Background
Phase III, open-label RCT. 322 patients with R/R DLBCL after first-line chemoimmunotherapy who were eligible for and intended to proceed to autologous SCT. Tested tisagenlecleucel (tisa-cel), a CD19-targeted CAR-T therapy, versus standard second-line platinum-based salvage chemotherapy followed by ASCT in responders. BELINDA was the first randomized comparison of CAR-T versus salvage chemo + ASCT in 2L DLBCL.
Interventions and follow up
Arm A: Tisagenlecleucel (tisa-cel) — leukapheresis → bridging chemotherapy (allowed) → lymphodepletion (fludarabine + cyclophosphamide) → single infusion of tisa-cel 0.6–6 × 10⁸ CAR-T cells
Arm B: Standard salvage chemoimmunotherapy (R-DHAP, R-ICE, R-GDP, or R-GemOx for 2–3 cycles) followed by ASCT in responding patients
Primary endpoint: Event-free survival (EFS)
Median follow-up: 10.0 months
Arm B: Standard salvage chemoimmunotherapy (R-DHAP, R-ICE, R-GDP, or R-GemOx for 2–3 cycles) followed by ASCT in responding patients
Primary endpoint: Event-free survival (EFS)
Median follow-up: 10.0 months
Results
EFS (primary): HR 1.07, 95% CI 0.82–1.40, P=.61 — not significant
12-month EFS: 28.4% vs 27.3% — nearly identical
CR rate at 3 months: 27.6% vs 27.7%
ORR: 46.3% vs 42.5%
mOS: Not reached in either arm — immature
12-month EFS: 28.4% vs 27.3% — nearly identical
CR rate at 3 months: 27.6% vs 27.7%
ORR: 46.3% vs 42.5%
mOS: Not reached in either arm — immature
Adverse events
CRS/ICANS (tisa-cel): CRS any grade 48.7%, grade ≥3 5.5%; ICANS any grade 8.6%, grade ≥3 3.3%
Hematologic: Grade ≥3 cytopenias 81.5% (tisa-cel) vs 65.4% (salvage)
Logistics: Bridging chemotherapy required in 83% of tisa-cel arm; median time from randomization to infusion 52 days
Hematologic: Grade ≥3 cytopenias 81.5% (tisa-cel) vs 65.4% (salvage)
Logistics: Bridging chemotherapy required in 83% of tisa-cel arm; median time from randomization to infusion 52 days
Conclusions
BELINDA was a major negative trial — tisa-cel did not improve EFS versus standard salvage + ASCT as second-line therapy in R/R DLBCL. This contrasted sharply with the simultaneously published ZUMA-7 (axi-cel) and TRANSFORM (liso-cel) trials, which both showed significant EFS improvements.
Key Limitations
Long manufacturing and infusion time (median 52 days) with high bridging chemotherapy use (83%) may have disadvantaged tisa-cel. Tisa-cel product characteristics (4-1BB costimulatory domain, less expansion than CD28-based products) may explain differential outcomes versus axi-cel (ZUMA-7). Salvage regimen heterogeneity in the control arm complicates cross-trial comparison. Early-relapsing vs primary-refractory subgroup differences not adequately addressed.
Clinical Context
BELINDA's negative result is a cautionary tale about CAR-T product heterogeneity — 4-1BB (tisa-cel) vs CD28 (axi-cel, liso-cel) constructs appear to have different efficacy in aggressive lymphoma. ZUMA-7 and TRANSFORM changed the 2L standard for transplant-eligible R/R DLBCL to CAR-T, specifically with axi-cel and liso-cel. FDA approved axi-cel (2022) and liso-cel (2022) for 2L DLBCL; tisa-cel is NOT approved in the 2L setting and remains approved for 3L+ DLBCL based on JULIET. ESMO and ASCO guidance reflect this distinction.