Real-World Comparison of Outcomes of Patients Treated with HMA Monotherapy Versus HMA-Ven – Insights from 1634 Patients Included within the Austrian Myeloid Registry of the AGMT Study Group

Background

Hypomethylating agents (HMA) combined with venetoclax (VEN) has become the 1^st line standard of care for many patients (pts) with acute myeloid leukemia (AML) and is currently being tested in pts with myelodysplastic syndromes (MDS).

Methods

Time-to-event endpoints were analyzed using the Kaplan-Meier method. Cox proportional hazards (CPH) models were applied to account for possible interrelations between predictors. Analyses were performed with SAS®9.4.

Results

1634 pts receiving 15099 cycles HMA (n=1416), and 2292 cycles of HMA-VEN (n=218) within the Austrian Myeloid Registry (NCT04438889) were identified. Among all pts treated with HMA vs HMA-VEN median follow-up (FU) was 10.8 vs. 8.4 months (mo) (p<.0001) 90 vs 67% (p<.0001) had died, mortality was 6 vs 6% at day 30 (p=1.00) and 11 vs 15% at day 60 (p=.048), 52 vs 42% (p=.0045) received the drugs as 1^st line therapy, 8 vs 12% proceeded to transplant (p=.023), median age was 74 vs 74 (p=.81), ECOG 1-2 was 76 vs 80% (p=.17), therapy-related disease was present in 15 vs 12% (p=.41), median bone marrow (BM) blast count was 13 vs 34% (p<.0001), median no. of cycles was 5 vs 4 (p=.034), median cycle duration was 28 vs 31 days for cycle 1 (p=.0021) and 28 vs 28 days for cycles 2-5 (p=ns), 49 vs 68% had a BM evaluation (BME) during therapy (p<.0001), median time to 1^st BME was 4.9 vs 1.0 mo (p<.0001), ELN22 response rates were 9 vs 17% (p=.0002) for complete remission (CR), 13 vs 23% (p<.0001) for CR/CRi, 16 vs 31% (p<.0001) for composite CR (cCR) and 19 vs 42% (p<.0001) for overall response rate (ORR), median time to best response was 5.5 vs 1.6 mo, p<.0001.

Use of ESA was higher (8 vs 5%, p=.0004), whereas G-CSF (17 vs 44%, p<.0001), prophylactic antibiotics (18 vs 37%, p<.0001), antifungals (10 vs 16%, p<.0001) and virostatics (18 vs 42%, p<.0001) were less often used among all HMA cycles vs HMA-VEN cycles.

Pts were hospitalized less often with HMA vs HMA-VEN (69 vs 83%, p<.0001), but median days in hospital (12 vs 11 days, p=.30), hospitalizations due to infections (37 vs 42%, p=.18), median duration of hospitalization for infection (13 vs 11 days, p=.12), rates of febrile neutropenia (32 vs 32%, p=1.00), the percentage of cycles in which adverse events G3-4 were reported (40 vs 39%, p=.35) and treatment emergent (worsening of baseline counts) G3-4 cytopenias were similar for both cohorts (anemia: 16 vs 12%, p=.072; neutropenia: 34 vs 32%, p=.64; thrombocytopenia: 23 vs 20%, p=.39; lymphopenia: 17 vs 21%, p=0.13).

For 1^st line AML pts receiving HMA (n=405) vs HMA-VEN (n=101) median age was 76 vs 73 yrs (p=0.012), ECOG was 0-1 in 68 vs 74% (p=0.22), t-AML was present in 15 vs 11% (p=.28), median BM blast count was 46 vs 52% (p=.067), median FU was 9.9 vs 9.1 mo (p=.58), 93 vs 59% (p<.0001) had died, mortality was 9 vs 5% at 30 day (p=.23) and 15 vs 9% at day 60 (p=.15). Median no. of cycles was 5 vs 7 (p=.38); ELN22 rates of CR (7 vs 21%, p<.0001), CR/CRi (13 vs 37%, p=.0001), cCR (13 vs 38%, p<.0001) and ORR (17 vs 50%, p<.0001) were lower in 1^st line AML HMA vs HMA-VEN cohorts, respectively.

Baseline variables with data missing in <100 pts (n=24) with p≤0.1 in univariate Cox regression (n=19) were included in the CPH model. After stepwise selection, 13 covariates remaining in the final model were used for multivariable adjustments. Median adjusted overall survival (OS) [95% CI] in the HMA vs HMA-VEN cohorts was 11.2 [10.6-12.2] vs 15.0 [12.5-17.6] mo for all pts (p=.0011, HR 1.401 [1.144-1.715]), and 11.2 [9.7-12.7] vs 16.3 [12.8-20.4] mo for 1^st line AML pts (p=.0012, HR 1.669 [1.226-2.274]).

Summary

This large prospectively collected real-world cohort validates HMA-VEN to have a similar safety/toxicity profile to HMA monotherapy but with sign. higher response rates and sign. longer adjusted OS. Most HMA-VEN cycles did not need to be delayed, which may be related to the prophylactic use of growth factors as well as antibiotics/antifungals/virostatics observed in our population. Pts were more often admitted to hospital for administration of HMA-VEN, but occurrence of G3-4 AE and hospitalizations for AEs were similar in both cohorts.

Compared with data from the pivotal phase 3 VIALE-A trial [Pratz K, AJH 2024], 1^st line AML pts in our cohort had similar baseline, treatment and toxicity characteristics, lower CR/CRi rates (37 vs 65%), but similar or better OS (16.3 vs 14.7 mo), despite the short median FU (9.1 vs 43.2 mo) and the lower percentage of events (deaths) at the time of analysis (59 vs 100%) observed in our cohort.