Outcomes with Real-World Use of Azacitidine and Venetoclax in Newly Diagnosed Acute Myeloid Leukemia Are Comparable to Long-Term Viale-a Results

Background: Based on the landmark VIALE-A trial of azacitidine + venetoclax (AZA-VEN) versus (vs) placebo-AZA, hypomethylating agent (HMA) therapy with VEN is the gold standard for frontline treatment in elderly/unfit patients (pts) with acute myeloid leukemia (AML). Due to improved response rates and overall survival (OS) with this regimen in elderly/unfit pts, there are ongoing trials to compare HMA-VEN to intensive chemotherapy in younger/fit pts. Despite improvements in outcomes for AML pts with this regimen and its expected increased use in fitter pts, the real-world usage of HMA-VEN is quite complex, and toxicities leading to dose modifications are common. As a result, dosing and administration of HMA-VEN can be inconsistent, and it is unclear how variability in dosing impacts outcomes for AML pts receiving this regimen. We hypothesized that toxicities and dose adjustments are observed frequently with AZA-VEN for elderly or unfit AML and that outcomes may be inferior to AZA-VEN treated pts with long-term follow-up on VIALE-A.

Methods: Newly diagnosed AML pts deemed unfit for intensive induction and initiated on frontline AZA-VEN at our center between 8/2018 and 6/2022 were included. Disease and pt characteristics, stem cell transplant (SCT) status, and survival outcomes were collected. European LeukemiaNet (ELN) 2017 criteria were used to assign AML risk. Treatment details for the first three cycles (C) of AZA-VEN, including dosing, treatment duration, and toxicities were retrospectively reviewed. OS was defined as time from initiation of AZA-VEN to death from any cause, with surviving subjects censored at time of last follow-up. Kaplan-Meier method was used to summarize OS between pts receiving <28 days VEN (including 21, 14, and 7 days) and ≥ 28 days (including 28 days or > if cycle length was extended). Outcomes in this cohort vs long-term follow up results of the VIALE-A AZA-VEN arm were compared with Fisher exact tests.

Results: 53 pts received frontline AZA-VEN and were included in this analysis. Median age at diagnosis was 72 (range, 24-83), and 58% were male. Race (92% White and 8% Black) and ethnicity (96% not Hispanic/Latino) were self-reported. ELN risk was 11% favorable, 47% intermediate, and 42% adverse. With 100% of pts on concomitant azole (91% on posaconazole), VEN 100 mg was the most common dose received in C1 (74%), C2 (69%), and C3 (64%). Despite VEN adjustments mostly due to toxicity in 42% of pts in C1, 54% in C2, and 60% in C3, 28 days was the most common duration received (49% in C1, 46% in C2, and 28% in C3) followed by 21 days; 14-day duration was provided in 4% of pts in C1 and increased to 28% in C2 and 20% of pts in C3. Among pts with dose adjustments due to toxicity, cytopenia was most common, indicated for 57% of toxicities in C1, 79% in C2, and 100% in C3. In C1 only, GI toxicity and tumor lysis syndrome were indicated for 14% and 19% of reported toxicities, respectively. 16 pts had bone marrow biopsy (BMBx) between C1 and C2, 17 between C2 and C3, and 2 after C3. Best response was available for 58% pts with 45% achieving CR, 26% CRi, 19% MLFS, and 10% refractory. With median follow-up of 34.6 months (mo) (range, 0.8 to 52), median OS was 15.2 mo (95% CI 6.8 to 33.7). No significant difference was seen in OS for C1 VEN duration <28 days vs ≥28 days (HR 1.67, P=0.15). Comparison of this cohort to VIALE-A VEN-AZA pts revealed the following: median follow-up 34.6 vs 43.2 mo, median age 72 vs 76, median OS 15.2 vs 14.7 mo, 24 mo OS 42.2% vs 37.5%, CR rate 45.2% v 38.8% (P=0.56), CR/CRi rate 71.0% vs 67.8% (P=0.69), and SCT 19% vs 1% (P<0.001).

Conclusions: With similar median follow up to VIALE-A, median OS and 24 mo OS were comparable in our cohort. Despite pts initially being deemed unfit and the median age of 72 in our cohort, 19% of pts received SCT compared to 1% in VIALE-A, highlighting the potential for consolidative SCT in select pts receiving upfront AZA-VEN. Despite variable dosing and duration of VEN in our cohort, no significant difference was observed in OS for those receiving 28 days vs <28 days. BMBx was performed in only 58% of pts, highlighting a pertinent aspect of care warranting further attention in real-world use of AZA-VEN. Based on these results, we seek to further optimize dosing strategies for AZA-VEN by evaluating interpatient variability through an ongoing prospective study of the genetic determinants of toxicity and response to AZA-VEN in newly diagnosed AML.