Difference-in-differences analysis of age-specific survival gains in acute myeloid leukemia following venetoclax approval.

Abstract

e23058 Background: Venetoclax in combination with a hypomethylating agent or low-dose cytarabine received accelerated approval from the FDA in 2018 for older or unfit patients with acute myeloid leukemia (AML). Prior real-world studies have sought to evaluate the impact of venetoclax approval, but they have used non-experimental statistical approaches that cannot disentangle venetoclax-specific effects from broader time-related improvements in AML outcomes. To address this limitation, we analyzed age-specific changes in AML survival using a quasi-experimental difference-in-differences (DiD) approach. Methods: We included all patients diagnosed with AML at age 20 or older between 2013 and 2022 in the Surveillance, Epidemiology, and End Results (SEER) 21 registry. Age was categorized into four groups (20-49, 50-64, 65-74, and ≥75), consistent with age ranges represented in venetoclax clinical trials, plus a younger comparison group. Parallel trends between age groups were assessed in the pre-venetoclax period (2013 – 2017), and DiD analyses compared outcomes before and after venetoclax approval, with 2018 excluded as a transition year. Primary outcomes included 1-year overall survival (OS) for 2013-2021 and use of cancer-directed agents for 2013-2022. Models were adjusted for race and sex. Wald tests were used for statistical inference. Results: Of the 59,676 patients with AML included, 15% were aged 20-49 years, 23% were 50-64, 27% were 65-74, and 35% were ≥75. Most patients were non-Hispanic White (68%) and male (55%), and 72% received cancer-directed agents. Pre-venetoclax trends in 1-year OS and cancer-directed agent receipt were parallel across age groups (Table). Following venetoclax approval, improvements in 1-year OS increased progressively with age: compared with 20–49-year-old patients, relative gains were 1.1% for ages 50–64 (p=0.5), 3.3% for ages 65–74 (p=0.02), and 4.2% for ages ≥75 (p<0.001). Similarly, post-approval increases in cancer-directed agent receipt were progressively larger with increasing age: compared with 20–49-year-old patients, relative gains were 1.9% for ages 50-64 (p=0.07), 4.7% for ages 65-74 (p<0.001), and 9.8% for ages ≥75 (p<0.001). Conclusions: Using a quasi-experimental DiD approach, we demonstrate that venetoclax approval was associated with greater improvements in 1-year OS and receipt of cancer-directed agents for older adults with AML. These findings suggest that the 2018 accelerated approval of venetoclax led to meaningful improvements in outcomes for patients with AML historically ineligible for standard induction chemotherapy. Age OS Slope (pre) (%/yr) p-value Treatment Slope (pre) (%/yr) p-value 1-yr OS (pre) 1-yr OS (post) %Tx (pre) %Tx (post) 20-49 1.06 - 0.43 - 75.2 75.4 90.7 90.1 50-64 0.59 0.37 0.22 0.68 54.7 55.8 85.1 86.2 65-74 -0.18 0.09 0.35 0.85 36.8 40.1 75.9 80.0 >75 0.74 0.54 1.41 0.07 18.3 22.6 49.8 59.0