Development of a RiskScore model related to cancer-associated fibroblasts for assessing patient prognosis and drug sensitivity in lung adenocarcinoma

Cancer-associated fibroblasts (CAFs) are central to the progression of lung adenocarcinoma (LUAD), yet their heterogeneity and prognostic relevance remain incompletely characterized. We performed scRNA‑seq analysis to delineate fibroblast subpopulations in LUAD. CAF‑related gene modules were identified using WGCNA based on bulk transcriptomic data from The Cancer Genome Atlas (TCGA). Differentially expressed genes (DEGs) between tumor and normal tissues were intersected with CAF‑related modules. Univariate, Lasso, and multivariate Cox regression were employed to select prognostic genes and construct a risk‑scoring model. Functional validation was conducted using LUAD cell lines via siRNA‑mediated knockdown, followed by proliferation, migration, and invasion assays. Immune infiltration was evaluated using ESTIMATE and MCP‑counter algorithms, and drug sensitivity was predicted via the pRRophetic package. Fibroblast cells were classified into three subtypes. Fibroblast_2 was identified as CAFs and was found to be mainly involved in extracellular matrix-relevant pathways. A total of 510 module genes and 202 CAF-related DEGs were screened. Subsequently, 8 key prognostic genes ( LATS2 , C1QTNF6 , EFNB2 , SERPINH1 , CEMIP , VEGFC , SLC24A3 , GAS7 ) were recognized to establish a RiskScore model, which exhibited strong robustness in prognostic prediction in LUAD. Silencing C1QTNF6 impeded the viability, migration, and invasion of LUAD cells. Compared to the low-risk group, the high-risk group displayed a significantly worse outcome and lower immune infiltration. The RiskScore was closely negatively correlated with 25 drugs. We established a CAF‑associated RiskScore model that can robustly reflect immune‑microenvironment features in LUAD and predict patients’ prognosis, offering a potential tool for risk stratification and therapeutic candidate prioritization.