Can the Gene Expression Profile of Patients with Acute Myeloid Leukemia Predict Complete Remission Following Induction Therapy?

Recent advances in sequencing technologies and genomics have led to the development of several targeted therapies such as BCL2 and Bromodomain and extra-terminal (BET) protein inhibitors for a more personalized treatment of patients with acute myeloid leukemia (AML), yet the majority of patients still receive standard induction chemotherapy. The molecular profiles of patients who are likely to respond to induction therapy and novel directed therapies remain to be determined. The expression of AML-related genes that are targeted by novel therapies such as BCL2 and BRD4, as well as functionally related genes and associated epigenetic modulators (TET2, EZH2, ASXL1, MYC) were analyzed in a series of 176 consecutive AML patients at multiple points during the disease course - diagnosis (Dx), post-induction (PI), complete remission (CR) and relapse (RL) - and their relationship with clinical variables and outcome investigated. Higher TET2 expression was observed PI and at CR compared to Dx, with significantly superior TET2 expression after induction therapy in the group of patients who reached CR compared to those who did not. Thus, the upregulation of TET2 at PI may be a marker of CR in AML patients. On the other hand, cells with high levels of MYC and BCL2 may be vulnerable to BRD4 inhibition.

BCL2 Expression at Post-Induction and Complete Remission Impact Outcome in Acute Myeloid Leukemia.

Advances in acute myeloid leukemia (AML) genomics and targeted therapies include the recently approved BCL2 inhibitor venetoclax. The association between BCL2 expression and patient outcome was analyzed in a series of 176 consecutive AML patients at diagnosis (Dx), post-induction (PI), complete remission (CR) and relapse (RL). Levels increased significantly at relapse (mean 1.07 PI/0.96 CR vs. 2.17 RL, p = 0.05/p = 0.03). In multivariate analysis, high BCL2-Dx were marginally associated with worse progression-free survival, while high PI levels or at CR had an independent negative impact on outcome (PI: HR 1.58, p = 0.014; CR: HR 1.96, p = 0.008). This behavior of high PI or CR BCL2 levels and increased risk was maintained in a homogeneous patient subgroup of age <70 and intermediate cytogenetic risk (PI: HR 2.44, p = 0.037; CR: HR 2.71, p = 0.049). Finally, for this subgroup, high BCL2 at relapse indicated worse overall survival (OS, HR 1.15, p = 0.05). In conclusion, high BCL2 levels PI or at CR had an independent negative impact on patient outcome. Therefore, BCL2 expression is a dynamic marker that may be useful during AML patient follow up, and BCL2 levels at PI and/or CR may influence response to anti-BCL2 therapy.