[Risk factors for invasive fungal infections during intensive chemotherapy of acute leukemia--retrospective study]

The incidence, outcome and risk factors for developing invasive fungal infection were retrospectively analyzed in 150 patients with acute leukemia during intensive cytostatic therapy.

Patients with and without the diagnosis of fungal infection were compared according to age, sex, diagnosis, stage of disease, type of therapy, antimicrobial prophylaxis, duration of febrile episodes, duration of antimicrobial therapy, duration of antifungal therapy, chest x-ray findings, results of surveillance cultures for fungal species isolation, clinical diagnosis at discharge from hospital, and autopsy findings. Clinical findings in patients with confirmed fungal infection on autopsy were analyzed separately.

The incidence of fungal infection according to clinical diagnosis was 38.5%. The incidence among patients who died during therapy at autopsy was 78.5%. The incidence of Candida and Aspergillus infections at autopsy was 40% and 60%, respectively. Specific incidence could not be determined during life. The mortality was 59% in the group of patients with fungal infection, and 43% in the group of patients without fungal infection. During the study, an increase in the rate of fungal infection as well as a trend to prolonged survival of these patients were observed. On multivariate analysis, independent risk factors associated with a greater incidence of fungal infection were duration of hospitalization (p=0.04), duration of granulocytopenia with granulocyte count less than 0.5x10(9)/L (p=0.05), number of febrile episodes (p=0.01), duration of febrile episode (p=0.001), intestinal decontamination (p=0.02), duration of antibiotic therapy (p=0.01), positive chest x-ray finding (p=0.001), and year of therapy (p=0.02). On univariate analysis, a greater incidence of fungal infections was also associated with younger age, acute lymphatic leukemia, newly diagnosed disease and second relapse of the disease. The occurrence of fungal infections showed no correlation with the type of therapy, number of chemotherapy cycles, type of fungal species isolated from particular locations and frequency of colonization at particular locations. However, the number of colonized locations and number of fungal species was two to three times greater in patients with than in those without fungal infection.

Fungal infections are becoming an increasing problem during intensive therapy of acute leukemia and contribute to poor therapy outcome. The diagnosis of fungal infection during life is extremely difficult and frequently late. There is the need of a more precise diagnostic test that would provide earlier diagnosis. The knowledge of risk factors is helpful in the diagnosis and therapy of fungal infections. The suspicion of fungal infection in patients at risk justifies the introduction of antifungal therapy and contributes to better therapeutic outcome.