Construction of methylation driver gene-related prognostic signature and development of a new prognostic stratification strategy in neuroblastoma

Background Aberrant DNA methylation is one of the major epigenetic alterations in neuroblastoma. Objective Exploring the prognostic significance of methylation driver genes in neuroblastoma could help to comprehensively assess patient prognosis. Methods After identifying methylation driver genes (MDGs), we used the LASSO algorithm and stepwise Cox regression to construct methylation driver gene-related risk score (MDGRS), and evaluated its predictive performance by multiple methods. By combining risk grouping and MDGRS grouping, we developed a new prognostic stratification strategy and explored the intrinsic differences between the different groupings. Results We identified 44 stably expressed MDGs in neuroblastoma. MDGRS showed superior predictive performance in both internal and external cohorts and was strongly correlated with immune-related scores. MDGRS can be an independent prognostic factor for neuroblastoma, and we constructed the nomogram to facilitate clinical application. Based on the new prognostic stratification strategy, we divided the patients into three groups and found significant differences in overall prognosis, clinical characteristics, and immune infiltration between the different subgroups. Conclusion MDGRS was an accurate and promising tool to facilitate comprehensive pre-treatment assessment. And the new prognostic stratification strategy could be helpful for clinical decision making.