Comparisons of clinical features and mortality of cryptococcal meningitis between patients with and without human immunodeficiency virus infection.

BACKGROUND: Cryptococcosis is a systemic infection caused by Cryptococcus neoformans, and cryptococcal meningitis can occur in patients with late-stage human immunodeficiency virus (HIV) infection and other forms of immunosuppressive status. This study was designed to compare clinical features and laboratory findings of cryptococcal meningitis in HIV-positive and HIV-negative patients.
METHODS: From January 1, 2000 to December 31, 2009, all patients aged more than 18 years hospitalized at National Taiwan University Hospital with a diagnosis of cryptococcal meningitis were analyzed retrospectively.
RESULTS: In total, 88 patients with cryptococcal meningitis were identified and 37 (42%) were HIV infected. Cryptococcal meningitis occurred in young (mean, 38 vs. 60; p < 0.001) and male (97% vs. 63%, p < 0.001) populations more frequently among HIV-positive group with higher Charlson comorbidity score (mean, 7 vs. 4; p < 0.001), higher initial complaint of cough (36% vs. 16%; p = 0.032), lower cerebrospinal fluid (CSF) white count (mean, 26 vs. 86; p = 0.024), lower total protein of the CSF (mean, 88 vs. 149; p = 0.012), higher percentage of serum latex agglutination cryptococcal antigen titer exceeding 1:512 (77% vs. 50%; p = 0.026), more extraneural involvement (70% vs. 49%; p = 0.046), more cryptococcemia (68% vs. 35%; p = 0.003), and higher proportion of normal brain images (44% vs. 13%; p = 0.003) than HIV-negative group. The all-cause mortality rates on Day 30 and Day 90 were 23.9% and 31.8%, respectively. The independent risk factors for Day 30 mortality were altered mental status, extraneural involvement, absence of lymphocyte predominance, and absence of leptomeningeal enhancement (odds ratio: 7.84, 9.71, 0.22, and 0.07, respectively; 95% confidence interval): 2.03-30.27, 2.01-46.94, 0.06-0.80, and 0.01-0.49, respectively). Those for Day 90 mortality were serum white count more than 11,000/μL, higher Charlson comorbidity score, and absence of normal brain images (odds ratio: 5.39, 1.40, and 0.09, respectively; 95% confidence interval: 1.22-23.72, 1.11-1.76, and 0.01-0.78, respectively).
CONCLUSIONS: The clinical features of cryptococcal meningitis between HIV and non-HIV patients have some divergences, including age, sex, underlying diseases, CSF parameters, extraneural site involvement, fungemia, and so on. We also identified risk factors for mortality of this disease. However, the mortality of cryptococcal meningitis was not different in HIV-positive and HIV-negative patients in terms of Day 30 and Day 90 mortality.