A New Online Dynamic Nomogram: Construction and Validation of an Assistant Decision-Making Model for Laryngeal Squamous Cell Carcinoma.

Background: Laryngeal squamous cell carcinoma (LSCC) is the most common type of head and neck squamous cell carcinoma. However, there are currently no reliable biomarkers for the diagnosis and prognosis of LSCC. Thus, this study aimed to identify the independent risk factors and develop and validate a new dynamic web-based nomogram that can predict auxiliary laryngeal carcinogenesis. Methods: Data on the medical history of 221 patients who were recently diagnosed with LSCC and 359 who were recently diagnosed with benign laryngeal lesions (BLLs) at the First Affiliated Hospital of Anhui Medical University were retrospectively reviewed. Using the bootstrap method, 580 patients were divided in a 7:3 ratio into a training cohort (LSCC, 158 patients; BLL, 250 patients) and an internal validation cohort (LSCC, 63 patients; BLL, 109 patients). In addition, a retrospective analysis of 31 patients with LSCC and 54 patients with BLL from Fuyang Hospital affiliated with Anhui Medical University was performed as an external validation cohort. In the training cohort, the relevant indices were initially screened using univariate analysis. Then, least absolute shrinkage and selection operator logistic analysis was used to evaluate the significant potential independent risk factors (P<0.05); a dynamic online diagnostic nomogram, whose discrimination was evaluated using the area under the ROC curve (AUC), was constructed, while the consistency was evaluated using calibration plots. Its clinical application was evaluated by performing a decision curve analysis (DCA) and validated by internal validation of the training set and external validation of the validation set. Results: Five independent risk factors, sex (odds ratio [OR]: 6.779, P<0.001), age (OR: 9.257, P<0.001), smoking (OR: 2.321, P=0.005), red blood cell width distribution (OR: 2.698, P=0.001), albumin (OR: 0.487, P=0.012), were screened from the results of the multivariate logistic analysis of the training cohort and included in the LSCC diagnostic nomogram. The nomogram predicted LSCC with AUC values of 0.894 in the training cohort, 0.907 in the internal testing cohort, and 0.966 in the external validation cohort. The calibration curve also proved that the nomogram predicted outcomes were close to the ideal curve, the predicted outcomes were consistent with the real outcomes, and the DCA curve showed that all patients could benefit. This finding was also confirmed in the validation cohort. Conclusion: An online nomogram for LSCC was constructed with good predictive performance, which can be used as a practical approach for the personalized early screening and auxiliary diagnosis of the potential risk factors and assist physicians in making a personalized diagnosis and treatment for patients.