Impact on Antifungal Susceptibility Patterns of Previous vs. Revised Clinical and Laboratory Standards Institute Breakpoints for Candida Species Isolated from Candidemia: Experience of Two Tertiary Care Institutions in Japan

Abstract

Candidemia has a high mortality rate. Identifying prognostic factors of candidemia, based on each regional data, is essential for better management. The Clinical and Laboratory Standards Institute (CLSI) recently revised Candida species-specific breakpoints (R-BP) for antifungal agents. Few studies have investigated the detection performance of resistance in Candida species by comparing the R-BP and previous species non-specific CLSI breakpoint (P-BP) among patients with candidemia. The primary objective was to investigate the impact of the R-BP on the antifungal susceptibility patterns of Candida species, while the secondary objective was to identify the prognostic factors of candidemia. A total of 193 Candida species isolated from 187 patients with candidemia between January 2007 and December 2016 were examined. Susceptibility based on CLSI M27-A3 was defined as the P-BP and based on species-specific CLSI M59 or M60 breakpoint was defined as the R-BP. Multivariate Cox's hazard analysis was performed to identify prognostic factors within 30 days of the diagnosis of candidemia. A significant difference was observed in the susceptibility rate to fluconazole (FLCZ) (P-BP; 93.0% vs. R-BP; 79.4%) and to voriconazole (VRCZ) (P-BP; 97.2% vs. R-BP; 91.0%). The susceptibilities of C. parapsilosis, C. glabrata, and C. tropicalis to azole antifungal agents were markedly lower with the R-BP. Based on the R-BP, anti-fungal therapy was regarded as inappropriate for approximately 10% of the patients. The 30-day mortality rate was 29.4%. In a multivariate Cox's hazard analysis, age, lung disease, C. albicans, and the absence of antifungal therapy were associated with a high mortality rate, whereas serum albumin, C. parapsilosis, surgical wards, the removal of central venous catheter (CVC), and follow-up blood culture tests to confirm the clearance of Candida species were associated with a lower mortality rate. Early initiation of antifungal therapy, removal of CVC, and follow-up blood culture tests are essential for improving the outcome. The R-BP efficiently detected non-susceptible strains to FLCZ and VRCZ, particularly in non-albicans Candida species. The present results support the importance of antifungal susceptibility tests and interpretations based on the R-BP among patients with candidemia.