Comparison of axi-cel versus liso-cel as 2nd line therapy for relapsed/refractory large B-cell lymphoma in real-life: A lysa study from the descar-T registry

Abstract Background Axicabtagene ciloleucel (axi-cel) and Lisocabtagene maraleucel (liso-cel), two anti-CD19 CAR T-cells, have been approved as 2nd line (2L) treatment in relapsed/refractory (R/R) large B-cell lymphoma (LBCL) for patients progressing within 12 months of first-line chemo-immunotherapy, based on the ZUMA-7 and TRANSFORM trials, respectively. In the ZUMA 7 trial, axi-cel demonstrated a median event-free survival (EFS) of 10.8 months and an estimated 4-year overall survival (OS) of 54.6% (Locke NEJM 2022, Westin NEJM 2023). In the TRANSFORM trial, liso-cel was associated with a median EFS of 29.5 months with a 3-year OS rate of 62.8% (Abramson Blood 2023; Kamdar ASCO 2024). Given the absence of a randomized comparison between axi-cel and liso-cel as 2L treatment for R/R LBCL, we conducted a comparative analysis of patients treated in France with these two products since their approval. Methods All data were collected through the French DESCAR-T registry designed by the LYSA/LYSARC to collect real-life data. Inclusion criteria were diagnosis of LBCL, 1 prior line of treatment, at least 6 months of follow up after infusion (Leukapheresis performed before September 2024). We conducted a comparison of axi-cel and liso-cel in a large cohort of 2L R/R LBCL patients treated in France using a propensity score (PS). Multiple imputations were used to handle missing variables. We used the stabilized inverse probability weighting (sIPTW) method to balance axi-cel and liso-cel populations. The PS was calculated for each patient with a logistic regression using the following variables: age, sex, refractory status at 1st line completion, time between 1st treatment end and leukapheresis; at leukapheresis, performance status (ECOG PS), Ann Arbor stage, center experience (time from 1st center leukapheresis and leukapheresis of patient), LBCL histology, LDH level; and at infusion, vein-to-vein time, response to bridging therapy, ECOG PS, LDH, CRP. Endpoints were best ORR and CRR (Lugano 2014), progression-free survival (PFS), OS, and toxicities (CRS, ICANS, intensive care unit transfer). OS and PFS were evaluated from leukapheresis. Results An initial cohort of 805 patients with LBCL (LBCL NOS n=663, PMBL n=38, transformed indolent NHL n=104) treated with axi-cel (n=663) or liso-cel (n=142) in 2L was analyzed. Most were male (63%) with refractory disease (73%). Characteristics were imbalanced regarding age (median 68y liso-cel vs 62y axi-cel), LBCL histology (PMBL 12% liso-cel vs 3% axi-cel), Ann Arbor stage (stage III/IV 69% liso-cel vs 76% axi-cel), center experience (more in liso-cel arm), and time from last treatment to infusion, longer with liso-cel (median 110 days liso-cel vs 90 days axi-cel). After sIPTW using axi-cel pts as reference, outcome was compared in a pseudo-population with balanced confounders: 647 intended for axi-cel and 155 for liso-cel, corresponding to 600 infused with axi-cel, 127 with liso-cel. No significant residual differences in baseline characteristics were observed (standard mean differences <0.1). Median follow-up from leukapheresis was 12.9 months for axi-cel, 7.7 months for liso-cel. OS and PFS were not significantly different between products in the intention to treat population, nor when analyzed from infusion. Regarding best CRR, no significant differences were observed (62.6% axi-cel vs 65.8% liso-cel, P=0.5). Best ORR was higher with liso-cel (79.8% axi-cel vs 90% liso-cel, P=0.008). Concerning toxicity, grade 3+ CRS was infrequent in both groups (5.1% axi-cel vs 3.1% liso-cel, P=0.346), but CRS of any grade was more frequent with axi-cel (93.9% vs 50.8%, P<0.001). Neurotoxicity and severe neurotoxicity were higher with axi-cel (grade 3+ ICANS 17.6% vs 0.2%, P<0.001). ICU transfers were more frequent after axi-cel (22.4% vs 5.3%, P<0.001). Conclusion After PS-weighting in a large French population with CAR T-cell therapy for 2L LBCL, there was no significant OS or PFS difference between axi-cel and liso-cel. However, axi-cel had a greater incidence of severe neurotoxicity. These results require confirmation in a larger liso-cel cohort with longer follow-up. Baseline differences reveal selection bias regarding CAR-T choice; liso-cel being more often proposed to older, frailer patients, able to wait longer for infusion. Axi-cel may be more effective in a subgroup with rapidly progressing, primary-refractory disease; further analyses are needed to define the optimal place of each product.