Biological Correlatives of Efficacy and Toxicity of BCMA Targeting CAR T Cell Therapy in a Real-World Setting

Le résumé fourni par la source

Introduction B cell maturation antigen (BCMA) targeting CAR T cells have impressive efficacy in multiple myeloma (MM) with two FDA-approved therapies, idecabtagene vicleucel (ide-cel) and ciltacabtagene autoleucel (cilta-cel), now standard of care for relapsed and refractory patients. Differences in the efficacy and toxicity profile of the two products exist, as well as intra-product heterogeneity in cellular composition. We hypothesized that CAR T cell expansion and BCMA antigen load would contribute to toxicity and response and sought to evaluate this in a real-world setting. Methods Peripheral blood mononuclear cells (PBMCs) and serum were isolated from the blood of consenting MM patients treated with ide-cel and cilta-cel and bio-banked on an IRB-approved protocol. Thawed PBMCs underwent flow cytometry to measure the % of CD3+ cells that were CAR+ (% CAR+) at D+14 post-infusion. Soluble BCMA (sBCMA) levels were measured in serum pre and D+28 post-infusion (Ella Protein Simple, Bio-Techne) to assess target antigen burden. Results Demographics Forty-five patients (ide-cel N=22, cilta-cel N=23) were included in the cohort, which was deliberately enriched for cases of movement and neurocognitive toxicity (MNT) and cranial nerve palsy (CNP) to investigate these complications. Median age was 68 years (range 48-85, ide-cel 68, cilta-cel 66), 69% were male (ide-cel 64%, cilta-cel 74%). Patients had received a median of 5 prior lines of therapy (both products), 76% were penta-refractory (ide-cel 73%, cilta-cel 78%), and 40% had prior BCMA targeting therapy (ide-cel 36%, cilta-cel 43%). Toxicity Eighty-four percent had cytokine release syndrome (CRS) (ide-cel 91%, cilta-cel 78%), 31% immune effector cell associated neurotoxicity (ICANS) (ide-cel 32%, cilta-cel 30%). High-grade CRS and ICANS were rare, with 2% CRS ≥ grade 3 (ide-cel n=0, cilta-cel n=1, 4%) and 7% ICANS ≥ grade 3 (ide-cel n=1, 5%, cilta-cel n=2, 9%) (ASTCT grading). Six patients had MNT (ide-cel n=1, cilta-cel n= 5), and 4 had CNP (ide-cel n=0, cilta-cel n=4). Two of 4 patients who had CNP also had MNT. All CNPs involved the peripheral facial nerve, and this was bilateral in 1 patient. % CAR+ T cells at D+14 was significantly higher in patients who subsequently developed MNT (median 97% vs. 49% p=0.0014) and CNP (median 94%, p=0.0092) and this difference was maintained when only cilta-cel patients were analyzed (MNT median % CAR+ 97% vs. 55% p=0.0085, CNP 94% p=0.0162). Median MNT onset was D+ 44 (range 20-126), median CNP onset was D +35 (range 23-77). Baseline sBCMA levels were similar in patients who developed MNT, compared to those who did not. Efficacy Sixty-nine percent had best overall response (BOR) ≥ partial response (PR) (ide-cel 64%, cilta-cel 74%), and median progression-free survival (PFS) was 9.6 months (ide-cel 7.5, cilta-cel 11.6 months, p=0.242). % CAR+ T cells at D+14 was associated with efficacy in cilta-cel patients but not ide-cel patients; BOR ≥PR (ide-cel p=0.6007, cilta-cel p=0.0015) and PFS ≥ 6 months (ide-cel p=0.5887, cilta-cel p=0.0009). Cilta-cel patients had a dramatically worse PFS (2 months vs. not reached, p=0.0002, median follow-up 12.3 months) if they received a prior anti-BCMA drug conjugate (n=7), and this was associated with low % CAR+ T cells at D+14 (36% CAR+ versus 81%, p=0.0029). Baseline sBCMA levels were higher in these patients suggesting antigen persistence and higher disease burden. In the entire cohort, baseline sBCMA levels were higher in patients who achieved Conclusion High % CAR+ T cells at D+14 is associated with atypical neurotoxicities, including MNT and CNP. These predominantly occurred in cilta-cel patients, but one case of MNT developed following ide-cel infusion, with high CAR expansion. This implicates BCMA targeting or antigen cross-reactivity in the pathogenesis of MNT. Percent of CAR+ T cells at D+14 was associated with efficacy in cilta-cel patients but not ide-cel patients; this may be due to the differential timing of peak expansion, or that this time point is not key to predicting response in different products. sBCMA levels were higher in primary refractory patients at baseline and D+28, suggesting complete antigen loss is a rare event at early time points.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.