The Use of Antifungal Prophylaxis Is Associated with Worse Outcomes in Patients with Newly Diagnosed AML Treated with Venetoclax/HMA: A Retrospective Analysis of Patients in the Flatiron Health Database

Introduction: Patients with AML are predisposed to the development of IFIs, which are highly morbid and often fatal when they occur. Antifungal prophylaxis (AFP) with mold-active azoles has been shown to reduce mortality in patients with AML receiving intensive chemotherapy¹ (IC) and is therefore regarded as an essential component of supportive care in this setting; however, the benefits of azole-based AFP in patients receiving less intensive-therapies, such as venetoclax and a hypomethylating agent (Ven/HMA), is less clear.^2,3 There is widespread variation in practice, both within academic and community practice, regarding the use of AFP, the optimal AFP regimen, and dose adjustment in the setting of Ven/HMA. Thus, we conducted a retrospective analysis of patients in a national de-identified electronic health record (HER) database who received Ven/HMA in the upfront setting to assess the practice patterns and the impact on antifungal prophylaxis on overall survival.

Patient population: Patient data was derived from a de-identified electronic health record (EHR) derived from participating academic and community healthcare centers. Patients included in this study received Ven/HMA as first line therapy for newly diagnosed AML from 2018 through 2023. We excluded patients for whom there were no record of receiving antibacterial or antiviral prophylaxis. Patients were classified as having received antifungal prophylaxis if they were administered or prescribed an antifungal agent within a window of -30 to +4 days of Day 1 of Ven/HMA.

Analysis: Baseline characteristics were summarized and compared according to receipt of AFP. Overall survival was compared between the two groups in a univariate analysis using a log-rank test and after adjusting for covariates using multivariate regression. For the multivariate analysis, variables associated with survival at a p-value of < 0.10 were included as covariates.

Results: 1,714 patients with AML who received first line Ven/HMA were identified. 150 of them were excluded because there was no documentation of them having received anti-infective prophylaxis. Of the 1,564 patients left, 420 (26.9%) of whom received any AFP with azoles. Among patients who received AFP, 172 (41%) received fluconazole, 128 (30%) received voriconazole, 52 (12%) received isavuconazole, and 68 (16%) received posaconazole. There were no differences in age, sex, race, insurance status, socioeconomic status, eastern cooperative group (ECOG) performance status, blast percentage, de novo percentage, European LeukemiaNet (ELN) 2017 risk, high risk mutations (TP53, ASXL1, RUNX1), receipt of cytoreduction, or HMA used. Patients treated at an academic center were more likely to have received AFP than patients treated at a community center (33% vs. 25%, p < .001). In univariate analysis, patients who received AFP had an increased risk of death (HR 1.17, 95% CI 1.01-1.34, P=.031) compared to patients who did not with a median OS of 10 months vs. 11 months, respectively. In a multivariate analysis with covariates including age, treatment setting, performance status, prior MDS or MDS/MPN, specific HMA used, and ELN risk, the use of AFP was still significantly associated with worse overall survival (HR 1.28, 95% CI 1.08-1.52, P=.004).

Conclusion: In a large national retrospective analysis, patients with newly diagnosed AML treated with Ven/HMA who received azole-based AFP had worse overall survival than patients who did not. These findings suggest that azole-based AFP is less beneficial or associated with an increased risk of adverse risk in patients receiving Ven/HMA. The former may be due to the lower risk of invasive fungal infections seen in patients receiving Ven/HMA,³ whereas the latter may be mediated by drug-drug interactions between venetoclax and AFPs.^4-6 The increased risk of adverse effects due to drug-drug interactions may be amplified by inconsistencies in appropriate dose-reduction practices. This study is limited by the fact that data on drug administration was missing for a large percentage of included patients, especially outpatients, increasing the risk of misclassification of AFP exposure. This further limited the ability to assess how venetoclax doses were modified. Finally, there may have been selection bias with some patients receiving antifungal prophylaxis because of additional patient or disease-related characteristics that could have impacted outcomes.