Prediction of survival for patients with neuroendocrine neoplasms (NENs) using multivariable cox regression and survival nomograms: Data from National Cancer Registration and Analysis Service (NCRAS, UK) Database

#4000

Introduction: Multiple nomograms for predicting survival of NENs from surveillance, epidemiology, and end results (SEER) database have been published. To the best of our knowledge, no survival nomogram has been published from NCRAS database

Aim(s): The main aim was to develop and validate survival nomograms and interactive web applications to predict survival of patients with NENs based on data from NCRAS database at 1, 3 and 5 years.

Materials and methods: 14,834 and 124,953 patients with NENs were extracted from NCRAS (2012 - 2018) and SEER (1975 – 2020) databases respectively. The included sites were appendix, cecum, colon, lung, pancreas, rectum, small intestine, and stomach. Multivariable cox regression models and associated nomograms for each site, based on sex, age, stage and grade (NET vs NEC) were developed. Variables which were not statistically significant were removed from some models. Internal validation and external validation against US SEER data were done for each model. Interactive web applications for predicting survival of each site were developed. Discrimination statistics (R2, and Dxy) were calculated. ROC curves and AUC were generated for each model at fixed observed time, and they provide the performance across all possible classification thresholds.

Conference:

Presenting Author:

Authors: Mortagy M, El Asmar M, Chandrakumaran K, Ramage J,

Keywords: SEER, NCRAS, Nomogram, Survival analysis, Cox regression, Interactive web application, neuroendocrine neoplasm,