Development of a novel scoring system based on endoscopic appearance for management of rectal neuroendocrine tumors.

BACKGROUND

The clinical significance of the endoscopic appearance of rectal neuroendocrine tumors (NETs) is poorly understood. We aimed to develop a novel scoring system based on endoscopic appearances to predict endoscopically advanced disease in patients with rectal NETs when initially diagnosed.

METHODS

Patients diagnosed with well-differentiated rectal NETs between January 2005 and December 2019 were retrospectively included. Logistic regression analyses were applied to study the relationship between endoscopic appearance and advanced disease. The whole dataset was randomly divided into training and validation sets, which were used to develop and validate a novel scoring system, respectively.

RESULTS

309 patients were included. The endoscopic appearance of rectal NETs was significantly associated with advanced disease ( < 0.001). A novel scoring system was developed based on endoscopic appearance, including tumor size, tumor shape, and mucosal surface, using the training set. The area under curve (AUC) of the scoring system to predict advanced disease was 0.953 (95 % confidence interval [CI] 0.915 - 0.991;  < 0.001) and 0.960 (95 %CI 0.905 - 1.000;  < 0.001) in the training and validation sets, respectively. Furthermore, the scoring system was significantly associated with tumor grade. Patients with high scores had significantly worse disease-free and overall survival than patients with low scores ( < 0.001).

CONCLUSION

This novel scoring system based on the endoscopic appearance of the primary tumor can help to accurately identify patients with endoscopically advanced disease who are not suitable for endoscopic resection. In addition, it is of great value in monitoring tumor recurrence and overall survival in patients with rectal NETs.