Venetoclax plus hypomethylating agents as effective bridge therapy to allogeneic stem cell transplant in relapsed/refractory acute myeloid leukaemia: A subanalysis of the <scp>AVALON</scp> study

Allogeneic haematopoietic stem cell transplantation (allo-HSCT) remains the only curative option for patients with relapsed or refractory (REF/REL) acute myeloid leukaemia (AML). However, many patients—particularly older or unfit individuals—are unable to reach transplant due to poor responses to salvage chemotherapy or toxicities. In recent years, the combination of hypomethylating agents (HMAs) with venetoclax (VEN) has demonstrated robust activity and manageable toxicity in both newly diagnosed (ND) and REF/REL AML.1, 2 In this context, we conducted a subanalysis of the AVALON, a retrospective Italian cohort evaluating real-world outcomes of HMA-VEN in AML, to investigate the feasibility and outcomes of HMA-VEN as a bridging therapy to allo-HSCT in the REF/REL setting. AVALON included 147 patients with REF/REL AML treated with HMA-VEN between 1 January 2015 and 1 April 2020, across 32 haematology centres.3 Twenty patients who received allo-HSCT prior to initiating HMA-VEN were excluded. Of the 127 remaining patients, 36 (28.3%) underwent allo-HSCT following HMA-VEN therapy. Transplanted patients were younger, fitter, more likely to have de novo or refractory AML, and to combine with decitabine compared to non-transplanted patients. Among the transplant cohort, 72.2% achieved a composite complete remission (cCR), including complete remission (CR), CR with incomplete haematological recovery (CRi) or morphologic leukaemia-free state (MLFS), prior to transplant (Table 1; Table S1). The median time from HMA-VEN initiation to allo-HSCT was 3.8 months [IQ–IIIQ: 2.5–5.3]. Sixteen patients (44.4%) received haploidentical transplants, and 28 (77.8%) underwent myeloablative conditioning (MAC). In the entire REF/REL cohort (n = 127), consolidation with allo-HSCT following HMA-VEN was associated with significantly longer overall survival (OS) and event-free survival (EFS). The hazard ratio (HR) for OS was 0.53 (95% confidence interval [CI]: 0.28–0.98; p = 0.045). When limiting the analysis to those who achieved cCR (n = 49), outcomes were even more favourable (HR for OS: 0.30, 95% CI: 0.11–0.84; p = 0.021). Overlapping results were obtained for EFS. Analyses used a Cox model with allo-HSCT as a time-dependent covariate, adjusting for age, fitness and European LeukemiaNet risk. In the allo-HSCT group (n = 36), median OS and EFS were 12.9 months (95% CI: 5.46–32.63) and 11.8 months (95% CI: 5.10–32.63), respectively, with a median follow-up of 20.3 months (Figure 1). In univariate analysis using the Cox model, the main factors associated with OS and EFS were disease status at treatment initiation and type of the conditioning regimen (Figure S1). Borderline associations were also observed with pretransplant response status and AML type (for OS only) (Figure S2). The 1-year cumulative incidence of relapse (CIR) post-transplant was 18.4% (95% CI: 7.2%–33.6%) in the overall transplanted population, while 22.1% (95% CI: 7.5%–41.5%) among patients transplanted in remission. One-year non-relapse mortality (NRM) was 34.2% (95% CI: 19.0%–50.1%), decreasing to 24.2% (95% CI: 9.5%–42.6%) in those transplanted in cCR. The 6-month cumulative incidence of grade II–IV acute graft-versus-host disease (aGVHD) was 16.7% (95% CI: 6.6%–30.6%), with grade III–IV aGVHD in 13.9% of patients. The incidence of chronic GVHD was 8.3% (95% CI: 1.4%–24.0%) at 1 year and increased to 14.0% (95% CI: 3.2%–33.0%) at 2 years, with moderate-to-severe forms remaining uncommon (4.5% and 11.0% respectively). One-year GVHD-free, relapse-free survival (GRFS) was 44.4%, with a median GRFS of 10.0 months (Figure S3). Among transplanted patients, 14 of 36 (38.9%) were alive and disease-free at last follow-up. The most common causes of death were relapse (n = 9), GVHD (n = 5) and infections (n = 4). Our study provides further evidence supporting the feasibility and effectiveness of HMA-VEN as a bridge to allo-HSCT in patients with REF/REL AML. In this real-world Italian cohort, about a quarter of the REF/REL AML patients treated with HMA-VEN were able to proceed to allo-HSCT, with promising post-transplant outcomes in terms of remission, survival and manageable toxicity. Our results show that 72.2% of patients undergoing allo-HSCT after HMA-VEN achieved CR or CRi prior to transplant, consistent with previous studies reporting high CR/CRi rates in the REF/REL setting with VEN-based regimens.1, 2 The median time from HMA-VEN initiation to transplant was 3.8 months, underscoring the regimen's ability to achieve timely disease control, a critical factor for successful transplantation. This short time interval may also reflect that 44% of patients received a transplant from a haplo-identical donor, which may be arranged more rapidly than matched unrelated donors' allo-HSCT. Several studies have reported outcomes of VEN-HMA used as a bridge to allo-HSCT, but few have focused on REF/REL AML patients,4-6 often grouping them with ND patients. This makes it challenging to assess post-HSCT outcomes for the REF/REL subgroup. Sandhu et al. analysed 19 ND and 13 REF/REL AML patients (median age 62 years) and found CR/CRi rates after Venetoclax plus azacitidine (VEN-AZA) of 68.8% for ND-AML and 31.3% for REF/REL AML. Post allo-HSCT, their cohort showed a 1-year OS of 62.5%, relapse-free survival (RFS) of 43.8%, CIR of 37.5% and NRM of 18.8%.4 Kennedy et al. studied 46 ND and 42 REF/REL AML patients (including 5 s transplants), with a median age of 67 years. They reported 1-year OS, CIR and NRM rates of 78%, 36% and 17% respectively.5 Bang et al. evaluated post-allo-HSCT outcomes in 10 ND and 40 REF/REL AML patients treated with VEN-HMA therapy, with a median age of 53 years at transplant. The proportions achieving CR/CRi or MLFS were 90% for ND and 92.5% for REF/REL AML. After a median follow-up of 13.7 months, 1-year OS, RFS, CIR and NRM probabilities were 63.7%, 59.3%, 28.5% and 12.2% respectively. In our study, the median age at transplant was 56 years, and 72.2% of patients achieved a response to VEN-HMA prior to allo-HSCT. Post-transplant, 1-year OS, EFS, CIR and NRM were 53.4%, 47.4%, 18% and 34.2% respectively. Our findings indicate that consolidation with allo-HSCT following HMA-VEN therapy significantly prolonged OS in both the overall REF/REL cohort and particularly in patients achieving CR/CRi. EFS was consistent with OS data, as expected in this population and reported in other studies.4-6 In univariate analysis, we observed a significant benefit for patients receiving MAC conditioning compared to those who received Reduced intensity conditioning (RIC). Another factor associated with OS and EFS was disease status at treatment initiation. Compared to other studies,4, 5 our transplant cohort had a younger median age, and 90% of the population was fit, which may explain why 78% had an MAC and nearly half of the population had a transplant from a haploidentical donor. The finding that the majority of patients in this cohort were treated with haploidentical allo-HSCT and received high-intensity conditioning regimens further supports the idea that HMA-VEN can enable the successful use of allo-HSCT in patients who might otherwise not be candidates for transplantation due to insufficient remission or high disease burden (especially refractory diseases). Compared to other studies,4-6 our NRM rate is somewhat higher, likely reflecting differences in patients' characteristics and the small sample size of our REF/REL cohort. However, in patients transplanted in CR/CRi, the NRM decreased to 24.2%, suggesting that disease control before transplant is a critical factor in minimizing transplant-related toxicity. The majority of NRM events were due to infections and GVHD, which remain leading challenges post-allo-HSCT. Interestingly, the cumulative incidence of grade II–IV aGVHD was relatively low (16.7% at 6 months), and chronic GVHD occurred in only 8.3% at 1 year, suggesting that pretreatment with HMA-VEN may not exacerbate GVHD risk and could offer an immunologically favourable profile, possibly due to limited toxicity and myelosuppression-focused action. Our study also revealed a low relapse incidence (18.4% at 1 year), particularly in those transplanted in remission (22.1%), highlighting that HMA-VEN may not compromise post-transplant disease control. This compares favourably with prior reports where relapse rates often exceed 30% after allo-HSCT in REF/REL AML.4-6 Moreover, the GRFS of 44.4% at 1 year provides a meaningful measure of quality-adjusted survival in this patient population. A strength of our analysis lies in its real-world setting, capturing outcomes across multiple centres and reflecting different transplant practices. However, there are several limitations. First, the retrospective design and inherent selection bias may limit the generalizability of our findings. Patients selected for transplant were younger and fitter, with a predominance of de novo AML and refractory rather than relapsed disease. Second, time to response and disease status assessments were not standardized, relying on investigator judgement Third, the cohort size, while one of the largest real-life European series to date, remains modest, and longer follow-up is needed to better define the durability of responses and late toxicity. In conclusion, despite the high NRM, our data confirm the utility of HMA-VEN as an effective bridge to allo-HSCT in REF/REL AML, because it brings an advantage in terms of relapse, especially considering that approximately 30% of patients in our cohort were not in CR/CRi at the time of allo-HSCT. Furthermore, 38.9% of patients alive and disease-free at last follow-up have an expected good quality of life, as demonstrated by the GRFS of 44%. These findings support the broader implementation of HMA-VEN in this setting, especially for patients traditionally considered ineligible for intensive salvage therapy. Future trials comparing HMA-VEN to intensive chemotherapy as a pretransplant regimen, and studies incorporating matched related donor monitoring and maintenance strategies post-transplant are warranted to further optimize outcomes in this high-risk group. E.T., E.P., L.C., F.P. and G.Mart. conceived the analysis and contributed to the definition of methods and end-points. E.P. performed statistical analysis. C.Z. and I.L. were involved in study database creation and revision. P.Z., C.P., C.V., V.F., A.C., E.B., N.F., L.F., L.M., E.A., M.B., F.F., F.L., C.S., I.M., C.C., J.N., A.D.R., G.Marc., F.G., G.Mart. and E.T. enrolled patients in the study. All the authors contributed in writing and reviewing the manuscript. This work was supported thanks to the contribution of Ricerca Corrente by the Italian Ministry of Health. The authors declare no conflicts of interest. The AVALON study (CT.gov: NCT04070807) was approved by the Romagna Ethics Committee on 10 April 2019 (Prot. 3371/2019) and subsequently by the ethics committee of each participating institution. It was also conducted in accordance with the ethical standards in the 1964 Declaration of Helsinki. All patients provided informed consent. In this work, generative artificial intelligence (AI) has not been used. Questions regarding data sharing should be addressed to the corresponding author; data will be shared upon request. Figure S1. Kaplan–Meier curves for overall survival (top panels) and event-free survival (bottom panels), by disease status (left) and conditioning regimen (right). Figure S2. Kaplan–Meier curves for overall survival (top panels) and event-free survival (bottom panels), by response status (left) and type of acute myeloid leukaemia (right). Figure S3. Kaplan–Meier curve for graft-versus-host disease (GVHD)-free, relapse-free survival (events: acute GVHD grade III–IV, systemic therapy-requiring chronic GVHD, relapse or death from any cause in the first post-HCT year). Table S1. Patient characteristics of relapsed or refractory acute myeloid leukaemia patients treated with hypomethylating agent-venetoclax receiving allo-haematopoietic stem cell transplantation, overall and by disease status before transplant. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.