Abstract
Constructing and validating two nomograms to predict the overall survival (OS) and cancer-specific survival (CSS) in cutaneous squamous cell carcinoma (CSCC) correlated with human papillomavirus (HPV) infection was the main goal of this study. We constructed predictive models for OS and CSS incidence in HPV infection-associated CSCC using information from 2238 patients in the Surveillance, Epidemiology, and End Results (SEER) database and screened the variables by LASSO regression, Cox univariate regression, and Cox multifactorial regression models, which were calibrated and validated by internal and external cohorts. Finally, all patients were categorized into intermediate-risk, low-risk, and high-risk groups based on the optimal threshold calculated from the total score. Multivariate analysis showed that HPV infection status, marital status, tumor metastatic stage, surgical status, radiotherapy status, lymph node biopsy, local lymph node dissection, primary tumor status, and bone metastasis were risk factors for OS and CSS. The C index, the time-dependent area under the receiver-operating characteristic curve, and the column-line diagrams of the calibration plot were among the excellent-performance metrics that were effectively displayed. Moreover, the decision curve analysis of the two nomograms consistently revealed their favorable net benefits spanning 1, 2, and 3 years. In addition, the survival curves indicate that each of the two risk classification systems clearly differentiates high, medium, and low risk groups. These meticulously crafted nomograms stand poised to serve as indispensable instruments in clinical practice, empowering clinicians to adeptly communicate with patients regarding their prognostic outlook over the forthcoming 1, 2, and 3 years.