First Report of the Prophylactic Administration of Ex Vivo Co-Stimulated Donor Lymphocyte Infusion (DLI) from Related and Unrelated Donors after Reduced Intensity Conditioning for High Risk Hematologic Malignancies

Enhancing the graft versus tumor (GvT) effect without graft versus host disease (GvHD) is critical to the success of reduced intensity (RIC) strategies for allogeneic transplantation. Despite establishing donor lymphohematopoiesis, relapse rates remain high. Standard DLI given at relapse or prophylactically to enhance donor chimerism has met with limited success in improving transplant outcomes. Co-stimulation of donor T-cells, using anti-CD3/CD28 coated beads to serve as artificial APCs (IND #6914), may reverse functional T cell tolerance, thereby restoring immune responsiveness, potentiate GVT, and maintain remission. Unlike standard DLI, ex vivo co-stimulation may indirectly enhance T cell activity by removing and activating T cells outside of a potentially tumor-induced immunosuppressive milieu. We report the preliminary results of a ‘first in humans’ feasibility trial of ex vivo co-stimulated (activated/expanded) DLI from sibling and unrelated donors given ‘prophylactically’ at 4 and 6 months after RIC, T-cell depleted, allogeneic transplantation for patients (pts) with high-risk hematologic malignancy.

Methods: 15 pts have undergone RIC with alemtuzumab (100 mg SQ divided over D -9 to D-6), Fludarabine (30 mg/m²/d, D -6 to -3), Busulfan (0.8mg/kg IV Q6h D-4,-3) with peripheral blood stem cell transplantation (PBSCT) and planned activated DLI (pADLI) at 4 months (1 × 10⁷ CD3+/kg recipient) and 6 months (1 × 10⁸ CD3+/kg recipient) post-transplant in the absence of GvHD or relapse. All pts received tacrolimus and methotrexate as GvHD prophylaxis with initiation of rapid tacrolimus taper at ~d+40 and discontinuation before the infusion of pADLI#1.

Results: 8 pts (AML, secondary n=2; CR2 n=3; CR1 n=1; ALL CR1 n=1; CR2 n=1) received grafts from HLA-identical sibling donors, 7 (AML, secondary n=3; CR2 n=1; CR1 n=1; MDS, n=1; ALL, CR2 n=1) from 6/6 matched unrelated donors. Donor hematopoiesis (□ 90%D) was established in all patients. The median is f/u 272 days, (range 25–424d). Exvivo co-stimulation and expansion was successful in all products, except in one instance where the second infusion fell below the targeted dose. Of the 11 infusions given to date (both fresh and cryopreserved), there has been no infusion-related side effects or toxicity, confirming the safety and feasibility of this novel strategy. 8 pts have received the first of 2 planned infusions of pADLI from their original donor. Of the remaining 7 pts, 4 infusions are upcoming, 1 was precluded by early relapse, and 2 were precluded by early non-relapse related mortality (1 GvHD with sepsis, 1 hepatic failure). Of the 8 pts who received their initial pADLI#1 at ~d+120 (range; d+119 – d +172), 3 have received a second infusion at ~d+180 (range; d+180 – d+237). 5 pts did not receive pADLI#2 because of relapse (n=3), GvHD (n=1), and transient uveitis in the absence of GvHD (n=1). Overall incidence of acute GvHD (aGvHD) has been low; only 1 pt developed aGvHD prior to pADLI#1, and only 1 pt developed aGvHD after pADLI #1. To date, no patient has developed chronic GvHD. T-cell functional assays performed on a DLI product from a CMV-positive unrelated donor before and after activation/expansion [analysis of proliferative capacity (CFSE assay) and interferon (IFN)-gamma secretion (ELISPOT)] suggest potential enhancement of immune function in the activated product. Using CMV, SEB, and autologous tumor as antigen, CD8 proliferation was increased after stimulation with all 3 antigens as well as media after activation of DLI, whereas a decrease in CD4 proliferation after activation was noted (except against SEB, as expected). Prior to activation/expansion, the number of effector cells secreting IFN-gamma in response to CMV or autologous tumor was below the threshold for a positive assay via ELISPOT (likely secondary to low levels of circulating CMV-specific T-cells); whereas a >10 fold increase in the number of effector cells/million was noted after activation/expansion in response to both CMV and autologous tumor. Serial functional analyses of recipient T-cells, before and after infusion of activated DLI, are underway.

Conclusion: Preliminary results of this trial demonstrates that RIC with PBSCT followed by ex vivo costimulated pADLI for poor prognosis hematologic malignancies is safe and feasible with potential for enhancing GvT without increasing GvHD.