Genetic Determinants of CAR T-Cell Therapy Outcomes: A Comparative Analysis of DLBCL and Glioblastoma

Abstract

CAR T-cell therapy has become a revolutionary option for patients with certain cancers, but its effectiveness remains uneven across cancer types. We are particularly interested in why CAR T treatment produces stronger and more durable responses in diffuse large B-cell lymphoma (DLBCL), compared to glioblastoma (GBM). We hypothesised that differences in the genetic mutations commonly found in each cancer may underlie these differences in treatment response. To investigate this, we are using cBioPortal, an open-access cancer genomics database, to analyze the mutation and clinical data of patients with DLBCL (n=2143) and GBM (n=881), respectively. Because patient-level genomic data from CAR T clinical trials is not directly available, we created a matching process to align real-world patient data from cBioPortal with the demographic and clinical characteristics reported in trial populations. Using this matched dataset, we performed a survival analysis to examine how specific genetic mutations may be associated with patient outcomes after CAR T therapy. Our findings suggest that while DLBCL patients with mutations in genes like PIM1 and IGHV2-70 had poorer survival, GBM patients with EGFR and PIK3R1 mutations faced even worse prognoses. These patterns, combined with factors like heterogeneity in cell-surface antigens and tumor location, help explain why CAR T therapy remains far more effective in blood cancers than in solid tumors like GBM. By identifying genetic profiles associated with survival outcomes, our study may inform future iterations of CAR T design and provide insights to predict the efficacy of CAR T therapy, especially for solid tumors.