RESUMO
PURPOSE: The purpose of this study was to understand the impact of population diversity and geographic variation on tumor mutation burden (TMB) scores across cancers and its implication on stratification of patients for immune checkpoint inhibitor (ICI) therapy. MATERIALS AND METHODS: This retrospective study used whole-exome sequencing (WES) to profile 1,233 Indian patients with cancer across 30 different cancer types and to estimate their TMB scores. A WES-based pipeline was adopted, along with an indigenously developed strategy for arriving at true somatic mutations. A robust unsupervised machine learning approach was used to understand the distribution of TMB scores across different populations and within the population. RESULTS: The results of the study showed a biphasic distribution of TMB scores in most cancers, with different threshold scores across cancer types. Patients with cancer in India had higher TMB scores compared with the Caucasian patients. We also observed that the TMB score value at 90th percentile (predicting high efficacy to ICI) was high in four different cancer types (sarcoma, ovary, head and neck, and breast) in the Indian cohort as compared with The Cancer Genome Atlas or public cohort. However, in lung and colorectal cancers, the TMB score distribution was similar between the two population cohorts. CONCLUSION: The findings of this study indicate that it is crucial to benchmark both cancer-specific and population-specific TMB distributions to establish a TMB threshold for each cancer in various populations. Additional prospective studies on much larger population across different cancers are warranted to validate this observation to become the standard of care.