Influence of Siltuximab for CAR T-Cell Related CRS/Icans on CAR T-Cell Reconstitution, Function, and Clinical Outcomes

Background:

Cytokine release syndrome (CRS) and immune effector cell associated neurotoxicity syndrome (ICANS) are potentially fatal complications after CAR T-cell infusion. It is essential to investigate biomarkers related to their onset and severity and to develop effective prophylactic and therapeutic strategies without affecting the infused cells. We report on the results of 20 pts treated on a ph-2 study of siltuximab for CAR-T-related CRS/ICANS, with a focus on its impact on CAR T-cell reconstitution and function (NCT04975555).

Methods

Adult patients (pts) with R/R Large B-cell lymphoma (LBCL), follicular lymphoma (FL), mantle cell lymphoma (MCL) and multiple myeloma (MM) planned to receive CAR T-cell were screened. Pts with persistent Grade 1 CRS/ICANS ≥ 12 hours or grade ≥2 CRS/ICANS were enrolled and received siltuximab 11mg/kg. A repeat dose was permitted if CRS/ICANS persisted ≥12 hours from the first dose. Pts requiring rescue tocilizumab or worsening CRS were considered non-responders. CRS/ICANS were assessed per the ASTCT consensus grading. The primary objective of this analysis is to evaluate the pattern of lymphocyte and CAR T-cell immune reconstitution (median and range) in pts who achieved complete remission of the primary disease versus those who did not, after siltuximab exposure. To assess the impact of siltuximab on CAR T-cells function we performed cytokine analysis on PBMCs isolated before and after siltuximab administration, used CAR T-cells isolated from siltuximab exposed patients in an in vitro killing assay with CD19 targets, and analyzed exhaustion markers (LAG3, PD1, and TIM3) on CAR T-cells using flow cytometry. Lymphocyte markers were assessed by standard flow cytometry, and the CAR expression was detected using the Miltenyi detection system. Cytokines were measured using the BD CBA assay, or the Isoplexis technology.

Results

Pts with CRS/ICANS non-responsive to siltuximab (N:5) had a significantly higher (P≥0.01, two tail T-Test) body temperature and elevation of CRP, procalcitonin, and IL6 vs. responders who required more than 1 dose (N:10), or 1 dose (N:4). The levels of IL2, IL4, IL10, TNF, IFNG, IL17A were not significantly elevated in any patient. CAR expression was detected in all pts. The median maximum number of CAR T-cells in the blood was 32/µl (range 1-1384). CAR T-cells expression was higher in pts who had CR of the primary disease 44/µl (4-1384) vs. 15/µl (1-733) versus those who did not. The median CD3+ lymphocyte count was 290 cells/µl (66-3082) vs. 254/µl (158-909), respectively. CAR T-cells had maximal expansion after 10±6.7 days post CAR T-cell infusion, and 8±6.8 days post CRS/ICANS (mean±SD). Isoplexis analysis on PBMC pre and post siltuximab showed that it did not hamper the single cell secretome assessed by a panel of 30 cytokines/chemokines. We tested activated T-cells isolated after siltuximab treatment from the PBMCs of four pts with NHL and demonstrated clearance of CD19+ targets in vitro. Expanded circulating CAR T-cells had higher levels of PD1 (18.4±30%), as compared with LAG3 (2.77±3.9%) or TIM3 (2±2%). To mimick in vivo antigen stimulation, activated T-cells from 4 pts were cultured overnight with or without 80 Gy irradiated targets (NALM6, Jurkat or Jurkat engineered to express CD19). We observed upregulation of LAG3 >PD1 >TIM3 exhaustion markers, vs. unstimulated activated T-cells or T-cells not expressing the CAR molecule.

Conclusions

Siltuximab was safe and effective for managing CRS/ICANS. Also, siltuximab did not hamper CAR T-cells expansion and function. Optimal expansion of T-cells may associate with superior tumor response. Antigen-driven upregulation of LAG3, PD1, and TIM3 markers can be amenable to therapeutic intervention to reduce T-cells exhaustion.