Genetic Makeup Predicts Cancer Patients’ Benefits and Side Effects from Immunotherapy

Viral CAR-T cell therapy has transformed treatment options for aggressive blood cancers, yet its benefits and risks vary widely among patients. A new study suggests that part of this variability is not random: it may be rooted in differences in people’s inherited genomes. The work, published in Science Immunology, links specific genetic variants to both therapeutic performance and treatment-related toxicity.

CAR-T cells are patient-specific immune cells engineered to recognize cancer targets. Because every CAR-T product is manufactured from donor or patient material and is therefore biologically distinct, the immune response they generate is strongly shaped by the starting cells. This study asks whether human genetic variation can predict how CAR-T cells will behave once infused.

Researchers analyzed whole-genome sequences from more than 200 patients with aggressive lymphoma drawn from two clinical trials of CAR-T cell therapy. By integrating genome-wide data with clinical outcomes, they searched for genetic “correlates” that track with efficacy and adverse inflammatory effects.

In one trial, variants that silence the STXBP2 gene in T cells were associated with increased toxicity. Consistently, engineered donor T cells lacking STXBP2 and/or expressing these silencing variants triggered inflammation in experimental settings, supporting a mechanistic connection between the gene and safety.

In both trials, variants in ADAMTSL3 correlated with lower risk of treatment-related toxicity, suggesting a protective genomic influence. Meanwhile, variants in PTPN22 were strongly associated with enhanced CAR-T expansion, a key biological marker that often predicts stronger antitumor activity.

Together, these findings indicate that multiple genes can shape distinct aspects of CAR-T performance—some affecting inflammatory safety pathways and others influencing proliferation and persistence of engineered immune cells. Rather than a single culprit, the results point to a network of inherited immune regulators.

The authors emphasize that this knowledge could change how donor T cells are selected in scalable CAR-T strategies, where one donor may supply material for many patients. It may also guide the design of next-generation engineered CAR-T cells with improved risk–benefit profiles.

Beyond CAR-T, the study implies that similar genomic logic could apply to other adoptive immune-cell therapies. As CAR-T products become more sophisticated, precision approaches that incorporate germline genetics may become a practical route to safer, more effective treatment.