Invasive carcinoma in colorectal adenomas: multivariate analysis of patient and adenoma characteristics

Endoscopy. 1997 Sep;29(7):626-31. doi: 10.1055/s-2007-1004268.


Background and study aims: The risk of invasive carcinoma developing in colorectal adenomas is influenced by a number of characteristics, relating both to the patients and to the adenomas, and by the composition of the sample analyzed. The aim of the present study was use a multivariate analysis to investigate the risk of invasive carcinoma in endoscopically and surgically removed adenomas.

Patients and methods: Between 1978 and 1993, more than 20,000 polyps were prospectively documented at the Erlangen Registry of Colorectal Polyps. A multivariate analysis of 11,188 adenomas detected at the first total colonoscopy was carried out in order to investigate the risks associated with size and site--both of which can be assessed by endoscopic inspection alone- and the extent to which these may be modified by other patient and adenoma characteristics, with an influence on the risk of invasive carcinoma in colorectal adenomas.

Results: The size of the adenoma proved to be the most important influencing factor. Invasive carcinoma was never found in 5027 small adenomas (< or = 5 mm). Adenomas in the right colon had a lower risk than those in the left colon or rectum. But with increasing adenoma size, the malignancy rate showed a right-sided shift, with a significant interactive effect of size and right-sided location. However, the risk determined by the size and site of the adenoma was significantly modified by a number of patient and adenoma characteristics, including histological type, presence of multiple adenomas, and patient age and sex.

Conclusions: The risk of invasive carcinoma in colorectal adenomas can only be adequately described by a complex model of the interactive effects of patient and adenoma characteristics on the main factors of size and site.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adenoma / pathology*
  • Colonic Neoplasms / pathology*
  • Female
  • Humans
  • Logistic Models
  • Male
  • Middle Aged
  • Multivariate Analysis
  • Neoplasm Invasiveness
  • Prospective Studies
  • Rectal Neoplasms / pathology*
  • Registries