A Multi-Faceted Approach to Explore the Role of Inflammatory CAFs, Providing Prognostic Value and Therapeutic Implications in Lung Adenocarcinoma.

Background: Lung adenocarcinoma (LUAD) is the most common subtype of lung cancer, where its complex tumor microenvironment (TME) significantly influences disease progression and treatment response. Inflammatory cancer-associated fibroblasts (iCAFs), as a key component of the TME, can promote tumor immune evasion and drug resistance. However, the characteristics of iCAFs in LUAD and their clinical significance have not been fully elucidated. Methods: The bulk RNA data and scRNA-seq data of LUAD from public databases were integrated to identify and characterize iCAFs subsets and screen their feature genes. iCAF-based signature (ICAFBS) was generated using multiple machine learning algorithms and confirmed in multiple independent cohorts. The relationship between ICAFBS and immune landscape as well as drug sensitivity was further analyzed. Finally, a series of functional studies were conducted to elucidate the role of PFN2 in LUAD cell lines. Results: The iCAF subtype identified by single-cell analysis was enriched in LUAD and closely linked to poor prognosis. Based on 145 iCAF characteristic genes, ICAFBS was finally screened out four key genes, MGP, LOXL2, FSTL3, and PFN2. ICAFBS demonstrated excellent prognostic predictive capabilities and was validated in multiple external datasets. Patients in the high ICAFBS group showed significant high expression of multiple immune checkpoints. Notably, silencing PFN2 inhibited cell viability and proliferation in LUAD cells, highlighting its potential as a therapeutic target. Conclusion: As a novel prognostic signature, ICAFBS can effectively predict the clinical outcome, immune landscape and treatment response of patients, providing an important reference for the individualized treatment of LUAD.