Impact of the number of bridging therapy lines on outcomes after axi-cel in LBCL: A lysa study from the descar-T registry

Abstract Introduction The optimal use of bridging therapy (BT) prior to anti-CD19 CAR T-cell therapy in large B-cell lymphoma (LBCL) remains uncertain. Response to BT has been associated with improved progression-free survival (PFS) (Rodie C et al., Blood Adv 2023), suggesting that disease control before infusion may enhance outcomes. However, whether administering additional BT lines to convert non-responders into responders improves outcomes remains unknown. To address this, we analyzed the impact of the number of BT lines on clinical outcomes in a real-world cohort. Methods We included adult patients with LBCL from the French DESCAR-T registry (NCT04328298) who had received ≥2 prior therapies, underwent leukapheresis between July 2018 and September 2024, received ≥1 BT line and were intended for axi-cel. Patients receiving only a holding therapy (i.e., ending before leukapheresis) were excluded. The primary endpoint was overall survival (OS) from leukapheresis. Secondary endpoints included OS and PFS from infusion, best overall response rate (ORR), complete remission rate (CRR, per Lugano 2014), and rates of CRS, ICANS, and persistent hematotoxicity (neutropenia, thrombopenia, and/or anemia lasting ≥30 days). To balance baseline characteristics, we used propensity score (PS) methods based on variables measured at the multidisciplinary tumor board (MTB) before leukapheresis. PS variables were selected through univariate and multivariate logistic regression. Final selection included age, sex, ECOG performance status, comorbidities, Ann Arbor stage, LDH, number of prior lines, time since last therapy, and year of MTB. Stabilized inverse probability of treatment weighting was applied to complete cases. Results A total of 777 patients were analyzed: 542 with diffuse large B-cell lymphoma, 150 with transformed indolent lymphoma, 43 with HGBL, 35 with PMBCL, and 7 with transformed Hodgkin lymphoma. Median age was 63 years (range, 18–80); 63% were male. One BT line was administered in 613 patients, while 164 received >1 line (2 lines: n=146; ≥3: n=8). Compared to the 1 BT line group, patients receiving >1 line more often had ECOG >0 (68% vs. 58%; p=0.01), aaIPI ≥2 (62% vs. 45%; p<0.001), elevated LDH (77% vs. 60%; p<0.001), stage III–IV disease (80% vs. 72%; p=0.02), and a shorter median time since last therapy (28 vs. 43 days; p<0.001). After a median follow-up of 25.3 and 26.7 months in the 1 and >1 BT line groups, crude 2-year OS from leukapheresis was higher in the 1 BT line group (54.9% vs. 37.2%; p<0.0001), as were OS from infusion (56.5% vs. 39.2%; p<0.0001) and PFS (43.3% vs. 30.5%; p=0.0015). Responders (CR or PR) to 1 BT line (n=212) had higher PFS (p=0.01) and OS (p=0.008) from infusion than those responding only after >1 line (n=51), with a 2-year PFS of 53.2% vs. 33.4% and 2-year OS of 64.9% vs. 47.2%. To limit confounding, we focused on patients in PD after their first BT line, with complete data and no holding therapy (n=257). All had undergone leukapheresis with intent to receive axi-cel, though not all proceeded to infusion. Among them, 46 (18%) received ≥1 additional BT line and 211 (82%) did not. Baseline characteristics were similar, except for a higher rate of transformed disease in the 1 BT line group (24% [50/211] vs. 9% [4/46]; p=0.02) and a longer median time since last therapy (41 vs. 35 days; p=0.04). Disease status improved in 33% of patients receiving >1 BT line (CR in 4, PR in 11). Axi-cel was infused in 93% of the 1 BT line group (all in PD) and 89% of the >1 BT line group. After PS-weighting, best ORR and CRR were 75.2% and 55.7% in the 1 BT line group vs. 65.9% and 41.6% in the >1 BT line group (p=0.21 and p=0.09, respectively). CRS and ICANS rates did not differ significantly, but persistent hematotoxicity was more frequent with >1 BT line (86.6% vs. 72.1%; p=0.047). One BT line was associated with improved 2-year OS from leukapheresis (40.9% vs. 19.1%; p=0.021), OS from infusion (43.3% vs. 18.3%; p=0.002), and numerically higher PFS (32.1% vs. 16.2%; p=0.12). Conclusion In this large real-world cohort of patients receiving axi-cel for LBCL after ≥2 prior lines, administering >1 BT line was associated with higher rates of persistent hematotoxicity and inferior OS compared to a single BT line, even in patients with PD after initial BT. While additional BT lines may improve disease status at infusion in a minority of patients, delaying CAR T-cell infusion for this purpose does not appear to improve survival.