Metabolic transcriptomic subtyping defines distinct immunogenomic and clinical landscapes in non–small cell lung cancer

This study aimed to evaluate the clinical significance of metabolic dysregulation and heterogeneity in patients with non-small cell lung cancer (NSCLC). We performed NanoString-based transcriptomic profiling of tumors from patients with advanced NSCLC receiving anti-PD-(L)1 therapy. Five distinct metabolic clusters were identified. Gene set analysis suggested differential enrichment of metabolic pathways across clusters, including immune signaling-dominant with minimal metabolic dysregulation (Cluster 1), amino acid metabolism-related (Cluster 2), hypoxia/autophagy-associated (Cluster 3), glycolytic/proliferative (Cluster 4), and mixed dysregulated metabolic features (Cluster 5). OXPHOS-related signatures were enriched in Clusters 3-5, whereas immune cell infiltration in the tumor was more prominent in Clusters 1-2. Survival outcomes differed significantly across clusters (progression-free survival, P = 0.035; overall survival, P = 0.008), with Cluster 1 showing the most favorable outcomes and Cluster 3 the worst following anti-PD-(L)1 therapy. Metabolic clusters were associated with clinicopathological and genomic features, including histology, smoking status, PD-L1 expression, and oncogenic driver alterations. In particular, RAS-RAF-MEK pathway gene alterations were more frequent in Clusters 3-5 than in Clusters 1-2. Together, metabolic transcriptomic profiling identifies clinically meaningful subtypes of NSCLC with distinct immunogenomic features and differential efficacy from anti-PD-(L)1 therapy.