Background: Few reports have described the clinical and microbiological features of cryptococcosis in immunocompetent patients.
Aim: To compare clinical presentations and outcomes of cryptococcosis in immunocompetent vs. immunocompromised patients.
Design: Retrospective case series.
Methods: All culture- or histology-confirmed cases (n = 46) of cryptococcosis in two acute hospitals in Hong Kong (1995–2005) were included. Clinical presentations, rates of fungaemia, cerebrospinal fluid (CSF) parameters and clinical outcomes were recorded.
Results: Twenty patients (43.5%) were apparently immunocompetent, 17 (37.0%) had predisposing factors other than HIV infection, and 9 (19.6%) were HIV-positive. Thirty-one (67.4%) presented with meningitis, four (8.7%) with pulmonary cryptococcosis, and 11 (23.9%) with extraneural, extrapulmonary cryptococcosis. Of the immunocompetent patients with retrievable isolates (n = 8), three (37.5%) were Cryptococcus gattii; all isolates (n = 6) from immunocompromised patients were Cryptococcus neoformans var. grubii. Immunocompetent patients more commonly presented with meningitis (80.0% vs. 47.1%, p = 0.03), and tended toward lower rates of fungaemia (10.0% vs. 35.3%, p = 0.06) and mortality (25.0% vs. 52.9%, p = 0.06). Death was associated with fungaemia (p = 0.01) and underlying malignancy (p < 0.01). In cryptococcal meningitis, immunocompetent patients had longer mean time from illness onset to presentation (34.4 vs. 12.6 days, p = 0.02), more intense inflammatory responses (CSF: white blood cells 108 vs. 35×109/l, p = 0.03; protein 1.61 g/l vs. 0.79 g/l, p = 0.07), less fungaemia (0% vs. 26.7%, p = 0.04) and more satisfactory clinical outcomes (81.3% vs. 46.7%, p = 0.04).
Discussion: A substantial proportion of patients with cryptococcosis are apparently immunocompetent. C. neoformans var. grubii and C. gattii are the common causes. Immunocompetent patients tend to present with localized, indolent neurological disease, with more intense inflammatory responses but better clinical outcomes.
Cryptococcus neoformans can infect not only HIV-positive patients and non-HIV-infected immunocompromised patients, but also apparently immunocompetent patients.1–3 With a declining incidence of AIDS-related cryptococcosis in the HAART era, and increasing use of immunosuppressants worldwide,1,,4 non-HIV-infected individuals (with or without predisposing factors) may become the predominant infected group.
In contrast to immunocompromised patients, few reports have described the variety of C. neoformans, clinical presentation, cerebrospinal fluid (CSF) parameters and prognosis associated with cryptococcal infections among immunocompetent individuals, especially in subtropical regions such as Hong Kong. Such knowledge can assist in the clinical management of cryptococcosis.
Clinical data collection and definitions
A retrospective analysis was performed. All adult cases of any form of cryptococcosis diagnosed in two major acute general hospitals in Hong Kong over a 10-year period (1995–2005) were traced. Cases were identified through the hospitals’ computerized Clinical Management System (searching ICD-9 code for ‘cryptococcosis’ or ‘cryptococcal meningitis‘) and the Laboratory Information System of the microbiology laboratories. Patients were included if: (i) clinical and/or radiological features indicated active cryptococcal infection; and (ii) infection was confirmed by culture or histology.
All patients’ charts and computerized laboratory and radiology records up to the year 2005 were reviewed. Clinical presentation (classified as ‘meningitis’, ‘pulmonary’, and ‘extraneural, extrapulmonary’ sites of infection),3 rate of fungaemia, CSF parameters, and clinical outcomes were recorded, as were predisposing factors: malignancies, cirrhosis, end-stage renal failure, autoimmune disorders, diabetes mellitus, chronic use of corticosteroid or other immunosuppressive therapies, idiopathic CD4 T-cell lymphopenia, and HIV infection.5 Patients without these predisposing factors were regarded as immunocompetent. HIV serology had been performed in the majority of our cases, especially when no obvious predisposing factor was noted. Hong Kong has a low HIV population seroprevalence of <0.1%.6
Clinical management included blood cultures and CSF examination if symptoms or signs suggested neurological disease, assuming the patient's condition allowed this. Antifungal therapy was initiated as soon as diagnoses were microbiologically confirmed; treatment included amphotericin B infusion, with or without flucytosine, followed by fluconazole, or fluconazole alone. Clinical outcomes were defined as ‘unsatisfactory’ if death occurred during the admission for acute cryptococcal infection, or if microbiologically confirmed relapse occurred after discontinuation of antifungal therapy after initial improvement. ‘Satisfactory’ outcomes were defined as survival during the admission period, and no relapse after stopping therapy.3,,7
All clinical specimens were obtained from patients during their hospitalization. Yeasts isolated from normally sterile sites were identified as Cryptococcus using a combination of chromagar (CHROMagar CANDIDA), microscopic morphology on cornmeal agar at 25°C, and the API 20C Aux (bioMerieux) according to the manufacturer's instructions. All blood cultures requested for fungal isolation were routinely done on Sabouraud dextrose agar, and were incubated at 35°C for 30 days before ‘no growth’ was reported. Cryptococcal antigen detection was performed using the Cryptococcal Antigen Latex Agglutination System (CALAS, Meridian BioScience) according to the manufacturer's instructions. The sera samples were pre-treated with pronase and heat deactivation to remove rheumatoid factor and other non-specific interference, to enhance detection of cryptococcal capsular polysaccharide antigens.
Retrieval of isolates
C. neoformans isolates obtained from patients during their hospitalization in the study period were saved on agar slopes and stored at room temperature. Isolates were cultured onto Sabouraud dextrose agar plates at 30°C for 48 h. Nineteen isolates from 14 patients were available, and were stored on subcultures for further analyses. Isolates previously identified as C. neoformans using the methods described above, had their identity reconfirmed by demonstrating capsule production using India ink and a positive urea test. A standard strain of C. neoformans var grubii (ATCC90112) was included as a control.
DNA was extracted by a method modified from Haynes.8 One loopful of cells was suspended in 500 µl lysis buffer (50 mM Tris–HCl (pH 8), 10 mM EDTA, 150 mM NaCl, 2% SDS) and incubated overnight at 65°C. The suspension was then boiled for 10 min and centrifuged at 5000 rpm for 1 min. We then purified 100 µl of the extract using GFX PCR DNA and a Gel Band Purification Kit (Pharmacia), according to the instructions of the manufacturer.
Subtyping by polymerase chain reaction and restriction fragment length polymorphism (PCR-RFLP)
PCR-RFLP was done according to a previously described protocol9 with slight modification. The primers IDPLB1 (5' TGA GCT TCA GGC GGA GAG AGG TTT GG 3') and IDPLB1R (5' AGG CTG GGT GGT GTT GTC GTC ACC 3') were used to amplify the phospholipase B gene (PLB1). PCR of the PLB1 gene was done in a final volume of 100 µl. Each reaction contained approximately 100 ng DNA, 1×PCR buffer (Amersham Biosciences), 0.2 mM of each nucleotide (Amersham Biosciences), 40 pmol of each primer and 4 U of Taq DNA polymerase (Amersham Biosciences). PCR cycling was 94°C for 3 min followed by 35 cycles of 94°C for 45 s, 62°C for 45 s, and 72°C for 2 min, and a final extension step at 72°C for 7 min. For RFLP analysis, 25 µl of the PCR products was digested with 10 U of AvaI or 15 U of HindIII in their 1×corresponding buffer and water in a final volume of 30 µl. The digestion was done overnight at 37°C. Separation was done on a 3% agarose gel and visualized by staining with ethidium bromide, observed under UV transillumination. Isolates with RFLP profile AvaI (A1) HindIII (H1) were identified as C. neoformans var. grubii (serotype A, molecular type VNI), while those with RFLP profiles AvaI (A5) HindIII (H4) or AvaI (A6) HindIII (H4) were identified as Cryptococcus gattii (serotype B/C, molecular type VGI or VGII).9 The standard strain of C. neoformans var. grubii (ATCC90112) was used as a control.
Data reporting and analysis
All continuous data were expressed as means ± SE. Data were analysed and compared between different patient groups using the χ2 test, Fisher's exact (two-tailed) test, or an unpaired t-test, as appropriate. SPSS for Windows (release 11.5) was used for the analyses, and the level of significance was set at 0.05 for all comparisons (two-tailed).
A total of 46 patients with cryptococcosis were identified during the 10-year study period. All met the inclusion criteria, and 45 had positive cultures of Cryptococcus from one or more body sites. Histological confirmations of the diagnosis were obtained in three patients with pulmonary cryptococcosis (one without culture result), one with vitritis, and one with a subcutaneous abscess. Twenty patients (43.5%) were apparently immunocompetent, 17 (37.0%) had one or more predisposing factors other than HIV, and 9 (19.6%) were HIV-infected. The mean ages of immunocompetent patients and those with predisposing factors were 52.8 ± 3.6 and 47.8 ± 2.7 years, respectively (p = NS); M:F ratios were 1.2:1 and 1.6:1, respectively. All HIV-infected patients in this series were male. Predisposing factors included: underlying malignancies (n = 5), cirrhosis (n = 4), end-stage renal failure (n = 2), autoimmune disorders (n = 5), diabetes mellitus (n = 9), chronic use of corticosteroid or other immunosuppressive therapies (n = 10), and idiopathic CD4 T-cell lymphopenia (n = 1). The case with idiopathic CD4 T-cell lymphopenia had HIV infection and other causes of T cell depletion excluded.10 Of the 20 immunocompetent patients, 17 tested negative for HIV, and were also followed-up for a median period of 14 months (range 3–48 months) with no evidence of new systemic disease. The remaining three patients were not tested for HIV, as the diagnosis of cryptococcosis was only known after death. CD4 T-cell counts were checked in three cases and were normal (Table 1, legend).
↵aCD4 T-cell counts were performed in three patients: 540, 970, 550 × 106/l. bCombination therapy with amphotericin B and flucytosine (n = 19): immunocompetent 10, predisposing factors +ve 9. cHepatocellular carcinoma (n = 1), nasopharyngeal carcinoma (n = 1), carcinoma of prostate (n = 1), carcinoma of lung (n = 1), lymphoma (n = 1). dHBV-related (n = 4). eDiabetes (n = 1), lupus nephritis (n = 1). fLupus erythematosus (n = 4), vasculitis (n = 1). gChronic use of corticosteroids and other immunosuppressive therapies for organ transplants. hHIV excluded by serology. iSepsis syndrome (n = 6). jCryptococcoma (n = 2), pneumonia (n = 2). kPeritoneal-dialysis-related (n = 1), cirrhosis/ascites (n = 1). lSkin abscess (n = 2), vitritis (n = 1). One patient who presented with peritonitis also had fungaemia. Am, amphotericin B; Flu, fluconazole.
Cryptococcus neoformans isolates were available in 14 of the HIV-negative patients (8 immunocompetent and 6 with predisposing factors) for PCR-RFLP subtyping analysis. Eleven patients were diagnosed as having C. neoformans var. grubii infections, based on RFLP (AvaI (A1) HindIII (H1), corresponding to serotype A, molecular type VNI of C. neoformans). Three patients were diagnosed as having C. gattii infections: two had RFLP pattern AvaI (A6) HindIII (H4) (serotype B/C, molecular type VGII); and one had AvaI (A5) HindIII (H4) (serotype B, molecular type VGI) (Figure 1). Identical RFLP patterns were observed on repeated isolates from the same patients. The RFLP patterns of these isolates and their respective identities are listed in Table 2.
Restriction fragment length patterns of the PLB1 gene after PCR amplification. Numbers above each lane represent isolate numbers and RFLP patterns (see table 2 and text). MK is a 100 bp standard. ATCC 90112 is the standard strain of C. neoformans var. grubii. aAvaI restriction patterns. A1 pattern corresponds to C. neoformans var grubii (serotype A); A5 and A6 patterns correspond to C. gattii (serotype B or C). bHindIII restriction patterns. H1 pattern corresponds to C. neoformans var grubii (serotype A); H4 pattern corresponds to C. gattii (serotype B or C).
↵a–c Specimens obtained from the same patients, individual patient identified by letter.
Of the immunocompetent patients, only two reported exposure to pigeon droppings prior to illness onset. None of the patients reported visits to rural areas, forests, and farms, or had relevant occupational/recreational exposures. The three patients with C. gattii infections had no travel history outside Guangdong Province, China, for at least 1 year before their illness. All three patients were apparently immunocompetent, and presented solely with meningitis. Two of these patients died, despite treatment. Isolates obtained from the other five (62.5%) immunocompetent patients were C. neoformans var. grubii. All six isolates (100%) obtained from patients with predisposing factors were of the variety grubii. Isolates from HIV-infected patients were not available for subtyping analysis.
From Table 1, a total of 31 (67.4%) patients presented with meningitis (with or without dissemination), 4 (8.7%) with pulmonary, and 11 (23.9%) with extraneural, extrapulmonary cryptococcosis (6 cryptococcaemia alone, 2 peritoneal, 2 cutaneous, 1 ocular infection). Apparently immunocompetent patients presented more commonly with meningitis when compared to non-HIV-infected patients with predisposing factors (80.0% vs. 47.1%; OR 4.5, 95%CI 1.1, 19.2; p = 0.03); a similar trend was observed when they were compared to all immunocompromised patients (80.0% vs. 57.7%; p = 0.07). Extraneural, extrapulmonary manifestations were observed in eight (47.1%) non-HIV-infected patients with predisposing factors, while only two (10.0%) immunocompetent patients had such manifestation (cryptococcaemia alone). There were only two cases of isolated pulmonary cryptococcosis among our immunocompetent patients. Also, cryptococcosis among immunocompetent patients tended to be associated with lower rates of fungaemia, when compared to non-HIV-infected patients with predisposing factors (10.0% vs. 35.3%; p = 0.06), and to all immunocompromised patients (10.0% vs. 34.6%; p = 0.04).
Amphotericin B and fluconazole were given in 71.7% and 13.0% of cases, respectively. Although treatment regimens and duration were individualized in non-HIV-infected cases, nearly all patients who received amphotericin B were given a mean dose of 0.4 mg/kg/day or above. HIV-infected patients were treated according to a specific protocol with an average daily dose of amphotericin B of 0.7 mg/kg or above.11 Amphotericin B treatment was not given in 5/6 patients who presented with ‘fungaemia alone’ (‘sepsis’ syndrome), as the diagnoses were only confirmed after the patient died. The only survivor was given amphotericin B treatment.
A trend towards lower mortality was observed in apparently immunocompetent patients when compared to non-HIV-infected patients with predisposing factors (25.0% vs. 52.9%; p = 0.06) (Figure 2). When HIV-infected patients were included, mortalities was 25.0% and 42.3%, respectively (p = NS). Fungaemia (p = 0.01), underlying malignancy (p < 0.01), chronic hepatitis B (p = 0.01), and diabetes mellitus (p = 0.01), but not age, sex, autoimmune diseases, or chronic use of corticosteroid or immunosuppressants (p > 0.05), were associated with death among non-HIV-infected patients.
Survival curves for immunocompetent cryptococcosis patients (black) and non-HIV-infected patients with predisposing factors (grey).
Cryptococcosis among HIV-infected patients (n = 9) presented predominantly with meningitis (7 patients, 77.8%), and fungaemia was also detected in three (33.3%). There were two (22.2%) deaths in this group.
Table 3 compares the clinical features and CSF profiles of cryptococcal meningitis in immunocompetent patients and patients with predisposing factors. Immunocompetent patients presented with more indolent disease (p = 0.02), and lower rates of fungaemia (p = 0.04). CSF showed higher initial white blood cell (WBC) counts (p = 0.03) and a trend towards higher protein levels (p = 0.07). Computed tomography (CT) brain scan showed evidence of hydrocephalus in five 'immunocompetent' patients and one patient with predisposing factors. Magnetic Resonance Imaging (MRI) of the brain showed meningeal enhancement in five 'immunocompetent' patients and none of those with predisposing factors. Interestingly, one immunocompetent patient presented with fever together with a manic disorder.12 Finally, immunocompetent patients had more satisfactory clinical outcomes when compared to all immunocompromised patients (81.3% vs. 46.7%, p = 0.04) or to non-HIV-infected patients with predisposing factors (81.3% vs. 37.5%, p = 0.04). Fungaemia was associated with unsatisfactory outcome (death/relapse) among immunocompromised patients (p = 0.03). A total of 29 patients received intravenous amphotericin B with (n = 19) or without (n = 10) flucytosine as induction therapy; mean daily doses of amphotericin B given in the first week of illness were similar (0.55 vs. 0.60 mg/kg/day) in those who had satisfactory vs. unsatisfactory clinical outcomes, respectively (p = NS). The three deaths in the immunocompetent group received amphotericin B treatment at a mean daily dosage of 0.5–0.6 mg/kg/day, while all five patients who received a treatment dosage ⩾0.7 mg/kg/day survived.
↵Data are numbers (%) or means ± SE, as appropriate. ap < 0.05; bp 0.1, both 'immunocompetent vs. predisposing factors (all). cAll three deaths occurred in patients who received average amphoterin dosage <0.7 mg/kg/day. CSF, cerebrospinal fluid (initial); fungaemia, positive blood culture; CSF/serum antigen titre, cryptococcal antigen titre determined by latex agglutination; higher amphotericin dose, mean dosage ⩾0.7 mg/kg/day in the first week of treatment; focal neurological deficit, cranial nerve palsies, hemiparesis; altered mental state, confusion, impaired conscious levels; hydrocephalus, confirmed by CT scan. Mean CSF total cell count among HIV-infected patients = 14 × 109/l. Onset to presentation, mean ± SE time from illness onset to presentation.
A substantial proportion of patients with cryptococcosis are apparently immunocompetent. They tend to present with indolent forms of meningitis; the CSF responses indicate more intense inflammation, fungaemia is infrequent, and clinical outcomes are more satisfactory. C. neoformans var. grubii and C. gattii are the causes of these diseases. Immunocompromised HIV-negative patients present with more acute forms of meningitis and the CSF showed less inflammatory responses; extraneural (mostly extrapulmonary in our series) manifestations are more common. Cryptococcus neoformans var. grubii is the most common cause of disease.
These findings concur with previous reports describing cryptococcosis in non-HIV patients (of whom 10–40% were immunocompetent),3 and suggest that host immune status may affect disease manifestation. A higher proportion (30%) of meningitis patients have no predisposing factor identified, compared to patients who present with disease involving other sites (10–15%), and rates of extraneural involvement and fungaemia tend to be lower in immunocompetent patients presenting with meningitis.3,7,13,,14 Cryptococcal meningitis in immunocompromised patients is a more acute illness.7,,13 CSF white-cell count is usually lower, and meningeal enhancement on CT scan of the brain is less frequently observed.2,,7 Cryptococcaemia, and other extraneural, extrapulmonary manifestations are associated with immunodeficiency and poorer prognosis.3,13–15
C. neoformans is a pathogenic yeast with classically three varieties, five serotypes and eight molecular types.9C. neoformans var. grubii (serotype A) has a worldwide distribution and typically causes disease in immunocompromised patients. C. gattii (serotype B/C), recently promoted to species level,16 is mainly found in tropical and subtropical regions, and produces pulmonary cryptococcoma and neurological disease that affects predominantly immunocompetent individuals.1,13,14,17,,18 It has been isolated from a variety of environmental sources.13,17,,19 This report is the first to suggest that C. gattii, in addition to C. neoformans var. grubii, can cause disease in Hong Kong, especially among immunocompetent individuals. However, our case numbers of C. gattii infection were too small (n = 3) to allow further analysis of its effect on disease manifestation and outcome.19 Interestingly, all three C. gattii cases presented with meningitis, of whom two died, suggesting higher virulence of this organism.19,,20 Inefficiency in binding to complement components and inhibition of polymorphonuclear leukocyte migration to the site of infection have been postulated as mechanisms for such virulence.21,,22 Subtyping of the organism can be considered in immunocompetent patients presenting with cryptococcal meningitis. C. neoformans var grubii can cause disease in both groups of patients, and is still the predominant agent in both immunocompetent and ‘immunocompromised’ patients. The basis of virulence of the variety grubii in immunocompetent hosts remains largely unknown,18,,23 as does whether C. gattii is a cause of disease in our HIV-infected patients.24,,25
Fungaemia is an important manifestation of cryptococcosis and predicts a poorer prognosis,26–28 and therefore every patient with cryptococcosis should have blood cultures taken for fungal isolation.2 Occasionally, cryptococcaemia presents as ‘sepsis syndrome’ without obvious evidence of neurological involvement.15 As suggested by this series, and by other reports, the diagnosis of isolated cryptococcaemia is often delayed, resulting in high mortality.15 Therefore a high index of suspicion is needed; and serum cryptococcal antigen assay, fungal blood culture and antifungal therapy within 48 h should be considered as soon as the diagnosis is suspected.15
The optimal treatment (both regimen and duration) for cryptococcosis in non-HIV-infected patients is largely unknown,3,7,11,,29 and current recommendations are derived from experience in HIV-infected hosts, due to the lack of randomized controlled trials.30 The treatment regimens in our series were heterogeneous, and do not allow direct comparison. Nevertheless, dosages of amphotericin B given in patients with ‘satisfactory’ and ‘unsatisfactory’ outcomes were similar. Moreover, in our study, as in others, mortality was associated with underlying diseases including malignancy, chronic hepatitis and diabetes.3,7,,15 These observations suggest that, as in many other invasive fungal infections, host immunity plays a major role in determining the outcome.3,,7 Dosage of amphotericin B, however, may also be important in the treatment of cryptococcosis in non-HIV-infected patients, as a higher dose of 0.7–1 mg/kg/day may be associated with a slightly lower mortality.15
Cryptococcosis continues to cause significant morbidity and mortality in immunocompromised as well as immunocompetent patients. C. neoformans var. grubii and C. gattii are the usual causes of disease among immunocompetent patients in subtropical regions. Clinical presentations and outcomes of cryptococcal infections are affected by host immune status. The effect of the infecting species on clinical manifestations in our locality warrants further study.
We are grateful to Ms Jenny Ho for her assistance in data collection.
. Epidemiology and host- and variety-dependent characteristics of infection due to Cryptococcus neoformans in Australia and New Zealand. Australasian Cryptococcal Study Group. Clin Infect Dis 2000;31:499-508.
. Involvement of Multiple Cryptococcus neoformans Strains in a Single Episode of Cryptococcosis and Reinfection with Novel Strains in Recurrent Infection Demonstrated by Random Amplification of Polymorphic DNA and DNA Fingerprinting. J Clin Microbiol 1995;33:99-102.
. Treatment of cryptococcal meningitis associated with the acquired immunodeficiency syndrome. National Institute of Allergy and Infectious Diseases Mycoses Study Group and AIDS Clinical Trials Group. N Engl J Med 1997;337:15-21.
. Fever and human immunodeficiency virus infection as sentinels for emerging mycobacterial and fungal bloodstream infections in hospitalized patients >/ = 15 years old, Bangkok. J Infect Dis 1999;180:87-92.
. Comparison of the efficacy of amphotericin B and fluconazole in the treatment of cryptococcosis in human immunodeficiency virus-negative patients: retrospective analysis of 83 cases. French Cryptococcosis Study Group. Clin Infect Dis 1996;22(Suppl. 2):S154-60.