Fungal and Bacterial Infections in Acute Myeloid Leukemia Patients Treated with Induction Regimens Including Fludarabine: A Retrospective Analysis of 224 Cases.

Abstract

Infections are the major cause of morbidity and mortality of acute myeloid leukemia (AML). Invasive Fungal Infections (IFIs) occur in at least 10 to 20% of the patients submitted to induction and consolidation treatments, are responsible for death during induction in up to the 5% of the cases and may cause a delay in consolidation and intensification therapy. Among the risk factors for IFIs it has been included the use of Fludarabine (Fluda), which can induce severe and prolonged immunosuppression. In this study we retrospectively analyzed the infections occurred in 224 newly diagnosed AML patients, aged at least 65 years, consecutively treated with an induction regimen including Fluda, Ara-C and idarubicine with or without etoposide (FLAI/FLAIE). During induction phase, 181/224 (81%) patients experienced a fever of undetermined origin (FUO), the incidence of Gram negative and positive sepsis was 16% (37/224) and 29% (65/224) respectively and 7/224 (3%) patients developed a possible/probable IFI. In 6/224 patients (3%) a proven IFI was found (4 aspergillosis and 2 candidiasis). We then collected the data of the incidence of infections during the first consolidation course (FLAI: n=70; high-dose Ara-C [HD-AC]: n=65; idarubicine and HD-AC: n=89). The overall incidence of FUO was 34% (76/224), the number of Gram negative and positive sepsis was 52/224 (23%) and 49/224 (22%), respectively and 2/224 (1%) patients developed a proven IFI (3 aspergillosis and 1 candidiasis). We subsequently evaluated the incidence of infections in the three different consolidation groups. No significant differences were observed in terms of FUO, Gram positive and negative bacteraemia/sepsis and possible, probable and proven IFIs, during consolidation with Fluda-based regimen and with HD-AC-based regimens. Interestingly, the overall incidence of IFIs during consolidation with FLAI was significantly lower than during consolidation with HD-AC-based treatment program (0% vs 9%; p=0.02). These data, even though retrospectively collected, suggest that Fluda-based chemotherapy is not followed by increased incidence of infections, in particular IFIs, in comparison with conventional non-Fluda based regimens commonly used for AML induction. In our series, Fluda-based consolidation chemotherapy caused a significantly lower incidence of IFIs compared to HD-AC-based consolidation. This may be related to the lower duration of neutropoenia in patients treated with FLAI with respect to those treated with HD-AC/HD-AC + idarubicine.