Dysplasia and cancer in ulcerative colitis: a soluble problem?

Abstract

Now that many of the problems associated with the recognition, classification and terminology of dysplasia in ulcerative colitis have been overcome, the clinical effectiveness of such a classification needs to be investigated. Because dysplasia has been defined as an unequivocally neoplastic transformation of the epithelium which is capable of giving rise to an invasive carcinoma, it follows that in waiting for the development of dysplasia in patients at risk, there will always be an associated and as yet undefined incidence of invasive carcinoma already being present. Nevertheless, it is currently inappropriate to consider proctocolectomy before such changes develop. It is emphasised that waiting for high grade dysplasia to develop may lead to an unacceptably high level of associated invasive carcinoma. Also, because dysplasia is frequently focal, the changes of detecting it on random biopsy are directly proportional to the number of biopsies taken, and if found in flat mucosa it may be difficult to confirm the presence of such dysplasia by re-biopsy. Colectomy at this stage is aimed at cancer prevention rather than early cancer detection. Conversely, targeted biopsies of endoscopically visible plaques, nodules or other abnormalities, which may already be the superficial part of invasive carcinomas, offer the best hope of early detection of carcinoma. However, a proportion of these may have metastasised already; this is much more likely to have occurred if such a lesion is found at the first colonoscopy rather than in a patient in whom several surveillance colonoscopies have been carried out. Finally, the initial stages of many colitic cancers are undetectable endoscopically or radiologically but may be biopsied accidentally. All of these factors suggest that unexpected carcinomas will occur in any long-term prospective study. Surprisingly, most will not be lethal and they can be kept to a minimum if proctocolectomy is considered once unequivocal dysplasia is demonstrated on biopsy, particularly in parts of the world where large bowel cancer is already common. Data are required which assess accurately the risk of an invasive carcinoma being present or developing when epithelium indefinite for dysplasia, low grade dysplasia or high grade dysplasia are present both with and without an endoscopic abnormality being the source of the biopsy, and when such biopsies are obtained at the first or at subsequent colonoscopies. Only then can the risks of colectomy be weighed adequately against the likelihood of carcinoma already being present.(ABSTRACT TRUNCATED AT 400 WORDS)