A Phase I Study of Nivolumab in Combination with Ipilimumab for the Treatment of Patients with High Risk or Refractory/Relapsed Acute Myeloid Leukemia Following Allogeneic Stem Cell Transplantation

Background: The prognosis for patients with AML and MDS who relapse after allogeneic stem cell transplant (allo-SCT) is dismal. AML is susceptible to immunotherapy, as evidenced by the success of allo-SCT. Enhancing graft-versus-leukemia effect of allo-SCT is a major area of focus. We investigated combination immune checkpoint inhibition (ICI) using ipilimumab (Ipi) and nivolumab (Nivo) in a phase I dose escalation trial enrolling AML/MDS patients who relapsed after allo-SCT.

Patients and Methods: For AML, evidence of relapse included bone marrow blast count ≥5% and/or presence of extramedullary disease and/or presence of marrow measurable residual disease (MRD) by multicolor flow cytometry. For MDS, evidence of relapse included appearance of dysplastic changes in the marrow, with or without increase in marrow blast count. Patients, ≥18 years-old, with a matched or haploidentical related donor, or 8/8 HLA matched unrelated donor (MUD) received Nivo (1mg/Kg) monotherapy administered on Day 1 and 15 of a 4 week cycle; ipilimumab (1mg/Kg) monotherapy administered on day 1 of a 3 week cycle; or combination Nivo (1 or 3 mg/Kg) + Ipi (1 or 3 mg/Kg), with Nivo and Ipi administered on day 1, followed by Nivo on days 15 and 29 of a 6 week cycle. Drug(s) were started > 42 days after stem cell infusion. Graft versus host disease (GVHD) prophylaxis was post-transplant cyclophosphamide 50 mg/kg on day +3 and +4, and tacrolimus ± MMF. Patients with history of >grade 2 GVHD were excluded. The primary outcome was to determine maximum tolerated dose (MTD) and dose limiting toxicity of monotherapy and combination Ipi + Nivo. ClinicalTrials.gov: NCT03600155.

Results: 29 patients, 10 females and 19 males, with a median age of 59 (27-76) years were enrolled between April 2019 and May 2023. Of the 29 patients, 20 (69%) had AML, 7 (24%) had MDS, and 2 patients (7%) had CMML. Twenty-six (90%) patients had marrow disease only, 1 patient (3%) had extramedullary disease, and 2 patients (7%) had involvement of both. Twenty-two (76%) patients with AML and MDS had morphologic relapse, and 7 (24%) AML patients had MRD by flow cytometry. Five patients (33%) had a history of grade I GVHD and 10 (67%) patients had a history of grade II GVHD; in all cases, symptomatic GVHD resolution was achieved prior to study initiation. The median time from administration of SCT to study drug infusion was 179 days (range, 54-987). Seventeen (59%) patients were on GVHD prophylaxis during the study. Twenty-seven patients (93%) had received post SCT cyclophosphamide.

Four patients (14%) developed GVHD on study. Of the 4 patients who developed GVHD, 2 (50%), 1 (25%) and 1(25%) patients developed grade II (skin and upper GI [UGI]), III (UGI and lower GI [LGI]) and IV (LGI) GVHD, respectively. Twenty-nine (100%), 28 (97%), 29 (100%), 16 (55%), and 7 (24%) patients experienced grades I, II, III, IV and V toxicities, respectively. Typical adverse events associated with ICI included: pneumonitis (n=3; 10%), 2 (7%) grade II and 1 (3%) grade III; colitis (n=3; 10%), 1 (3%) grade II, 1 (3%) grade III and 1 (3%) grade IV; hepatitis (n=6; 21%), 2 (7%) grade 1, 2 (7%) grade II, 1 (3%) grade III and 1 (3%) grade IV; endocrinopathy (n=3; 10%), 1 (3%) grade I, 1 (3%) grade II, and 1 (3%) grade III; dermatitis (n=4; 14%), 1(3%) grade 1, and 3 (10%) grade III. Dose limiting toxicities were noted at combination nivo 1 mg/Kg + Ipi 3 mg/Kg, mandating Ipi dose reduction to 1mg/Kg.

During the study period, median follow-up was 65 days (range, 9-277 days); 13 patients experienced disease progression, GVHD, or death. Median PFS, estimated using Kaplan-Meier method, was 109 days (95%CI: 48 - NA days). Nine (31%) patients achieved stable disease or remission: 4 patients on Nivo 1 mg/Kg arm, 2 patients on Ipi 1 mg/Kg arm, 2 patients on Nivo + Ipi 1 mg/Kg arm, and 1 patient on the Nivo 3 mg/Kg + Ipi 1 mg/Kg arm. Patients with MDS had a longer PFS (HR = 0.15; CI, 0.02-1.18; p=0.03) vs. patients with AML. Of patients with extramedullary disease, 1 patient had CR of skin lesions and the second patient had a mixed response; both patients received Nivo 1mg/Kg + Ipi 1mg/Kg arm.

Comprehensive immune profiling using CyTOF enabled the identification of distinct stem cell subsets and myeloid and lymphoid populations in bone marrow aspirates and circulating blood.

Conclusion: This study met its primary objective of determining the MTD of the combination regimen with encouraging safety and efficacy outcomes. Our results warrant further study of this regimen.