Background/aims: To date, risk factors for severe dysplasia in colorectal adenomas have been studied mainly in endoscopically removed material. The aim of this study was to determine whether there are any differences between adenomas removed endoscopically and surgically in terms of patient and polyp characteristics.
Methodology: Between 1978 and 1993, 5092 consecutive patients with 9874 colorectal adenomas attending the Medical and Surgical Departments of the University of Erlangen were prospectively documented in the Erlangen Registry of Colorectal Polyps, and statistically examined using regression analysis.
Results: 5995 adenomas were removed endoscopically from 3343 patients. 3879 adenomas were surgically resected in 1749 patients. 313 (9.4%) of the former, and 218 (12.5%) of the latter, patients harboured adenomas with severe dysplasia. The size of the adenoma and the tubulovillous or villous structure were the leading risk factors, both in endoscopically and surgically removed adenomas. Adenomas located in the right colon were at a lower risk in both endoscopic and surgical specimens. The risk factors in endoscopically removed adenomas can be described by a model of independent variables. However, the risk of severe dysplasia in surgically resected adenomas needs to be described in an interactive model. Villous structure act as a confounder which modifies the risk profile by increasing the risk associated with multiplicity and right-sided location.
Conclusions: The risk of severe dysplasia in colorectal adenomas needs to be described in different models, depending on the method of adenoma removal-either endoscopic or surgical. A global model describing the colorectal dysplasia-carcinoma sequence will depend on the composition of the endoscopically or surgically removed adenomas and will be a more complex model than the clear and simple description applicable to purely endoscopic studies.