BCMA CAR T-cell expansion dynamics and toxicity after idecabtagene vicleucel and ciltacabtagene autoleucel for multiple myeloma

Abstract

Abstract Introduction: B-cell maturation antigen (BCMA)-targeted chimeric antigen receptor (CAR) T-cell therapies for relapsed/refractory multiple myeloma (RRMM) have shown deep and durable responses. However, real-world data on early CAR T-cell expansion dynamics and its association with treatment-related toxicities remain incomplete. This study aims to assess the early expansion of standard-of-care (SOC) idecabtagene vicleucel (ide-cel) and ciltacabtagene autoleucel (cilta-cel) and the association with toxicities and outcomes in the first 30 days.Methods: We analyzed retrospectively 76 RRMM patients (pts) treated with SOC ide-cel (n=19) and cilta-cel (n=57) and had CAR-T quantification available. Peripheral blood BCMA CAR-T cells were quantified by flow cytometry using BCMA protein/long peptide conjugated with PE. Patient demographics and characteristics were summarized through descriptive statistics and are reported in medians and interquartile ranges (IQR). Fisher's exact test and Wilcoxon rank sum test were used for categorical and continuous variables.We correlated peak inflammatory markers (ferritin, C-reactive protein [CRP]), absolute lymphocyte count (ALC), and absolute BCMA CAR-T peak (CD4+/CD8+), with the appearance of cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), delayed toxicities, and IEC-Enterocolitis. Response was assessed by the IMWG criteria. Kaplan-Meier method was used for time-to-event analysis. Results: The median age was 64 yrs (55-72), 67% had high-risk cytogenetics, and 14% had extramedullary disease (EMD). The median prior lines of therapy (LOT) was 4, and 58% received SOC CAR-T as late-line (≥4 LOT). Pts in the ide-cel group were older (73 vs 63 yrs), had received more prior LOT (5 vs 4, p=0.02), and had a higher proportion with ECOG ≥2 (21.1% vs 3.8%, p=0.049) vs those in the cilta-cel group. Any grade CRS was comparable (89% vs 84%; p=0.72), but with an earlier median onset in the ide-cel cohort (1 vs 7 days; p <0.0001). Any grade ICANS was significantly higher in the ide-cel group (26.3% vs 7.0%, p=0.038).For pts treated with ide-cel, peak CAR-T level reached a median of 600 cells/µL (IQR 293-1038). The expansion levels showed no significant correlation with CRS or ICANS grade ≥2 (p= 0.25 and p= 0.31, respectively). Similarly, no correlation was found with peak ALC, ferritin, or CRP for these toxicities or delayed toxicities. No cases of IEC-enterocolitis were reported in this group. 68% achieved ≥VGPR and showed a slightly higher median peak CAR-T level (641 vs 547 cells/µL) vs those with <VGPR, though not reaching statistical significance (p=0.57). Pts in the cilta-cel cohort had a significantly higher CAR-T peak level (1281 cells/µL; 460-2616; p=0.017) vs the ide-cel group. Furthermore, in the cilta-cel group, higher peak ALC was significantly associated with the development of grade ≥2 CRS (3.9 vs 1.6 K/uL, with and without, respectively; p=0.014), grade ≥2 ICANS (5.6 vs 1.8 K/uL; p=0.034), and any delayed toxicity (3.9 vs 1.8 K/uL; p=0.020). Additionally, the occurrence of any delayed toxicity was also associated with a significantly higher peak ferritin (5768 vs 990 ng/mL; p=0.029), whereas CRP was not. Pts who developed grade ≥2 CRS or ICANS had numerically higher CAR-T peak levels (1727 and 3756 cells/µL, respectively) compared to those with lower grade/no CRS/ICANS (1155 and 1277 cells/µL), though not statistically significant. Notably, the 2 pts who developed IEC-enterocolitis demonstrated markedly higher peak levels vs those without (26,152 vs 1281 cells/µL). However, only one had persistence of high circulating CAR-T levels beyond day 120 post-infusion. ≥VGPR was achieved in 82% of the pts who had a lower peak ALC than those who achieved <VGPR (1.8 vs 3.7 K/uL, p=0.047). Conclusion: We demonstrate that the two CAR-T products differ not only in toxicity profile but also in CAR-T expansion dynamics, with cilta-cel associated with a higher peak. We did not observe a significant correlation between CAR-T levels and most specific toxicities within subgroups, likely due to small sample size. However, there were notable trends for certain toxicities within the cilta-cel group for expansion, ALC, and inflammatory markers. These trends suggest that additional immune mechanisms may underlie these toxicities beyond the CAR expansion alone, and larger studies are warranted to confirm these findings to refine possible early predictors of toxicity.