Impact of putative cell of origin on CD19+CAR-T outcomes in patients with large B-cell lymphoma

Abstract

Abstract Background CAR-T therapy has demonstrated improved survival in patients with relapsed, refractory large B cell lymphoma (R/R LBCL). While the putative cell of origin (COO) (germinal center B-cell [GCB] vs non-GCB) impacts prognosis and treatment responses in LBCL as well as with autoHCT (Iqbal et al. CLML 2022), the impact of COO on CAR-T outcomes remains elusive. Additionally, while survival following axi-cel therapy (CAR-T in second line) was prolonged compared to SOC for both COO subtypes, the studies were not designed to detect differences in outcomes based on COO (Locke et al. NEJM 2022). Methods In this retrospective analysis, we examined the impact of COO in patients with LBCL receiving commercial CD19+CAR-T between April 2019 and October 2024, stratified by lines of therapy (LOT; 2L vs 3L+). Axicabtagene ciloleucel (axi-cel) and lisocabtagene maraleucel (liso-cel) were included in 2L and axi-cel, liso-cel and tisagenlecleucel (tisa-cel) in third-line and beyond (3L+) setting. COO stratification was based on Hans algorithm (Hanset al. Blood 2004). Response to CAR-T was assessed per Lugano PET/CT criteria. Progression-free survival (PFS) was the primary endpoint. Cox proportional hazards models were used for multivariable analysis. Results Of 344 patients, 59% (n=203) patients had GCB and 41% (n=141) had non-GCB subtype. Overall, 63.1% (n=217) patients received bridging and 99.4% (n=342) FluCy-based lymphodepletion. Axi-cel accounted for 79.1% (n=272) CAR-T, liso-cel 17.2% (n=59), and tisa-cel 3.8% (n=13); and 75% (n=257) patients received CAR-T in 3L+ vs 25% (n=87) in 2L. Median interval between diagnosis and CAR-T was 16.6 months, 48% (n=165) had HCT-CI>3, 59.9% (n=172) had elevated LDH, 9.5% (n=25) bulky disease (>10cm), 15.7% (n=54) ECOG>2, and 45.5% (n=154) IPI>3. Compared to 3L+, CAR-T recipients in 2L had higher HCT-CI (p=0.035), median platelet count (p=0.0007), hemoglobin (Hb) (p=0.0025), and albumin at apheresis (p=0.012). There were no significant differences in response rates (best response; response rates at d+30, d+100, 6 months, 1 year) between 2 COO subtypes or by LOT. Rates of CRS, ICANS (all-grade and grade>3) and NRM were not significantly different between 2 COO subtypes or by LOT, except all-grade CRS, which was greater with 2L vs 3L+ (93.1% vs 80.5%; p=0.007). Of 132 patients who died, primary cause of death was disease progression in 69.5% patients, infections (11.5%), secondary malignancy (6.1%), ICANS (3.1%), and CRS (0.8%). Median PFS in GCB cohort was 13.8 months (95%CI, 8.28-28.6) compared to 6.7 months (95%CI, 6.01-17.1) in non-GCB cohort. In univariate analysis, there was a trend towards superior PFS in GCB arm vs non-GCB arm at 1, 2, and 3 years (p=0.066). In multivariate analysis adjusted for LOT, HCT-CI, prior overall response rate (ORR), and ECOG, PFS was significantly superior in GCB arm vs non-GCB arm (HR, 1.40; 95%CI, 1.04-1.89; p=0.028). Median OS in GCB arm was 24.1 months (95%CI, 18.6-34.7) compared to 18.3 (95%CI, 12.75-25.23) in non-GCB arm. In univariate analysis, there was a trend towards superior OS in GCB vs non-GCB arm at 1, 2, and 3 years (p=0.056). In multivariate analysis adjusted for LOT, prior ORR, and ECOG, OS was significantly superior in GCB arm compared to non-GCB (HR, 1.46; 95%CI, 1.03-2.07; p=0.032). While there was no significant difference in median PFS (p=0.49) and OS (p=0.82) between 2 COO arms with CAR-T in 2L, median PFS (13.5 vs 6.2 months; p=0.0176) and OS (28.6 vs 14 months; p=0.028) were significantly superior in the GCB arm with CAR-T in 3L+ compared to non-GCB arm. Similarly, there were no significant differences in PFS or OS at 1, 2, and 3 years between the 2 COO subtypes in 2L, whereas these were superior in GCB arm, compared to non-GCB, when CAR-T was given in 3L+. In multivariate analysis, CAR-T in 2L vs 3L+ was associated with superior PFS (HR, 1.65, 95%CI, 1.11-2.46; p=0.013); however, there was no significant difference in OS (HR, 1.23, 95%CI, 0.78-1.93; p=0.366). Conclusion These results demonstrate that GCB subtype has a superior PFS (HR 1.4) and OS (HR 1.5) compared to non-GCB from CD19+CAR-T. Additionally, GCB subtype derive a significant survival benefit when CAR-T is given in 3L+ setting. Consistent with existing literature, CAR-T in earlier lines (2L vs 3L+) confers a PFS advantage (HR 1.65). Further studies are needed to better define the impact of COO on CAR-T outcomes and elucidate CAR-T resistance mechanisms in non-GCB subtype LBCL.