Things (like AML salvage) go better with VEN(etoclax)

In this issue of Blood Advances, Unglaub et al¹ report a retrospective analysis in patients with relapsed/refractory (R/R) acute myeloid leukemia (AML), suggesting that nonintensive chemotherapy (non-IC) plus venetoclax (VEN) is a superior bridge-to-transplant therapy compared with treatment of physician’s choice (TPC) of varying intensities not containing the B-cell leukemia/lymphoma 2 inhibitor. The authors compared outcomes in 37 patients with R/R AML who received VEN-based salvage (mainly with azacitidine [Aza]) with 90 patients from the German Study Alliance Leukemia AML registry (SAL) treated with non–VEN-containing salvage. The results demonstrated superior overall response and event-free survival rates, as well as trends favoring the percentage of patients bridged to transplant and overall survival in the VEN-containing salvage group. The authors used propensity matching to ensure similarity in host- and disease-related features in the 2 groups, thus bolstering their conclusion that hypomethylating agent (HMA) + VEN regimens were safe and effective in the AML pre–allogeneic stem cell transplant (alloSCT) setting. These results, although not overly surprising, provide support for an important treatment strategy in this challenging AML population.

VEN is an oral B-cell leukemia/lymphoma 2 inhibitor that promotes apoptosis in response to cytotoxic stress in AML and other neoplastic cells. The addition of this drug to HMA or low-dose cytarabine (LDAC) is now considered the standard for the treatment of previously untreated adults with AML aged >74 years (or younger patients with significant comorbidities) based on the VIALE-A and VIALE-C trials.^2,3 In VIALE-A, the overall survival for those randomized to VEN plus Aza was superior to those assigned to Aza alone (14.7 vs 9.6 months; P < .001). Similarly, VIALE-C demonstrated better results with VEN plus LDAC than LDAC alone (7.2 months vs 4.1 months; P = .11 in planned analysis, respectively; 8.4 months vs 4.1 months in unplanned analysis, with 6 months additional follow-up). Although the survival benefit of adding VEN to non-IC in this population seems clear-cut, the magnitude of the benefit varied greatly according to certain disease features. For example, patients with IDH1/2 mutant AML had greater relative benefit from the combination than those with FLT3-ITD or TP53 mutations.

The success of HMA/VEN in the older “chemotherapy-unfit” population has prompted its evaluation in other settings, including initial therapy in younger patients, posttransplant maintenance, and salvage therapy. Retrospective comparisons of HMA/VEN vs standard initial IC, including daunorubicin, and those using propensity matching to decrease heterogeneity, suggest that (1) HMA/VEN is less toxic than standard intensive therapy and (2) is at least as, if not more, effective in certain subgroups (eg, patients with AML whose blasts harbor mutations typical of myelodysplastic syndrome).^4-6 A prospective randomized trial of HMA/VEN vs 3+7 chemotherapy is ongoing (#NCT04801797). Furthermore, a high rate of deep remissions has been observed in trials adding VEN to intensive initial regimens such as fludarabine/cytarabine/granulocyte-colony stimulating factor/idarubicin.⁷ 

What is the role of VEN when added to non-IC in the R/R setting (table)? The relatively favorable tolerance of the regimen makes it attractive as a “bridge to transplant,” in which a less toxic regimen could have clear advantages by potentially reducing posttransplant mortality. A multicenter retrospective study comparing Aza/VEN salvage with “no Aza/VEN” demonstrated a comparable overall response rate (55% vs 57%; P = .852); the side effect profiles in the 2 arms were similar.⁸ Although the authors concluded that Aza/VEN did not represent an effective salvage therapy, in contrast to the Unglaub report, the differences may have been due to the lack of propensity matching, leading to more favorable patients in the No Aza/VEN arm, inclusion of only first salvage patients, and no stated plan for consolidative transplant. Stahl et al evaluated factors that predicted response to VEN-based therapy in R/R AML and noted that AMLs with NPM1 mutations had relatively high response rates, whereas those with TP53, KRAS/NRAS, and SF3B1 mutations experienced worsened overall survival than the average patient.⁹ However, given the lack of prospective randomized trials comparing HMA/VEN with IC in R/R AML, Unglaub et al tackled this issue using a retrospective analysis, apparently demonstrating the superiority of HMA/VEN to standard salvage in the pre-alloSCT setting.

The conclusions of the report by the SAL investigators are provocative but ideally need to be confirmed by a prospective randomized trial. The cohorts were of small size, and the control TPC group was highly heterogeneous with respect to disease biology and salvage therapy. Many of the TPC patients received single-agent non-IC, which has been previously shown to lead to lower rates of measurable residual disease (MRD) clearance than HMA/VEN. Because patients who achieve an MRD-negative state before allogeneic hematopoietic cell transplant fare better than those who have higher disease burdens, deep remission should be the goal of any salvage therapy. Propensity matching is a useful but imperfect tool to ensure cohort equivalence. Moreover, the lack of incorporation of mutationally targeted therapy in the TPC group was a limitation that may have produced bias in favor of VEN. For example, R/R AML studies with gilteritinib vs IC in FLT3-mutant patients, enasidenib in mutant IDH2, revumenib in KMT2A rearranged or NPM1-mutant disease, and ivosidenib or olutasidenib in IDH1-mutant patients are promising and could (either alone or in combination with other agents) be a very effective bridge to transplant in the correct context. To complicate matters, recent data suggest that it might be advisable to proceed to alloSCT quickly after relapse is noted, using either no or minimal chemotherapy.¹⁰ Lastly, the now frequent use of VEN in the upfront setting may lead to a need to re-evaluate the data reported by Unglaub et al, because patients who relapse after prior VEN may not respond as well to HMA/VEN in the R/R setting as those in the current study.

Overall, the article by Unglaub et al provides useful data in favor of a VEN-based regimen vs salvage chemotherapy in patients with R/R AML. It will be important to build on these data, particularly in an era when VEN is given upfront, and targeted therapies are often used.

Conflict-of-interest disclosure: E.S.W. reports advisory board membership with Curis and Rigel Pharmaceuticals. R.M.S. reports advisory board membership with AbbVie, Amgen, Aptevo, AvenCell, BerGen Bio, Bristol Myers Squibb, Cellarity, CTI BioPharma, Curis Oncology, Daiichi Sankyo, ENSEM, Epizyme, GlycoMimetics, GlaxoSmithKline, Hermavant, Kura Oncology, Lava Therapeutics, Ligand Pharma, Redona Therapeutics, Rigel Pharmaceuticals, Syntrix, and Takeda; data and safety monitoring board membership with Aptevo, Epizyme, Syntrix, and Takeda; and research funding to Dana-Farber Cancer Institute for protocols, on which he is a local principal investigator, from AbbVie, Janssen, Novartis, and Syndax.