Development and validation of a prognostic nomogram based on the log odds of positive lymph nodes (LODDS) for parotid gland cancer

Abstract

Abstract Objective The focus of this research was to explore the predictive usefulness of the logarithm ratio of positive lymph nodes (LODDS) in patients with parotid gland tumors and try to develop a clinical prognosis model. Methods A total of 924 patients with n1/n2 stage parotid tumors were retrospectively studied. The researchers looked at the link between clinicopathological characteristics, AJCC N stage, LODDS, and overall survival (OS). The association between overall survival and numerous variables was investigated using Cox regression, and a clinical prediction model was developed. In addition, the likelihood ratio (LR), Harrell consistency index (C index), area under the curve of receiver operating characteristic (ROC-AUC) and Akaike information criteria were used to assess the model's prediction ability (AIC). Results The ideal LODDS cutoff value was − 0.56 based on a training set of 645 patients. Patients' age, tumor size, T stage, radiation, and LODDS were all found to be independent factors affecting their survival in a Cox multivariate analysis. After examination, the prediction model based on effective prognostic parameters performed well: LR = 154.4, AIC = 4045.1, CI = 0.693 and ROC-AUC of 3-5-7 years: 0.744, 0.752, 0.815, which outperformed AJCC TNM model; DCA curve shows higher clinical practicability. Consistent results were obtained in the validation cohort of 278 patients. Conclusion In patients with parotid tumor, LODDS is an independent prognostic factor, and those with LODDS less than − 0.56 had a better prognosis. The established prediction model and nomogram demonstrate outstanding prediction performance as well as broad applicability.