Frontiers in Oncology (Oct 2020)
Integration of the Tumor Mutational Burden and Tumor Heterogeneity Identify an Immunological Subtype of Melanoma With Favorable Survival
Abstract
The tumor mutational burden (TMB) has been reported as a predictive marker of the response to immune checkpoint inhibition (ICI) therapy in previous melanoma clinical trials. However, the TMB alone is not sufficient to accurately predict immunotherapy benefit. Additional biomarkers are needed for better stratification of immunotherapy-sensitive patients. In the present study, mutation data and survival information of patients with melanoma were collected from several immunotherapy studies, and tumor heterogeneity was estimated using mutant-allele tumor heterogeneity (MATH). The benefit score was defined as the ratio between the TMB and tumor heterogeneity, and optimal critical values were selected to group patients and evaluate their response to ICI treatment. The benefit score significantly improved the performance of stratifying the overall survival of patients compared with the TMB alone as a predictor in two independent cohorts (p = 0.0068 vs. p = 0.1 and p = 0.045 vs. p = 0.13), in which patients were treated with Ipilimumab and Nivolumab, respectively. In another cohort of patients with melanoma receiving mixed ICI treatment, the benefit score was also positively associated with higher overall survival (p = 0.022) and outperformed the TMB alone, with a significance of p = 0.089. The benefit score showed a positive correlation with clonal TMB, a reported immunotherapy marker, and exceeded it in immunotherapy response prediction. Besides, a high benefit score was found to be associated with higher proportions of natural killer cells, lower proportions of M2 macrophages and elevated CD8 T cells, all of which favor ICI therapy. In summary, tumor heterogeneity combined with the TMB showed superior efficacy in predicting the response to ICI therapy. This might further help to delineate the mechanisms of immunotherapy in patients with melanoma.
Keywords