Real world outcomes of sequencing bispecific antibodies (BsAb) in patients with relapsed / refractory follicular lymphoma (R/R FL): A multi-center cohort study from cubic consortium

Abstract Background: BsAb have significantly improved outcomes for patients (pts) with multiply relapsed FL. However, with the availability of several therapeutic options for R/R FL, the appropriate sequencing of BsAb is unclear. Hence, we sought to evaluate the real-world utilization patterns of BsAb and how sequencing affects their efficacy and toxicity. Methods: This is a retrospective multicenter observational study of patients with R/R FL who received BsAb between 2023 and 2025 at 11 US medical centers. Patients with transformed FL were excluded. Descriptive analysis was conducted to calculate the overall response rate (ORR) and complete response rate (CRR) stratified by prior bendamustine (benda) use, number of prior lines of therapy (LOT) and POD24 status - defined as progression of disease within 24 months from first line treatment. The primary endpoints were progression-free survival (PFS) and overall survival (OS), both measured from the start of BsAb. Univariable Cox proportional hazard models were used to evaluate the overall survival (OS) and progression-free survival (PFS) in different subgroups. Variables with p<0.1 in the univariable models were included in multivariable Cox models. Exploratory analyses of PFS and OS were conducted among pts who received CAR-T after BsAb. Results: Among the 86 pts who received a BsAb, the median age was 68 years (range 33 – 92) with 53% males, 87% White, 79% with ECOG performance status of 0-1, 71% with stage 3-4 disease and 42% with FLIPI 3-5 (high-risk). Mosunetuzumab was the most utilized BsAb (n=77; 89%) followed by epcoritamab (n=7; 8.1%) and glofitamab (n=2; 2.3%). Median prior LOT were 3 (range 1-14) with POD24 in 42% of pts and 44.2% refractory to the last LOT. All pts had prior exposure to anti-CD20 monoclonal antibody therapy, 66.3% to prior anthracycline, 57% to prior benda and 39.5% to prior lenalidomide. Among response evaluable pts (n=81), the best ORR and CRR to BsAb were 86.4% (n=70) and 70.4% (n=57), respectively. ORR was similar irrespective of prior CAR-T (vs not, 86.8% vs 84.6%), prior benda exposure (vs not, 86.7 % vs 86.1%) or by prior number of LOT (≤3 vs ≥4, 89.7% vs 84.6%). The CRR was higher in pts without prior CAR-T (75% vs 46%) and without prior benda (77.8% vs 64.4%) but similar regardless of number of prior LOT (≤3 vs ≥4, 69% vs 71.1 %). Pts with POD24 had lower ORR (78.1% vs 91.8%) and CR (68.8% vs 71.4%) compared to non-POD24. Median follow up was 12.1 months (range 0 – 35 mos). Median PFS and OS were not reached. 1-year PFS and OS estimates were 61.6% and 89.4%, respectively. In the univariable Cox model, PFS was not significantly inferior in pts with prior CAR-T (HR 1.68, 95% CI (0.73-3.86)), prior benda exposure (HR: 1.68, 95% CI (0.77-3.66)), POD24 status (HR: 1.45, 95% CI 0.68-3.08), or ≥4 prior LOT (HR 1.01, 95% CI (0.44-2.32)). Similar patterns were seen for OS in univariate analyses. In the multivariable analysis, high LDH prior to BsAb was associated with slightly worse OS (HR 1.004, 95% CI (1.002–1.006)) and PFS (HR: 1.005, 95% CI (1.001 – 1.008)). 14 pts received CAR-T prior to BsAb and 8 pts received CAR-T after BsAb. Among pts who received CAR-T after BsAb (n=8), PFS from time of BsAb was significantly inferior (HR: 8.9, 95% CI (3.4-22.3)) compared to pts who did not receive CAR-T. In contrast, there was no significant difference in PFS among patients with CAR-T prior to BsAb compared with those without prior CAR-T exposure (HR 0.37, 95% CI (0.13 – 1.08)). OS was not significantly different irrespective of the timing of CAR-T and BsAb. The rate of any grade cytokine release syndrome post BsAb was 34.9% (n = 29) with no grade 3 or higher events. Immune effector cell-associated neurotoxicity syndrome occurred in 2.4% (n=2), all with grade 1. Infections occurred in 27.4% (n=23) pts following BsAb therapy. Conclusion: We report, for the first time, the real-world outcomes of sequencing BsAb in pts with R/R FL. Prior CAR-T use and benda exposure were associated with similar ORR but lower CRR. No significant differences in PFS or OS were observed based on prior benda or the number of LOT. Notably, prior CAR-T did not impact PFS to subsequent BsAb therapy. The subgroup of patients who received CAR-T after BsAb demonstrated inferior PFS, suggesting a high risk population. However, these results need to be interpreted with caution due to small sample size. Our findings warrant further investigation in a larger cohort with longer follow-up.