Manufacture of Anti-BCMA CAR T Cell Is Feasible after Salvage Autologous or Allogeneic Hematopoietic Stem Cell Transplantation

Background: Commercially available anti-BCMA CAR T cells have revolutionized treatment for patients with relapsed/refractory multiple myeloma (MM). However, most patients ultimately relapse. We therefore evaluated strategies to combine salvage autologous (autoHCT) or allogeneic hematopoietic stem cell transplant (alloHCT) with commercial CAR T cells. We hypothesized that early consolidation with CAR T cell therapy after salvage autologous (3-6 months) or relapse from allogeneic cell therapy would be safe and feasible. In addition, CAR T cell quality as defined by pre-infusion phenotypic markers of memory is associated with improved patient response. We further hypothesized that CAR T cells produced from good quality autologous T cells (those derived from hematopoietic progenitor cells collected early in the disease course) or allogeneic T cells (having never seen chemotherapy) would result in better expansion and persistence translating into deeper remission and longer remission duration when compared to CAR T cells produced from T cells collected at the time of refractory relapse. We also hypothesized that autologous or allogeneic stem cell transplantation may modulate T cell phenotypes. Methods: This phase II single institution feasibility study (NCT05393804) included two cohorts of 15 patients each who had received at least 4 prior lines of therapy: (1) patients collected for idecabtagene vicleucel (Ide-Cel) or ciltacabtagene autoleucel (Cilta-Cel) 3-6 months after salvage high dose melphalan and autoHCT with stem cells cryopreserved after initial therapy and (2) after relapse following an alloHCT. Leukapheresis, lymphodepletion, and post CAR T infusion management was done as per standard of care. Peripheral blood (PB) and marrow samples were collected to evaluate expansion, persistence, cytokines, immune reconstitution, and disease response. CAR T bag wash samples were collected to evaluate T cell phenotypes. Results: To date, 11 patients have been enrolled in cohort 1 (median age 65 (52-76), 73% male, 82% cilta-cel), with median time from autoHCT to apheresis of 106 days (84-140). Pre-auto status was POD in 6/11, SD in 3/11 and PR in 2/11 and improved to CR in 3/11, VGPR in 3/11, PR in 4/11, and SD in 1/11 pre apheresis. Median absolute lymphocyte count pre-apheresis was 1.1 (range 0.3 - 2.9). Ten patients were infused with 1 patient receiving an out of specification (OOS) product, and one patient requiring re-collection of T cells due to an OOS product. Prior to infusion 7/10 had POD, 1/10 had SD, and 2/10 remained in VGPR. Day 28 responses was CR in 1/10, VGPR in 4/10, PR in 3/10, SD in 1/10, and POD in 1/10, with grade 1 CRS in 7/10, grade 2 CRS in 1/10, and no ICANS. In cohort 2, 10 patients have enrolled (median age 52 (42-67), 90% male, 60% cilta-cel) with 8/10 infused with 1 OOS product infused and one patient requiring recollection who has not yet been infused with the 2nd product that met release criteria. Consistent with high rates of successful CAR T cell manufacture, no difference was observed in median absolute CD3+ T cell count in either cohort 1 or 2 compared to a control standard-of-care population (622 vs 947 vs 783 cells/uL, ns). Although limited pre-infusion CAR T cell specimens were available (n=7 for cohorts 1 and 2, n = 24 for SOC), we observed a relative increase in proportion of central memory populations in both the CD4 (46% vs 13%, p = 0.014) and CD8 (37% vs 13%, p = 0.035) T cell compartments. Conclusion: CAR T manufacturing was feasible with rare products out of specification. Salvage autoHCT in cohort 1 improved disease control pre apheresis, but timing from apheresis to CAR is important as the majority of patients progressed during manufacturing. CAR T cell products generated after autologous or allogeneic transplant may have distinct phenotypic properties. Further analyses are ongoing to evaluate expansion and persistence.