Tumor mutational burden predicts neoantigen profiles and immunotherapy response in microsatellite stable tumors across different cancer types

Olesia Kondrateva^1,2,3*Tugce Bilgin Sonay^1,2Inti Zlobec⁴Maria Anisimova^1,2

• ¹Zurich University of Applied Sciences, Winterthur, Switzerland
• ²Swiss Institute of Bioinformatics, Lausanne, Vaud, Switzerland
• ³University of Zurich, Zürich, Zürich, Switzerland
• ⁴University of Bern, Bern, Bern, Switzerland

Immunotherapy has shown positive response in many patients with microsatellite instable (MSI-H) tumors, but its effectiveness in microsatellite stable (MSS) tumors remains limited. We hypothesize that tumor mutational burden (TMB) could help identify patients with MSS tumors who can benefit from immunotherapy. We analyzed the molecular characteristics, including mutational landscape, mutational signatures, immune cell profiles and neoantigen load of MSS tumors with high TMB using data from colorectal cancer datasets (TCGA-COAD and TCGA-READ). After that, we extended these findings across other cancer types with MSI classification, further supporting the potential of using TMB as a biomarker for predicting immunotherapy response in MSS tumors. Our results show that MSS tumors with TMB greater than 50 mutations per megabase have POLE gene mutations, which lead to hypermutation. These hypermutated tumors have immune cell signatures that are more similar to MSI-H tumors, rather than MSS tumors with low TMB. We also found that MSS tumors with high TMB have a much higher number of neoantigens compared to low-TMB MSS tumors, suggesting they may respond better to immunotherapy.