ABCL-289: Matching-Adjusted Indirect Comparison (MAIC) of Axicabtagene Ciloleucel (Axi-Cel) and Lisocabtagene Maraleucel (Liso-Cel) in Relapsed or Refractory (R/R) Large B-Cell Lymphoma (LBCL) After Two or More Prior Lines of Therapy

Context Axi-cel and liso-cel are autologous, anti-CD19, chimeric antigen receptor T-cell therapies for the treatment of R/R LBCL. Both have demonstrated high and durable responses in single-arm trials, but comparative evidence is lacking. Objective To estimate relative treatment effects with axi-cel versus liso-cel for the treatment of R/R LBCL. Methods A logistic propensity score model was used to estimate weights for the individual patient data in ZUMA-1 (axi-cel) to match the mean baseline characteristics observed in TRANSCEND-NHL-001 (liso-cel). The following variables were matched: ECOG score, best response to last therapy, use of bridging therapy, LBCL subtype, number of prior therapies, prior autologous stem cell transplant, tumor burden, age, and lactate dehydrogenase. Relative treatment effects (with 95% confidence interval [CI]) were estimated using outcomes from TRANSCEND-NHL-001 and the weighted population in ZUMA-1. Outcomes included response rate, duration of response (DoR), progression-free survival (PFS), overall survival (OS), cytokine release syndrome (CRS), and neurological events (NEs). Base–case analyses were conducted using ZUMA-1 pivotal cohorts 1 and 2 (N=101) and published outcomes for liso-cel. Scenario analyses included ZUMA-1 additional safety management cohorts 4 (N=41) and 6 (N=40). Results In the base–case analysis, the estimated odds ratios (ORs) for objective response (2.18 [0.96–4.98]) and complete response (1.84 [0.97–3.50]) were higher with axi-cel, but these differences were not statistically significant. Axi-cel was associated with a statistically significant improvement in OS (hazard ratio [HR]: 0.53 [0.34–0.83]) and PFS (HR: 0.61 [0.40–0.93]). DoR results were in favor of axi-cel, although not statistically significant (HR: 0.63 [0.37–1.06]). Compared to liso-cel, axi-cel was associated with a higher rate of grade ≥3 CRS (OR: 3.64 [1.04–12.76]) and grade ≥3 NEs (OR: 3.45 [1.65–7.19]), but differences were smaller in scenarios including the expanded safety management cohorts (OR: 1.71 [0.47–6.19] and 2.90 [1.54–5.47], respectively). Differences in OS and grade ≥3 CRS did not reach statistical significance in the scenario analysis. Conclusions Our findings suggest that axi-cel safety outcomes have improved over time, but grade ≥3 NEs remained higher than liso-cel. Although we found no difference in response rates, axi-cel showed higher OS and PFS. Axi-cel and liso-cel are autologous, anti-CD19, chimeric antigen receptor T-cell therapies for the treatment of R/R LBCL. Both have demonstrated high and durable responses in single-arm trials, but comparative evidence is lacking. To estimate relative treatment effects with axi-cel versus liso-cel for the treatment of R/R LBCL. A logistic propensity score model was used to estimate weights for the individual patient data in ZUMA-1 (axi-cel) to match the mean baseline characteristics observed in TRANSCEND-NHL-001 (liso-cel). The following variables were matched: ECOG score, best response to last therapy, use of bridging therapy, LBCL subtype, number of prior therapies, prior autologous stem cell transplant, tumor burden, age, and lactate dehydrogenase. Relative treatment effects (with 95% confidence interval [CI]) were estimated using outcomes from TRANSCEND-NHL-001 and the weighted population in ZUMA-1. Outcomes included response rate, duration of response (DoR), progression-free survival (PFS), overall survival (OS), cytokine release syndrome (CRS), and neurological events (NEs). Base–case analyses were conducted using ZUMA-1 pivotal cohorts 1 and 2 (N=101) and published outcomes for liso-cel. Scenario analyses included ZUMA-1 additional safety management cohorts 4 (N=41) and 6 (N=40). In the base–case analysis, the estimated odds ratios (ORs) for objective response (2.18 [0.96–4.98]) and complete response (1.84 [0.97–3.50]) were higher with axi-cel, but these differences were not statistically significant. Axi-cel was associated with a statistically significant improvement in OS (hazard ratio [HR]: 0.53 [0.34–0.83]) and PFS (HR: 0.61 [0.40–0.93]). DoR results were in favor of axi-cel, although not statistically significant (HR: 0.63 [0.37–1.06]). Compared to liso-cel, axi-cel was associated with a higher rate of grade ≥3 CRS (OR: 3.64 [1.04–12.76]) and grade ≥3 NEs (OR: 3.45 [1.65–7.19]), but differences were smaller in scenarios including the expanded safety management cohorts (OR: 1.71 [0.47–6.19] and 2.90 [1.54–5.47], respectively). Differences in OS and grade ≥3 CRS did not reach statistical significance in the scenario analysis. Our findings suggest that axi-cel safety outcomes have improved over time, but grade ≥3 NEs remained higher than liso-cel. Although we found no difference in response rates, axi-cel showed higher OS and PFS.