Safety and Efficacy of Decitabine Plus Ipilimumab in Relapsed or Refractory MDS/AML in the Post-BMT or Transplant Naïve Settings

Background: CTLA-4 blockade with ipilimumab (IPI) has modest activity in hematologic malignancies relapsed post allogeneic hematopoietic cell transplantation (allo-HCT) (Davids M, NEJM). Adding decitabine (DAC) to IPI may increase efficacy of this approach in myelodysplastic syndromes (MDS)/acute myeloid leukemia (AML), and we further asked the extent to which alloreactivity contributed to this activity. Here, we present safety (Table 1) and clinical activity (Figure 1) from a phase I, multicenter, investigator-initiated study (CTEP 10026) of DAC plus IPI in pts with R/R MDS/AML with (Arm A) and without (Arm B) prior HCT.

Methods: Pts ≥ 18 y with R/R MDS (≥ 5% blasts), R/R AML or untreated secondary AML were eligible. Prior HMA was permitted but trial was later amended to exclude those with disease progression on HMA therapy within 12 weeks of study entry. Pts post allo-HCT (Arm A) were required to be > 3 mo from donor cell infusion, have no prior acute GVHD Gr III or higher, be > 2 weeks off of immunosuppression and have donor T-cell chimerism ≥ 20%. Treatment was given in 28-day cycles with a lead-in single agent DAC (cycle 0) followed by combination DAC + IPI for up to a year. DAC infusion was given at 20 mg/m2 days 1-5. IPI infusion was given at 3 mg/kg (DL0), 5 mg/kg (DL1) or 10 mg/kg (DL2) on day 1 of cycles 1-4, 5, 7, 9 and 11. DLT was defined as ≥ Gr 3 non-hematologic adverse event (AE), acute GVHD, and steroid-refractory immune-related AE within 8 weeks of first IPI dose.

Results: At data cut-off, 32 of 39 enrolled pts received at least 1 dose of DAC+IPI; 4 of 7 did not move on to C1D1 due to AML-related complications. Median age was 72 y (range 29-85); 25 had AML and 7 had MDS. Median number of prior regimens was 2 (range, 0-5). Twenty of 32 pts had received prior HMA therapy. A median of 3 IPI doses were given (range, 1-8). NGS performed on screening marrows using an 88-gene panel revealed mutations most commonly in ASXL1 (n=9), RUNX1 and STAG2 (n=6 each), and DNMT3A, NRAS and TP53 (n=5 each).

Arm A (Post allo-HCT): Sixteen pts received DAC+IPI across the 3 DLs (7 at DL0, 3 at DL1, 6 at DL2; dose-expansion is open at DL2). The most common treatment emergent Gr 3/4 AEs were febrile neutropenia (n=3) and AST elevation, lymphocytopenia, neutropenia, thrombocytopenia, leukopenia (n=2 each). Seven of 16 pts experienced immune-related AE, including 1 with late-onset acute GVHD Gr III (colon/liver), 4 with chronic GVHD (mild, n=2; moderate, n=1; severe n=1), and 2 with Gr 2/3 ipi-induced colitis. All immune-related AEs were reversible with steroids, except for one case of steroid refractory Gr III acute GVHD complicated by fatal septic shock (DLT at IPI 10 mg/kg). In total, 4 of 16 evaluable pts achieved response, including 3 with CR and 1 with marrow CR; 2 of 4 responders had immune-related AE. For these 16 pts, the median OS is 12.8 mo (95% CI: 7.6-NA).

Arm B (transplant-naïve): Sixteen pts received DAC+IPI across the 3 DLs (4 at DL0, 3 at DL1, 6 at DL2; 3 are in dose-expansion at DL2). The most common treatment emergent Gr 3/4 AEs were lymphocytopenia, thrombocytopenia and leukopenia (n=5 each); neutropenia (n=4); and anemia, fatigue, lung infection, and febrile neutropenia (n=3 each). In 6 of 16 pts that had ≥ Gr 2 immune-related AEs, all were reversible with steroids (including Gr 3 colitis (n=1), Gr 3 dermatitis (n=2), Gr 2 pneumonitis (n=1), Gr 3 hypophysitis/Gr 2 arthritis (n=1)), except for a case of Gr 2 dermatitis (resolved)/Gr 4 immune thrombocytopenia (refractory to steroids, transfusions and IVIG) (n=1; DLT at IPI 10 mg/kg). No study treatment-related deaths were observed. Objective responses were detected in 8 of 16 evaluable pts with 3 CR, 2 CRi, and 3 marrow CR. Only 2 of 8 responders had immune-related AEs. For these 16 pts, the median OS is 18.3 mo (95% CI: 11.7-NA).

Overall, 2 pts discontinued study treatment due to AE. Dose expansion continues at DL2 (MTD) in Arm A and Arm B to further establish safety and tolerability. Correlative science including immunophenotyping is ongoing and will be presented.

Conclusion: In this ongoing phase 1 study in pts with R/R or secondary MDS/AML, DAC plus IPI had expected AE profile and exhibited encouraging clinical activity. The rate of immune-related toxicity was frequent but similar to prior observations with IPI 10 mg/kg for melanoma. Further, high clinical activity was observed amongst pts who did not receive allo-HCT, suggesting that an alloreactive environment may not be required to benefit from CTLA-4 blockade.