CBX3:IL1RN Reflects Distinct Cellular States That Defines the Clinical Outcome of Oral Squamous Cell Carcinoma

<h4>Background</h4>Oral squamous cell carcinoma (OSCC) exhibits heterogeneous therapeutic outcomes owing to various tumor microenvironments (TMEs). Previous studies have classified OSCC according to its molecular characteristics; however, clinical practice is limited by technical complexity.<h4>Methods</h4>A total of 33 patients from five single-cell datasets of primary OSCC were integrated for CMS classification. The correlation between CMS classification and prognosis was analyzed with integrated bulk transcriptomics. Top scoring pairs (TSPs) algorithm was used to identify the gene pair that effectively captures the information of CMS classification. The selected gene pair was ultimately validated in independent single-cell data, spatial transcriptomics and immunofluorescence assays.<h4>Results</h4>OSCC patients were stratified into two distinct molecular subtypes (CMS1 and CMS2) harboring diverse prognostic levels, therapeutic vulnerabilities, microenvironmental characteristics, and malignant behaviors. Furthermore, CBX3 and IL1RN were identified as signature genes whose expression ratio (CBX3:IL1RN) effectively captures the critical features of CMS classification. Patients whose CBX3:IL1RN > 1 exhibited characteristics of CMS1 subtype with worse outcomes and more active malignant behaviors, vice versa. Further analyses implicated epithelial-mesenchymal transition (EMT) as a potential mechanism through which this ratio may modulate tumor progression. Finally, our finding was validated with our clinical samples and confirmed the ratio of CBX3:IL1RN was related to T and N stages.<h4>Conclusion</h4>This study presents a simplified and clinically applicable OSCC classification system based on CBX3:IL1RN. And provide a practical tool for OSCC subtyping with implications for prognosis prediction and personalized treatment strategies.