Impact of TP53 mutation on CD19-chimeric antigen receptor T-cell therapy in large B-cell lymphoma

Abstract

Abstract Introduction: Although CD19-targeted chimeric antigen receptor T-cell (CD19-CAR-T) therapy is highly effective in relapsed or refractory large B-cell lymphoma (R/R LBCL), over half of patients ultimately experienced disease progression (NEJM 2018:380:45-56; Lancet Oncol 2019:20:31-42; Lancet 2020:396:839-852). Thus, identification of patients who benefit most from CD19-CAR-T therapy remains a significant challenge. TP53 genomic alteration has been established as the prognostic biomarker in LBCL (Blood 2012:120:3986-3996); thus, this study hypothesized that TP53 mutation could also predict treatment outcomes in R/R LBCL patients receiving CD19-CAR-T therapy. Methods: This retrospective study included adult R/R LBCL patients treated with CD19-CAR-T therapy from Ruijin Hospital (China) between January 2018 and September 2024. Genomic mutation was detected by next-generation sequencing (NGS) with a fixed panel covering coding sequence (CDS) region of 152 high pathogenic genes of LBCL in tumor tissues. Outcomes included progression-free survival (PFS), overall survival (OS), and safety. To investigate the role of TP53 mutation, efficacy and safety data reported in this study were stratified by TP53 mutation status. Results: A total of 139 patients were included, among whom 52 had TP53-mutated LBCL (TP53mut group) and 87 had TP53-wild type LBCL (TP53wt group). Baseline characteristics were generally balanced between the TP53mut and TP53wt groups. In the TP53mut group, the proportion of patients treated by either CD28- or 41BB-costimulatory domain CAR-T products was 63.5% and 36.5%; and in the TP53wt group was 51.7% and 48.3%, respectively. There were comparable proportions of male patients (53.8% vs. 59.8%) and patients aged over 60 (48.1% vs. 47.1%) between the groups. Diffuse LBCL was the most common pathological subtype (98.1% vs. 94.3%). In the TP53mut group, 50.0% of patients exhibited coexpression of MYC/BCL2 (double expressor) and 9.6% were double-hit/triple-hit lymphoma; same indicators in the TP53wt group were 39.1% and 8.5%. Before the infusion, 36 (69.2%) patients in the TP53mut group and 56 (64.4%) in the TP53wt group had stage III-IV disease. 26.9% of patients with TP53mut and 33.3% with TP53wt were heavily pretreated (≥3 lines); 39 (75.0%) and 62 (71.3%) in the respective groups had refractory diseases. Eight (9.2%) patients with TP53wt had received prior autologous stem cell transplantation, while none with TP53mut had. With a median follow-up of 25.1 months by the data cutoff (July 1, 2025), the median PFS in the TP53mut group was 41.1 months (95% confidence interval=12.7-69.5) and in the TP53wt group was not reached (p=0.36). The median OS has not yet been reached in these two groups (p=0.78). The choice of CAR-T product was notable, thus, we also assessed the clinical outcomes stratified by TP53 status, according to the type of CAR-T product categorized by the costimulatory domains. TP53 mutation status did not affect median PFS (not reached [NR] vs. NR; 2-year rate, 64.7% vs. 61.2%) or median OS (NR vs. NR; 2-year rate 76.5% vs. 79.4%) for patients treated with CD28 CAR-T products. In patients receiving 4-1BB products, TP53 mutation was associated with poorer median PFS (6.3 months vs. NR, p=0.024; 2-year rate 35.5% vs. 60.5%) but not with median OS (30.8 months vs. NR; 2-year rate 59.3% vs. 66.5%). Grade ≥2 cytokine release syndrome occurred in 38.5% of patients with TP53mut and 47.1% with TP53wt (p=0.32). TP53 genomic alterations were also not correlated with grade ≥2 immune effector cell-associated neurotoxicity syndrome (3.8% vs. 3.4%; p=0.90). Conclusion: These findings suggested that TP53 mutation status did not significantly impact the efficacy or safety of CD19-CAR-T therapy in the overall population. Further investigation in prospective studies is warranted.