Development and validation of a novel fibroblast scoring model for lung adenocarcinoma.

The interaction between cancer-associated fibroblasts (CAFs) and the tumor microenvironment (TME) is a key factor for promoting tumor progression. In lung cancer, the crosstalk between CAFs and malignant and immune cells is expected to provide new directions for the development of immunotherapy. In this study, we have systematically analyzed a single-cell dataset and identified interacting genes between CAFs and other cells. Subsequently, a robust fibroblast-related score (FRS) was developed. Kaplan-Meier (KM) and ROC analyses showed its good predictive power for patient prognoses in the training set comprising of specimens from the cancer genome atlas (TCGA) and in three external validation sets from the Gene Expression Omnibus (GEO). Univariate and multivariate Cox regression analyses suggested that FRS was a significant prognostic factor independent of multiple clinical characteristics. Functional enrichment and ssGSEA analyses indicated that patients with a high FRS developed "cold" tumors with active tumor proliferation and immunosuppression capacities. In contrast, those with a low FRS developed "hot" tumors with active immune function and cell killing abilities. Genomic variation analysis showed that the patients with a high FRS possessed a higher somatic mutation burden and copy number alterations and were more sensitive to chemotherapy; patients with a low FRS were more sensitive to immunotherapy, particularly anti-PD1 therapy. Overall, these findings advance the understanding of CAFs in tumor progression and we generated a reliable FRS-based model to assess patient prognoses and guide clinical decision-making.