A novel lysosome-related prognostic signature associated with prognosis and immune infiltration landscape in oral squamous cell carcinoma.

Background: Predicting the outcome of oral squamous cell carcinoma (OSCC) is challenging due to its diverse nature and intricate causes. This research explores how lysosome-associated genes (LRGs) might forecast overall survival (OS) and correlate with immune infiltration in OSCC patients. Methods: We analyzed OSCC patients' LRGs' mRNA expression data and clinical details from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO). Through univariate Cox regression, we pinpointed LRGs with prognostic potential. A signature comprising 12 LRGs linked to prognosis was developed via the Least Absolute Shrinkage and Selection Operator (LASSO) in a training dataset. Patients were classified as higher or lower risk based on their risk scores, and the prognostic independence of the risk score was assessed using multivariate analysis. The model's robustness and precision were confirmed through bioinformatics in the GEO test set. Differential gene expression analysis between risk groups highlighted functional disparities, while various immune evaluation methods elucidated immune differences. Results: The prognostic framework utilized 12 LRGs (SLC46A3, MANBA, NEU1, SDCBP, BRI3, TMEM175, CD164, GPC1, SFTPB, TPP1, Biglycan (BGN) and TMEM192), showing that higher risk was associated with poorer OS. This set of genes independently predicted OS in OSCC, linking LRGs to cellular adhesion and extracellular matrix involvement. Initial assessments using ssGSEA and CIBERSORT suggested that the adverse outcomes in the higher-risk cohort may be tied to immune system deregulation. Conclusion: Twelve-LRGs signature has been identified for OSCC prognosis prediction, offering novel directions for lysosome-targeted therapies against OSCC.