Optimal Therapy for Secondary Acute Myeloid Leukemia after Prior Hypomethylating Therapy

Background: Patients (pts) with secondary acute myeloid leukemia (sAML) emerging from myelodysplastic syndrome (AML-MR) with extensive prior hypomethylating agent (HMA) exposure remain a high unmet need with significantly worse survival (4.2 months) and complete remission (CR) rate (32%) than sAML without prior HMA independent of age, cytogenetic risk, or treatment intensity. Because these pts are often specifically excluded from (and therefore vastly underrepresented in) AML clinical trials, the optimal therapy for these individuals is uncertain. For instance, although venetoclax (Ven) + HMA therapy is typically offered to these pts, the VIALE-A trial specifically excluded sAML pts with any prior HMA therapy. Here, we performed a single-center retrospective analysis of the outcomes of these pts following intensive chemotherapy (IC) vs low intensity (LI) regimens for sAML diagnosis, focusing on clinical outcomes, specifically response rates, overall survival (OS), and ability to proceed to allogeneic stem cell transplantation (alloSCT) in first or second CR.

Patients and Methods: Medical records from pts treated at Roswell Park from June 2005 to January 2024 were retrospectively reviewed to identify adult pts with newly diagnosed sAML emerging from previously documented MDS who had received at least one cycle of prior single agent HMA (azacitidine (Aza), decitabine (Dec)). Induction regimens for sAML were divided into IC (7+3, CLAG+/-M, FLAG, CPX-351) vs LI (Ven-based, targeted agent, other) regimens.

Results: We identified 105 pts meeting these criteria. Median age was 71.7 years (range 31-91), with younger pts (median 67.6 years) in the IC group and older pts (median age 73.7 years) in the LI group. Overall, 72 (68.6%) were men. Ninety pts were Caucasian, 6 African American, 1 Asian, 8 other/unknown ethnicity. Half of pts (52/105, 50%) received an intensive chemotherapy (IC) induction, and half (53/105) received a low intensity (LI) induction. Median months of prior HMA therapy was 11.1 (range 0.03-88.7) with fewer (5.2 months) in the IC group than LI (8.3 months). The majority (86/105, 82%) received prior Aza; the rest (19/105, 18%) received prior Dec. Most (65/105) were fit with ECOG PS of 0 (10%) or 1 (52%). Almost half (50/105 pts, 47.6%) had an adverse ELN risk stratification: 25/53 (47.2%) for LI and 25/52 (48.1%) for IC cohort. Among those receiving IC, the overall response rate (ORR=CR/CRi/CRh) was 60%, with 14 CR (27%) and 17 (33%) CRi/CRh. Pts receiving LI induction had low response rates: 11 ORR (20%) with 5 CR (9%), 6 CRi/CRh (10%), and 3 PR. Among the 11 pts achieving CR/CRi/CRh with LI therapy, all had prior Aza therapy, and all subsequently responded to a Ven-based regimen: 3 Ven+LDAC, 4 Ven+Dec, 4 Ven+Aza. ORR for pts receiving Ven-based LI was 50% vs 3.2% for non-Ven LI, respectively. Out of 105 pts, 19 proceeded onto alloSCT after first induction therapy: 17/19 received upfront IC, and 2/19 received LI (Ven-based). Only 2/22 (9%) of pts receiving Ven-based LI underwent alloSCT after upfront therapy. An additional 10 pts received a second-line therapy for sAML induction and then proceeded onto alloSCT; of these, 7/10 received IC, and 3/10 received LI (Ven-based). The median OS of all 105 patients was 7.7 months (range 0.6-133.5 months). The 30-day mortality rate was 6.67%, and the 60-day mortality rate was 18.1%. Median OS of pts receiving IC vs LI was significantly longer (9.8 vs 5.3 months, p=0.02). Median OS for all pts undergoing alloSCT for sAML was longer than pts receiving chemo alone (13.3 vs 5.5 months, p=0.0001). Median OS for Ven-based LI regimens (n=22) was 5.3 months (range 1.6-38.8) vs. 4.0 months (range 0.6-133.5) for non-Ven-based LI regimens (n=31, p=0.327).

Conclusion: In this single-center retrospective analysis, older adults (median age 71.7 years) with secondary acute myeloid leukemia emerging from myelodysplastic syndrome (AML-MR) following prior HMA exposure benefited most from intensive chemotherapy over low intensity therapy with overall response rates of 60%, prolonged median survival, and high rates of subsequent alloSCT. Lower intensity therapy with non-Ven based regimens was ineffective. Pts receiving Ven-based LI for AML-MRC had an ORR of 50% but short median OS (5.3 months) and low rates of subsequent alloSCT. Randomized prospective trials to determine the optimal induction regimen (IC vs Ven-based LI) followed by alloSCT for all eligible pts are warranted.