Outcome of secondary acute myeloid leukemia treated with hypomethylating agent plus venetoclax (HMA‐Ven) or liposomal daunorubicin‐cytarabine (CPX‐351)

Abstract

Patients presenting with secondary AML (sAML) defined as AML evolving from (i) prior myelodysplasia (MDS) or myeloproliferative neoplasm (MPN), (ii) as a consequence of previous exposure to a proven leukemogenic chemotherapeutic agent have inferior outcomes with “7 + 3” regimen with CR rates ranging only between 30–40. Population based studies show that 25% of AML are classifiable as sAML and the incidence of sAML is set to rise due to increasing numbers of cancer survivors and improved survival due to introduction of novel therapies for solid tumors.1 Patients with sAML have inferior outcomes due to older age/ comorbidities, adverse cytogenetics and multidrug resistance phenotype, hence new approaches are needed to improve poor outcomes in this high-risk AML subgroup. Note, CPX-351 is a liposomal formulation of cytarabine and daunorubicin in a 5:1 molar ratio and leads to enhanced intracellular delivery of the synergistic drug ratio in vivo to cells resulting in enhanced killing of leukemic blasts. A phase 3 study comparing CPX-351 to 7 + 3 regimens demonstrated superior CR rates (47.7% vs. 33.3%; p = .016) and overall survival (OS) benefit (9.56 vs. 5.95 months; p = .003) in older patients (≥60 years) with sAML.2 The combination of a 7-day course of azacytidine with venetoclax or placebo was studied in a multicenter phase 3 trial in patients presenting with AML who were unfit for intensive chemotherapy regimens due to age ≥ 75 years or comorbidities. At median follow up of 20.5 months, the composite CR rate was 66.4% versus 28.3% (p < .001) and median OS was 14.7 versus 9.6 months(p < .001) in favor of HMA + Ven arm.3 On subgroup analysis of sAML patients in this trial, the median OS was 16.4 months in the HMA + Ven group showing activity in this subgroup of high-risk patients. Based on these data, we compared the outcomes of patients with sAML who received frontline treatment with either HMA + Ven or CPX-351 at our institution. Methods: We retrospectively analyzed records of consecutive patients with previously untreated sAML who presented to City of Hope between 2018–2020. Patients were treated for remission induction with either HMA + Ven or CPX-351 based on physician discretion based on assessment of comorbidities and performance status. Approved doses of HMA were used as follows: Decitabine was given intravenously at 20 mg/m2 for 5 (n = 12) or 10 (n = 18) doses. The venetoclax dose was 400 mg oral daily with dose reduction for drug–drug interaction with azoles. Induction dose of CPX-351 was per guidelines (100 mg/m2 of cytarabine and 44 mg/m2 of daunorubicin) given on days 1, 3 and 5. Complete remission (bone marrow blasts <5%) was defined as (i) with hematologic recovery (CRh; absolute neutrophil count ≥1000/μl and platelet counts ≥100 000/μl) (ii) without hematologic recovery (CR) in patients with bone marrow blasts <5% but without recovery of blood counts. Minimal residual disease (MRD) assessment was done on the day 28 bone marrow aspirate, using multiparametric flow cytometric assessment of leukemia associated immunophenotype with lower limit of sensitivity being 0.01%. Patient selection: Patients were eligible for response assessment for HMA-Ven if they met inclusion criteria and received any duration of treatment on the assigned arm. In the HMA + Ven arm, 31 patients were identified based on these criteria and analyzed for clinical outcomes response. Two patients died due to sepsis caused by multidrug resistant Escherichia coli and disseminated mucor respectively (induction mortality) and one patient died of unknown cause (lost to follow up)-all treated patients were included for response assessment. One patient received HMA + Ven for undifferentiated leukemia and was thus excluded from final analysis (N = 30). In the CPX-351 arm, 20 patients met eligibility criteria and were studied for clinical outcomes. Results: We identified 50 patients with sAML eligible for analysis of whom 30 were treated with HMA-VEN, and 20 received CPX-351 as induction therapy. Patients' demographic, ELN-AML risk and molecular features are summarized in Table 1. The CPX-351 and HMA-Ven groups were balanced with regards to their demographic profile, cell counts at diagnosis and bone marrow blasts. Patients in HMA + Ven group had significantly poor Karnofsky Performance Status (KPS) and adverse ELN classification compared to CPX-351 arm. Of the 30 cases of sAML in HMA+ Ven, nine (30%) arose due to prior chemotherapy (Hodgkin Disease n = 2; paraganglioma, desmoid tumor, breast cancer, DLBCL, multiple myeloma, ALL and ovarian cancer, one each) while 21 (70%) patients had AHD. In CPX-351 group, six (30%) cases of sAML arose from prior chemotherapy (n = 2 DLBCL, n = 2 breast cancer, T lymphoblastic lymphoma, and colon cancer one each) and 14 (70%) from AHD. The mean number of treatment cycles were three (range 0.5–18) in HMA + Ven arm and 1.45 (1–3) in CPX-351 arm. The composite CR rate (CR + CRh) in patients treated on HMA-Ven group was 70% versus 50% in CPX-351 group (p = .154). In the HMA-Ven group, of the 21 responding patients, seven (33%) achieved CRh and while four of 10 responding patients in CPX-351 arm achieved CRh (40%). Among patients who achieved remission with available MRD data, MRD negative remission was achieved in 14 patients (66%) in HMA-Ven group and in five (50%) patients in CPX-351 group (p = .122). In the HMA-Ven group, 14 patients (66%) achieved remission after one cycle of therapy and remaining seven (34%) required two cycles of therapy to achieve CR. In the CPX-351 group, nine patients (90%) achieved CR with one cycle of therapy. Eleven (37%) of patients in HMA + Ven and 12 patients in CPX-351 group (65%) underwent allo-HCT (p = .049). With median follow up of 9.9 months (0.6–41.2) in HMA + Ven and 11.3 months (1.8–40.1) in CPX-351 groups in responding patients, the LFS was 13.5 months (5.3-NA) and 16.1 months (0.9-NA) in CPX-351 and HMA + Ven groups (p = .418), while the OS for all patients was 11.3 months (4.2-NA) and 10.1 months (5.2–20; p = .761) respectively. In patients who underwent HCT, with median duration of follow up was 17.5 months (5.2–29.3) and 17 months (5.3–40.1) respectively in HMA + Ven and CPX-351 arms. The LFS was 16.2 months (9.3-NA) and 24.3 months (10.8-NA) in HMA + Ven arms and CPX-351 group (p = .836), while OS was 20 months (9.3-NA) and 24.3 (10.8-NA; p = .974) respectively (Figure 1) . In the HMA-Ven group (n = 30), AML-with AHD (MDS), was seen in 20 patients of which four patients were exposed to HMA (20%) prior to addition of venetoclax and two of these patients achieved remission. In the CPX-351 group, AML-AHD (MDS) was seen in 14 patients of whom four patients had received prior HMA (58%) therapy and remission was achieved in four patients of these patients. Patients with sAML without prior HMA exposure were more likely to respond to HMA + Ven compared to CPX 351 with CR rates of 87% versus 42%: (p = .009) respectively. The TP53 mutation was seen in 10 patients in HMA + Ven group and five patients in CPX-351and the CR rates were 60% for HMA + Ven and 45% for CPX-351(p = NS). Note, OS was improved in patients who underwent HCT (n = 7) compared to non HCT (n = 8) patients with TP53 mutation 19.7 versus 4.2 months (p = .018). Twenty-one (70%) died in HMA-Ven group with relapsed AML being most common cause of death (N = 16) followed by sepsis (n = 4) and congestive heart failure in one patient. In CPX-351 group, 13 patients (65%) died of: AML (n = 8), sepsis (n = 3), Covid-19 pneumonia (n = 1) and unknown cause in one patient. Discussion: In this retrospective review, we compared the clinical outcomes of patients presenting with sAML treated with either CPX-351 or HMA-VEN in the frontline setting. Our study showed high CR rates in both HMA-Ven and CPX-351 arms (70% vs. 50%; p = .154) and no difference in OS/LFS between the groups-supporting use either regimen in sAML patients for upfront therapy. Notably, significantly more patients in HMA + Ven group had poor KPS and ELN adverse risk AML compared to CPX-351 group suggesting that this regimen could be prioritized in these patient subgroups. Both groups had similar clinical outcomes in terms of MRD negative remission, LFS and OS. Patients with sAML benefit from Allo-HCT consolidation after achievement of remission compared to chemotherapy consolidation.4 We noted similar OS benefit of allo-HCT in our retrospective review in patients treated with newer regimens using HMA + Ven and CPX-351 arms (Table 1). In the pivotal phase 3 study using CPX-351 vs. 7 + 3 in sAML, 30% patients underwent allo-HCT in remission with superior allo-HCT outcomes in patients who received CPX-351.2 In our study a significantly higher number of patients underwent allo-HCT after CPX-351 induction compared to HMA + Ven group (65% vs. 35%; p = .049). The higher transplantation rate may be secondary to selection bias as patients in HMA + Ven had lower KPS indicating less fit patient population. We did not see any difference in clinical outcomes of OS/LFS in patients post allo-HCT after induction with either regimens partly due to similar MRD remissions in both groups. Published studies have shown either CPX-351 or HMA + Ven 2, 5 can be used as an effective bridge to transplant and the decision to offer either therapy should be based on baseline performance status and co-morbid conditions at diagnosis. In our study, patients with AML-AHD, HMA naïve patients achieved higher CR with HMA + Ven compared to CPX-351 (87% vs. 42%; p = .009) suggesting prior HMA exposure may reduce CR rates in sAML unlike the situation in MDS where HMA + Ven responses are not affected by prior HMA exposure.6 Fifteen patients with the TP53 mutation were seen in our study and induction with either regimen resulted in similar OS/LFS(Table 1). Clear benefit of Allo-HCT was noted in TP53 mutated patients with improvement in OS from 4.2 month in non HCT patients to 19.7 months in patients who were able to undergo HCT (p = .018; Table 1). Finally, it is important to point out those patients with AML-MRC (MDS related changes) is a subgroup that has shown to benefit from CPX-351, and we did not have patients with this subtype of AML in our study potentially resulting in lower response rates in CPX-351 group. In conclusion, our report shows upfront therapy with either HMA + Ven or CPX-351 results in encouraging response rates in sAML compared to patients historically treated with the 7 + 3 regimen. This data supports development of prospective trials in sAML to determine the most effective frontline regimen in this high-risk AML subgroup. A.S.: has research funding from Bristol Myers Sqibb, serves on advisory board of Syros and advisor for Kadmon, A.A.: consultancy for Seattle Genetics, M.A.L.: consultancy for Neximmune, Riegel pharma and Jazz pharmaceuticals, H.A. consultancy for Incyte corporation, E.B.: Consultancy for Gilead, Mustang and Roche, Research funding Merck, Amgen and Kite. S.S.D.: Research funding Ansun Biopharma, Karius, Shire/Takeda, Gilead, Chimrex, R.N.: Advisory board, Merck, Magenta therapeutics, Kadmon Napajen pharma: consultancy Viracor, ASS serves on the speaker bureau for Stemline, Amgen and Celgene, and on advisory boards for Stemline and Amgen. G.M. is a member of the speakers' bureau for AbbVie and Novartis and serves on advisory board with Janssen Pharm. I.A. advisory boards AbbVie, Amgen, KiTE pharmaceuticals, Agios, consultant for Autolus and Amgen, speaker for Jazz Pharmaceuticals. V.P. advisory boards for AbbVie, Jazz Pharmaceuticals, Pfizer, Genentech; member of speakers' bureau for Jazz, Amgen, Novartis and AbbVie. Consultant for Dova, Amgen. The remaining authors declare no conflict of interest. Data available on request due to privacy/ethical restrictions