Early clinical and microbiological predictors of outcome in hospitalized patients with cryptococcal meningitis.

BACKGROUND

Cryptococcal meningitis causes high mortality in immunocompromised and immunocompetent patients. The objective of this study was to identify early predictors of clinical outcome, available at the first days of hospitalization, in patients with cryptococcal meningitis in a tertiary center in Brazil.

METHODS

Ninety-six cases of cryptococcal meningitis with clinical, epidemiological and laboratory data, and identification and antifungal susceptibility of the strains were analyzed. Quantitative CSF yeast counts were performed by direct microscopic exam with a Fuchs-Rosenthal cell counting chamber using an institutional protocol. Univariable and multiple analyses using logistic regression were performed to identify predictors, available at the beginning of hospitalization, of in-hospital mortality. Moreover, we performed a secondary analysis for a composite outcome defined by hospital mortality and intensive care unit transfer.

RESULTS

The species and the antifungal susceptibility were not associated with the outcomes evaluated. The variables significantly associated with the mortality were age (OR = 1.08, 95% CI 1.02-1.15), the cerebrospinal fluid (CSF) yeasts count (OR = 1.65, 95% CI 1.20-2.27), systemic arterial hypertension (OR = 22.63, 95% CI 1.64-312.91) and neurological impairment identified by computed tomography (OR = 41.73, 95% CI 3.10-561.65). At the secondary analysis, CSF yeast count was also associated with the composite outcome, in addition to the culture of Cryptococcus spp. from bloodstream and cerebral toxoplasmosis. The associations were consistent with survival models evaluated.

CONCLUSIONS

Age and CSF yeast count were independently associated with in-hospital mortality of patients with cryptococcal meningitis but Cryptococcus species identification and antifungal susceptibility were not associated with the outcomes. Quantitative CSF yeast counts used in this study can be evaluated and implemented in other low and middle-income settings.