Dose-Escalation Study of Azacitidine Followed By High-Dose Cytarabine (HiDAC) and Mitoxantrone (Mito) for Remission Induction in High-Risk Acute Myeloid Leukemia (AML)

Background: High-risk AML including relapsed and/or refractory (R/R) disease remains largely incurable. There is no standard therapy for R/R AML but HiDAC based regimens are commonly employed. We hypothesized that the use of azacitidine (AZA), a DNA methyltransferase inhibitor and epigenetic modulator, followed by HiDAC and Mito would sensitize leukemia cells to chemotherapy.

Methods: We performed a phase I dose escalation trial in cohorts of patients (pts) with high risk AML that combined increasing doses of AZA with a fixed dose/schedule of HiDAC/Mito. The primary endpoint of the study was to determine the maximum tolerated dose (MTD) of AZA when given in combination with HiDAC/Mito. AZA was given daily on days 1-5 followed by single doses of HiDAC 3 gm/m2 and Mito 30 mg/m2 on days 6 and 10. Dose reductions were made to 2mg/m2 and 20 mg/m2 for HiDAC and Mito, respectively, for pts >70. Cohorts of 3-6 pts were treated in a dose escalation design to a maximum AZA dose of 75 mg/m2. Dose limiting toxicity (DLT) was defined as any grade 4 or greater non hematologic toxicity, grade 3 non hematologic toxicity lasting >7 days, or persistent bone marrow aplasia in the absence of bone marrow involvement for >56 days. Once a dose level had been declared tolerable, additional pts could be enrolled at that level to provide additional toxicity and efficacy experience while further dose escalation was ongoing.

Results: As of July 2015, 46 pts had enrolled. AZA dose levels were 37.5mg/m2 (n=18), 50 mg/m2 (n=16), 75mg/m2 (n=12). Median age = 66 (range, 21-83). R/R AML = 25 (54%), de novo t-MN =10 (22%), de novo AML arising from antecedent myeloid neoplasm = 6 (13%), untreated de novo AML in pts age >60 = 5 (11%). 17 (37%) of pts were in first relapse, 4 (9%) were beyond first relapse, and 4 (9%) were primary refractory. Median number of prior therapies was 1 (range of 0-4 prior therapies). Prior therapies included cytarabine and anthracycline =21 (46%), HiDAC =14 (30%), AZA =9 (20%) and allogeneic stem cell transplant (alloSCT) =9 (20%). Molecular/genetic group profiles by European Leukemia Net (ELN) criteria included adverse =21 (46%), intermediate I =10 (22%), intermediate II =7 (15%), favorable =7 (15%) and unknown in 1 patient. Pts in the favorable ELN group all had R/R AML, AML from antecedent myeloid neoplasm, or age> 60.

Myelosuppression was the most common toxicity and universal. Both median and average times to count recovery (defined as ANC >1000, platelet count >100,000, and independent of RBC transfusions) from initiation of AZA (Day 1) was 42 days and from initiation of HiDAC (Day 6) was 37 days. Common toxicities included 38 cases of febrile neutropenia (87%) and pneumonia (20%). This was similar to our prior published retrospective experience with only the HiDAC/Mito regimen; 64% had neutropenic fever and a 20% had pneumonia (Larson S et al, Leuk Lymphoma 2012). There was 1 DLT from reversible acute liver failure at the 37.5mg/m2 dose level in a pt with relapsed t-MN and prior chemoradiotherapy for both breast cancer and mantle cell lymphoma. 30 day and 60 day induction mortality was 2% (1 pt with refractory AML died from sepsis). Pts in the highest AZA dose cohort had similar rates of hematologic and non-hematologic toxicity as well as time to count recovery as lower dose cohorts.

45 pts were evaluable for response: CR =19 (42%), CRi= 7 (16%), ORR (CR/CRi/PR) =27 (60%). Among pts with R/R AML, CR = 8 (32%), CRi = 5 (20%); among pts with untreated de novo AML and age >60, CR = 4 (80%). In first relapse, CR = 7 (41%) with an ORR =11 (65%). Response rates were similar at all 3 AZA dose levels (Table 1). All 7 pts with favorable ELN risk entered CR including 5 with R/R disease. Overall, 18 (40%) pts proceeded to alloSCT after treatment.

Conclusions: The AZA/HiDAC/Mito regimen is feasible and tolerable. AZA dose escalation was feasible up to 75 mg/m2/day. It is associated with significant clinical activity with an ORR of 60% and allowed 40% of pts to proceed to alloSCT. The recommended phase II dose is 75 mg/m2 of AZA given daily on days 1-5 followed by 3g/m2 HiDAC and 30mg/m2 Mito on days 6 and 10. A randomized study is required to assess the relative contribution of AZA to HiDAC/Mito for remission induction in high risk AML. Correlative laboratory studies to analyze the effects of AZA on deoxycytidine kinase and related enzymes involved in cytarabine metabolism are ongoing.