Abstract:2589 Background: Tumor mutation burden (TMB), calculated by whole-exome sequencing (WES) or large NGS panels, has an important association with immunotherapy responses. Elucidating the underlying biological mechanisms of high TMB might help develop more precise and effective means for TMB and immunotherapy response prediction. Meanwhile, the landscape of TMB across different cancer types and its association with other molecular features have not been well investigated in large cohorts in China. Methods: Cancer patients whose fresh tissue (n = 1556), formalin-fixed, paraffin-embed (FFPE) specimen (n = 1794), and pleural fluid (n = 84) were profiled using 295- or 520-gene NGS panel. The association of the TMB status with a series of molecular features and biological pathways was interrogated using bootstrapping. Results: TMB, measured by 295- or 520-cancer-related gene panels, were correlated with WES TMB based on in silico simulation in the TCGA cohort. We compared the TMB landscape across 11 cancer type groups and found the highest average TMB in lung squamous cell carcinoma, whereas the lowest TMB was established in sarcoma. High microsatellite instability, DNA damage response deficiency, and homologous recombination repair deficiency indicated significantly higher TMB. The independent predictive power for TMB of twenty-six biological pathways was tested in 10 cancer groups. FoxO signaling pathway most commonly correlated with low-TMB; significant association was identified in four cancer groups. In contrast, no pathway was significantly correlated with high-TMB in more than two cancer groups. Overall, we discovered that the underlying pathways which may be the main drivers of TMB status varied greatly and sometimes had an opposite association with TMB across different cancer types. Moreover, we developed a 14- and 22-gene signature for TMB prediction for LUAD and LUSC, respectively, with only 10 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.

Predicting Tumor Mutational Burden From Lung Adenocarcinoma Histopathological Images Using Deep Learning

LUADpp: an Effective Prediction Model on Prognosis of Lung Adenocarcinomas Based on Somatic Mutational Features

Prediction of Tumor Mutation Load in Colorectal Cancer Histopathological Images Based on Deep Learning

Predicting gastric cancer tumor mutational burden from histopathological images using multimodal deep learning

Predicting Tumor Mutational Burden from Liver Cancer Pathological Images Using Convolutional Neural Network.

[Multi-classification prediction model of lung cancer tumor mutation burden based on residual network]

Deep learning-based approach to reveal tumor mutational burden status from whole slide images across multiple cancer types

Decoding tumor mutation burden and driver mutations in early stage lung adenocarcinoma using CT-based radiomics signature

Histopathological Images and Multi-Omics Integration Predict Molecular Characteristics and Survival in Lung Adenocarcinoma

Prediction of Target-Drug Therapy by Identifying Gene Mutations in Lung Cancer With Histopathological Stained Image and Deep Learning Techniques

Identification and Validation of Efficacy of Immunological Therapy for Lung Cancer From Histopathological Images Based on Deep Learning

Tumor Mutation Burden Analysis in a 5,660 Cancer Patient Cohort Reveals Cancer Type-Specific Mechanisms for High Mutation Burden.

Classification and mutation prediction from non–small cell lung cancer histopathology images using deep learning

Computational Staining of Pathology Images to Study the Tumor Microenvironment in Lung Cancer

Pan-cancer Computational Histopathology Reveals Tumor Mutational Burden Status Through Weakly-Supervised Deep Learning.

Development of Prognostic Biomarkers by TMB-Guided WSI Analysis: A Two-Step Approach

Predicting EGFR Mutation Status in Lung Adenocarcinoma on CT Image Using Deep Learning.

A novel transformer-based aggregation model for predicting gene mutations in lung adenocarcinoma

Predicting EGFR mutation status in lung adenocarcinoma on computed tomography image using deep learning

Tumor Mutation Burden–Related Histopathologic Features for Predicting Overall Survival in Gliomas Using Graph Deep Learning