A Putative XIST–miRNA–ZNF662 ceRNA Axis with Diagnostic and Prognostic Potential in Oral Squamous Cell Carcinoma

Oral squamous cell carcinoma (OSCC) remains a major cause of cancer-related morbidity and mortality, and reliable biomarkers for early diagnosis and risk stratification are still lacking. Long non-coding RNAs (lncRNAs) and microRNAs (miRNAs) can regulate gene expression through competing endogenous RNA (ceRNA) interactions, but OSCC-specific ceRNA axes with clinical relevance are still poorly defined. We integrated lncRNA, miRNA, and mRNA expression data from six OSCC-related datasets in the Gene Expression Omnibus with in silico interaction predictions to construct an OSCC-focused ceRNA network and examine its association with survival. The resulting network comprised 8 mRNAs, 22 miRNAs, and 12 lncRNAs. Within this network, we identified a previously unrecognized XIST–miRNA–ZNF662 axis that has not been characterized in OSCC. ZNF662 was consistently downregulated in tumors, and higher ZNF662 expression was associated with improved survival in an independent head and neck squamous cell carcinoma cohort. Components of the XIST–miRNA–ZNF662 axis also showed excellent diagnostic performance for distinguishing OSCC from normal samples across (Gene Expression Omnibus) GEO datasets, highlighting a ceRNA module with promising diagnostic and prognostic potential that could be explored further in non-invasive biofluids.