SOURCE beyond first-line: A survival prediction model for patients with metastatic esophagogastric adenocarcinoma after failure of first-line palliative systemic therapy.

Abstract

4037 Background: Prediction models for survival may aid shared decision making between physicians and patients. Prior models have been developed that predict survival for patients with potentially curable esophagogastric cancer and patients with metastatic esophageal cancer who start first-line therapy (the SOURCE models). The aim of this study was to develop and internally validate a registry-based clinical prediction model, called SOURCE beyond first-line, for survival of patients with metastatic esophagogastric cancer after failure of first-line palliative systemic therapy. Methods: Patients with unresectable or metastatic (synchronous or metachronous) esophageal or gastric cancer who received first-line systemic therapy (N = 1067) between 2015-2017 were selected from the Netherlands Cancer Registry. Follow-up data were retrieved in 2019. Patient, tumor and treatment characteristics at primary diagnosis and at progression of disease, were used to develop the prediction model. A Cox proportional hazards regression prediction model was developed through forward and backward selection using Akaike’s Information Criterion. The model was internally validated through 10-fold cross-validations to assess performance on unseen data. Model discrimination (C-index) and calibration (slope and intercept) were used to evaluate performance of the complete and cross-validated models. Results: The final model consisted of 10 patient, tumor and treatment characteristics. The C-index was 0.75 (0.74-0.77), the calibration slope 0.99 (0.98-0.99) and the calibration intercept 0.02 (0.01-0.02). Internal cross-validation of the model showed that the model performed adequately on unseen data: C-index 0.79 (0.76-0.80), calibration slope 1.02 (1.00-1.04) and calibration intercept -0.01 (-0.01-0.02). Conclusions: The SOURCE beyond first-line model predicted survival with fair discriminatory ability and good calibration, and is a valuable addition to the existing SOURCE prediction models. In the future this model will be integrated in an online decision support tool that can be used in clinical practice to aid personalized treatment.