Establishment and validation of an individualized nomogram to predict distant metastasis in chondrosarcoma patients: a population-based study

Abstract

BackgroundDistant metastasis is a significant factor influencing chondrosarcoma (CHS) patients’ treatment and prognosis. We aimed to establish a consistent and effective nomogram to better predict distant metastases of CHS individuals.MethodsThe Surveillance, Epidemiology and End Results (SEER) database was used to obtain the demographics and clinicopathological characteristics of CHS patients from 2010 to 2018. Independent risk factors were identified via univariate and multivariate logistic regressive analysis. A nomogram that predicts metastasis risk was established based on the training cohort, and its accuracy was validated through the validation cohort. The performance of this predictive model was assessed by the receiver operating characteristic (ROC) curve and Harrell’s concordance index (C-index). Finally, decision curve analysis (DCA) was conducted to test its clinical reliability.ResultsData of 1,066 patients were extracted, of these, 66 cases (6.19%) were with distant metastasis at initial diagnosis. The following features were shown to be linked to an increased risk of metastasis: high-grade tumor, T3 stage, and large tumor size; whereas unmarried and use of surgery were independent protective factors. Marital status, tumor grade, T stage, use of cancer-directed surgery and tumor size were incorporated to develop the novel nomogram. The ROC curves showed the effectiveness of the nomogram with the high area under the curves, the C-indices were 0.931 and 0.951 in the internal and external validation, respectively. The calibration plots indicated a good consistency and agreement of the nomogram, while the DCA illustrated that the nomogram had favorable potential clinical applicability due to great positive net benefit with wide ranges of the threshold probabilities.ConclusionsThis work developed a novel nomogram for predicting distant metastasis in CHS patients, which might assist clinicians to determine the optimal treatment plan by precisely predicting individualized metastatic risk.