P514: treatment patterns and real-world outcomes of molecular subgroups in patients with acute myeloid leukemia receiving frontline venetoclax-based therapy

Background: Venetoclax (VEN) combined with azacitidine (AZA) represents the standard of care frontline therapy for patients with newly diagnosed acute myeloid leukemia (AML) not suitable for intensive induction chemotherapy. Treatment (tx) patterns including the dosing and duration of VEN in the community setting and subsequent tx following relapse remain unknown. Overall survival (OS) and event free survival (EFS) in specific molecularly defined subgroups (i.e., NPM1, IDH1/2, and FLT3-ITD/TKD mutated) have been reported in clinical trials, however these outcomes have not been examined using real-world (rw) data. A better understanding of rw tx patterns and outcomes in this population is needed. Aims: To describe tx patterns and outcomes of adult patients with AML receiving initial VEN-based therapy including VEN dosing, reasons for tx discontinuation, subsequent therapies received, and outcomes in patients with NPM1, IDH1/2, and FLT3-ITD/TKD mutated AML. Methods: This retrospective analysis utilized the COTA rw oncology database comprised of data abstracted from electronic health records of partnered healthcare provider sites in the United States. Adult patients (age ≥ 18 years) who received first-line (1L) VEN-based tx for AML and had available dosing information were included. Patients with missing or unknown data for key study dates were excluded. Subgroups included: academic vs. community setting, R/R vs. non-R/R, and presence vs. absence of specific mutations (i.e., NPM1, IDH1, IDH2, and/or FLT3-ITD). R/R was defined as having initiated 2L therapy after VEN-based tx. The index date for the study population was the date of initiation of 1L tx. The index date for analyses of the R/R population was the date of initiation of 2L tx. Patient characteristics, tx patterns, and outcomes were examined overall and by subgroups. Results: A total of N=331 patients met inclusion criteria of which 184 (55.6%) were male, 246 (74.3%) were White, and 267 (80.7%) were treated in community practices. Median follow up time from diagnosis estimated using the reverse-KM method was 24.3 months (95% CI: 21.8, 26.8). In patients with available molecular data for the given marker, 8.8% (N=19/217) had mutations in IDH1, 19.7% (N=43/218) had mutations in IDH2, 11.0% (N=24/219) had mutations in FLT3-ITD, and 19.7% (N=41/208) had mutations in NPM1. A total of 148 patients (44.7%) had VEN interruptions of ≥ 7 days during 1L. The most common reason for discontinuation of VEN-based tx was toxicity (20.8%), followed by AML progression (8.8%). Following 1L VEN-tx, 115 patients experienced R/R disease, of which 30 (26.1%) received intensive chemotherapy, 70 (60.9%) received low-intensity regimens (i.e., alternative regimens containing AZA or decitabine), and 10 (8.7%) received investigational tx. Median OS for the population was 13.9 months (95% CI: 11.8, 16.6), and differed by mutation status (13.1 months for IDH1 positive patients, 42.0 months for IDH2 positive patients, not reached for FLT3-ITD positive patients, and 42.0 months for NPM1 positive patients). Summary/Conclusion: Median OS for this study was comparable to results in the VIALE-A trial, despite the community-based nature of these data, with IDH2 or NPM1-mutated patients experiencing improved outcomes. In patients with R/R AML following 1L VEN-based tx, no specific tx pattern was observed, with a minority of patients enrolling on investigational therapy. These data highlight the need for novel tx for R/R AML patients following 1L VEN-based tx.Keywords: Acute myeloid leukemia, Venetoclax, Clinical outcome