e15170 Background: Tumor mutational burden (TMB) is an impressive predictive biomarker for immune checkpoint inhibitors therapy, and immunotherapy is one of the most promising methods for cancer treatment. However, the profile of TMB in various types of cancers was still poorly understood. Methods: In this study, genomic profiling of DNA was performed using next generation sequencing from a 539 genes panel in tumor tissues. The TMB was defined as the numbers of SNVs including synonymous and nonsynonymous mutations, and InDels per megabase in coding regions of sequenced genome. TMB-H was defined as highest mutation load quintile (top 20%) in each cancer type. The values of TMB of 874 patients were compared among eight main tumor types including 174 patients with liver cancer, 32 patients with bile duct cancer, 54 patients with gastric cancer, 119 patients with colorectal cancer, 27 patients with pancreatic cancer, 32 patients with melanoma, 25 patients with glioma and 411 patients with lung cancer cases by multiple independent samples nonparametric Kruskal-Wallis H test via SPSS v22.0. Results: The median values of TMB in liver cancer, bile duct cancer, gastric cancer, colorectal cancer, pancreatic cancer, melanoma, glioma and lung cancer were 10.29, 8.09, 11.03, 10.29, 5.88, 7.35, 5.88 and 7.35 Muts/Mb, respectively. The cut-off values of TMB-H in corresponding eight solid tumors were 16.18, 19.12, 19.85, 16.18, 10.29, 13.24, 10.29 and 18.38 Muts/Mb, respectively. Based on nonparametric Kruskal-Wallis H test, there were significant differences of median TMB values across eight independent tumor types (χ2 = 26.752, p = 3.693×10−4). Significant differences between patients with colorectal cancer and pancreatic cancer ( p = 0.005) or lung cancer ( p = 0.011) were observed via pairwise comparisons. Conclusions: Our study proved the presence of significant differences across eight cancer types, deepened the knowledge of the cancer-specific of TMB, provided useful information in immunology therapy for East Asian patients with solid tumors. As a retrospective study with a relatively small population, the conclusions of this study needed to be verified with a larger sample.