Safety and efficacy of out-of-specification ciltacabtagene autoleucel (cilta-cel) in Relapsed/Refractory multiple myeloma (RRMM)

Abstract Background: Cilta-cel is an anti-B-Cell maturation antigen (BCMA) CAR T-cell therapy approved for RRMM. Manufacturing failure (MF) and out-of-specification (OOS) products are not infrequent. Although OOS products may be administered to patients under the expanded access program (EAP), their real-world incidence and impact on outcomes in RRMM patients remain poorly characterized. Methods: We retrospectively reviewed patients undergoing cilta-cel apheresis from February 2022 through April 2025. OOS product was defined by failure to meet one or more FDA product release specifications but eligible for EAP use. MF was defined as insufficient product yield or failure to meet EAP criteria. Products with non–patient-related issues (e.g., transport delays, contamination) were excluded. Results: Of 212 patients undergoing apheresis, 18.9% (n=40) experienced OOS/MF (OOS:16%, n=34; MF:2.9%, n=6) events. In the overall OOS/MF population, median age was 64 years (range 35–81) with 60% of patients being male and 65% had high-risk cytogenetics. The most common reasons for OOS included low cell dose (52.9%), low viability (38%), low CAR-T expression (14.7%), low IFN-γ release (11.8%), and low IL-2 (5.9%). Nine patients had 2 criteria for OOS product. Very low cell yield/production failure (n=5) and high vector copy number (n=1) were causes for MF. Median absolute lymphocyte count (ALC) at apheresis was 1.1 ×10⁹/L (range: 0.2–5.2) with a median time from last therapy to apheresis of 34.5 days (7–661). Median lines of prior therapy (LOT) in the OOS group was 5 (2-17), and 6.5 (5-13) for MF group. Most patients (82.5%) had ≥4 prior LOTs. 65% received bridging therapy. Sixteen patients (40%) received bispecific antibodies (bsAbs; teclistamab: n=11; talquetamab: n=8, other: n=5) prior to apheresis for a median of 7 cycles (range: 1–34). Median time from last bsAb administration to apheresis was 43 (range: 13-472) days. 67% (4 of 6) of patients with MF had prior bsAbs compared to 38% (13 of 34) with OOS. Thirty-four patients had prior autologous transplant and 22 had prior alkylating agents (cyclophosphamide: n=18, bendamustine: n=4). The median time from the last dose of cytotoxic chemotherapy to apheresis was 223 days (21-2798). OOS/MF rates by year were 34.8% (2022), 17.6% (2023), 20% (2024), and 8.3% (Jan-April 2025). FDA approval for cilta-cel in earlier lines (2L+) and widening of release specifications occurred in April 2024, yet OOS/MF rates remained similar in the 12 months before and after this change (15% vs. 16%). Of the 34 OOS cases, 17 received an OOS product after the first apheresis and 6 after one re-collection. Two patients received in-specification products after remanufacturing from backup cells; one patient achieved an in-specification product after a third apheresis attempt two years after her prior collections. Five patients died before infusion, and three declined the OOS product. Among 6 MF cases, four underwent re-collection, all of which resulted in repeated MF. Among the 23 patients infused with an OOS product, the median follow-up was 6.9 months (6 days-2.6 years). Cytokine release syndrome occurred in 78% (all events grade 1–2); immune effector cell neurotoxicity syndrome occurred in 13%, with one grade 3 event. Fourteen patients received tocilizumab, and 5 received high-dose steroids. Only one patient experienced prolonged grade 3 neutropenia and thrombocytopenia beyond day 90. No patients died within 30 days of infusion. Eighteen patients were evaluable for response at day 90, and 5 patients had insufficient follow-up. The objective response rate (ORR) at day 90 was 89%, including 8 CR/sCR, 3 VGPR, and 5 PR. Progression-free survival (PFS) and overall survival (OS) rates at 6 months were 89% (95%CI: 61–97) and 88% (95%CI: 59–97), respectively. The median PFS and OS were not reached. Conclusions: In this real-world cohort, OOS/MF events were frequent but with a trend towards decreasing incidence over time. Despite failure to meet FDA release criteria, infused OOS products resulted in encouraging efficacy and manageable toxicity. Additional possible risk factors for OOS/MF and relative efficacy compared to an in-specification control cohort with longer-term follow-up will be presented at the meeting. The authors thank Johnson & Johnson and Legend Biotech USA Inc for their review.