Background
Standard second-line therapy for relapsed LBCL was salvage chemoimmunotherapy and ASCT, with poor outcomes in early-relapsing disease. TRANSFORM evaluated lisocabtagene maraleucel (liso-cel), a CD19-directed CAR-T with defined CD4:CD8 ratio, versus standard salvage → ASCT in primary refractory or early-relapsed LBCL.
Interventions and follow up
Arm A (SOC): Salvage chemoimmunotherapy (R-DHAP, R-ICE, or R-GDP, 2-3 cycles) → ASCT if responsive
Arm B: Lisocabtagene maraleucel (liso-cel, anti-CD19 CAR-T, 100 × 10⁶ cells) single infusion after lymphodepletion (fludarabine + cyclophosphamide); bridging permitted
Population: N=184 (92/arm), LBCL refractory or relapsed ≤12 months after 1L
Primary endpoint: event-free survival (EFS)
Median follow-up: 17.5 months
Arm B: Lisocabtagene maraleucel (liso-cel, anti-CD19 CAR-T, 100 × 10⁶ cells) single infusion after lymphodepletion (fludarabine + cyclophosphamide); bridging permitted
Population: N=184 (92/arm), LBCL refractory or relapsed ≤12 months after 1L
Primary endpoint: event-free survival (EFS)
Median follow-up: 17.5 months
Results
Median EFS: 10.1 vs 2.3 months (HR 0.35, 95% CI 0.23–0.53, P<.0001)
Median PFS: 14.8 vs 5.7 months (HR 0.40, P<.0001)
Complete response: 66% vs 39%
Median OS: not reached vs 29.9 months (HR 0.53, P=.021); 43% of SOC patients received ASCT
Median PFS: 14.8 vs 5.7 months (HR 0.40, P<.0001)
Complete response: 66% vs 39%
Median OS: not reached vs 29.9 months (HR 0.53, P=.021); 43% of SOC patients received ASCT
Adverse events
CRS: any grade 49%, Grade ≥3 1%
ICANS/neurologic: any grade 30%, Grade ≥3 4%
Hematologic: Grade ≥3 cytopenias ~50%
Other: notably lower Grade ≥3 CRS and ICANS than axi-cel; treatment-related deaths 1 (liso-cel) vs 3 (SOC)
ICANS/neurologic: any grade 30%, Grade ≥3 4%
Hematologic: Grade ≥3 cytopenias ~50%
Other: notably lower Grade ≥3 CRS and ICANS than axi-cel; treatment-related deaths 1 (liso-cel) vs 3 (SOC)
Conclusions
Liso-cel demonstrated superior EFS, CR rate, and OS over salvage + ASCT in second-line LBCL, with a favorable toxicity profile (lower Grade ≥3 CRS/ICANS), providing a second CAR-T option for this setting.
Key Limitations
Small sample (N=184); shorter follow-up than ZUMA-7; only 43% of SOC patients received ASCT; no head-to-head with axi-cel; vein-to-vein time ~29 days requiring bridging therapy; subgroup analyses limited by sample size.
Clinical Context
FDA approved liso-cel for 2L LBCL in June 2022; EMA followed. With ZUMA-7, two CAR-T products have phase III support for 2L use. Liso-cel's lower neurotoxicity (Grade ≥3 ICANS 4% vs 21% for axi-cel) makes it attractive for higher-risk or resource-limited settings; ASCO/ESMO guidance endorses 2L CAR-T for early-relapse/refractory LBCL.