The AML Hub uses cookies on this website. They help us give you the best online experience. By continuing to use our website without changing your cookie settings, you agree to our use of cookies in accordance with our updated Cookie Policy

Predicting response and eliminating resistance to venetoclax

Mar 29, 2021

Acute myeloid leukemia (AML) is a heterogeneous disease, both clinically and genetically. Upregulation of members of the anti-apoptotic BCL2 family (i.e., BCL2 and MCL1) in AML result in a poor prognosis and resistance to treatment. The approval of venetoclax, a potent BCL2-selective BH3 mimetic, has changed the treatment outcomes of patients with AML; however, its activity is limited in relapsed or refractory and/or secondary AML, both as monotherapy and when combined with hypomethylation therapy.

Previous studies have shown that resistance to venetoclax may develop due to the upregulation of other anti-apoptotic members of the BCL2 family, including MCL1, BCL-xL and BCL-w. In this study, Haijiao Zhang et al. investigated the biomarkers to predict response to venetoclax by combining clinical parameters, whole-exome sequence (WES) data, and RNA-sequencing data with primary AML sample venetoclax response data from the Beat AML cohort. The results were published in Nature Cancer.1

Clinical and genetic factors affecting venetoclax sensitivity

Clinical characteristics, chromosome translocations, and distribution of venetoclax sensitivity were compared within the data obtained from the Beat AML trial (with WES, RNA-sequencing, and clinical information). There was a significant association between venetoclax resistance and

  • low bone marrow blasts (p < 0.0001) and peripheral blood (PB) blasts (p = 0.0002);
  • elevated monocyte (p < 0.0001) and neutrophil counts (p < 0.0003);
  • the French–American– British (FAB) transformed AML subtype (p = 0.065); and
  • FAB subtypes M4 and M5 (p = 0.021).

AML samples with high blast counts, PML-RARA translocation, or from the FAB M3 subset were found to be sensitive to venetoclax.

The association between common AML mutations and venetoclax sensitivity was also investigated. KRAS, PTPN11, and SF3B1 mutations were found to be relatively resistant to venetoclax and show reduced sensitivity to ABT-737 (an inhibitor of BCL2), BCL-w, and BCL-xL.

When the weighted correlation network analysis Beat AML gene expression ‘brown cluster’ was compared in the most venetoclax sensitive and resistant samples, there was a high degree of gene overlap (20.8–27.4%). Pathway analysis indication that these common genes were associated with the immune system, innate immune function, and neutrophil degranulation.

BCL2A1 overexpression

There was a positive correlation between BCL2A1 expression and area under the curve (AUC) values of venetoclax (which represented venetoclax sensitivity), with the strongest association compared to 16 other common BCL2 family genes. The overexpression of BCL2A1 was associated with

  • less apoptosis with different concentrations of venetoclax.
  • relative resistance to venetoclax combinations and other BCL2 inhibitors.

In addition, venetoclax-resistant AML samples showed relatively high BCL2A1 expression.

Targeting BCL2A1 was shown to promote apoptosis in U937 cells and inhibit cell growth in a variety of AML cell lines, indicating a possible on-target effect. Of note, there was no impact on the colony-forming ability of CD34+ hematopoietic stem and progenitor cells (HSPC). Knockdown of BCL2A1 induced sensitivity to venetoclax in primary AML cells and Molm13 cells.

CLEC7A or CD14 overexpression

CLEC7A (encoding CD369) and CD14, cell surface markers for granulocytes and monocytes and expressed at low levels in normal HSPCs, were found to be associated with resistance to venetoclax when expressed at higher levels. High expressions of both genes were observed in the FAB M4 and M5 subsets, while expression was low in AML samples with PML-RARA translocation or those in the M3 subset.

In order to validate these findings, a venetoclax sensitivity assay was done in leukemic blasts positive or negative for CD369 and CD14. Blasts positive for CD369 and positive/negative for CD14 showed higher venetoclax AUCs compared with blasts negative for both. There was also a correlation between CLEC7A and CD14 expression and BCL2A1 expression. Of note, KRAS mutation was also associated with higher expression of CLEC7A and CD14.

KRAS mutations

In this analysis, AML samples with KRAS mutations were associated with higher venetoclax AUCs compared with wild-type KRAS, and this observation was further evaluated comparing overexpression of wild-type and G12D KRAS in three different AML cell lines. G12D KRAS was shown to induce resistance to venetoclax in all cell lines including venetoclax combinations, and two BCL2, BCL-w, and BCL-xL inhibitors (ABT-263 and ABT-737) but not to an MCL1 inhibitor (AZD5991) or other single agents.

The in vitro analysis, performed to identify the potential mechanisms behind venetoclax resistance mediated by KRAS mutation, showed that

  • G12D KRAS expression was associated with reduced BCL2 and increased BCL2A1 transcripts.
  • KRAS-mutated AML samples showed increased BCL2A1 mRNA, and increased MCL1 along with significantly reduced BCL2, pBCL2, and BAX at the protein level.
  • G12D KRAS was associated with upregulation of CD40, which was found to induce venetoclax resistance.
  • there were no apparent differences in BCL2 family proteins and CD40 in G12D NRAS cells.

Overall, KRAS mutation was sensitive to MCL1 inhibition, and induces venetoclax resistance via upregulation of MCL1.

PTPN11 mutations

The association between venetoclax resistance and PTPN11 mutations was explored in mouse BM lineage-negative HSPCs and three human AML cell lines. Cells transduced to express A72D PTPN11 were resistant to venetoclax compared with HSPCs transduced to express an FLT3-ITD infusion. High-dose venetoclax (1,000 nM) was not effective in killing A72D PTPN11-expressing cells. In addition, PTPN11 mutation did not induce venetoclax resistance in CTS cells, where MCL1 expression level was low. Combinations of venetoclax, ABT-263, and ABT-737 could eliminate the resistance to some extent, while AZD5991 could fully eliminate the resistance based on in vitro and in vivo data.

Venetoclax resistance induced by PTPN11 mutations could be eliminated with AZD5991, a MCL1 inhibitor.

Overcoming venetoclax resistance by adding a MCL1 inhibitor AZD5991

Investigators investigated the potential of venetoclax combined with AZD5991 in eliminating resistance induced by BCL2A1 overexpression, KRAS, and PTPN11 mutations. This combination was shown to promote full sensitivity for KRAS- and PTPN11-mutated cells, and a partial recovery of resistance mediated by BCL2A1 overexpression with a robust cytotoxic effect and synergy. The combination of venetoclax and azacitidine/cytarabine did not circumvent the resistance of venetoclax.

An in vivo model was designed to further evaluate this finding, and mice were randomly given vehicle, venetoclax, AZD5991, or a combination. Venetoclax alone was associated with a slower leukemic progression, while the combination was able to significantly reduce the leukemia burden and prolong overall survival. Investigators suggested that the cytotoxic effect of the combination may be attributed to robust MCL1 and pMCL1 elimination.


The results of this study indicated that BCL2A1 expression was strongly correlated with resistance to venetoclax. Overexpression of BCL2A1 was not only associated with resistance to venetoclax monotherapy, but also to combinations of venetoclax and other BCL2-targeted therapies, suggesting an immediate unmet need for BCL2A1-targeted therapies for the treatment of AML. Clinical and genetic factors that may be related to venetoclax resistance included AML with high monocyte and neutrophil counts, the FAB M4 or M5 subset of AML, high levels of monocyte and neutrophil markers (CD14 and CLEC7A), and transformed AML, all of which may also be correlated with BCL2A1 expression. For predicting response to venetoclax-based approaches, blast and monocyte/neutrophil counts, as well as CD14 expression, may be considered good candidates as these are routinely monitored in clinical practice. Consistent evaluation of the FAB subtype and including CD369 into immunophenotyping panels may also help in predicting the response. KRAS and PTPN11 mutations were also associated with resistance to venetoclax. The combination of venetoclax and AZD5991, an MCL1 inhibitor, was found to be effective in eliminating resistance to venetoclax both in vitro and in vivo.

  1. Zhang H, Nakauchi Y, Köhnke T, et al. Integrated analysis of patient samples identifies biomarkers for venetoclax efficacy and combination strategies in acute myeloid leukemia. Nature Cancer. 2020;1:826–839. DOI: 10.1038/s43018-020-0103-x