Pan-cancer genomic analyses reveal prognostic and immunogenic features of the tumor melatonergic microenvironment across 14 solid cancer types

We performed comprehensive genomic analyses of the melatonergic system within the tumor microenvironment and their clinical relevance across a broad spectrum of solid tumors. RNA-seq data from The Cancer Genome Atlas (TCGA) of 14 solid tumors representing 6658 human samples were analyzed. The tumor melatonergic system was characterized by the rates of melatonin synthesis and metabolism using a two-gene expression model (melatonin synthesis/metabolism Index). We calculated three indexes according to different melatonin metabolism isoenzymes (Index-I [ASMT:CYP1A1], Index-II [ASMT:CYP1A2], and Index-III [ASMT:CYP1B1]). Samples of each cancer type were classified into two subgroups (high vs low) based on median values. Clinical outcomes, mutational burden, and neoepitope abundance were analyzed and compared. We found that the ability of the tumor microenvironment to synthesize and accumulate melatonin varied across cancer types and negatively correlated with tumor burden. Kaplan-Meier survival analyses and multivariable modeling showed that the three indexes played different roles across different cancers and harbored prognostic values in breast cancer (adjusted hazard ratio [AHR] _Index-II  = 0.65 [0.44-0.97]; P = 0.03), cervical cancer (AHR _Index-I  = 0.62 [0.39-0.98]; P = 0.04), lung squamous cell carcinoma (AHR _Index-III  = 0.75 [0.56-0.99]; P = 0.04), melanoma (AHR _Index-I  = 0.74 [0.55-0.98]; P = 0.04), and stomach adenocarcinoma (AHR _Index-III  = 0.68 [0.41-0.94]; P = 0.02). We further investigated its clinical relevance with tumor immunogenic features (mutational burden and neoantigen abundance), which may predict immunotherapy benefits. We observed significant negative correlations with mutational burden in the majority of tumors (P < 0.05), except cervical cancer, pancreatic adenocarcinoma, and thyroid carcinoma. Our study provides a systematic overview of the oncostatic values of the melatonergic system and highlights the utilization of this simple and promising gene signature as a prognosticator and potential predictor of response to immunotherapy.