Incidence of Invasive Fungal Infections in Acute Leukemia Patients Utilizing Micafungin Prophylaxis Compared to Second-Generation Azole Prophylaxis
Morris, H; Moorman, MT; Mackey, MC; Erba, HP; LeBlanc, TW; Rizzieri, DA; Barak, I; Li, Z; Zapolskaya, T
Published in: Blood
Introduction/Background:Standard therapy for both acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL) is myelosuppressive, and patients are expected to be neutropenic for a prolonged period. Due to the high risk of infection, antimicrobial prophylaxis is warranted and should be continued throughout neutropenia. The incidence of documented invasive fungal infections ranges from 12 to 24% in patients with AML and approximately 6.5% in patients with ALL. Guidelines currently include posaconazole as a category 1 recommendation for antifungal prophylaxis, with echinocandins and other azole antifungals as category 2B. However, posaconazole can be subject to drug-drug interactions and can increase the risk of hepatotoxicity, both of which can be problematic in acute leukemia patients undergoing chemotherapy. For patients unable to receive posaconazole or another second-generation azole antifungal, micafungin is utilized.To our knowledge, there are no studies comparing second-generation azole antifungals to micafungin as prophylaxis in both AML and ALL during initial or relapsed/refractory induction therapy. The aim of this retrospective study is to compare the incidence of invasive fungal infections in patients with acute leukemia in the setting of micafungin or second-generation azole antifungal prophylaxis during initial and/or relapsed/refractory induction therapy at an academic medical center.Methods:Retrospective, single-center study from June 22, 2013 to June 22, 2018. Detailed chart reviews were performed by hand.Results:The incidence of invasive fungal infections within 100 days of chemotherapy was 11.2% in patients who received a second-generation azole and 2.7% in patients who received micafungin (P=0.7618). The incidence of invasive fungal infections within 100 days in patients who received a high intensity AML regimen was 11.2%, low intensity AML regimen 0.9%, and ALL regimen 1.8% (P=0.5378). The incidence of invasive fungal infections within 30 days of chemotherapy was 4.5% in patients who received a second-generation azole and 2.2% in patients who received micafungin (P=0.2447). The most common reason for micafungin prophylaxis was drug-drug interactions with standard ALL regimens that utilize vincristine. Of those with a documented fungal infection, 14 patients (6.3%) had a proven invasive fungal infection, and 17 patients (7.6%) had a probable/possible invasive fungal infection. The most common site of infection was in the chest (9.8%), followed by fungemia (2.2%). Other sites of infection include the brain, sinuses, liver/spleen, and skin. Beta-D-glucan was collected in 17 patients, and galactomannan antigen was collected in 26 patients. Three patients had a positive beta-D-glucan, and 1 patient had an indeterminate beta-D-glucan. Two patients had a positive galactomannan antigen. Fusarium species accounted for 4 positive cultures in patients. For those patients who did experience an invasive fungal infection, 23 out of the 31 patients (74.2%) were on posaconazole, 6 patients on micafungin (19.4%), and 2 on voriconazole (6.5%). Ten out of the 14 proven fungal infections (71.4%) occurred while on posaconazole prophylaxis, and 3 of these patients had posaconazole levels <0.7 mcg/mL. Twenty-one patients (67.7%) with documented infections had a neutrophil nadir < 0.1x109/L. The majority (93.5%) of those with breakthrough infections while on antifungal prophylaxis were switched to another antifungal.Conclusions:Micafungin is a reasonable alternative therapy for use in patients who are unable to receive second-generation azoles as antifungal prophylaxis during induction therapy for acute leukemia. Some invasive fungal infections occurred while patients had sub-therapeutic posaconazole levels, however others had breakthrough fungal infections with therapeutic levels. Our conclusions are limited by the small sample size and non-randomized, retrospective nature of this analysis.
