LAUSR

Abstract Background In this study, we examined the TMB landscape of 5,660 pan-cancer cases in Chinese population, using NGS panel. We established cancer-specific and histology-specific biological pathways associated with the TMB status. In addition, as a proof on concept, an unsupervised algorithm was conducted using stepwise logistic regression to generate TMB-predicting signatures from both lung adenocarcinoma and lung squamous cell carcinoma. Methods Patients: 5,660 Chinese cancer patients across 11 cancer types Panel: Targeted sequencing was performed on tissue samples using a panel consisting of 295 or 520 cancer-related genes, spanning 1.4 and 1.6Mb of human genome, respectively. An average sequencing depth of 1,000X and 10,000x were achieved for tissue and plasma samples, respectively. Tumor mutation burden: calculated as the ratio of mutation count to the size of coding region of the panel, excluding CNV, fusions, large genomic rearrangements and mutations occurring on the kinase domain of EGFR and ALK. Results Across the 11 cancer types included in the analysis, lung squamous cell carcinoma had the highest average TMB, whereas sarcoma has the lowest TMB. High microsatellite instability, DNA damage response deficiency, and homologous recombination repair deficiency indicated significantly higher TMB.The independent predictive power for TMB 26 biological pathways was tested in 11 cancer types. Mismatch repair, DNA damage response, homologous recombination repair, and PI3K-AKT signaling pathway were most commonly correlated with high-TMB. In contrast, ERBB signaling pathway, and adhesion-related pathways were most commonly correlated with low-TMB. Moreover, we developed a 23- and 16-gene signature for TMB prediction for LUAD and LUSC, respectively, with 12 genes shared by both signatures, indicating a histology- specific mechanism for driving high- TMB in lung cancer. Conclusions The findings extended the knowledge of the underlying biological mechanisms for high TMB and might be helpful for developing more precise and accessible TMB assessment panels and algorithms in more cancer types. Legal entity responsible for the study The authors. Funding Burning Rock Biotech. Disclosure All authors have declared no conflicts of interest.