First-line salvage therapies in relapsed/refractory large B-cell lymphoma after second- or third-line CD19-directed CAR T-cell therapy

Abstract Introduction: CD19-directed CAR T-cell (CAR-T) therapy is increasingly deployed in earlier lines of treatment for relapsed/refractory large B-cell lymphoma (r/r LBCL), including 2nd and 3rd line settings. This shift raises important questions about post-CAR-T management, yet existing data largely reflect heavily pretreated patients, limiting relevance to current practice. In this evolving context, the clinical features, salvage strategies, and outcomes following relapse after early-line CAR-T remain poorly defined. Objective To determine the effectiveness of first-line salvage therapies and associated survival outcomes in patients with r/r LBCL relapsing after as2nd or 3rd line CAR-T therapy. Methods This international, multicenter retrospective cohort study included LBCL patients from five institutions, who received CAR-T therapy as 2nd or 3rd line therapy. Response was assessed per Lugano criteria. One-year progression-free survival (PFS) and overall survival (OS) were calculated from the initiation of salvage therapy. Prognostic factors were evaluated using univariable and multivariable Cox regression models. Results:Among 557 LBCL patients treated with 2nd line (31%) and 3rd line (69%) CAR-T, the 1-year OS and PFS was 76% (95% CI 73-80) and 54% (49-58), respectively; 18% were refractory to CAR-T therapy. Cumulative incidence (CI) of relapse/progression at 1-year post-CAR-T was 42% (37-46), with a median time to relapse of 2.9 months (IQR 1.2-6.0). Notably, the complete responses (CR) rates and relapse or progression CI were comparable between 2nd and 3rd line CAR-T (71% vs. 66% and 37% vs. 44%, respectively). Of 232 patients who relapsed or progressed following CAR-T, 182 (78%) received subsequent treatment. The median age of the treated cohort was 63 years (IQR 52-73); 52% had received axi-cel, 24% tisa-cel, 18% liso-cel, and 6% point of care CD19-CAR-T product. Most patients had received CAR-T as 3rd line therapy (72%), with the remainder treated in the 2nd line setting. Overall response rate (ORR) to post-CAR-T treatment was 49% (34% CR, 15% PR). With a median follow up of 22 months (IQR 8-40), 1-year OS from post-CAR-T treatment was 47% (95% CI 39-56) and PFS was 29% (22-38). In multivariable Cox regression adjusting for age, disease transformation, CAR-T product, and best response to CAR-T, the line of prior CAR-T therapy was not associated with OS (HR 1.29, 95% CI, 0.75–2.23) or PFS (HR 0.95, 95% CI, 0.61–1.47). One-year OS rates following salvage therapy was comparable between 2nd and 3rd line groups (44% vs. 48%, p=0.76), as was PFS (23% and 31%, p=0.66). Post-CAR-T treatments were heterogeneous, with the most common being involved-site radiation therapy (ISRT)- (n=39) followed by bispecifics- (n=37), bendamustine- (n=23), and polatuzumab- (n=21) containing regimens. Checkpoint inhibitors-containing regimens were administered to 14 patients. These categories were not mutually exclusive and therefore not directly comparable. Among patients treated with polatuzumab-containing regimens, 14% had prior exposure to the same agent; while bispecific-containing regimens had 27% previous exposure to polatuzumab, and 3% to bispecifics. Response rates to systemic treatments varied by regimen-containing group: polatuzumab- and checkpoints inhibitors-containing regimens demonstrated the highest ORR (57%), followed by bispecifics- containing regimens (56%). At 1-year from post-CAR-T treatments OS was approximately 50%, reaching the highest with checkpoint inhibitors- (70%), followed by bispecifics- (60%), and the lowest with anthracycline/platinum based chemotherapies-containing regimens (33%). However, lower PFS (around 30%) was observed, ranging from 51% with bispecifics-, followed by checkpoint inhibitors- (44%), to bendamustine-containing regimens (8%). Conclusions:This large, multicenter study is among the first to evaluate first-line salvage therapies in patients with LBCL relapsing after 2nd or 3rd line CD19 CAR-T. Salvage treatments yielded meaningful responses and survival, regardless of CAR-T therapy line. Immunotherapy-based regimens demonstrated encouraging activity, though longer follow-up is needed to confirm durability of response and survival benefit. These findings may guide future post–CAR-T treatment considerations and support further exploration of immunotherapy-based approaches in this high-risk population.