Extracellular Matrix-Related Prognostic Signature for Head and Neck Squamous Cell Carcinoma via Multi-Algorithm Survival Modeling.

Abstract

Background: Head and neck squamous cell carcinoma (HNSCC) has a poor prognosis, and the extracellular matrix (ECM) plays a key role in tumor progression, emerging as a potential biomarker for prognosis and therapy.

Objectives: To develop and validate an ECM-related prognostic signature (ECMS) and assess its association with immune features and therapeutic response in HNSCC.

Design: Retrospective, multi-cohort bioinformatics and experimental study integrating transcriptomic analysis, machine learning, and molecular biology validation.

Methods: ECM-related genes were identified from transcriptome data. An ECMS was constructed using 10 machine learning algorithms with 101 algorithm combinations and evaluated across training, internal, and external validation cohorts. An integrated nomogram combining ECMS with clinical variables was developed for prognosis prediction. Immune infiltration and treatment responses were analyzed. qRT-PCR validated gene expression in 15 paired HNSCC and adjacent normal tissues, and molecular experiments confirmed key gene functions.

Results: Twenty-three ECM genes were significantly associated with prognosis. The ECMS demonstrated moderate and consistent predictive performance across datasets. The nomogram provided a potential tool for clinical outcome prediction. Significant differences in immune cell infiltration and immune checkpoint gene expression were observed between high- and low-risk groups. qRT-PCR confirmed elevated expression of key ECM genes, including WNT7A, in tumor tissues, and functional assays showed that WNT7A promotes HNSCC cell proliferation, migration, and invasion.

Conclusion: This study developed an ECMS with potential prognostic value, which may complement existing clinical variables for outcome prediction in HNSCC.