Predictors of presence, multiplicity, size and dysplasia of colorectal adenomas. A necropsy study in New Zealand.

Three hundred and thirty six forensic necropsy specimens of large bowel were examined in order to identify subject related variables that independently predicted the following adenoma characteristics: presence, size (largest), multiplicity and high grade dysplasia. The variables were age, gender, body mass index, race (European origin versus Maori/Polynesian) and presence of hyperplastic (metaplastic) polyp(s). Subjects included 303 New Zealanders of European origin (M = 185, F = 118) yielding 149 adenomas and 251 hyperplastic polyps and 33 Maori/Polynesians (M = 25, F = 8) yielding five adenomas and one hyperplastic polyp. Independent predictors of adenoma presence as determined by regression analysis were age (p = 0.0001), presence of hyperplastic polyps (p = 0.0001) and male gender (p = 0.05). Models were poor at explaining variation in size, multiplicity, and dysplasia. Larger adenomas occurred more frequently in subjects with multiple adenomas (p = 0.03) and multiple adenomas were probably associated with hyperplastic polyps (p = 0.09) and male gender (p = 0.09) in Europeans. High grade dysplasia was more frequent in women (p = 0.05) and possibly in subjects with hyperplastic polyps (p = 0.2). Body mass index and ethnicity did not predict any adenoma characteristics, but hyperplastic polyp prevalence was influenced by European origin (p = 0.04) and to a lesser extent by body mass index (p = 0.08) as well as presence of adenoma (p = 0.0002) and age ( = 0.005). The association of hyperplastic polyp with presence, multiplicity but not size of adenoma and with a high risk group for colorectal cancer (New Zealanders of European origin) suggests that the hyperplastic polyp serves as a marker for a factor which influences neoplastic evolution at the stages of initiation/transformation but not promotion. Fifty nine per cent of individuals with adenoma(s) did not have hyperplastic polyp(s) emphasising that the last would serve only as a marker of populations and not individuals at high risk of bowel cancer. Low intracolonic butyrate may be the factor linking the expression of the two types of polyp.