Real-world overall survival in patients with Acute Myeloid Leukemia treated with ivosidenib plus azacitidine in the USA: Trinetx-based analysis

Abstract Introduction In patients with acute myeloid leukemia (AML) who are unfit for intensive chemotherapy, the combination of ivosidenib and azacitidine (IVO+AZA) is approved in the USA and Europe for patients with isocitrate dehydrogenase 1 (IDH1) mutations, based on the results of the pivotal, randomized, phase 3 AGILE trial. However, real-world data on outcomes of IVO+AZA treatment are limited. As part of the REAL-IDH umbrella initiative (A Retrospective, Longitudinal, Real-World Evidence AML/mIDH1 AML Cohort Study Using Routinely Collected Health Data to Gather Information on AML Management), this study aims to describe baseline characteristics and overall survival (OS) in patients with AML treated with IVO+AZA using data from the TriNetX Dataworks Platform. Methods This retrospective cohort study utilized the USA-based TriNetX Dataworks Platform to identify adult patients (≥18 years) with newly diagnosed AML treated with IVO+AZA. Data were extracted from electronic medical records from a network of healthcare organizations and other data providers (e.g., cancer registries). Data were collected on adult patients with an AML diagnosis between January 1, 2019 and December 31, 2024. Data cut off was July 18, 2025. Patients were included if front-line IVO and AZA were initiated within the same 30-day period. Patients were excluded if they received systemic anticancer therapy within 12 months prior to IVO+AZA initiation. In the TriNetX database, any patient counts that are ≤10 are masked and thus numerically unknown. We assessed patient demographics, treatment duration, and OS (analyzed using the Kaplan–Meier method). OS was calculated from initial AML diagnosis date until death. As a comparison cohort, results from patients treated with venetoclax + AZA (VEN+AZA) were also retrieved from the same TriNetX database. Results Between 2019 and 2024, out of 45,481 patients included in the database with a diagnosis of AML, 23 (0.051%) patients were treated with IVO+AZA and were included in this analysis. Mean (SD) age was 71.7 (7.7) years, 12 (52.2%) patients were male and 17 (73.9%) were white in patients receiving IVO+AZA. Allogeneic stem cell transplantation (AlloSCT) was performed in ≤10 (≤43.5%) patients receiving IVO+AZA anytime post-index date, and 16 (69.6%) did not receive AlloSCT. The median follow-up time was 21.2 months (interquartile range [IQR]: 15.1–23.5 months) in the IVO+AZA population. The median duration of treatment was 3.5 months (mean, 6.5 months; SD, 11.3) and the median OS from initial AML diagnosis and from treatment initiation was not reached in the IVO+AZA cohort. One-year OS rate was 79.6% and 79.7% in the IVO+AZA cohort from AML diagnosis and from treatment initiation, respectively. A total of 2,228 (4.9%) patients received VEN+AZA during the same time period. As mutational status was not available and the cohorts are not directly comparable, only descriptive analyses were performed on the VEN+AZA patient population. The median OS from initial AML diagnosis and VEN+AZA treatment initiation was 15.1 months and 14.3 months, respectively. One-year OS rate was 52.2%. Conclusion This is the first report of a real-world cohort of patients treated with IVO+AZA from a comprehensive robust US database. Although limited by small numbers of patients and the absence of molecular data confirmation, OS outcomes are consistent with the results seen in the AGILE study. Results from those treated with VEN+AZA were consistent with results from the VIALE-A study. These preliminary findings support the utility of IVO+AZA in clinical practice, but further real-world evidence generation is warranted, including within the framework of REAL-IDH.