Results from a Global Randomized Phase 3 Study of Guadecitabine (G) Vs Treatment Choice (TC) in 302 Patients with Relapsed or Refractory (r/r) Acute Myeloid Leukemia after Intensive Chemotherapy (ASTRAL-2 Study)
Background: Guadecitabine (G) is a next generation subcutaneous (SC) hypomethylating agent (HMA) resistant to degradation by cytidine deaminase which results in prolonged in vivo exposure to the active metabolite decitabine. We conducted a large global randomized phase 3 study (ASTRAL-2) of G vs TC in AML patients who were either refractory to or relapsed (r/r) after prior anthracycline-based intensive chemotherapy with or without hematopoietic cell transplant (HCT).
Methods: r/r AML patients were randomized to G (60 mg/m2 SC/d for 10 days in first 1-2 cycles followed by 5-day cycles Q 28 days) vs Treatment Choice (TC). TC were preselected prior to randomization to either low intensity (LI) treatment; high intensity (HI) chemotherapy; or Best Supportive Care (BSC). LI choices were other HMAs of azacitidine or decitabine, or low-dose Ara-C (LDAC) at their standard doses. HI choices were high-dose Ara-C (HiDAC), MEC, or FLAG± Ida combination chemotherapy at standard doses. Primary endpoint was overall survival (OS) based on ITT analysis with secondary endpoints including 12 and 24-month survival rates, complete response (CR), event-free survival (EFS), and safety. P values for secondary endpoints and subgroups are nominal as there was no adjustment of p values for multiple analyses.
Results: 302 patients were randomized to G (148) or TC (154). Preselected TCs were mainly LI (77%) predominantly HMAs (86% of the patients randomized to LI), or HI (21%), with only 6 patients (2%) in the BSC subset. Baseline variables were well balanced across the 2 treatment arms. For G vs TC respectively, age ≥65 y in 51.4% vs 40.3% with median age 65y vs 63y, ECOG PS 2 in 15.5% vs 20.8%, poor risk cytogenetics in 44.6% vs 42.2%, refractory AML in 44.6% vs 33.1%, prior HCT in 18.2% vs 26%, a majority of patients were in second or subsequent relapse after ≥ 2 prior therapies (54.7% vs 56.5%).
Median number of treatment cycles was short (3 cycles for G vs 2 cycles for TC). Median follow up was 21.6 months. Most common causes of treatment discontinuation were disease progression (35.2% for G vs 38.1% for TC), or death (15.2% for G vs 18.4% for TC).
Median OS on G was 6.4 months vs 5.4 months for TC and not statistically significant (OS HR 0.88, 95% CI 0.67, 1.14, log rank p value 0.3). There was no significant difference in OS between G and each of the LI and HI preselected subsets. However, several other planned prospective subgroups favored G with OS HR 95% CI upper limit ≤ 1.0 including patients <65y (HR 95% CI 0.47, 0.97, p 0.032) ; ECOG PS 0-1 (HR 95% CI 0.57, 1.0, p 0.049); refractory AML (HR 95% CI 0.38, 0.89, p 0.013); lower peripheral blood (PB) disease burden of ≤ 30% PB blasts (HR 95% CI 0.46, 0.92, p 0.015); and those who received at least 4 cycles in either treatment arm (HR 95% CI 0.36, 0.95, p 0.031). The 12- and 24-month survival rates for G vs TC respectively were 32% vs 26%; and 19% vs 10%. Median EFS was short with 3 months for G vs 2.4 months for TC; log rank p 0.005. CR rate was 12.8% for G vs 7.1% for TC (p 0.051). CR + CR with partial hematologic recovery (CRh) rate was double for G with 16.9% compared to 7.8% for TC (p 0.007). Composite CR (CRc) or CR+ CR with incomplete hematologic recovery (CRi) rate was 27% for G vs 14.3% for TC (p 0.003).
Adverse events (AEs) of grade ≥3, regardless of relationship to treatment, were 89% on G vs 84% on TC. Most common Grade ≥3 AEs for G vs TC respectively were febrile neutropenia (38.6% vs 38.1%); neutropenia (32.4% vs 17%); thrombocytopenia (28.3% vs 29.9%); anemia (21.4% vs 24.5%); pneumonia (18.6% vs 20.4%); and sepsis (11.7 vs 10.9%). None of the differences were significant except for neutropenia (p 0.003).
Summary/Conclusions: In this randomized study in r/r AML after intensive chemotherapy, G did not significantly improve OS compared to standard of care TC composed mainly of LI treatment with other HMAs. The data suggest that G may be better than TC in in some of the secondary endpoints (24-month survival rate, EFS, CR, CR+CRh, and CRc). Prospective subgroup analyses of OS suggest that younger (<65 y), more fit patients (PS 0-1); with lower PB disease burden (PB blasts ≤ 30%), and those who could receive at least 4 cycles may benefit from G. The results of secondary endpoints and subgroup analyses should be interpreted with caution. Grade ≥3 AEs were the expected hematological and infection AEs with no significant differences between G and TC except for significantly higher incidence of neutropenia with G.