Glucocorticoids as a risk factor for infection and adverse outcomes in non-HIV and non-transplant patients with cryptococcal meningitis.

ABSTRACT

BACKGROUND: Cryptococcal meningitis (CM), an opportunistic fungal infection affecting immunocompromised hosts, leads to high mortality. The role of previous exposure to glucocorticoids as a risk factor and as an outcome modulator has been observed, but systematic studies are lacking. OBJECTIVE: The primary aim of this study is to evaluate the impact of glucocorticoid use on the clinical outcomes, specifically mortality, of non-HIV and non-transplant (NHNT) patients diagnosed with CM. METHODS: We queried a global research network to identify adult NHNT patients with CM based on ICD codes or recorded specific Cryptococcus CSF lab results with or without glucocorticoid exposure the year before diagnosis. We performed a propensity score-matched analysis to reduce the risk of confounding and analysed outcomes by glucocorticoid exposure. We used a Cox proportional hazards model for survival analysis. RESULTS: We identified 764 patients with a history of glucocorticoid exposure and 1267 patients without who developed CM within 1 year. After propensity score matching of covariates, we obtained 627 patients in each cohort. The mortality risk in 1 year was greater in patients exposed to prior glucocorticoids (OR 1.3, CI 1.2-2.0, p = 0.002). We found an excess of 45 deaths among CM patients with previous glucocorticoid use (7.4% increased absolute risk of dying within 1 year of diagnosis) compared to CM controls without glucocorticoid exposure. Hospitalisation, intensive care unit admission, emergency department visits, stroke and cognitive dysfunction also showed significant, unfavourable outcomes in patients with glucocorticoid-exposed CM compared to glucocorticoid-unexposed CM patients. CONCLUSIONS: Previous glucocorticoid administration in NHNT patients seems to associate with 1-year mortality after CM adjusted for possible confounders related to demographics, comorbidities and additional immunosuppressive medications. Serial CrAg screening might be appropriate for higher-risk patients on glucocorticoids after further cost-benefit analyses.