptococcomas on their initial scan
developed cryptococcomas in the basal ganglia on
follow-up scans at 17 and 60 days. The basal ganglia
cryptococcomas were bilateral in the majority of patients
(n = 7), often fairly symmetrical, and varied from several
millimetres (> 3 mm) to several centimetres in size. They (a)

(b)
Fig. 2 Axial T2-weighted images demonstrating multiple dilated V - R
Fig. 1 - Axial T2-weighted image showing multiple dilated V - R spaces spaces in the mid-brain (a) which show enlargement with disease over
in the basal ganglia region (arrow). an 18-day interval (b).

© 1996The Royal College of Radiologists, Clinical Radiology, 51,842-850.


MRI IN CNS CRYPTOCOCCOSIS IN AIDS 845

(a)

(b) (c)
Fig. 3 - The T2-weighted axial images in this patient with co-existant cerebral atrophy demonstrate hyperintense cryptococcomas in the right side of
the pons (a), together with poorly defined cryptococcomas (arrow) in image (b) and multiple dilated V - R spaces (curved arrow) in the cerebral white
matter (e).

further lesions around the fourth ventricle and in the brain cryptococcomas also developed cryptococcomas
stem (Fig. 3a, b and c) but without the characteristic basal around the fourth ventricle. None of the cryptococcomas
ganglia involvement detected in the other eight patients. showed enhancement after intravenous Gd-DTPA.
One other patient with progressive basal ganglia Two patients with basal ganglia cryptococcomas also
© 1996 The Royal College of Radiologists, Clinical Radiology, 51, 842-850.
846 CLINICAL RADIOLOGY

(a)

(b) (c)
Fig. 4 - The T2-weighted axial image (a) is from an examination performed 8 months prior to CNS cryptococcosis. The repeat MRI at the time of
diagnosis of CNS cryptococcosis showed an acquired right fronto-temporal arachnoid cyst (white arrow) and several hyperintense well defined
cryptococcomas in the basal ganglia (b). The cryptoeoccomas are seen on the Tl-weighted coronal images as well defined intermediate low signal
lesions (e). Note the presence of peritrigonal white m~/tter hyperintensity (black arrow) on image (b) indicative of HIV/viral encephalopathy.

had arachnoid cysts. In one patient a right fronto- morphological changes seen on M R I during the course
temporal arachnoid cyst was thought to represent an of appropriate treatment, a more detailed analysis was
acquired collection of C. neoformans within the subar- undertaken of the follow-up scans performed on nine of
achnoid space as it had not been present on cranial MRI the ten patients who had dilated V R spaces and/or
performed 8 months previously (Fig. 4a, b and c). In the cryptococcomas. In these nine patients the median
other patient no previous scans were available for com- number of follow-up scans performed per patient was 2
parison, and we cannot be certain whether the right (range 1-5) and they were acquired between 3 and 148
temporal arachnoid cyst was incidental and congenital days following the initial scan. Review of the first follow-
in origin or acquired secondary to CNS cryptococcosis. up scans performed at a median time interval of 21 days
To further characterize the chronology ! i
of the (range = 3-94 days) from the initial M R I showed
© 1996 The Royal College of Radiolo~sts, Clinical Radiology, 51,842-850.
MRI IN CNS CRYPTOCOCCOSIS IN AIDS 847

(b)
(a)

(c)

(a)
Fig. 5 - T2-weighted axial (a and c) and Tl-weighted coronal (b and d) images demonstrating the lag phenomenon in the radiological appearances
compared with CSF/blood parameters in CNS cryptococcosis. At the time when images (a) and (b) were acquired the blood C R A G titre was
1 : 128 000. Images (c) and (d) were acquired 32 days later by which time the blood C R A G titre had fallen to 1 : 64000 on treatment, indicating disease
response to therapy; the bilateral basal ganglia cryptococcomas increased in size during this time with marked expansion of the caudate nuclei despite
improvement in the patient's clinical condition.

progression of the dilated V - R spaces and cryptococco- had two or more follow-up scans performed, four
mas in all but one patient, who had a decrease in size of patients had progression of cryptococcomas on the
the bilateral basal ganglia and periventricular lesions on second follow-up scan despite a falling CSF (n = 3) or
the follow-up M R I at 78 days. In the seven patients who blood (n -- 1) C R A G titre (Fig. 5a, b, c and d). One
© 1996 The Royal College of Radiologists, Clinical Radiology, 51, 842-850.
848 CLINICAL RADIOLOGY

patient showed improvement on MRI (at 76 days) Autopsy Correlation


despite a rising CSF CRAG, but decreasing blood
One patient with dilated V - R spaces in the basal
CRAG titre. Two patients showed no change in appear-
ganglia and brainstem, who developed progressive non-
ance of the cryptococcomas from the previous scan (one
enhancing basal ganglia and periventricular cryptococ-
of these patients had already shown improvement on the
comas on MRI over a period of 5 weeks, had a post-
first follow-up scan at 78 days) despite a falling CSF
CRAG titre in one patient. Overall, when all the avail- mortem 2 weeks after his last MRI. Pathological
examination revealed thin and transparent lepto-
able follow-up scans were reviewed, in total four of the
meninges and a microcystic appearance of the basal
nine patients showed eventual improvement and a
ganglia extending to the thalamus and midbrain.
decrease in size of cryptococcomas, improvement being
Histology confirmed cryptococcal meningo-encephalitis
first detected on MRI at a median time interval of 77
and Cryptococcomas.
days (range 38-106 days) from the initial scan. In three
At post-mortem in the second patient with minimal
patients whose scans were judged to show overall
non-specific high signal in the caudate nuclei, macro-
response to treatment, a mixed response to treatment
scopic examination of the brain showed normal lepto-
was seen on three follow-up scans. Not all the crypto-
meninges and dura, together with mild ventricular
coccomas responded to treatment at the same rate in
dilatation. Histology revealed a diffuse nodular cytome-
these cases; although the majority of lesions decreased in
galovirus (CMV) encephalitis in addition to cryptococcal
size, others remained unchanged and one or two
meningo-encephalitis.
increased in size.

Dural/Leptomeningeal Enhancement DISCUSSION


These abnormalities were not detected in any patient.
Cryptococcal meningitis is the most common mani-
festation of CNS cryptococcosis often with a chronic
Atrophy and White Matter Hyperintensity insidious course [14]. Several studies have shown that
MRI is superior to CT in identifying abnormalities in
Nineteen patients had evidence of HIV or viral ence-
CNS cryptococcosis, but both imaging modalities under-
phalopathy; eight of these nineteen patients had atrophy
estimate the number of lesions detected when compared
only, two patients white matter hyperintensity only and
with pathological examination [6,15,16]. The spectrum
nine patients had both features present. The atrophy
of imaging abnormalities detected on MRI in CNS
ranged in degree from mild to moderate. Six patients
cryptococcosis reflects the pathological mechanism of
had no evidence of either atrophy or white matter
invasion by the fungus. In CNS cryptococcosis the
hyperintensity.
leptomeningeal involvement and associated inflamma-
tory reaction is mild and results in the production of a
Enhancing Mass Lesions mucoid material within the subarachnoid space. This
process may extend to involve the perivascular (V-R)
One patient had multiple enhancing mass lesions in spaces which are extensions of the subarachnoid space
the cerebral hemispheres and basal ganglia on MRI surrounding the perforating arteries into the brain
performed 6 weeks prior to the diagnosis of CNS cryp- parenchyma. With cryptococcal infection these potential
tococcosis. The signal characteristics of these lesions spaces become dilated and filled with the mucoid
were those of mixed/intermediate signal on T2-weighted material, inflammatory cells and organisms. This is
images, with surrounding high signal oedema, and they manifest on the MRI as punctate,'hyperintense round/
were of low signal on Tl-weighted images with foci of oval lesions on T2-weighted images usually less than
high signal haemorrhage. These mass lesions showed 3 mm in size with a propensity for the basal ganglia
patchy and peripheral ring enhancement with Gd- and brainstem as we found in this study. With disease
DTPA; follow-up scans showed response of the original progression these V - R increase in size, resulting in
mass lesions to anti-toxoplasma drug therapy, with the the development of cryptococcomas, which may extend
development of new superadded, well-defined, non- into the brain parenchyma. Pathologically these
enhancing cryptococcomas in the basal ganglia. These cryptococcomas are composed of gelatinous masses of
changes were interpreted as treated toxoplasma cryptococcal organisms interspersed by acid mucopoly-
abscesses and concurrent C. neoformans infection. A saccharide that is produced by the organism [17].
further three patients developed multiple enhancing In this study we have reviewed the cranial MRI scans
mass lesions typical of cerebral toxoplasma infection or of a larger cohort of immunocompromized patients with
lymphoma on subsequent follow-up at 5 months, and at AIDS and CNS cryptococcosis than previously
18 months in the other two patients. described, many of whom underwent Gd-DTPA
enhanced serial MRI while receiving appropriate anti-
fungal treatment to characterize the chronological
Non-specific Findings
changes that occur on MRI in this CNS infection.
One patient had a virtually normal scan at presenta- None of the previous studies have concentrated on this
tion, with minimal high signal in the heads of both latter feature [6,7,8,18].
caudate nuclei on the T2-weighted images but with no In our study, ten of the cohort of twenty-five patients
corresponding abnormality on the Tl-weighted images. had features on MRI attributable to C. neoformans
This patient did not have follow-up scans to assess the meningitis; in keeping with previous studies [6,7,8,18]
significance of this finding but died 3 months later and we found that dilated V - R spaces and cryptococcomas
had an autopsy. developed most commonly, but not exclusively, in the
© 1996 The Royal College of Radiologists, Clinical Radiology, 51, 842-850.
MRI IN CNS CRYPTOCOCCOSIS IN AIDS 849

basal ganglia, with the lentiform nuclei most commonly patients with AIDS has not been previously described in
involved. The basal ganglia cryptococcomas were nearly any detail. We have studied the chronological changes
always bilateral. None of the cryptococcomas in our seen on MRI in this CNS infection on appropriate
study showed evidence of enhancement, although antifungal treatment and shown that in general there is
enhancement of cryptococcomas in patients with AIDS poor correlation between the MRI appearances and
has been described, albeit infrequently. Tien et al. blood and CSF disease parameters, with a decrease in
reviewed the CT scans of twenty-nine immunocompro- the latter often preceding radiological improvement. In
mized patients (twenty-eight of whom had AIDS) with addition cryptococcomas may show a mixed response to
CNS cryptococcosis, 10 of whom had concomitant M RI, treatment, with some increasing in size while others are
four with intravenous Gd-DTPA. Enhancement of mili- diminishing.
ary leptomeningeal nodules and parenchymal cryptococ- In conclusion CNS cryptococcosis produces a wide
comas on MRI was detected in two of four patients. No variety of MRI appearances with dilated V - R spaces as
histology was available in these patients so exclusion of the earliest manifestation, with progression to intrapar-
other concomitant pathology was not possible. Only one enchymal cryptoeoccomas reflecting the distribution of
patient in Tien's study had a follow-up MRI, after these dilated perivascular spaces. Contrast enhancement
antifungal treatment, which showed complete resolution of cryptococcomas or leptomeninges is uncommon due
of the previously noted abnormalities, but the time to the unique characteristics of this organism and the
interval between the two scans was not specified [8]. profoundly immunosuppressed state of this group of
Leptomeningeal nodules, or enhancement of the lep- patients. One or two dilated V - R spaces may be present
tomeninges or dura, was not detected in any of our in normal individuals, particularly the elderly, but may
patients. Consistent leptomeningeal enhancement also occur in cerebral atrophy, which is common in HIV
caused by CNS cryptococcosis in AIDS has been demon- infection [24], and in this study was a feature in seventeen
strated on MRI in only one study of eight patients in patients; thus they are not specific for CNS cryptococ-
which the authors used a higher dose of Gd-DTPA cosis but suggest the diagnosis in the appropriate clinical
(double that conventionally used) and performed setting. In addition a normal MRI, or one showing only
delayed scans to amplify signal abnormalities due to evidence of HIV or viral encephalopathy, does not
mild leptomeningeal inflammation [18]; seven of the exclude CNS cryptococcosis, as the typical features of
eight patients exhibited leptomeningeal enhancement. this infection occurred in only 40% of our patients. Of
In addition, one patient showed enhancement of those patients who had developed cryptococcomas on
parenchymal cryptococcomas, while another had follow-up MRI, initial radiological progression was seen
enhancement of an intraventricular choroid plexus despite appropriate treatment and falling CSF and serum
cryptococcoma. It is not clear from this study whether CRAG titres. Thus it would appear that although MRI
these latter findings were pathologically confirmed [18]. is useful as part of the initial investigation protocol in
None of the patients.in the present study had choroid suspected C. neoformans meningitis serial imaging has
plexus cryptococcomas and possibly the discrepancy probably only a minimal role in monitoring response to
between the frequency of leptomeningeal enhancement therapy unless the clinical scenario raises the possibility
detected by us and by Andreula et al. in their study may of developing hydrocephalus or concurrent intracranial
be accounted for by the different dose of Gd-DTPA used. pathology.
C. neoformans is surrounded by a polysaccharide
capsule which may protect it from the host inflammatory
response even in immunocompetent patients [19] but it Acknowledgements. This work was supported by the Medical
may also have immunosuppressive properties per se. Research Council of Great Britain under grant number: SPG8915593.
We would like to thank the pathologists Professor F. Scaravilli and
This, together with the profoundly immunosuppressed Professor S. Lucas for performing the autopsies.
state of patients with AIDS, may account for the lower
incidence of leptomeningeal or lesion enhancement
occurring in CNS cryptococcosis compared with bacter-
ial meningitis or brain abscesses [20-22]. In the study by REFERENCES
Mathews et al. [6], pathological examination of the brain 1 Gabuzda DH, Hirsch MS. Neurologic manifestations of infection
parenchyma adjacent to the cryptococcomas showed with human immunodeficiency virus. Clinical features and
no evidence of an inflammatory exudate, necrosis or pathogenesis. Annals of Internal Medicine 1987;107:383 391.
2 Levy RM, Bredesen DH, Rosenblum MI. Neurologic manifesta-
haemorrhage in any case. The meninges of the brains tions of the acquired immunodeficiency syndrome (AIDS): experi-
examined in this study were only mildly inflammed with ence at UCSF and review of the literature. Journal o f Neurosurgery
mucoid material present in the subarachnoid spaces [6]. 1985;62:475-495.
One patient in their study had evidence of abnormal 3 Zuger A, Louie E, Holzman RS, et aL Cryptococcal disease in
patients with the acquired immunodeficiency syndrome. Annals of
leptomeningeal enhancement on MRI, but the authors Internal Medicine 1986;104:234 240.
found this to be secondary to an area of adjacent focal 4 Chuck SL, Sande MA. Infections with Cryptococcus neoformans in
cerebritis due to Nocardia [6]. acquired immunodeficiency syndrome. New England Journal of
To our knowledge we have described the first observa- Medicine 1989;321:794-799.
tion of an acquired arachnoid cyst developing during the 5 Eng RHK, Bishburg E, Smith SM, et al. Cryptococcal infections in
patients with acquired immune deficiency syndrome. American
course of the C. neoformans meningitis in patients with Journal o f Medicine 1986;81:19-23.
AIDS who have undergone MRI. Acquired arachnoid 6 Mathews VP, Alo PL, Glass JD, et al. AIDS-related CNS crypto-
cysts have been described as a consequence of meningitis coccosis: radiologic-pathologic correlation. American Journal of
[23] and in our patient we suspect it represented a focal Neuroradiology 1992;13:1477-1486.
7 Wehn SM, Heinz ER, Burger PC, et al. Dilated Virchow-Robin
collection of mucoid material Within the subarachnoid spaces in cryptococcal meningitis associated with AIDS: CT and
space. MR findings. Journal o f Computer Assisted Tomography
Follow-up of C. neoforrnans meningitis on MRI in 1989;13:756-762.

O 1996 The Royal College of Radiologists, ClinicalRadiology, 51, 842-850.


850 CLINICAL RADIOLOGY

8 Tien RD, Chu PK, Hesselink JR, et al. Intracranial cryptococcosis imaging of cryptococcal meningoencephalitis. Neuroradiology
in immunocompromised patients: CT and MR findings in 29 cases. 1991 ;33:443-446.
American Journal of Neuroradiology 1991 ;12:283-289. 17 Garcia CT, Weisberg LA, Lacorte WSJ. Crytococcal intracerebral
9 Braffman BH, Zimmerman RA, Trojanowski JQ, et al. Brain MR. mass lesions: CT - pathologic considerations. Neurology
Pathologic correlation with gross and histopathology. 1. Lacunar 1985;35:731 734.
infarction and Virchow-Robin spaces. American Journal of Neuro- 18 Andreula CF, Burdi N, Carella A. CNS cryptococcosis in AIDS:
radiology 1989;9:621-628, American Journal of Radiology 1988; spectrum of MR findings. Journal of Computer Assisted Tomogra-
151:551-558. phy 1993;17:438-441.
10 Heier LA, Bauer CJ, Schwartz L, et al. Large Virchow-Robin 19 Hassin GB. Torulosis of central nervous system. Journal of Neuro-
spaces: MR-clinical correlation. American Journal of Neuro- pathology and Experimental Neurology 1947;6:44-60.
radiology 1989;10:929 936. 20 Mathews VP, Kuharik MA, Edwards MK, et al. Gd-DTPA-
11 Vinters HV, Anders KH. Neuropathology of AIDS. Boca Raton, enhanced MR imaging of experimental bacterial meningitis: evalua-
FL: CRC Press 1990:61 65. tion and comparison with CT. American Journal of Neuroradiology
12 Jarvik JG, Hesselink JR, Kennedy C, et al. Acquired immunodefi- 1988;9:1045 1050.
ciency syndrome: magnetic resonance pattern of brain involvement 21 Chang KH, Han MH, Roh JK, et al. Gdf-DTPA-enhanced MR
with pathologic correlation. Archives of Neurology 1988;45: 731- imaging of the brain in patients with meningitis: comparison with
736. CT. American Journal of Neuroradiology 1990; 11:69 76.
13 Olsen WL, Longo FM, Mills CM, et al. White matter disease in 22 Enzmann DR, Britt RH, Yeager AS. Experimental brain abscess
AIDS: findings at M R imaging. Radiology 1988;169:445 448. evolution: computed tomographic and neuropathologic correlation.
14 Dismukes WE. Cryptococcal meningitis in patients with AIDS. New Radiology 1979;133:113 122.
England Journal of Medicine 1987;317:334 431. 23 Starkman SP, Brown TC, Linell EA. Cerebral arachnoid
15 Post MJD, Sheldon J J, Hensley CT. Central nervous system disease cysts. Journal of Neuropathology and Experimental Neurology
in acquired immunodeficiency syndrome: prospective correlation 1958; 17:484-500.
using CT, MR imaging and pathologic studies. Radiology 24 Levy RM, Rosenbloom S, Perrett LV. Neurologic findings in
1986;158:141-148. AIDS: a review of 200 cases. American Journal of Neuroradiology
16 Takasu A, Taneda M, Otuki H, et al. Gd-DTPA-enhanced MR 1986;7:833-839.

© 1996 The Royal College of Radiologists, Clinical Radiology, 51, 842-850.

You might also like