Neuroendocrine tumours of the small intestine (SI-NETs) are rare and heterogeneous. There is an unmet need for prognostication of disease course and to aid treatment strategies. A previously developed nomogram based on clinical and tumour characteristics aims to predict disease-specific survival (DSS) in patients with a SI-NET. We aimed to validate the nomogram and identify predictors of survival. Four hundred patients with a grade 1 or 2 SI-NET were included, between January 2000 and June 2016. Predicted 5- and 10-year survival was compared to actual DSS. Multivariable analysis identified predictors for actual DSS. We found that in low-, medium- and high-risk groups 5-year nomogram DSS vs. actual DSS was 0.86 vs. 0.82 (p < 0.001), 0.52 vs. 0.71 (p < 0.001) and 0.26 vs. 0.53 (p < 0.001), respectively. Ten-year nomogram DSS vs. actual DSS was 0.68 vs. 0.69 (p < 0.001), 0.40 vs. 0.50 (p < 0.001) and 0.20 vs. 0.35 (p < 0.001), respectively. Age, WHO-performance score of 2, Ki-67 index ≥10, unknown primary tumour, CgA > 6x ULN and elevated liver tests were identified as independent predictors for a worse DSS. This shows that the nomogram was able to differentiate, but underestimated DSS for patients with a SI-NET. Improvement of prognostication incorporating new emerging biomarkers is necessary to adequately estimate survival.
Keywords: Neuroendocrine tumours; nomogram; prediction model; prognosis; small-intestine.