Adoptive Transfer of Treg-Depleted Donor Th1 and Th2 Cells Safely Accelerates Alloengraftment After Low-Intensity Chemotherapy

Abstract 521

Ex-vivo culture of murine donor CD4⁺ T cells using rapamycin, co-stimulation, and IL-4 yielded a defined T cell population (T-rapa cells) that favorably modulated the balance between GVHD, graft rejection, and GVT effects. To translate these findings, we conducted a multi-center clinical trial (NCT0074490) to evaluate T-rapa cell therapy after allogeneic HCT. T-rapa cells were manufactured by ex vivo culture of donor CD4⁺ T cells using CD3/CD28 co-stimulation in media containing IL-4, IL-2, and rapamycin. T-rapa cells had a mixed Th2/Th1 phenotype with minimal Treg content (intra-cellular flow, n=48 products; median transcription factor expression: 11.5% [GATA-3], 5.1% [T-bet], and 0.1% [FoxP3]). Median T-rapa cell cytokine secretion (pg/ml; re-stimulation at harvest) was 1.3 [IL-4], 20.6 [IL-5], 9.7 [IL-10], 23.7 [IL-13], 34.7 [IFN-g], and 17.1 [IL-2]. Patients received an HLA-matched sibling, T cell-replete, G-CSF mobilized allograft, and GVHD prophylaxis of cyclosporine plus short-course sirolimus (to d14 post-HCT). Two protocol arms evaluated T-rapa cell therapy after induction chemotherapy and outpatient, low-intensity preparative chemotherapy (Table I). First, patients (n=25) were accrued to arm A to evaluate T-rapa infusion at d +14 post-HCT; subsequently, accrual was initiated to arm B (n=25) to evaluate T-rapa infusion on d0 of HCT. Arm A was then expanded to n=40 patients. Patients accrued to arms A and B were similar for recipient age, high-risk malignancy diagnosis, chemotherapy refractoriness, and prior regimen number (Table I). Most recipients were not in remission at the time of HCT. High-risk NHL was the most frequent diagnosis (25/65 patients), followed by non-high-risk NHL (11/65), AML/MDS (8/65), myeloma (7/65), CLL (6/65), Hodgkin's disease (5/65), and CML (3/65). Arm A and B recipients had similar mean donor myeloid cell chimerism at d +14, +28, and +100 (arm A, 43%, 74%, and 89%; arm B, 50%, 62%, and 84%). At d +14, arm A and B recipients also had mixed donor T cell chimerism (mean values, 60% in each arm; Table I). At d +28 and +100, T cell chimerism increased in arm A to 80% and 89%; in arm B, these values increased to only 67% and 69%. Four recipients on arm B had < 10% donor T cell chimerism at d +100; in contrast, the lowest donor T cell chimerism value observed at d +100 on arm A was 36%. T-rapa therapy on arm A was relatively safe as there was: no engraftment syndrome, a 10% rate of acute grade II to IV GVHD, a 67% incidence of chronic GVHD, and no transplant-related mortality (Table I). On arm A, 37.5% (15/40) of recipients are in sustained complete remission, with a median survival probability of 63.6% at 24 months post-HCT. Therefore, pre-emptive donor lymphocyte infusion with ex-vivo manufactured T-rapa cells that express a balanced Th2/Th1 effector phenotype represents a novel approach to safely accelerate alloengraftment and harness allogeneic GVT effects after low-intensity conditioning.