Development and Validation of an m6A-Derived Prognostic Signature in Lung Adenocarcinoma

Lung adenocarcinoma (LUAD) exhibits extensive molecular heterogeneity and poor prognosis, necessitating novel epigenetic biomarkers. N6-methyladenosine (m6A) modification, a pivotal epigenetic regulator of RNA, is frequently dysregulated in LUAD, yet its systematic roles in clinical prognosis and the tumor microenvironment (TME) remain poorly elucidated.

Transcriptomic profiles and clinical data from large-scale cohorts were integrated and analyzed. Unsupervised consensus clustering based on 47 m6A-related genes (MRGs) was performed to distinguish distinct molecular subtypes. A prognostic model was developed via LASSO-Cox regression algorithm and further validated in multiple independent cohorts. Immune infiltration, tumor mutational burden (TMB), immunotherapy response (TIDE/IPS), and chemotherapy sensitivity were analyzed, complemented by single-cell RNA sequencing.

Two distinct m6A-related gene clusters (mRGclusters A and B) were identified. Patients in cluster B exhibited inferior survival outcomes, higher m6A pathway activity, and significant enrichment of cell cycle-related pathways. The eight-gene signature successfully stratified patients into high-risk (HR) and low-risk (LR) subgroups. Patients in the HR group presented significantly worse overall survival (P < 0.05), higher TMB levels, and upregulated expression of multiple immune checkpoint genes such as LAG3 and PDCD1. CSMD3 mutations were capable of improving the survival of HR patients by facilitating the infiltration of natural killer cells and follicular helper T cells. The signature independently predicted prognosis (AUC: 0.70-0.84) and treatment response: LR patients favored immunotherapy (lower TIDE, higher IPS), while HR patients were sensitive to chemotherapy (e.g., Bosutinib, Tozasertib).

This transcriptome-derived m6A-associated prognostic model can effectively predict clinical survival outcomes and therapeutic response in LUAD patients. Combined with immune landscape, genomic mutation profiles and single-cell transcriptomic evidence, this signature provides a reliable basis for personalized risk stratification and rational treatment choice.