Addition of venetoclax to hypomethylating agent offers limited incremental benefit for patients with treatment-naïve TP53-mutated myelodysplastic syndromes

Abstract

Abstract Introduction Myelodysplastic syndromes (MDS) with TP53 mutations (mTP53) are characterized by suboptimal treatment response and dismal overall survival (OS). Safety and activity of hypomethylating agent (HMA) and BCL-2 inhibitor venetoclax (Ven) combination (HMA-Ven) was shown in a phase 1b trial (Garcia et al. Blood, 2025), but the phase 3 randomized VERONA study did not meet its primary endpoint per recent report. Nonetheless, the comparative effectiveness of Ven addition specifically in mTP53-MDS remains unclear. Methods We conducted a retrospective study of clinical, genomic, and outcome data on patients (pts) with mTP53-MDS treated with frontline HMA or HMA-Ven at Vanderbilt Ingram Cancer Center. Treatment response was assessed with International Working Group 2023 criteria. Overall response included complete remission (CR), CR with partial hematologic recovery (CRh), CR with limited count recovery (CRL), partial remission, and hematologic improvement. Results Our study included 76 pts with mTP53-MDS treated with HMA (n=54) or HMA-Ven (n=22). The median age at diagnosis was 68 (range, 27-90) years [HMA-Ven vs. HMA: 67.5 vs. 68.5; p= 0.83]. Most pts (n=51; 67%) had MDS with biallelic TP53 inactivation (WHO 2022; HMA-Ven, n=19 & HMA, n=32). ECOG performance status was similar between treatment groups. Baseline median bone marrow (BM) blast percentage was higher in HMA-Ven vs. HMA cohorts (11% vs. 3.8%; p=0.004); 18 pts had 5-9% (IB1) BM blasts (HMA-Ven, n=6; HMA, n=12) and 23 pts had ≥10% (IB2) blasts (HMA-Ven, n=11; HMA, n=12). There were no significant differences between cohorts in frequencies of higher-risk (IPSS-R/IPSS-M) disease, complex karyotype, or deletion 17p. Overall, 22 pts (29%) had two TP53 mutations and median maximum TP53-VAF was 37.6%. Therapy-related MDS was more frequent in HMA-Ven cohort (64% vs. 37%; p=0.03). HMA-based treatment was initiated at a median of 28 days from diagnosis. The HMA used included azacitidine (HMA: n=18; HMA-Ven: n=11), decitabine (HMA: n=15; HMA-Ven: n=7), and decitabine-cedazuridine (HMA: n=21; HMA-Ven: n=4). In cycle 1, 13 pts received 7-14 days of Ven and 6 received Ven for >14 days in evaluable HMA-Ven treated pts. Pts received a median of 4 (1-26) cycles of HMA-based therapy. Overall, 38 pts (50%) had treatment interruption in first 2 cycles, with no difference between subgroups. Post-cycle 1 treatment modifications included HMA dose reduction (n=14), decrease in Ven dose/duration (n=5), dose reduction in both HMA and Ven (n=3), and Ven discontinuation (n=2); these occurred due to cytopenias (62%), disease progression/death (24%), and infections (9%), with no significant difference between groups. Rates of neutropenic fever (46%) and significant bleeding (14%) in first 2 cycles were similar between cohorts. In response-evaluable cohort (n= 52), there were no significant differences in overall response rate (ORR) or composite CR (CR+CRh+CRL) rates between HMA-Ven and HMA cohorts (45% vs. 35%, p=0.40; and 36% vs. 24%, p=0.27, respectively). In pts with increased blasts (IB1/IB2; n=41), ORR trended higher with HMA-Ven vs. HMA (53% vs. 33%; p=0.20). Median number of cycles till best response was significantly less for HMA-Ven vs. HMA (1 [1-2] vs. 3 [1-8]; p <0.001). At best response, fewer pts on HMA-Ven were independent of RBC transfusion (17% vs. 41%; p=0.17) and platelet transfusion (25% vs. 44%; p=0.31). Overall, 18 pts (HMA-Ven, n=3; HMA, n=15) proceeded to allogeneic stem cell transplant (allo-SCT). Transformation to acute myeloid leukemia (AML) was noted in 25 pts (33%), with no significant difference between HMA-Ven and HMA in AML transformation rates (42% vs. 32%; p=0.41) or median leukemia free survival (mLFS) (8.2 vs. 11.8 months [mo]; p=0.42). Median OS (mOS) was not improved with HMA-Ven vs. HMA (9.7 vs. 12.4 mo; p=0.37), but with a consistent trend of improved mOS in pts with IB2 (n=22) treated with HMA-Ven (9.7 vs. 3.7 mo; p=0.87). In a multivariable Cox regression model, allo-SCT was the only independent predictor of improved mOS (HR 0.37; p=0.01). Conclusions In our single institution cohort, HMA-Ven did not significantly improve response rates, mLFS, or mOS compared to HMA alone in pts with treatment-naïve mTP53-MDS. However, pts with increased blasts and mTP53 appeared to respond better to HMA-Ven. Allo-SCT should be considered for all eligible pts with mTP53-MDS, as it remains the single independent predictor of improved OS in this subgroup.