Competing risk model for prognosis of small cell neuroendocrine lung carcinoma based on SEER database.

backgroundIn recent years, the incidence of small cell neuroendocrine lung carcinoma (SCNLC) shows an upward trend. Yet there is still a lack of competing risk models for SCNLC. Therefore, this study aims to establish a competing risk model to predict the specific mortality of patients with SCNLC.

methodsPatients diagnosed with SCNLC from 2000 to 2018 were selected from the Surveillance, Epidemiology, and End Results (SEER) database. Univariate and multivariable Fine-Gray regression models were used to select independent predictors of SCNLC-specific mortality. A competing risk model was constructed to predict the cumulative 3- and 5-year specific mortality of SCNLC. C-index, ROC curve, and calibration curve were used for internal validation, and the clinical utility was evaluated by decision curve analysis.

resultsA total of 27,758 patients were included in the analysis. The results of feature selection showed that patient’s age, race, marital status, tumor grade, histology, size, stage, and number of tumors were independent risk factors for specific death of patients. The factor that had the greatest impact on SCNLC-specific mortality was stage IV tumor (HR [95%]: 1.51 [1.31, 1.73]; P < 0.0001), followed by brain metastasis among distant metastases (HR [95%]: 1.37 [1.25, 1.49]; P < 0.0001). Based on the model’s C-index and the area under the ROC curve (AUC), the results indicate that this model can effectively predict SCNLC-specific mortality and has good clinical value.

conclusionThe competing risk model for patients with SCNLC was successfully established and validated internally. The validation results suggest that the model is accurate and practical in the prediction of SCNLC-specific mortality and plays an important role in clinical decision-making and health management for patients with SCNLC.