Idecabtagene Vicleucel (ide-cel, bb2121) Responses Are Characterized By Early and Temporally Consistent Activation and Expansion of CAR T Cells with a T Effector Phenotype

Background: Patients (pts) with relapsed and refractory multiple myeloma (RRMM) experience unsatisfactory outcomes with established treatment modalities. In the pivotal phase 2 KarMMa study (NCT03361748), idecabtagene vicleucel (ide-cel, bb2121) demonstrated frequent, deep, and durable responses in triple-class exposed pts with RRMM, with an overall response rate (ORR) of 73% and a complete response rate of 33% (Munshi et al. J Clin Oncol. 2020;38[suppl, abstr]:8503). The overall safety profile of ide-cel was also manageable. The median time to onset and duration of cytokine release syndrome (CRS) were 1 d and 5 d, respectively, and the median time to onset and duration of neurotoxicity (NT) were 2 d and 3 d; the frequencies of higher-grade CRS and NT were low. Evaluated here are the T cell phenotypes, soluble factors, and cytokines associated with ide-cel activation, CRS, NT, and tumor responses over time in pts who received ide-cel in the KarMMa study.

Methods: After longitudinal sampling of peripheral blood post-ide-cel infusion in the KarMMa study (N=128), plasma was analyzed for levels of proinflammatory cytokines and inflammation-related soluble factors; serum was evaluated for soluble BCMA (sBCMA) as a peripheral surrogate measure of tumor burden, and peripheral blood mononuclear cells were assessed by flow cytometry for memory phenotypes of CAR T cells. Associations between these features over time after ide-cel infusion were evaluated in the context of ORR, ongoing response at 9 mo, and grade ≥2 CRS and NT. Response at 9 mo was selected for analysis because this visit was proximal to the median progression-free survival (PFS) reported in all ide-cel-treated pts in KarMMa. The 9-mo responders were defined as pts with assessments >8 mo postinfusion and no progression before 10 mo postinfusion.

Results: The levels of both CD4⁺ and CD8⁺ populations of CAR T cells increased to a greater degree postinfusion in responders and were skewed towards a higher fraction of CD8⁺ cells through peak expansion. Cell expansion in responders was characterized by an increased proportion of T_EM in CAR T cells (CCR7⁻/CD45RA⁻) for both CD4⁺ and CD8⁺ subsets. Congruent with dominant T_EM expansion, characteristic T_EM-associated proinflammatory cytokines, such as IFN-ɣ and IL-6, were consistently upregulated early after infusion. Peak IFN-ɣ and IL-6 levels occurred a median of 4 d postinfusion, and 90% of pts (5^th−95^th percentile) had IFN-ɣ and IL-6 peaks 21 d and 15 d postinfusion, respectively, which was in line with the observed early onset of CRS and NT. Pharmacodynamic responses, shown by decreases in sBCMA after infusion, also occurred consistently early after infusion, and the sBCMA nadir occurred in 90% of pts 7 d−85 d postinfusion (5^th−95^th percentile; median, 31 d). Early sBCMA clearance below the limit of detection of the assay was associated with longer responses, and median PFS was significantly longer in pts with undetectable sBCMA vs detectable sBCMA at 2 mo (12.3 mo [95% CI, 11.6−17.7] vs 2.9 mo [95% CI, 1.9−3.1]; P<0.0001). Several inflammation-related soluble factors that may modulate T cell effector functions were negatively correlated with ongoing responses at 9 mo postinfusion. Higher peak levels of both CD4⁺ and CD8⁺ CAR T cells and a higher proportion of T_EM CAR T cells were associated with pts experiencing grade ≥2 CRS or NT. These observations were consistent with the effector-dominant profile of ide-cel in responders and the reported associations between overall CAR T cell activation and expansion, efficacy, and safety. No difference was observed in the CD4:CD8 ratio during initial expansion in pts with grade ≥2 CRS or NT vs that in pts with grade <2 CRS or NT.

Conclusions: Postinfusion expansion of ide-cel was characterized by a T_EM-dominant phenotype with expansion of both CD4⁺ and CD8⁺ CAR T cell populations. The relative magnitudes of the increase in both populations were substantially greater in responders vs nonresponders, with a bias toward a higher proportion of CD8⁺ CAR T cells in responders. The induction of effector cytokines followed by return to baseline levels occurred predictably and early after infusion and was consistent with both the observed early onset and resolution of CRS and early sBCMA clearance in pts who received ide-cel in the KarMMa study.