An analysis of treatment-naïve patients with acute myeloid leukemia (AML) receiving venetoclax plus azacitidine demonstrates that the European LeukemiaNet (ELN) AML genetic risk classification systems do not provide clinically meaningful stratification of overall survival (OS) for patients who are ineligible for intensive chemotherapy. The authors of the study, which was published in Blood, also provide a four-gene prognostic model that can stratify OS outcomes of patients with AML treated with venetoclax-azacitidine.1
“We developed a novel alternative mutational model based on three prognostic risk signatures, defined by the mutational status of only four genes corresponding with higher, intermediate, and lower benefit with venetoclax-azacitidine. The major advantage of this risk model is that it is simple and thus widely applicable to patients not only in clinical trials but also in the community setting,” said study author Hartmut Döhner, MD, medical director of the Department of Hematology/Oncology at Ulm University in Germany.
The ELN AML genetic risk classification systems and recommendations are based on cytogenetics and molecular disease criteria. These systems were developed in 2017 and 2022 using responses from younger patients with AML who received intensive induction chemotherapy with or without hematopoietic cell transplantation. Dr. Döhner and his colleagues hypothesized that these risk stratification systems, which categorize patients into one of three prognostic groups — favorable, intermediate, and adverse — may not be applicable to older patients being treated with venetoclax-azacitidine.
The authors conducted an exploratory post hoc pooled analysis of 279 patients treated with venetoclax-azacitidine in the phase III VIALE-A trial2 and the phase I trial of testing the same combination3 and 113 patients treated with placebo-azacitidine to understand the ability of the ELN 2017 and 2022 classification systems.
According to the ELN 2022 classification, in the venetoclax-azacitidine and placebo-azacitidine groups, 35 (12.5%) and 13 (11.5%) patients had favorable-risk AML, 41 (14.7%) and 15 (13.3%) had intermediate-risk AML, and 203 (72.8%) and 85 (75.2%) had adverse-risk AML, respectively.
Using ELN 2022 criteria, the median OS with venetoclax-azacitidine versus placebo-azacitidine was 39.0 versus 11.0 months (hazard ratio [HR] = 0.32) in the favorable-risk group, 15.2 versus 9.1 months (HR=0.51) in the intermediate-risk group, and 12.7 versus 9.3 months (HR=0.64) in the adverse-risk group. However, when the authors analyzed only patients who received the venetoclax-azacitidine combination, OS outcomes were similar for patients who had intermediate- and adverse-risk AML (median OS of 15.2 and 12.7 months, respectively). Similar results were observed by the authors using the ELN 2017 criteria. Overall, these results suggest that the ELN risk criteria do not distinguish outcomes well among patients with AML treated with venetoclax-azacitidine.
With these data in mind, researchers developed a prognostic tool using 29 genetic aberrations (six cytogenetic and 23 pooled or individual genes), including seven genes that were not part of the ELN — TET2, DNMT3A, IDH1, IDH2, FLT3-TKD, KRAS, and NRAS. They found that the patients could be categorized into one of three venetoclax-azacitidine benefit groups based on the mutational status of four of these genes: TP53, FLT3 internal tandem duplication (FLT3-ITD), NRAS, and KRAS. The higher-benefit group had AML with wild-type TP53 and KRAS/NRAS and without FLT3-ITD. The intermediate-benefit group had AML with wild-type TP53 and the FLT3-ITD and/or KRAS/NRAS mutation, and the lower-benefit group had TP53-mutated AML. Following the combination therapy, the median survival of these three groups was 26.5, 12.1, and 5.5 months, respectively.
Most patients (52.0% [145/279]) were in the higher-benefit group, with 25.4% (71/279) in the intermediate- and 22.6% (63/279) in the lower-benefit group.
“The initial phase I/II trials of venetoclax as a single agent and in combination with a hypomethylating agent or low-dose cytarabine suggested that AML with IDH1/2 mutations or NPM1 mutations is particularly sensitive to venetoclax-based therapy. However, in our statistical model, neither IDH nor NPM1 genes entered the predictive molecular signature for response to venetoclax-azacitidine,” Dr. Döhner said.
Dr. Döhner also highlighted the finding that concurrent mutations in signaling genes, such as FLT3-ITD, NRAS, and/or KRAS, had a significant effect on outcomes among IDH- and NPM1-mutated AML. “For example, NPM1-mutated AML without concurrent signaling gene mutations had a median OS of 39.0 months, whereas those patients with signaling mutations [had a median OS of] only 9.9 months. Signaling mutations have previously been shown to confer resistance to venetoclax and likely explain why IDH and NPM1 mutation as single genes were not identified in the predictive signature. Nevertheless, the effect size of concurrent signaling mutations on outcome was surprising.”
A limitation of the current study is the number of patients with AML included, as well as the smaller amount of data on unfit, older patients with AML in general, compared to younger patients with AML who receive intensive chemotherapy.
Along with the recently published 2024 ELN genetic risk classification for patients receiving less-intensive therapies,4 which also incorporates patients’ mutational status, these “genetic signatures provide a basis for consensus stratification of patients receiving less-intensive therapies,” Dr. Döhner said.
Further testing, validation, and refinement, including testing on real-world cohorts, of this proposed genetic prognostic tool is necessary for its widespread use in AML.
The Germany and Austria-based AMLSG Registry study (NCT01252485)5 has recently initiated analyses of a real-world cohort of approximately 500 patients treated with venetoclax-azacitidine or venetoclax-decitabine to further evaluate and validate the new prognostic signature for these regimens.
Any conflicts of interest declared by the authors can be found in the original article.
References
- DöhnerH, Pratz, KW, DiNardo CD, et al. Genetic risk stratification and outcomes among treatment-naive patients with AML treated with venetoclax and azacitidine [published online ahead of print, 2024 Aug. 12]. Blood. doi: 10.1182/blood.2024024944.
- DiNardo C, Jonas B, Pullarkat V, et al. Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med. 2020;383(7):617-629.
- DiNardo CD, Pratz K, Pullarkat V, et al. Venetoclax combined with decitabine or azacitidine in treatment-naive, elderly patients with acute myeloid leukemia. Blood. 2019;133(1):7-17.
- Döhner H, DiNardo CD, Appelbaum F, et al. Genetic risk classification for adults with AML receiving less-intensive therapies: the 2024 ELN recommendations [published online ahead of print, 2024 Aug. 12]. Blood. doi: 10.1182/blood.2024025409.
- Registry study on patient characteristics, biological disease profile and clinical outcome in acute myeloid leukemia and related neoplasms, and higher risk myelodysplastic syndrome - the Biology and Outcome (BiO)-Project (AMLSG BiO). NCT01252485.
Perspective
Q: What are the most important results of this study?
A: For many years, at the time of AML diagnosis, we have assessed patient disease risk according to genetic risk classification systems, most recently being the ELN AML genetic risk classification systems. These genetic risk groups (favorable, intermediate, and adverse risk) were derived from the outcomes of patients with AML who were treated using intensive chemotherapy regimens. Meanwhile, over the last decade, several new treatment regimens have been developed for patients with AML, notably including the combination of venetoclax-azacitidine for patients who are unfit for intensive chemotherapy or those older than age 75.
The current study by Dr. Döhner and colleagues explores whether the ELN risk groupings — and subsequent implications for patient responses to treatment and OS — are still applicable when a patient is treated with venetoclax-azacitidine, rather than intensive chemotherapy.1 Surprisingly, in the cohort studied, the authors found that the ELN risk groups don’t do a good job of predicting which patients will do well or poorly when using venetoclax-azacitidine. Building on this, the authors propose a new risk stratification tool, based on the presence or absence of four common mutations in AML — TP53, FLT3-ITD, NRAS, and KRAS — and resulting in three risk categories correlating with a low, intermediate, and high chance of survival benefit from venetoclax-azacitidine. This suggests that when we discuss prognosis with our patients, it doesn’t just matter which mutations are present but also which treatment strategy we plan on using.
Q: Were any of the study results surprising?
A: This study is a secondary analysis of the VIALE-A phase III trial, which showed a significant survival advantage for venetoclax-azacitidine compared to azacitidine alone.2 Surprisingly, however, this study suggested that typical ELN risk groupings didn’t do a very good job of predicting how well patients respond to the venetoclax-azacitidine combination. Indeed, using the most recent ELN 2022 update, there was effectively no difference in outcomes between patients with ELN adverse risk and those with ELN intermediate risk, making it hard to prognosticate for the great majority of adult patients with AML.
Q: Do these results have actionable implications for you as a clinician for patients with AML?
A: This study suggests to me that if I’m meeting a patient with a new AML diagnosis and feel that they are a candidate for venetoclax-azacitidine, I should not be using typical ELN stratification to define their risk. Rather, I need to use an alternative risk grouping, such as the one proposed here, to provide accurate treatment counseling about prognosis and the likelihood of response to venetoclax-azacitidine.
Q: What are the limitations of the study, and what else needs to be addressed before this new score is implemented in AML clinics?
A: Venetoclax-azacitidine has become a standard AML treatment regimen, and this study is an important step in understanding which patients benefit from this combination. But this study does not help identify which patients should get venetoclax-azacitidine instead of, say, intensive chemotherapy or chemotherapy incorporating a targeted therapy. Given the molecular heterogeneity of AML, it would be important to replicate these findings in larger cohorts of patients to confirm the results and the new prognostic grouping. It would also be helpful to build upon these findings in mutational subgroups. For instance, how do NPM1 mutations interact with other poor-risk mutations, and what effect might this have on prognostic category? Such co-mutation analyses were likely underpowered in this study. It would be similarly interesting to more extensively evaluate the prognosis associated with karyotypic abnormalities with venetoclax-azacitidine. Lastly, an important question that this study raises but is not designed to address is how, as a clinician, should I use genetic risk to decide whether to treat with intensive chemotherapy or with venetoclax-azacitidine for those patients who may be candidates for either therapy. In other words, is the patient more likely to respond to intensive chemotherapy or venetoclax-azacitidine?
Andrew Brunner, MD
Assistant Professor of Medicine
Massachusetts General Hospital
Boston, Massachusetts
Interview by Anna Azvolinsky, PhD
References
- DöhnerH, Pratz, KW, DiNardo CD, et al. Genetic risk stratification and outcomes among treatment-naive patients with AML treated with venetoclax and azacitidine [published online ahead of print, 2024 Aug. 12]. Blood. doi: 10.1182/blood.2024024944.
- DiNardo C, Jonas B, Pullarkat V, et al. Azacitidine and venetoclax in previously untreated acute myeloid leukemia. N Engl J Med. 2020;383(7):617-629.
