Glofitamab Monotherapy in Patients with Non-Hodgkin B-Cell Lymphoma after Failing CAR T-Cell Infusion: Primary Analysis of the Bicar Study, a Phase II Lysa Study

Introduction.

Chimeric antigen receptor (CAR) T-cells have improved outcomes of patients (pts) with relapsed/refractory diffuse large B-cell lymphoma (R/R DLBCL). However, 40% to 60% of those pts will experience relapse or progression with a median overall survival (mOS) shorter than 6 months (mo) according to the French DESCAR-T registry study. Glofitamab is a 2:1 CD20xCD3 bispecific antibody that demonstrated efficacy in pts with R/R DLBCL. We report the results of the primary analysis of a phase II multicenter study using glofitamab in pts with R/R non-Hodgkin B-cell lymphoma after CAR T-cell infusion.

Patients and Methods

Pts with a biopsy-proven DLBCL (cohort 1) or non-DLBCL (cohort 2), with no metabolic response, progression or in relapse at least one month after CAR T-cell infusion, received obinutuzumab (1000 mg) 3 days before (Day -3, D-3) the first glofitamab dose. Intravenous glofitamab was administered with step-up dosing on D1 (2.5mg), D3 (10mg) and D8 (30mg) of Cycle (C) 1 and at 30mg on D1 of C2-11 (21-day cycles). The primary endpoint was defined for cohort 1 with a null hypothesis of 6.3 mo of mOS and a hypothesis of improvement to 12.6 mo (hazard ratio [HR]: 0.5). No hypothesis testing was planned for cohort 2. On Dec 22nd, 2022, enough events had been observed allowing the analysis of the primary endpoint based on data exported on June 2nd, 2023.

Results.

As of June 2 ^nd, 2023, 67 pts were enrolled, and 63 pts (44 pts in cohort 1, 19 pts in cohort 2) received ≥1 dose of study treatment. All pts in cohort 1 had a biopsy proven DLBCL whereas pts in cohort 2 had follicular lymphoma (n=6), mantle cell lymphoma (n=5), transformed follicular lymphoma (n=4), transformed marginal zone lymphoma (n=2), primary mediastinal B-lymphoma (n=1), or t-Waldenström macroglobulinemia(n=1). In combined cohort 1 and cohort 2, median age was 65 years (range: 33-77) and 84.1% of pts had Ann Arbor stage III/IV. Median number of prior therapies received was 3 (range: 2-6; ≥3 prior therapies: 87.3%; IPI ≥3: 58.7%). Prior CAR T-cell therapy was axi-cel (n=29, 46%), tisa-cel (n= 25, 39.7%), brexu-cel (n=5, 7.9%) or investigational CAR T-cell (n= 4, 6.3%). Thirteen pts (20.6%) were refractory (no response to CAR T-cells), 50 pts (79.4%) were in relapse/progression, among these, 15 (30%) relapsed/progressed between 1-3 mo, 18 (36%) between 3-6 mo and 17 (34%) >6mo after CAR T-cell infusion. The median number of glofitamab cycles administered was 5 (range: 1-12). Nine pts (14.3%) experienced cytokine release syndrome (CRS) (grade [G]1: 3 pts, G2: 6pts) and 2 pts experienced neurologic events (NE) G2. Neutropenia G ≥3 was observed in 22 pts (33.3%), thrombocytopenia G ≥3 in 9 pts (11.1%) and anemia G ≥3 in 7 pts (11.1%). Eighteen pts (28.6%) experienced serious adverse events (SAE) G ≥3, mainly infection (17 pts, among them 9 pts related to COVID-19). Thirty-five pts (55.5%) permanently discontinued glofitamab: 28 pts (44.4%) for progression, 4 pts for death and 3 pts for toxicity (musculoskeletal pain, hepatitis, genital infection). With a median follow-up of 9.7 mo (95% CI: 8.1-11.8), for cohort 1 the mOS was 17.6 mo (90% CI: 8.3-19.7), the lower limit of the 90% CI was ≥6.3 and so the null hypothesis could be rejected, and the primary endpoint was met. The mOS for cohort 2 was not reached (NR) (90% CI: 6.4-NR). According to central review, overall best metabolic response was 65.9% (CMR: 36.4%) and 57.9% (CMR 52.6%) in cohort 1 and 2, respectively. The median duration of CMR was 19.7 mo (95% CI: 4.9-19.7) and NR (95% CI: 2.8-NR) for cohort 1 and 2, respectively. Median progression-free survival was 4.9 mo (95% CI: 2.6-19.7) and 4.1 months (95% CI:1.4-NR for cohort 1 and 2, respectively).

Conclusion.

Glofitamab demonstrated a significant increase in OS when used in pts in first relapse/progression after CAR T-cells. In this context, glofitamab resulted in deep and durable responses and manageable safety.