Background: Guadecitabine (G) is a next-generation hypomethylating agent (HMA), administered as a small-volume subcutaneous (SC) injection, designed with the potential to overcome pharmacokinetic resistance to first-generation HMAs (decitabine and azacitidine). Preliminary guadecitabine phase 2 data in r/r MDS showed an overall survival of almost 12 months (Garcia-Manero,G, et al; Lancet Haematol http://dx.doi.org/10.1016/S2352-3026 [19]30029-8) leading to the ASTRAL-3 study. Aims: Compare overall survival of guadecitabine to that of TC consisting of low-dose cytarabine (LDAC), intensive chemotherapy (IC), or best supportive care (BSC). Secondary objectives included multiple standard assessments of response and safety. Methods: Subjects with r/r MDS or Chronic Myelomonocytic Leukemia (CMML) who had progressed or failed to respond after 6 full cycles of standard HMA therapy were randomized 2:1 between guadecitabine (60 mg/m2) SC Days 1-5 every 28 days vs. a pre-selected TC (BSC, IC, or LDAC). Stratification was by disease, disease burden (>10% bone marrow blasts or not, geography, and pre-selected TC option). Primary endpoint was overall survival (OS) and the study was designed with ~90% power to detect a hazard ratio of 0.68 (approximately 2.8 month difference in median survival). Results: 417 MDS/CMML subjects were randomized to G or TC (G:277, TC: 140). Demographics and disease status were reasonably balanced across the groups (see Table 1 below) as was percentage of commonly associated genetic mutations (TET2, DNMT3A, SF3Baa, RUNX1, and TP53). Median guadecitabine exposure was 4 cycles and 3 cycles for TC. Neither overall survival (G: median 9.1 months, TC: median 8.3 months, p=0.61, HR:0.94 with 95% CI: [0.74-1.19]) nor leukemia-free survival (LFS) (G: median 5.7 months, TC: median 5.9 months; p=0.38) demonstrated a significant difference between the two groups and subgroup analyses did not suggest a difference between guadecitabine and any of the different TC options or different genetic mutations. Transfusion dependence for 8 weeks was similar with 32.1% and 22.4% of the G group and 37.9% and 20.0% of the TC group being platelet transfusion or red blood cell transfusion independent, respectively. Safety was consistent with known profiles of the respective agents with the G group having a slightly higher overall incidence of adverse events (AE) (99.3% vs 92.6% for TC) and grade 3 or higher AE (92.2% vs 70.5% for TC). The AEs with the highest incidence in subjects who received guadecitabine were febrile neutropenia (38.5 vs 18.9% for TC), pneumonia (34.4% vs 18.9% for TC), neutropenia (34.1% vs 15.6% for TC), and thrombocytopenia (32.2% vs 21.3% for TC). Image:Summary/Conclusion: This large, global, randomized phase 3 study did not demonstrate superiority of guadecitabine over Standard TC in MDS/CMML patients who were refractory or relapsed following full course of prior HMA treatment.
Read full abstract