Impact of invasive aspergillosis occurring during first induction therapy on outcome of acute myeloid leukaemia (SEIFEM-12B study)

Background: Acute myeloid leukaemia (AML) patients are at high risk of invasive aspergillosis (IA) after first induction chemotherapy (CHT). Although IA risk factors have been identified, few data are available on impact of IA, occurring during induction phase, on overall AML outcome. Patients and results: The end point of this multicentre, case-control, study was to evaluate whether IA, occurring after first induction CHT, can affect treatment schedule and patient's outcome. We identified 40 AML patients (cases) who developed IA during first induction phase, 31 probable (77.5%) and 9 proven (22.5%). These cases were matched with a control group (80 AML) without IA, balanced according to age, type of CHT, AML characteristics and cytogenetic-molecular risk factors. The overall response rate to induction CHT was the same in the 2 groups. In the 40 cases with IA, the overall response rate to antifungal treatment was favourable (80%) but it was significantly affected by the achievement of leukaemia complete remission (CR) with induction CHT. In fact, in cases with AML responsive to induction CHT, responses of IA to antifungal therapy were 96% compared to 21% in cases of AML not responsive to induction treatment (P <.0001). The adherence to the schedule and full doses of CHT were reported in 35% of cases (14/40) and in 76% of controls (61/80) (P =.0001; OR 6.7; 95% CI 2.7-16.6). After first induction CHT, a significant higher number of cases (15/40; 37.5%) compared to controls (9/80; 11%) could not receive additional cycles of CHT (P =.0011, OR 4.8; 95% CI 1.9-12.3). The IA-related mortality was 22.5%. The median OS of cases was significantly worse than OS of controls with a difference of 12.3 months (12.1 vs 24.4 months, P =.04). However, the occurrence of IA during first induction phase did not have a significant impact on the OS of cases who achieved a CR of AML with induction CHT which are able to proceed, despite the IA, with their therapeutic program, achieving the same OS as the control group with AML in CR (P = ns). Conclusions: These data show that IA during first induction CHT can delay the subsequent therapeutic program and has a significant impact on OS, specifically in AML patients who did not achieved a CR of AML with the first course of CHT.