Safety and immunomodulatory effects of siltuximab prophylaxis prior to standard of care CD19 directed chimeric antigen receptor T-cell (CD19.CART) therapy for B-cell lymphomas: Final phase I trial results

Abstract Introduction: CD19.CART is highly effective for B-cell non-Hodgkin lymphomas (NHL). However, cytokine release syndrome (CRS) and immune effector cell-associated neurologic toxicity syndrome (ICANS) remain a challenge. Interleukin-6 (IL-6) is a mediator of CRS/ICANS with CD19.CART. Siltuximab (siltux), an IL-6 antagonist, has efficacy in treating CRS/ICANS (Bajwa, Blood Advances, 2025) and has been shown to reduce checkpoint inhibitor toxicity while improving anti-tumor T-cell differentiation (Hailemichael, Cancer Cell, 2022; Speake et al. JCI, 2022). We hypothesized that prophylactic (ppx) siltux prior to CD19.CART would be safe, mitigate severe CRS/ICANS, and lead to unique CART differentiation. Methods: This phase I, investigator-initiated, single-center trial evaluated the safety of siltux before CD19.CART for patients (pts) with NHL. Adult pts received a standard of care CD19.CART and a single dose of siltux (11mg/kg) 1 hour prior to CART infusion. Ppx corticosteroids were not permitted. The primary endpoint was estimation of safety utilizing a Pocock-type stopping boundary with 5% probability of crossing the boundary when the true rate of dose-limiting toxicity (DLT) is 20%. Secondary endpoints included: incidence of grade > 3 CRS/ICANS, objective response rate (ORR), progression-free survival (PFS), and overall survival (OS). Exploratory endpoints included the evaluation of CART phenotype by high dimensional spectral flow cytometry and cytokine analysis by 32 plex semi-custom Luminex panel. A time-matched cohort of pts treated with SOC CD19.CART within 6 months of study pts who did not receive siltux ppx was utilized as a control. Results: 10 pts were treated from 1/2023 to 8/2023. Median age was 71 years (yrs) (range 53-79). Diagnoses included diffuse large B-cell lymphoma (n=6), high-grade B-cell lymphoma (3), and grade 3B follicular lymphoma (1). Median prior lines of therapy was 3 (1-5). CD19.CART product was axicabtagene ciloleucel (4 pts), tisagenlecleucel (4), and lisocabtagene maraleucel (2). At infusion, 7/10 pts had stage III/IV disease, 7/10 had elevated lactate dehydrogenase, 5/10 had elevated ferritin, 6/10 had elevated C-reactive protein, and 4/10 had revised international prognostic index > 3. No DLTs were observed. CRS occurred in 6/10, max grade (gr) 2 CRS (5 pts), gr 1 CRS (1 pt), and no gr > 3 CRS. Median time to CRS: 2.5 days (2-5); time to CRS resolution: 1.5 days (1-3). ICANS occurred in 4/10: gr 1 (1 pts), gr 2 (2 pts), and gr 4 (seizure; 1 pt). The latter was not on seizure ppx prior to ICANS. Median time to ICANS: 7 days (6-13); time to ICANS resolution: 1 day (1-4). ORR was 80% (95% CI, 44-97) with 70% complete responses (CR). Median follow-up was 1.9 yrs (1.5-2.2), 2-yr PFS and OS were 60% (95% CI, 25-83) and 70% (95% CI, 33-89). Three deaths occurred and were related to disease progression. High-dimensional flow cytometry analysis of 10 siltux ppx pts vs 26 non-siltux ppx control pts at D+14 post-CART revealed broad differences in CAR+ T-cell phenotype and function. For example, increased CAR+ CD8+ memory T-cells (CD45RO+/CD28+) were found with 13.8% in siltux vs. 7.7% (p=.03), and a trend towards less senescent effector CAR+ CD8+ T-cells (Granzyme B+/CD57+) with 25.5% in siltux vs. 38.8% (p=.08). Longitudinal cytokine analysis demonstrated improved inflammatory parameters typically associated with CRS/ICANS with siltux ppx. On D+14, median cytokine levels (log10pg/ml) were lower with siltux: TNF-a (1.32 vs 2.05, p=.02), IFN-g (.98 vs 1.56, p=.04), CXCL10 (2.43 vs 3.12, p=.04), and MIP1-alpha (p=.02). No significant differences were observed for IL-2 (1.06 vs 1.07, p=.47) or the homeostatic cytokines IL-15 (.902 vs 1.02, p=.32) and IL-7 (.667 vs .716, p=.61). IL-6 could not be accurately measured due to siltux bound to plasma IL-6; however, increased IL-6 receptor-alpha was consistent with lower effective IL-6 signaling with siltux ppx.Conclusion: Siltux was safe as ppx prior to CD19.CART. No gr > 3 CRS occurred and only 1 pt developed gr > 3 ICANS which fully resolved. CD19.CART ORR and CR rate were excellent and responses were durable. In addition, siltux ppx resulted in a favorable phenotypic and functional differentiation of CAR+ T-cells post-infusion and reduction in cytokines typically associated with CRS/ICANS. This is the first study to demonstrate the ability of IL-6 blockade in human subjects to alter CD19.CART cell differential expansion.