Background: Older patients with acute myeloid leukemia (AML), those with multiple comorbidities, and/or poor performance status are generally considered ineligible for intense chemotherapy and are often treated with combination therapy of hypomethylating agents (HMAs) (azacitidine or decitabine) and venetoclax (VEN) as the standard of care. However, clinical outcomes and treatment modifications in patients receiving HMA+VEN in the real-world data are not well understood. The purpose of this study is to describe patient characteristics, treatment modifications, and outcomes among adults with AML initiating HMA+VEN in the U.S.

Methods: This was a retrospective cohort study using the ConcertAI Oncology Dataset. Adults diagnosed with AML on or after January 2020, who initiated HMA+VEN as first-line (1L) therapy, were included. First-line HMA+VEN was defined as the initiation of both agents within 30 days of each other as the first treatment following diagnosis. The data cut-off date was April 2025. Data were extracted from unstructured and structured electronic medical records. Patients were followed from the HMA start (index date) to death or end of record, whichever occurred first. Treatment modifications, including hold (i.e., delay in next cycle), dose decrease/increase, frequency decrease/increase, and discontinuations, and reasons for treatment modifications were assessed based on physician documentation. Real-world complete remission (CR) including complete hematologic recovery, incomplete hematological recovery (CRi), or partial hematologic recovery (CRh) was examined based on physician documentation at any point between the index date and the end of VEN+HMA regimen + 30 days. The Kaplan-Meier method was used to evaluate real-world overall survival (rwOS), defined as the time from HMA initiation to death. Patients were censored at their last medical record.

Results: Among the 266 eligible patients, most were White (84%) and male (57%), with 47% age 75 years or older at the index date. Approximately 37% received care in a community setting, 41% in integrated delivery network (i.e., network practices with multiple specialty providers), and 22% in an academic setting. Almost half (47%) had non-AML comorbidities captured by the Charlson Comorbidity Index. While 25% of patients with a measurement had an ECOG performance status of ≥2, the most common status (46%) was 1.

Most patients were treated with azacitidine plus venetoclax (61%), while 39% received decitabine plus venetoclax. 37% of patients received their first HMA+VEN therapy as an inpatient, 35% as an outpatient, and 27% had an undocumented setting. Of patients with an ELN 2024 risk classification (n=160, 60%), 69% were classified as favorable risk, 8% as intermediate risk, and 23% as adverse risk.

Most patients (n=229, 86%) had at least one treatment modification during the follow-up period, with 192 patients having treatment modifications to both HMA and VEN. A total of 202 patients (76%) had at least one treatment modification to HMA, with treatment holds (46%) and HMA discontinuations (41%) being the most common modifications. Neutropenia (31%) and pancytopenia (26%) were the most frequently cited reasons. A total of 219 patients (82%) had VEN modifications including treatment holds in 55% and VEN discontinuation in 40%, primarily due to neutropenia (33%) and pancytopenia (33%). Overall, 94 (35%) patients had treatment hold to VEN as the first treatment modification, and in 46% of cases, the HMA was also modified.

Approximately 52% of patients attained CR/CRi/CRh on 1L HMA+VEN. A total of 56 patients (21%) received a subsequent line following a relapse/refractory event after 1L HMA+VEN. Median rwOS was 19.5 months (95% CI: 14.9, 25.6) with 136 death events recorded.

Conclusions: This real-world study showed that 1L combination of HMA+VEN was effective in managing AML patients that are ineligible for intense chemotherapy. However, treatment modifications due to cytopenia were common. The longer observed rwOS is likely reflective of a younger, high-functioning patient population with favorable ELN risk profile, subsequent therapies used (including stem cell transplant), and/or improvements in dose optimization in clinical practice. Future subgroup and sensitivity analyses are planned to further investigate how patient characteristics and treatment modifications influence clinical outcomes.

This content is only available as a PDF.
2025
Sign in via your Institution