Autologous stem cell transplantation (ASCT) remains an important consolidative therapy for patients with recurrent non-Hodgkin lymphoma (NHL), but is limited by the high incidence of NHL relapse. We report a Phase I clinical trial of ASCT followed by CD19-specific CD28-costimulatory chimeric antigen receptor (CD19:28z-CAR) T cells, with the goal of reducing NHL relapse rates. Safety and feasibility were the primary objectives, with CAR T cell persistence and expansion in the myeloablative ASCT setting as secondary objectives.This study examines safety and feasibility for two manufacturing platforms that differed in the T cell subset composition used for CAR engineering. Initially, the T cell population for CAR transduction was central memory (Tcm)-enriched: participants' peripheral blood mononuclear cells (PBMC) were depleted for CD14+ monocytes, CD25+ Tregs, and CD45RA+ naïve and stem-memory T cells, after which they were selected for CD62L+ Tcm (Wang et al. Blood;127:2980). Based on comparative preclinical data, a second arm was added to the trial to evaluate a Tn/mem-derived manufacturing platform that enriched central memory, naïve, and stem memory T cells (no CD45RA+ depletion). Either Tcm- or Tn/mem-enriched T cells were activated with CD3/CD28 beads, transduced with lentiviral vector encoding the CD19:28z-CAR, and expanded ex vivo.This phase I trial used the toxicity equivalence range design (Blanchard and Longmate. Contemp Clin Trials; 32;114) with an equivalence range for DLTs of 0.20-0.35 and a target toxicity rate of 0.25. The first 3 participants on each arm were followed one at a time, with later accrual in cohorts of 3. Twenty-three participants underwent ASCT and received CD19:28z-CAR T cells 2 days post stem cell infusion at the assigned dose level (DL): 17 on the Tcm arm (DL 50 million [M] CAR+ T cells [n=3], 200 M [n=5], 600 M [n=9]); 6 on the Tn/mem arm (DL 200 M). Participants were followed for dose limiting toxicity (DLT) for 28 days. Table 1 shows results by arm and DL. Both arms demonstrated safety and feasibility. There was no delayed hematopoietic reconstitution on either arm. On the Tcm arm, the only DLT was at DL 600 M (1 of 9 at 600 M). The Tn/mem arm was opened at 200 M and 6 participants were treated with no DLTs. The dose was not escalated as the protocol management team had seen activity at the 200M level in 2 other trials using the Tn/mem product.Tcm Arm: Fourteen of 17 participants (82%) had a diagnosis of diffuse large B-cell lymphoma (DLBCL) and 3 had mantle cell lymphoma. The mean age of the participants on the Tcm arm was 57 (35-75). The median number of prior chemotherapy regimens was 2 (1-5). The median progression-free survival (PFS) was 34.6 months 95% CI [21.8, undefined]. Seven of 17 participants (41%) have progressed, 1 died in remission of unrelated intracranial hemorrhage (6%), 7 (41%) remain in CR and are still in follow-up, and 2 are lost to follow-up (12%). All 17 participants achieved a CR or a continuing CR after ASCT and T cells.Tn/mem arm: Five of 6 participants (83%) had a DLBCL diagnosis, and 1 was NHL not otherwise specified. The mean age of the participants was 50 (40-72). The median number of prior chemotherapy regimens was 2.5 (1-3). The median follow-up time for the Tn/mem arm was 12 months, with median PFS not yet reached. One of 6 (17%) has progressed, 4 (66%) remain in CR and are still in follow-up, and 1 is lost to follow-up (17%). Five of 6 (83%) participants achieved a best response of CR or continuing CR after therapy.Several differences were observed between the manufacturing platforms. Since the Tn/mem production platform has fewer depletion steps, it resulted in a higher product yield, which shortened the ex vivo expansion period by 4.1 days (95% CI [1.5%, 6.6%]) from 18.9 days (15-24) for Tcm to 14.8 days (12-18) for Tn/mem (P<0.005). Notably in the ASCT minimal disease burden setting, the Tn/mem-derived CD19:28z-CAR T cell products exhibited significantly higher in vivo CAR T cell expansion compared to Tcm products at the 200M DL (Figure 1). We conclude that although both Tcm- and Tn/mem-enriched CD19CAR T cell therapies are safe, the Tn/mem product is more promising due to its 1) shorter production time, 2) higher cell yield, and 3) better in vivo expansion, despite the low antigen drive in these patients post-salvage and ASCT therapy. Longer follow-up for the 2-year PFS secondary objective will indicate if improved Tn/mem expansion impacts tumor control. DisclosuresWang:Mustang Therapeutics: Other: Licensing Agreement, Patents & Royalties, Research Funding. Budde:Mustang Therapeutics: Consultancy, Other: Licensing Agreement, Patents & Royalties, Research Funding. Brown:Mustang Therapeutics: Consultancy, Other: Licensing Agreement, Patents & Royalties, Research Funding. Forman:Mustang Therapeutics: Other: Licensing Agreement, Patents & Royalties, Research Funding.
Read full abstract