Candidemia in Patients with Hematological Malignancies: The Role of Prophilaxis and the Importance of Local Epidemiology for Treatment

Introduction: Candidemia is a serious condition with a high mortality rate in patients with hematological malignancies. It is thus important to understand the associated risk factors, as well as the need to establish adequate prophylaxis and early, effective therapy. The objective of this study was to determine the incidence of candidemia in hospital patients with hematological malignancies; to describe its clinical features and the risk factors associated with infection and with a poor outcome.

Materials and methods: An electronic database was used to identify cases with a positive blood culture for Candida spp in patients with hematological malignancies admitted to the Hematology Ward of Hospital Universitario La Paz between January 2000 to March 2008. The clinical history of each identified case was reviewed. SPSS 15.0 was used for the statistical analysis. Univariant analysis was carried out using χ².

Results: Forty seven patients were identified, with an annual incidence of 1%. The species identified were Candida parapsilopsis in 46% of cases (n = 22) and Candida albicans in 21.3% (n = 10); the remainder was distributed amongst C. guillermondi, C. tropicalis and C. krusei. The underlying hemalogic malignancies were non-Hodgkin lymphoma (34%, n = 16), multiple myeloma (19%, n = 9) and acute myeloid leukemia (17%, n = 8). 48.9% of patients underwent stem cell transplantation (45.3% allogeneic and 54.7% autologous). No significant association was found between the underlying hemalogic malignancy and the species of Candida that was isolated. The antifungals used in treatment were liposomal amphotericin in 48.9% of cases, fluconazole in 12.7%, caspofungin in 4.2% and voriconazole in 4.2%, with combined therapy in 30% of patients. MIC50 and MIC90 for fluconazole against C. parapsilopsis were 4 and 32, respectively, and 0.03 and 8, respectively against C. albicans. MIC90 against the other species was 0.03. MIC50 and MIC90 for amphotericin were 0.03 and 1, respectively, against C. albicans, C. parapsilopsis and C. krusei. Voriconazole, itraconazole and caspofungin were found to have an MIC90 of 0.03 against all species of Candida. Thirty seven point eight percet of patients were already receiving antifungal prophylaxis at the time of diagnosis of candidemia, although 90% of cases of C. albicans candidemia were not on prophylaxis (p<0.05). In terms of risk factors, 76.6% of patients had a central venous catheter, 78.8% were undergoing chemotherapy, 95.6% were receiving concomitant, broad-spectrum antibiotics, 21.7% were diabetic, 46.8% were receiving parenteral nutrition of which more than half (59.1%) were associated with C. parapsilopsis, 26.7% had a serious associated mucositis, 60.5% had less than 0.2 ×109/L neutrophils and 98% had less than 1.5 × 109/L neutrophils, 37% had kidney failure. Eight patients (17%) died as a result of candidemia: 4 from C. albicans, 2 from C. parapsilopsis, 1 from C. glabrata and 1 from C. krusei. Of the patients with C. albicans, 33% died, compared to 11.8% of those with other species of Candida (p<0.05).

Conclusion: Candida parapsilopsis was found to be the main causative species of candidemia in our centre, with a markedly high MIC50 and MIC90 for fluconazole, probably related to fluconazole prophylaxis. These findings highlight the importance of understanding the epidemiology of each centre when planning treatment and establishing an effective scheme of prophylaxis in high-risk patients to avoid the mortality associated with this type of infectious complication