print.50472555

7/23/2019 print.50472555

1/11

2010 ISHAM DOI: 10.3109/13693780903437876

Medical MycologyJune 2010, 48, 570579

Cryptococcal meningitis in non-HIV-infected patients in a

Chinese tertiary care hospital, 19972007

LI-PING ZHU, JI-QIN WU, BIN XU, XUE-TING OU, QIANG-QIANG ZHANG & XIN-HUA WENG

Department of Infectious Diseases, Huashan Hospital, Fudan University, Shanghai, China

Information remains sparse about non-HIV patients with cryptococcal meningitis in the

era of triazole therapy. Particularly of interest are the clinical manifestations and prog-

nosis of the infection in these previously healthy patients. We retrospectively reviewed

154 non-HIV-infected patients with cryptococcal meningitis who presented in our hos-

pital from 1997 to 2007. We compared the clinical features and outcomes between pre-

disposed and otherwise healthy hosts.The number of cases per year showed a steady

increase over time.The majority of patients were otherwise apparently healthy (103

patients, 66.9%) and predisposing factors were identied in only 51 (33.1%) patients.

Corticosteroid medication accounted for the most common underlying factor in these

cases (n

21). Morbidity was appallingly high, with seizures in 28.6%, cranial nervespalsies in 51.5% and cerebral herniation in 19.5%. Despite these complications, overall

mortality during 1 year was 28.7% (41/143), close to that reported from other centers

with non-HIV patients. Death attributed to cryptococcosis occurred in 19.6% (28/143)

patients with most receiving amphotericin B as a component of their initial therapy.

Among surviving patients who had lumbar punctures at weeks 2 and 10, those given

amphotericin B for initial therapy achieved higher rates of overall response than those

receiving initial uconazole therapy at either week 2 (84.4% of 96 patients vs. 33.3% of

24 patients, P0.001) or week 10 (85.0% of 93 patients vs. 66.7% of 24 patients, P

0.041). In multivariate analysis, coma, cerebral herniation, and initial antifungal therapy

without amphotericin B were independently correlated with both increased overall and

attributable mortality, while advanced age (60 years) was correlated with increased

overall mortality only. Patients with apparently normal immune status were overallyounger than those who were immunocompromised. In addition, previously healthy

patients for whom diagnosis was delayed had more severe disease, experiencing more

brain herniation, coma, seizures, hydrocephalus and more surgical shunt procedures. On

the other hand, immunocompromised patients were more commonly found to have high

fever and brain parenchymal involvement. However, both groups had a similar treatment

response and 1-year survival.

Keywords cryptococcal meningitis, HIV seronegativity

ingitis [1]. The incidence of HIV-associated cryptococcal

meningitis has been decreasing in recent years owning to

the advent of highly active antiretroviral therapy (HAART)[24]. As the proportion of cases in previously healthy

patients increases, additional attention is needed to better

understand the diagnosis and management problems in

these patients [57].

The natural history of cryptococcal meningitis in patients

with AIDS had been well documented in the pre-HAART era.

Some studies have also described the disease in HIV-negative

Received 19 May 2009; Received in nal revised form 15 September

2009; Accepted 25 October 2009

Correspondence: Li-Ping Zhu, Department of Infectious Diseases,

Huashan Hospital, 12 Central Urumqi Road, Shanghai, China. Tel: +86

(0)21 52888292; fax: +86 (0)21 62489015; E-mail: [email protected]

Introduction

Cryptococcus neoformans(C. neoformans var. grubiiand

C. neoformans var. neoformans) and C. gattiiare respon-

sible, on a worldwide basis, for most cases of fungal men-

7/23/2019 print.50472555

2/11

2010 ISHAM,Medical Mycology, 48, 570579

Non-HIV cryptococcal meningitis in China 571

patients [811]. But comparative data relative to predisposed

and normal patients in the non-HIV population has been very

limited. Here, we present a single-center study of cryptococ-

cal meningitis in non-HIV-infected patients, with a compre-

hensive analysis of clinical features and outcomes, together

with comparisons between predisposed and normal hosts.

Methods

This retrospective study was carried out in Huashan Hos-

pital, a 1,088-bed tertiary health care hospital located in

Shanghai, China. Records of all patients with denite or

probable cryptococcal meningitis admitted during an

11-year period (19972007) were examined. Demographic

data, underlying diseases, clinical manifestations, labora-

tory ndings, cranial radiology, antifungal treatments and

outcomes were recorded.

Diagnosis of cryptococcal meningitisA denite diagnosis of cryptococcal meningitis was made

if the patient met at least one of the following criteria; (1)

positive culture of Cryptococcus neoformans from cere-

brospinal uid (CSF), (2) positive CSF India Ink smear of

centrifuged sediment for Cryptococcus, (3) compatible his-

topathology (5- to 10-m encapsulated yeasts observed in

brain tissue). Probable cryptococcal meningitis was diag-

nosed in 5 patients who presented with the clinical syn-

drome of meningitis and positive cryptococcal antigen

titer in CSF, albeit without microbiological or patho-

logical documentation. A positive test on undiluted CSF

using the latex-cryptococcus antigen detection system

(Immuno-Mycologics, Inc., USA) was considered diag-

nostic. The API 20C AUX system was used for identica-

tion of isolates, without additional conrmation and tests

for C. gattiiwere not done.

CD4+T lymphocytopenia

During hospitalization, 72 out of 154 patients were tested for

CD4+ T cell counts, by Becton-Dickinson FACS Calibur

Flow Cytometer System and IMK kit (BD Multitest). The

normal range of CD4+ T cell counts varies depending on

population. In America, CD4+lymphocytopenia was dened

as CD4

+

T lymphocyte count

300/mm

3

(or

20%)) [5,12].Though large-scale surveys on the distribution of CD4+T

cell count in healthy Chinese are not available, several stud-

ies have suggested that the average absolute CD4+ T cell

counts are approximately 100/mm3 lower in Chinese than

North Americans [13]. Therefore, the cut-off value of 200/

mm3was adopted in our study, i.e., patients with CD4+T cell

count 200/mm3 and no immunocompromising conditions

were considered to be idiopathic CD4+T lymphocytopenic.

Normal and predisposed hosts

Patients with one or more predisposing factors such as

underlying immunocompromising diseases (cirrhosis,

chronic kidney diseases, autoimmune diseases, solid malig-

nancies, hematologic malignancies, splenectomy, and

SOT (solid organ transplantation)), corticosteroid or other

immunosuppressive medications, or idiopathic CD4+ Tlymphocytopenia were classied as predisposed hosts. We

also included diabetes mellitus, though we are aware that

this common condition is considered to be a controversial

predisposing factor [14]. Patients with no identiable risk

factors were classied as previously healthy hosts.

Antifungal therapy

Treatment given initially and continued for at least 1 week

was dened as initial therapy, whether or not additional

agents were added later.

Outcomes

Treatment response was evaluated at week 2 and 10 after

initiation of antifungal therapy according to the criteria

described by Segal et al. modied as follows [15]; (1) suc-

cessful response being survival within the 2- or 10-week

period of observation, improvement or resolution of attribut-

able symptoms and signs of disease, normalization or

improvement in CSF chemistry and cell count, and negative

CSF culture; (2) failure was considered in patients with

stable or worsening attributable symptoms and signs of dis-

ease, persistently positive results of cultures of CSF speci-

mens, or death regardless of attribution; and (3) patients who

survived but did not have a lumbar puncture at that particu-lar time point were evaluated based only on clinical symp-

toms and signs. After discharge, patients were followed up

as outpatients or via telephone. For patients who died during

follow-up, time and cause of death were recorded. Deaths

that occurred within 1 week of observation, or occurred later

but without evidence of other causes (such as worsening of

underlying diseases) were attributed to cryptococcosis.

Statistical analysis

Categorical variables were analyzed by 2 test. Continuous

variables were analyzed by rank-sum test. Log-rank test

was used in univariate analysis for prognostic factors of1-year survival, and Cox proportional hazards for identi-

cation of independent prognostic predictors. P values

.05 were considered statistically signicant.

Results

A total of 149 patients with proven cryptococcal meningitis

and ve probable cases were included in this investigation.

7/23/2019 print.50472555

3/11


572 Zhu et al.

Basic information

Patients were from 8 provinces and districts of China.

Fifty-nine (38.3%) were admitted during 19972002, and

another 95 (61.7%) during 20032007. The number of

cases diagnosed each year increased over time (nptrend

analysis: P 0.003, Fig. 1). The median time from symp-

tom onset to diagnosis was 32 days (range, 1~2,890 days).Ninety-four (61.0%) patients were male and ages of all

patients ranged from 9~75 years (median, 38.5 years).

HIV negativity

Of the 154 patients, 141 (91.56%) were conrmed to be

non-HIV infected by negative serum HIV antibody and/or

normal CD4+ cell counts. HIV status was not available in

13 patients case les.

Predisposing factors

Predisposing factors were identiable in 51 (33.1%) of154 patients, these included 9 patients with idiopathic

CD4+ lymphocytopenia. In addition, 43 patients had

one or more of the following immunocompromising

conditions; corticosteroids in 21, autoimmune diseases in

17, cirrhosis in 15, diabetes mellitus in 14, immuno-

suppression in 13, chronic kidney diseases in 11, splenec-

tomy in two, lung cancer in one, hematologic malignancy

in one, and kidney transplantation in one patient. The

remaining 103 (66.9%) patients were apparently healthy

(Table 1).

Signs and symptomsThe most common symptoms at presentation were head-

ache (154, 100%), fever (125, 81.2%), and vomiting (98,

63.6%). Meningismus was present in 110 (71.4%) patients.

Altered mental status was observed in 72 (46.8%) patients,

including 44 (28.6%) who experienced at least one seizure

during the course of disease. Cranial nerve palsies were

observed in 51.5% (80/154), specically involving the

abducent nerve in 29.9% (46/154), 8th nerve in 24.7%(38/154), optic nerve in 21.4% (33/154), 7th nerve in

7.8% (12/154), 3rd nerve in 3.9% (6/154), olfactory nerve

in 1.3% (2/154), and 5th nerve in 0.7% (1/154). Upper or

lower limb weakness was seen in 18.2% (28/154) of

patients. During the course of disease, 19.5% (30/154)

experienced cerebral herniation.

CSF ndings

Before antifungal treatment, CSF opening pressure was

documented in 105 patients, ranging from 80~300

mmH2O (median pressure, 300 mmH

2O), and 81.9%

(86/105) of the patients had elevated CSF pressure beyond

200 mmH2O. The median white blood cell (WBC) count

in CSF was 86/mm3 (124 specimens; range, 01,030/mm3),

and 85.5% (106/124) patients had pleocytosis (CSF

WBC10/mm3). The median protein concentration in CSF

was 718 mg/dl (124 specimens; 206,018 mg/dl) with

81.5% (101/124) above 450 mg/dl. The median CSF glu-

cose concentration was 1.7 mmol/l (123 specimens; range,

1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

0

5

10

15

20

25

Normal Predisposed Total

year

casenu

mber

Fig. 1 Number of cryptococcal meningitis diagnosed per year.

Table 1 Demographic features and underlying diseases of 154 non-HIV

patients with cryptococcal meningitis.

Items n %

Sex

Male 94 61.0

Age (years)

30 39 25.33059 104 67.5

60 12 7.8

Time to diagnosis

1 week 3 2.0

1 week4 months 132 85.7

4 months 19 12.3

Denite cases 149 96.8

CSF culture (+)a 76 78.4

Indian ink smear (+)b 131 88.5

Histopathology (+)c 5 3.3

Predisposing factors 51 33.1

Corticosteroids 21 13.6

Autoimmue diseases 17 11.0

Liver cirrhosis 15 9.7

Diabetes mellitus 14 9.1

Immunosuppression 13 8.4Chronic kidney diseases 11 7.1

Splenectomy 2 1.3

Solid malignancy 1 0.7

Hematologic malignancy 1 0.7

Kidney transplantation 1 0.7

Idiopathic CD4+lymphocytopeniad 9 12.5

Extraneural involvement

Pulmonary cryptococcosis 14 9.1

Cryptococcemiae 4 33.3

The total number of cases examined was 154 unless otherwise specied.aIn 97 patients. bIn 148 patients. cIn 5 patients. dCD4+lymphocyte count

tests were taken in 72 patients. eBlood cultures were performed in 12

patients before starting antifungal treatment.

7/23/2019 print.50472555

4/11



sites included occipital lobe (n5), insula (n5), brain-

stem (n5) and thalamus (n5).

Pulmonary cryptococcosis, cryptococcemia, and

skin cryptococcosis

Pulmonary cryptococcosis was diagnosed in 14 (9.1%) ofthe 154 patients based on imaging and response of the

lesions to antifungal treatment. The infections were proven

in two patients by lung histopathology. Of the other 12

patients who had blood culture before antifungal therapy,

4 (33.3%) were positive for Cryptococcus. One patient

had cryptococcal skin lesions proven by histopathology of

a skin biopsy. In addition, neither urine nor sputum was

cultured in our patients.

Initial antifungal therapy and outcome

Twelve (7.8%) of the patients died or were discharged in less

than 1 week after commencement of antifungal therapy(Table 2).In 142 patients who were treated for more than 1

week, initial therapy involved intravenous amphotericin B

alone for 14 (9.9%) patients, amphotericin B and ucytosine

combination for 32 (22.5%), amphotericin B and uconazole

combination in 12 (8.5%), and a triad of amphotericin B,

ucytosine and uconazole for 38 (26.8%). In some patients,

intravenous uconazole alone (20 patients, 14.1%) or in

combination with ucytosine (4 patients, 2.8%) was adopted

as the initial treatment. The remaining 22 (15.5%) patients

were treated with other antifungal drugs, including amphot-

ericin B colloidal dispersion, itraconazole, etc (Fig. 2).

0.567.5mmol/l), 57.7% (71/123) below 2.0 mmol/l.

Cultures of CSF after the rst lumbar puncture was posi-

tive for Cryptococcus in 76/97 (78.4%) patients , and

Indian Ink smear was positive in 131/148 (88.5%) patients.

Ten patients were smear negative but culture positive, and

16 patients were culture negative but smear positive. Eight

(5.2%) patients had negative culture and India ink smeardespite positive cryptococcal antigen test results at the time

of diagnosis. Among these eight patients, two had a posi-

tive CSF smear afterwards, and one had positive histopa-

thology. At time of diagnosis, CSF cryptococcal antigen

tests were positive in all 88 specimens from 88 patients,

titers ranging from 1:8 to 1:2560 (median, 1:1280). All

of the 154 patients had positive CSF cryptococcal antigen

when tested later.

Cranial imaging

Before antifungal treatment, 65 and 58 patients had MRI

or CT scans, 76.9% (50/65) and 41.4% (24/58) yieldingabnormal results, respectively.

Among CT scan results, hydrocephalus was found in 10

of 58 patients (17.2%), and local lesions in 16 patients

(27.6%). MRI scans detected hydrocephalus in 9 (13.9%)

and local lesions in 45 (69.2%) of 65 patients. Local lesions

were characterized by decreased density in CT, low signal

in T1, high signal in T2 and FLARE in MRI, with or with-

out enhancement. Common sites involved included the

frontal lobe (n39), parietal lobe (n35), basal ganglion

(n33), periventricular region (n27), temporal lobe

(n 16), and cerebellum (n14). Other less common

Table 2 Patients not evaluated for initial antifungal therapy.

No. Age Sex

Days to

diagnosis

Underlying

conditions HIV

Mental status

at presentation

Hospitalization

days

Antifungal drugs

administered Outcome

1 54 M 124 Cirrhosis, CD4+

lymphocytopenia

- Consciousness 1 Non Died (hepatic failure)

2 62 M 30 Non - Coma 1 Non Transferred

3 43 F 420 Non - Coma 1 Non Died (cerebral herniation)

4 67 F 28 Lung cancer

(postoperative)

- Confused 4 Itra+AmB Died (cerebral herniation)

5 26 M 24 Non NA Consciousness 3 ABCD+Flu+5FC Died (cerebral herniation)

6 34 F 30 Non NA Confused 2 AmB+5FC Died (cerebral herniation)

7 26 M 143 Non - Consciousness 1 AmB+5FC Died (cerebral herniation)

8 44 F 871 Non - Coma 1 AmB+Flu+5FC Died (cerebral herniation)

9 36 M 50 Non - Drowsiness 7(5FC+Flu)

4d/(AmB+Flu) 1d/

(AmB+Itra) 2d

Died (cerebral herniation

& respiratory failure)

10 43 M 30 Non NA Confused 7 ABCD+5FC Died (cerebral herniation)

11 26 M 10 Non NA Confused 7 (Flu+5FC) 5d/

(AmB+Flu+5FC)

2d

Died (cerebral herniation)

12 16 M 32 Non - Confused 6 (Flu+5Fc) 6d Died (cerebral herniation)

Itra, itraconazole; AmB, amphotericin B deoxycholate; ABCD, amphotericin B colloidal dispersion; Flu, uconazole; 5FC, ucytosine; NA, not

available.

7/23/2019 print.50472555

5/11


574 Zhu et al.

Amphotericin B was given intrathecally in 63 (40.9%)

patients during hospitalization. Ventricular drainage or

ventricular-peritoneal shunt was carried out in 27 (17.7%)

patients. The median dose of intrathecal amphotericin B

therapy was 4.25 mg (0.145.9 mg).

Survival

Patients were followed up for a median time of 878.5 days

(ranging 05,472 days). Eleven (7.1%) were lost to follow

up and 41 died within 1 year after antifungal therapy. The

remaining 102 patients were alive and followed up for at

least 1 year. The all-causes and cause-specic mortality at

1-year follow-up were 28.7% (41/143) and 19.6% (28/143),

respectively. All deaths due to cryptococcal meningitis

occurred within 10 months of antifungal therapy, and

82.1% (23/28) occurred within 3 months.

In patients receiving amphotericin B-based, uconazole-

based and other antifungal treatments, the 1-year cause-

specic mortalities were 7.9% (7/89), 33.3% (8/24) and10.5% (2/19), respectively. The 1-year all-cause mortality

was 9.1% (4/44), 20.0% (9/45), 45.8% (11/24), and 31.6%

Treatment outcomes were evaluated in 153 patients at

week 2 (one patient was discharged and lost to follow-up

before week 2), and 149 patients at week 10 (ve patients

were discharged and lost to follow-up before week 10).

The proportion of patients who successfully responded,

regardless of antifungal therapy, was 68.0% (104 of 153

patients) at week 2 and 74.5% (111 of 149 patients) atweek 10. Successful response rates in patients subdivided

by different initial antifungal therapy are shown in Table 3.

Response rates favored amphotericin B treated patients at

both week 2 and week 10. Specically, successful response

rates were 84.4% (81/96) and 33.3% (8/24) at week 2

(P0.001), 85.0% (79/93) and 66.7% (16/24) at week 10

(P 0.041) for patients receiving amphotericin B and

uconazole-based treatments, respectively.

In 32 patients who received amphotericin B and ucyto-

sine combination as initial therapy, the daily dosage of

amphotericin B ranged from 0.311.19 mg/kg (average

0.47 mg/kg), for a median duration of 92 days (7245 days),

reaching an accumulating dosage of 756,120 mg (median,2,020 mg), and notably, the majority of these patients (n

31) received a daily dosage lower than 0.7 mg/kg.

Fig. 2 Initial therapies and subsequent changes in patients with cryptococcal meningitis. AmB, amphotericin B; Flu, uconazole; 5FC, ucytosine; Itra,

itraconazole; ABCD, amphotericin B colloidal dispersion; ADR, adverse drug reactions. Reasons for changing drugs are shown in brackets following

duration of initial therapies. Addition or removal of 5FC was not listed as therapy changes.

7/23/2019 print.50472555

6/11



Comparison between two host groups

There were 103 (66.9%) patients in the normal group, and

51 (33.1%) in the predisposed group. In predisposed

patients, it took around 27 weeks (25th75th percentile:

1550 days) from initial onset of symptoms to diagnosis,

whereas in the normal group it took longer, about 311

weeks (25th75th percentile: 2276 days, P = 0.006).

Patients in the latter were generally younger than the pre-

disposed group (median age: 35 vs. 48 years, P 0.001).

The two groups of patients did not differ signicantly in

other baseline characteristics such as sex, time of admis-

sion, and initial antifungal therapy. As for clinical fea-

tures, patients who were apparently healthy were moreprone to seizure (34.0% vs. 17.7%, P 0.035) and cere-

bral herniation (25.2% vs. 7.8%, P 0.010) compared to

predisposed hosts. Patients in the predisposed group were

more frequently found to have high fever (41.2% vs.

16.5%, P 0.001), and parenchymal lesions in cranial

MRI (88.9% vs. 61.7%, P 0.034), but less frequently

received ventricular drainage or shunting procedures

(5.9% vs. 23.5%, P0.007). Of the rst lumbar puncture,

no signicant difference in pressure, protein, glucose con-

centration or WBC count was found between members of

the two groups and all had similar CSF antigen titer

proles (Table 6).

The week 2 and week 10 successful response rate was

similar between the patients in the predisposed group

and those of the normal group (week 2, 68.6% vs. 67.7%,

P 0.902; week 10, 82.0% vs. 70.7%, P 0.135). One-

year survival did not differ between the two groups either

(all-causes mortality: 26.5% in predisposed vs. 29.8%

in normal, P 0.689; cause-specic mortality: 14.3% in

predisposed vs. 22.3% in normal, P 0.295).

(6/22), respectively. There was a signicant difference in

mortality between the amphotericin B-based group and the

uconazole-based group (P 0.001). The cause-specic

mortality in patients given amphotericin B therapy with

uconazole combination was signicantly higher than

patients given amphotericin B with or without ucytosine

(15.6% of 45 patients vs. 0% of 38 patients, P 0.008).

Prognostic factors

Univariate analysis of both overall and cause-specic mor-

tality showed that, diagnostic delay of more than 4 months,

and solid malignancies were signicantly associated withdeath in 1 year after treatment. Patients who presented with

altered mental status (coma, seizure), and cerebral hernia-

tion also had poorer prognoses. Initial treatment was also

strongly related to 1-year survival of patients. Inclusion of

amphotericin B or ucytosine as part of the initial therapy

signicantly improved patients survival (Table 4). Of note,

intrathecal amphotericin B therapy was also signicantly

correlated to 1-year survival. In addition, several factors

were related to all-causes mortality, but not cause-specic

mortality, i.e., age 60 years, hematologic malignancy,

ventricular drainage or shunting, and local lesions revealed

by cranial CT.

In multivariate analysis, factors positively associated with

both all-causes and cause-specic mortality for 1 year

included non-amphotericin B-based initial antifungal therapy,

delayed diagnosis ( 4 months), and cerebral herniation.

Advanced age (60 years) and coma were associated with

all-causes mortality only (Table 5). CSF parameters, such as

pressure, WBC, glucose and antigen titer were not analyzed

in the multivariate analysis because of missing data.

Table 3 Treatment response in cryptococcal meningitis patients with different initial antifungal therapies.

Initial therapy n total

2 weeks 10 weeks

Overall response CSF Overall response CSF

Evaluated Success Examined Culture (+) Evaluated Success Examined Culture (+)

AmB5FC 46 46 41 (89.1) 23 1/20 45 38 (84.4) 24 0/24

AmB 14 14 13 (92.9) 4 0/20 13 11 (84.6) 5 0/5

AmB+5FC 32 32 28 (87.5) 20 1/20 32 27 (84.4) 20 0/20

AmB+Flu5FC 50 50 40 (80.0) 34 1/34 48 41 (85.4) 31 0/31

Flu5FC 24 24 8 (33.3) 17 4/17 24 16 (66.7) 17 1/17

Other treatments 22 22 15 (68.2) 19 1/19 21 16 (76.2) 15 0/15

Total 142- 142 104 (73.2) 93 7/93 138 111 (80.4) 87 1/87

1. Thirteen patients who received less than 1 week of antifungal treatment were not included.

2. Success was dened as survival within the 2- or 10-week period of observation, improvement or resolution of attributable symptoms and signs of

disease, normalization or improvement in CSF chemistry and cell count, and negative CSF culture; patients who survived but did not have a lumbar

puncture at that particular point were evaluated based merely on clinical symptoms and signs.

3. AmB, amphotericin B deoxycholate; 5FC, ucytosine; Flu, uconazole; Other treatments, amphotericin B colloidal dispersion, itraconazole, etc.

4. Three out of four patients who had positive blood culture at the time of diagnosis were reexamined afterwards, and all turned negative. One patient

was not rechecked due to quick death.

7/23/2019 print.50472555

7/11


576 Zhu et al.

Discussion

Since the introduction of uconazole, there has been little

information about management and prognosis of cryptococ-

cal meningitis in non-HIV infected patients. To our knowl-

edge, the present study is one of the largest case series of

cryptococcal meningitis in non-HIV patients [7,8]. How-

ever, the retrospective nature of the study has its own aws

of bias in our analysis, especially on therapeutic outcomes.

In our series, the number of patients enrolled per

year steadily increased over its 11-year span. The 1.57:1

male/female ratio is consistent with the reported male

predominance [2,5,8]. The median age of 38.5 years

at presentation in our patient population is similar to the

average age reported in non-HIV patients (41.556 years),

and in the HIV population (median age, 36 years)

[2,5,8,10,11,16,17].

Table 4 Factors associated with 1-year mortality by Log-rank tests.

All-causes mortality Cause-specic mortality

Factors Total

No. of

death Expected P

No. of

death Expected P

Male 94 27 24.67 0.4574 19 16.92 0.4218

Age 60 years 12 8 2.38 0.0002* 3 1.74 0.3217

Time to diagnosis 4 months 20 10 4.36 0.0042* 7 3.06 0.0168*Pulmonary cryptococcosis 14 5 3.96 0.5808 1 2.67 0.2819

Underlying conditions

Cirrhosis 15 4 4.11 0.9557 4 2.77 0.4338

Chronic kidney failure 11 4 2.99 0.5427 1 2.08 0.4367

Diabetes mellitus 14 3 4.34 0.4958 1 2.8 0.2396

Autoimmune diseases 17 4 4.95 0.6476 1 3.34 0.1713

Hematologic malignancy 1 1 0.11 0.0071* 0 0.1 0.7471

Solid malignancy 1 1 0.04 0.0000* 1 0.04 0.0000*

Splenectomy 2 0 0.64 0.4184 0 0.42 0.5135

Corticosteroid 21 7 5.79 0.5873 1 3.7 0.1073

Immunosuppressant 13 5 3.42 0.3713 1 2.39 0.3465

Transplantation 1 1 0.28 0.1752 0 0.20 0.6528

Idiopathic CD4lymphocytopenia 9 0 1.58 0.1778 0 0.79 0.3403

Apparently healthy 103 13 14.22 0.6891 7 9.63 0.2950

Clinical manifestations

Altered mental status 55 26 17.68 0.0086* 19 12.23 0.0097*Coma 16 14 2.41 0.0000* 10 1.88 0.0000*

Seizure 44 1 10.76 0.0034* 14 7.52 0.0056*

Cerebral herniation 30 21 5.56 0.0000* 17 4.09 0.0000*

Pretreatment cranial imaging

Hydrocephalus (CT or MRI) 16 4 3.79 0.9066 7 5.26 0.4041

Local lesions (in MRI) 45 10 8.10 0.2410 14 11.42 0.1824

Local lesions (in CT) 16 7 3.35 0.0219* 7 4.54 0.1832

Antifungal treatment

AmB based 104 15 30.42 0.0000* 8 20.45 0.0000*

Including 5FC 86 17 24.58 0.0155* 11 16.62 0.0302*

Intrathecal AmB 63 9 18.62 0.0025* 7 12.54 0.0347*

Ommaya inplantation 27 13 7.03 0.0132* 7 4.90 0.2944

Success response at 2 weeks 153 14 31.09 0.0000* 5 20.80 0.0000*

Success response at 10 weeks 149 12 34.50 0.0000* 4 23.01 0.0000*

AmB, amphotericin B deoxycholate; 5FC, ucytosine.Success was dened as survival within the 2- or 10-week period of observation, improvement or resolution of attributable symptoms and signs of disease,

normalization or improvement in CSF chemistry and cell count, and negative CSF culture; patients who survived but did not have a lumbar puncture

at that particular point were evaluated based merely on clinical symptoms and signs.

*P0.05.

Table 5 Multivariate analysis of factors affecting 1-year cause-specic

mortality

Factors entered P RR (95% CI)

Coma 0.056 2.51 (0.986.47)

Non-AmB-based initial therapy 0.000 8.87 (3.5322.25)

Cerebral herniation 0.000 7.08 (2.9616.95)

Time to diagnosis120 days 0.000 6.30 (2.4116.53)

1. Parameters entering the initial model included sex, age 60 years,time to diagnosis 4 months, pulmonary cryptococcosis, autoimmune

diseases, hematologic malignancy, solid malignancy, corticosteroid,

transplantation, apparently healthy, altered mental status (0 awake,

1drowsiness, 2light coma, 3 deep coma); coma, seizure, cerebral

herniation, non-AmB-based initial therapy, inclusion of 5FC in initial

treatments, intrathecal AmB treatments, ommaya implantation.

2. Only statistically signicant (P0.05) predictors of 1-year all-causes

mortality are listed.

3. AmB, amphotericin B deoxycholate.

7/23/2019 print.50472555

8/11



Data from America, France, and Thailand have indi-

cated that a relatively low proportion (1732%) of non-

HIV infected patients were non-immunocompromised

[2,810,18,19]. In our study, however, about two thirds of

patients had apparently normal immune systems. This is

consistent with studies from Hong Kong and Taiwan,

which reported that normal hosts accounted for up to

5567% of all the patients [2022]. Two studies from

Shanghai stated that 7276% of cryptococcal meningitis

cases were found in apparently healthy patients [23,24].

A study in the 1970s from Singapore, a country where

Chinese is the predominant ethnic group, reported that

96% (24/25) of the patients were healthy [25]. Interest-

ingly, in an earlier case series of cryptococcal meningitis

from Malaysia in the 1970s, 20 out of the 30 patients were

Chinese [26]. These observations together highly suggest

that in the Chinese population, cryptococcosis patients are

predominantly immunocompetent; and raise the possibility

that the Chinese population may be more susceptible than

other ethnic groups to cryptococcal meningitis.

Although it may appear counterintuitive, the prognosis

in previously healthy patients is not better than in immu-

nocompromised patients [2,5,2729]. Previous studies

have shown that cryptococcal meningitis in immunocom-

petent individuals had a longer mean time from illness

onset and more intense inammatory response. They were

more commonly associated with papilloedema, hydroceph-alus, focal decits, seizures and cryptococcomas [19]. Our

study also revealed that the duration from onset of symp-

toms to diagnosis of cryptococcal meningitis in immuno-

competent patients was longer than with predisposed

patients. The longest delay of diagnosis was 2,890 days, in

a female patient who presented with chronic afebrile head-

ache and hydrocephalus. Immunocompetent patients in our

study were less frequently found to have high fever and the

lack of fever may have contributed to a limited suspicion

of infectious diseases. The clinical manifestations noted in

our immunocompetent patients were overall more severe

in that they presented with more coma, seizure, cerebral

herniation, hydrocephalus and surgical intervention. Theincidence of these extremely morbid complications in our

previously healthy patients was appallingly high, indicat-

ing either delayed referral or a more severe natural course

of the infection in our Chinese patients. Approximately

half of patients in both groups had cranial nerve palsies

which is remarkably higher than reported elsewhere. Paren-

chymal lesions in MRI seemed to be less frequently found

in healthy hosts, but the lesions may not be exclusive to

cryptococcal meningitis.

Corticosteroid medication was the most common under-

lying condition in our patients. Our study included only

one patient with kidney transplantation and one with hema-

tologic malignancy in our compromised group, which isquite different from the high proportion of patients with

solid organ transplantation and hematologic malignancy

reported in other studies [8].

Nine of 54 (12.5%) of our patients had CD4+ counts

below 200/mcl at time of diagnosis with no underlying

cause, including a negative HIV serology. While idiopathic

CD4+lymphocytopenia is a relatively rare condition, it has

been characterized as a predisposing factor for some oppor-

tunistic infections including cryptococcosis. Zonioset al.

found only 53 cases in their review of all reports of cryp-

tococcosis in patients with idiopathic CD4+lymphocytope-

nia over a 12-year span [5].The proportion of idiopathicCD4+lymphocytopenia in our patients was exceptionally

high but seemed to be in accordance with the nding of

Jiang et al., i.e., the average level of CD4+ T cells appears

to be signicantly lower in the Chinese population [13].

In the present study, 32 patients received amphoteri-

cin B and ucytosine as primary therapy for a median

duration of 13 weeks. The average amphotericin B dos-

age was 0.47 mg/kgd (0.311.19 mg/kgd), lower than

Table 6. Comparison between predisposed hosts and normal hosts.

N/Total/ Median (%/Range)

PPredisposed hosts Normal hosts

Male 28/51 (54.9) 66/103 (64.1) .272

Age 48 (14~67) 35 (9~75) .0001

Time to diagnosis 30 (1~124) 40 (6~2890) .0062

Initial therapy

Not evaluated 2/51 (3.9) 10/103 (9.7) .207

AmB based treatment 31/51 (60.8) 73/103 (70.9) .208

Flu based treatment 11/51 (21.6) 13/103 (12.6) .150

Other treatments 9/52 (17.7) 12/103 (12.6) .402

Symptoms

Fever>39C 21/51 (41.2) 17/103 (16.5) .001*

Coma 4/51 (7.8) 12/103 (11.7) .466

Seizure 9/51 (17.7) 35/103 (34.0) .035*

Cerebral herniation 4/51 (7.8) 26/103 (25.2) .010*

Cranial nerve involvement 24/51 (47.1) 56/103 (54.4) .393

CSF

CSF WBC 63 (0~756) 100 (0~1030) .0611

CSF cryptococcal antigen 1280 (10~>1280) 1280 (1~>1280) .7805

Cranial imaging

(prior to treatment)

Parenchymal lesions (CT) 4/13 (30.8) 12/45 (26.7) .771Parenchymal lesions

(MRI)

16/18 (88.9) 29/47 (61.7) .034*

Hydrocephalus

(CT and MRI)

2/27 (7.4) 14/67 (20.9) .115

Surgical procedure 3/51 (5.9) 24/103 (23.3) .007*

Prognosis

Success at week 2 35/51 (68.6) 69/102 (67.7) .902

Success at week 10 41/50 (82.0) 79/99 (70.7) .135

1-year all-causes

mortality

13/49 (26.5) 28/94 (29.8) .6891

1-year cause-specic

mortality

7/49 (14.3) 21/94 (22.3) .2950

NOTE. AmB, amphotericin B deoxycholate; Flu, uconazole; Other

treatments, amphotericin B colloidal dispersion, itraconazole, etc.; CSF,

cerebrospinal uid; WBC, white blood cell. *P< 0.05.

7/23/2019 print.50472555

9/11


578 Zhu et al.

0.7 mg/kgd. The overall success rates were comparable

to those reported by Bennett and Pappas (78.3% at week

2, 83.3% at week 10) [8,9]. The 1-year all-causes mortal-

ity was also as low as 9.4%, and there was no death

attributable to cryptococcal meningitis. The long dura-

tion of initial therapy may possibly explain the favorable

response and low mortality in our patients with lowerdoses of amphotericin B. However, whether a lower dos-

age of amphotericin B (0.7 mg/kgd) for a longer time

course is efcacious as initial therapy in non-HIV patients

remains to be explored in future studies.

In our study, the treatment response rates for patients

receiving uconazole as initial therapy were remarkably

inferior to the response of patients receiving amphotericin

B therapy. These results were noted at both week 2 and week

10. Both the all-cause and specic mortality were higher

than with patients receiving amphotericin B-based thera-

pies. Not including amphotericin B in initial therapy had

also been identied as the strongest independent factor pre-

dicting poor prognosis. Our data supports the concept thatuconazole is also not suitable for initial treatment for HIV-

negative patients, as it is in HIV-infected patients [30,31].

In our analysis of prognostic factors, the risk of death

in patients not receiving amphotericin B-based initial ther-

apy was about 79 times higher than those given amphot-

ericin B. Whether a patient was immunocompromised or

not did not affect the survival. Other factors that might be

related to patients long-term prognosis were; age 60

years, delays in diagnosis, coma, cerebral herniation, etc,

consistent with factors reported by others. But we have not

been able to include some other factors previously reported

for poor prognosis, including increased opening pressure,

positive results of India ink preparations, and higher cryp-tococcal antigen titers [8,16,18,32].

Acknowledgments

We thank Dr. John Bennett for his critical review and valu-

able suggestions about this manuscript.

Declaration of interest:The authors report no conicts of

interest.

References

Mitchell TG, Perfect JR. Cryptococcosis in the era of AIDS 1001 years after the discovery of Cryptococcus neoformans. Clin Microbiol

Rev1995; 8: 515548.

Mirza SA, Phelan M, Rimland D,2 et al. The changing epidemiology

of cryptococcosis: an update from population-based active surveil-

lance in 2 large metropolitan areas, 19922000. Clin Infect Dis2003;

36: 789794.

Dromer F, Mathulin-Pelissier S, Fontanet A,3 et al. Epidemiology of

HIV-associated cryptococcosis in France (19852001): comparison of

the pre- and post-HAART eras.AIDS2004; 18: 556562.

Antinori S, Ridolfo A, Fasan M,4 et al. AIDS-associated cryptococ-

cosis: a comparison of epidemiology, clinical features and outcome

in the pre- and post-HAART eras. Experience of a single centre in

Italy.HIV Med2009; 10: 611.

Zonios DI, Falloon J, Huang C, Chaitt D, Bennett JE. Cryptococcosis5

and idiopathic CD4 lymphocytopenia. Medicine (Baltimore) 2007;

86: 7892.

Pukkila-Worley R, Mylonakis E. Epidemiology and management of6

cryptococcal meningitis: developments and challenges. Expert Opin

Pharmacother2008; 9: 551560.

Singh N, Dromer F, Perfect JR, Lortholary O. Cryptococcosis in solid7

organ transplant recipients: current state of the science.Clin Infect

Dis2008; 47: 13211327.

Pappas PG, Perfect JR, Cloud GA,8 et al. Cryptococcosis in human

immunodeciency virus-negative patients in the era of effective azole

therapy. Clin Infect Dis2001; 33: 690699.

Bennett JE, Dismukes WE, Duma RJ,9 et al. A comparison of ampho-

tericin B alone and combined with ucytosine in the treatment of

cryptococcal meningitis.N Engl J Med1979; 301: 126131.

Dismukes WE, Cloud G, Gallis HA,10 et al. Treatment of cryptococcal

meningitis with combination amphotericin B and ucytosine for four

as compared with six weeks.N Engl J Med 1987; 317: 334341.

Dromer F, Mathoulin S, Dupont B, Brugiere O, Letenneur L. Com-11

parison of the efcacy of amphotericin B and uconazole in the treat-ment of cryptococcosis in human immunodeciency virus-negative

patients: retrospective analysis of 83 cases. French Cryptococcosis

Study Group. Clin Infect Dis 1996; 22: S154160.

Centers for Disease Control. Unexplained CD412 +T-lymphocyte deple-

tion in persons without evident HIV infection. United States. Morb

Mortal Wkly Rep1992; 41: 541545.

Jiang W, Kang L, Lu HZ,13 et al. Normal values for CD4 and CD8

lymphocyte subsets in healthy Chinese adults from Shanghai. Clin

Diagn Lab Immunol 2004; 11: 811813.

Chayakulkeeree M, Perfect JR. Cryptococcosis.14 Infect Dis Clin N Am

2006; 20: 507544.

Segal BH, Herbrecht R, Stevens DA,15 et al. Dening responses to thera-

py and study outcomes in clinical trials of invasive fungal diseases: My-

coses Study Group and European Organization for Research and Treat-

ment of Cancer consensus criteria. Clin Infect Dis2008; 47: 674683.Diamond RD, Bennett JE. Prognostic factors in cryptococcal menin-16

gitis. A study in 111 cases.Ann Intern Med1974; 80: 176181.

Dromer F, Mathoulin S, Dupont B, Laporte A. Epidemiology of cryp-17

tococcosis in France: a 9-year survey (19851993). French Crypto-

coccosis Study Group.Clin Infect Dis1996; 23: 8290.

Dromer F, Mathoulin- Plissier S, Launay O, Lortholary O. Determi-18

nants of disease presentation and outcome during cryptococcosis: the

CryptoA/D study.PLos Med 2007; 4: e21.

Kiertiburanakul S, Wirojtananugoon S, Pracharktam R, Sungkanu-19

parph S. Cryptococcosis in human immunodeciency virus-negative

patients.Int J Infect Dis 2006; 10: 7278.

Lui G, Lee N, Ip M,20 et al. Cryptococcosis in apparently immunocom-

petent patients. QJM2006; 99: 143151.

Lu CH, Chang HW, Chuang YC. The prognostic factors of cryptococ-21

cal meningitis in HIV-negative patients. J Hospital Infect1999; 42:313320.

Shih CC, Chen YC, Chang SC, Luh KT, Hsieh WC. Cryptococcal22

meningitis in non-HIV-infected patients. QJM 2000; 93: 245251.

Wu MH, Xu ZY, Weng XH,23 et al. A clinical analysis of 74 cases of

cryptococcosis.Acta Academiae Med Shanghai1987; 14: 457459.

Yao Z, Liao W, Chen R. Management of cryptococcosis in non-24

HIV-related patients.Med Mycol 2005; 43: 245251.

Tjia TL, Yeow YK, Tan CB. Cryptococcal meningitis.25 J Neurol

Neurosurg Psychiatry 1985; 48: 853858.

7/23/2019 print.50472555

10/11



Saag MS, Graybill RJ, Larsen RA,30 et al. Infectious Diseases Society

of America. Practice guidelines for the management of cryptococcal

disease. Clin Infect Dis2000; 30: 710718.

Bicanic T, Harrison T, Niepieklo A, Dyakopu N, Meintjes G.31

Symptomatic relapse of HIV-associated cryptococcal meningi-

tis after initial uconazole monotherapy: the role of uconazole

resistance and immune reconstitution. Clin Infect Dis 2006; 43:

10691073.

Husain C, Wagener MM, Singh N.32 Cryptococcus neoformans

infection in organ transplant recipients: variables influencing

clinical characteristics and outcome. Emerg Infect Dis 2001; 7:

375381.

Richardson PM, Mohandas A, Arumugasamy N. Cerebral cryptococcosis26

in Malaysia.J Neurol Neurosurg Psychiatry1976; 39: 330337.

Wu G, Vilchez RA, Eidelman B,27 et al. Cryptococcal meningitis: an

analysis among 5521 consecutive organ transplant recipients. Transpl

Infect Dis2002;4: 183188.

Ecevit IZ, Clancy CJ, Schmalfuss IM, Nguyen MH. The poor progno-28

sis of central nervous system cryptococcosis among nonimmunosup-

pressed patients: a call for better disease recognition and evaluation of

adjuncts to antifungal therapy.Clin Infect Dis 2006; 42: 14431447.

White M, Cirrincione C, Blevins A,29 et al. Cryptococcal meningitis:

outcome in patients with AIDS and patients with neoplastic disease.

J Infect Dis 1992; 165: 960963.

This paper was rst published online on Early Online on 15 April

2010.

7/23/2019 print.50472555

11/11

Copyright of Medical Mycology is the property of Taylor & Francis Ltd and its content may not be copied or

emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.

However, users may print, download, or email articles for individual use.

print.50472555

Documents