394. Outcomes in Patients With Disseminated Noncentral Nervous System Cryptococcus

BACKGROUND: Differentiating between localized and disseminated cryptococcal disease is key to the management of this infection, since induction therapy with amphotericin B and flucytosine is warranted in the latter. We compared mortality in disseminated Cryptococcus with non-central nervous system (CNS) involvement, with those with CNS involvement and localized pulmonary disease. METHODS: Demographics, predisposing factors, presentation, laboratory values, treatment and outcome data were collected retrospectively on patients hospitalized at an academic tertiary-care hospital for cryptococcal infection from 2002 to 2017. Outcomes were compared between three patient groups based on extra-pulmonary and CNS involvement. Survival analysis was performed using univariate and multivariate Cox Regression with censoring at 90 days. RESULTS: Of 312 patients identified, 63 (20%) had pulmonary, 154 (49.2%) CNS and 95 (30.4%) had disseminated non-CNS disease. At day 90, 38 (40%) from the disseminated non-CNC group had died, compared with 37 (24%) in the CNS disease and 13 (20.6%) in the pulmonary groups. After adjusting for age ≥55 years, organ transplant, end-stage liver disease (ESLD) and AIDS, 90-day mortality risk was higher in the disseminated non-CNS group compared with the pulmonary (HR 2.97 [95% CI 1.55, 5.7]; P = 0.001) and the CNS disease group (1.84 [1.16, 2.93]; P = 0.009) (Figure 1). Median [IQR] time to diagnosis was 10 [4, 19] days and not significantly different between groups (P = 0.752). Induction therapy for ≥2 weeks was more common in the CNS disease (64.3%) that in the pulmonary (33.3%) or disseminated non-CNS disease group (38.7%) (P = 0.01). Median duration of azole therapy in days was longer (315 [61, 750]) in the CNS disease than in the disseminated non-CNS (184 [23.5, 403.5]) or the pulmonary group (214 [86, 415]) (P = 0.04). CONCLUSION: Patients with disseminated cryptococcal disease without CNS involvement have higher risk for mortality than those with CNS disease. However, management of patient’s disseminated non-CNS cryptococcosis was similar to those with localized pulmonary infection. DISCLOSURES: All authors: No reported disclosures.