Pre-Emptive Donor-Derived Innate Immune Cell Add-Back after Haploidentical Hematopoietic Stem Cell Transplantation to Reduce High Rate of Infection, Gvhd, and Non-Relapse Mortality

Introduction Post-transplant cyclophosphamide has allowed alloHSCT using haploidentical donors. PTCy is very effective at reducing GVHD-causing T cells, but it has the unwanted effect of eliminating Natural Killer cells. We have developed a new reduced-intensity haploHSCT transplant conditioning regimen based on an a regimen developed for fully-matched donors with lymphoma: bendamustine 130 mg/m2 IV daily, and fludarabine 30 mg/m2 IV daily, for 3 days, +/- rituximab. We made additional modifications designed to improve NK immune recovery: 1.) no mycophenolate, 2.) no routine G-CSF, 3.) early tacrolimus taper starting on day +60, 4.) pre-emptive NK-enriched donor lymphocyte infusion on Day +8. Objectives Conduct a prospective clinical trial to evaluate the safety of bendamustine-fludarabine conditioning, PTCy, NK cell add-back, and short course tacrolimus for patients with myeloma or lymphoma undergoing haploHSCT. Assess clinical outcomes over the first year after transplantation. Efficacy Endpoints: Death or engraftment failure at D+30, death or severe chronic extensive GVHD at D+100. Test rate of survival, engraftment, and GVHD. Measure reconstitution of adaptive and innate lymphocyte subsets. Measure NK KIR repertoire recovery and assess for ‘missing self’. Methods This is a Phase I, single center, single cohort, open-label, proof-of-concept clinical trial to evaluate the safety of the BFRHaplo regimen (Figure 1) and preemptive CD56-selected DLI following PTCy for adults with myeloma or lymphoma. Results We have safety data on six patients (Table 1). Recovery of neutrophils and platelets is prompt. No primary or secondary graft failures have occurred. No patients have developed Grade III-IV or steroid-refractory acute GVHD. No organ toxicities, and no alkylator-alkylator toxicity from bendamustine – PTCy, have been observed. No instances of either veno-occlusive disease or transplant-related thrombotic microangiopathy, and no deaths, serious infections, or re-hospitalizations. One patient relapsed and is alive without GVHD receiving salvage therapy. We used single-step CD56 selection to prepare CD56-enriched donor lymphocyte infusions (Table 2). The one-step CD56 selection method has the attractive property of preserving CD3+/CD56+ cells, including NK/T and gdT. The target cell dose (> 1 million CD56+ cells/kg and Conclusion The trial will establish whether BFRHaplo, and the infusion of cytotoxic NK and some T cells (