A nomogram to predict cancer-specific survival of ocular melanoma

To develop a nomogram that accurately predicts cancer-specific survival (CSS) in patients with ocular melanoma (OM).

Patient data for individuals diagnosed with OM between 2004 and 2015 were obtained from the Surveillance, Epidemiology, and End Results (SEER) database. The patients were then randomly divided into a training group and a validation group. Multivariate Cox proportional hazards regression analysis was utilized to identify significant variables. Subsequently, an independent variable-based nomogram was developed. Finally, multiple methods were used to verify and evaluate the performance of the nomogram.

A total of 5,053 eligible patients diagnosed with OM were randomly divided into a training group (n = 3,537, 70%) and a validation group (n = 1,516, 30%). Age, race, primary site, histological type, American Joint Committee on Cancer (AJCC) stage, SEER stage, surgery, radiotherapy and chemotherapy were identified as independent prognostic factors for predicting the CSS of OM. These factors were incorporated into the development of a nomogram. Compared with AJCC, the nomogram is better than AJCC in concordance index (C-index), receiver operating characteristic (ROC) curve, net reclassification index (NRI), integrated discrimination improvement index (IDI), calibration curve and decision curve analysis (DCA).

We developed a nomogram that can be used to predict the 3, 5 and 8 years cancer-specific survival (CSS) rates for patients diagnosed with OM. The nomogram offers a more precise and personalized approach for predicting patient survival outcomes while assisting ophthalmologists in devising improved clinical management strategies and treatment plans.