Initial treatment and outcome of Candida glabrata versus Candida albicans bloodstream infection

Abstract

Candida glabrata is a common cause of bloodstream infection (BSI) and exhibits decreased susceptibility to fluconazole. We sought to determine whether patients with C. glabrata infection were at increased risk of inappropriate initial therapy and mortality compared with the more fluconazole-susceptible species Candida albicans by performing a matched case-control study using the Prospective Antifungal Therapy Alliance registry of invasive fungal infections. C. glabrata BSI patients were matched to those with C. albicans BSI by age, sex, and underlying illness after screening all C. glabrata patients entered into the registry from March 2004 through September 2007. Of 161 patients with C. glabrata BSI included and matched to 161 C. albicans patients, those with C. glabrata were less likely to receive an adequate dose of fluconazole as initial therapy (12% versus 52%, P < 0.05) and more likely to receive an echinocandin (44% versus 26%, P < 0.05) or inadequately dosed fluconazole (32% versus 8%, P < 0.05) as initial therapy. Although time to initiation of therapy did not differ by species (P = 0.2), time to receipt of adequate therapy was longer for those with C. glabrata BSI (P < 0.001). Overall, C. glabrata patients were more likely to receive inadequate initial therapy (34% versus 11%, P < 0.05), but 4-week mortality was no different between groups (30% for C. glabrata versus 29% for C. albicans, P = 0.80). We found hematologic malignancy, age greater than 60, the presence of a central venous catheter at diagnosis, mechanical ventilation, and dialysis dependence to be independent predictors of 4-week mortality. The lack of difference in mortality between species may reflect the overriding importance of host variables and/or a difference in virulence by species: further study is needed to investigate these hypotheses.