Expanded Antigen-Specific Donor Regulatory T Cells for Gvhd Prevention

Minor histocompatibility antigen (mHA)-specific alloreactive donor T cells cause graft vs. host disease (GVHD) in matched related donor allogeneic hematopoietic cell transplantation (HCT). In a translational phase I trial, we expanded and infused mHA-specific donor regulatory T cells (Treg) in the context of sirolimus-based pharmacologic prophylaxis to examine safety and preliminary efficacy of this targeted GVHD prevention approach.We employed a 3+3 phase I design using escalating Treg in 4 dose levels: 0.5 × 105/kg, 1 × 105/kg, 2 × 105/kg, and 4 × 105/kg. Dose-limiting toxicity (DLT) included grade 4-5 related infusion reaction, grade 4-5 unexpected organ toxicity, grade III-IV acute GVHD, or treatment-related death after Treg infusion. Secondary and exploratory measures examined acute and chronic GVHD, survival outcomes, and Treg clone (TCR-Seq) expansion in culture, and in-vivo longevity and expansion post-HCT.Included subjects were age 18-70, had hematologic malignancy in remission, met HCT eligibility, and had HLA-matched sibling donor. HCT patients underwent apheresis for generation of dendritic cells (DC), and donors underwent apheresis followed by single-step Miltenyi CD25+ Treg enrichment. The DC:Treg co-culture (rapamycin, IL-2 and IL-15, GMP conditions) for 12 days was followed by final Treg infusion on day -2 after meeting release criteria (fig 2A). On day 0, patients received unmanipulated donor PBSC graft. Sirolimus was started day -2, tacrolimus on day 0, and other therapy followed institutional standards.A total 15 subjects were included (N=3 each per dose levels 1-3, and N=6 in dose level 4). No DLT were observed, and 4 × 105/kg Treg (level 4) was identified as the MTD. 1 other subject (level 4) did not have adequate Treg expansion for treatment, and 2 subjects had adequate Treg expansion but were not treated (1 relapse, 1 sepsis). Adverse events reflected usual post-HCT toxicity, and there was no evidence of prolonged engraftment, poor donor chimerism, increased infectious risk, or excess relapse. Treg exceeded activated CD4 Tcon post-HCT (Fig 1). At time of analysis, median follow up for survivors was 41.7 months (range 14.5-72.8). The day 100 cumulative incidence of grade II-IV acute GVHD was 13% (95% CI 2-35%). NIH moderate/severe chronic GVHD by 1 year was 6.7% (95% CI 0.36-27%) and by 3 years was 20% (95%CI 4.4-44%). Three relapse events (ALL N=2, AML N=1) and one non-relapse death (dose level 1, pneumonia amidst GVHD therapy) occurred. Overall survival was 73% (95% CI 54-100%). Donor Treg clones expanded in culture (Fig 2B), maintained lineage fidelity in culture and post-HCT (Fig 2C), and demonstrated post-HCT persistence and in-vivo expansion (Fig 3A, B).This translational trial provides data on mHA-specific expanded donor Treg as a novel GVHD prevention strategy, and demonstrates that expanded donor Treg clones can persist and expand through one year post-HCT.