S_mmary Congenital hypertrophy of the retinal pigment epithelium (CHRPE) and multiple mandibular osteomata are markers of familial adenomatous polyposis (FAP). We have assessed their prevalence in non-polyposis familial colorectal neoplasia. Multiple mandibular osteomata were present in 1 29 (3%) patients with familial colorectal neoplasia. CHRPE was present in 11 33 (33%) patients with familial colorectal neoplasia compared with 3 36 (8%) with sporadic disease (P = 0.01) and 4 32 (12.5%) control subjects (P = 0.04). Seven patients with familial colorectal neoplasia had multiple areas of CHRPE compared with one with sporadic disease (P = 0.02) and one control subject (P = Berk et al., 1988;Chapman et al., 1989;Burn et al.. 1991;Giardiello et al., 1991;Morton et al., 1992). Not infrequently, normal individuals have one or two areas of CHRPE (Chapman et al., 1989;Burn et al., 1991), and therefore it is thought to be the presence of multiple areas which is of significance. The gene for FAP has been localised to 5q21 (Bodmer et al., 1987), and a variety of polymorphic DNA markers are available (Nakamura et al., 1988;Meera Khan et al., 1988;Dunlop et al., 1990Dunlop et al., , 1991 family syndrome. Morton et al. (1992) found CHRPE in five of ten individuals who were members of five hereditary nonpolyposis colorectal cancer families, however none of these individuals met the authors' criteria for a positive test. Sondergaard et al. (1985) identified multiple mandibular osteomata in 8 of 31 (26%) individuals with familial colorectal cancer, however these individuals were all members of two large families.The aim of this study was to assess the incidence of CHRPE and mandibular osteomata in patients with familial colorectal neoplasia and to compare this with the incidence in patients with sporadic colorectal neoplasia and a control population of unaffected individuals.
Patients and methods
RecruitmentThree groups of patients were recruited (Table I).Group I (familial colorectal neoplasia, n = 34) Forty-eight patients under follow-up by the Department of Surgery, University of Nottingham, were identified as having a firstdegree family history of colorectal cancer. They were contacted by letter and asked if they would participate in the study. Thirty-four patients agreed to do so. All these patients were under review in the colorectal cancer clinic, however on subsequent review of their histology two were found to have had large adenomas (one 3 cm villous adenoma and one 2 cm adenoma with severe dysplasia). Thirty-three patients also had a first-degree family history of colorectal cancer. On verification of the relatives' diagnoses, one individual's relative was found to have a 3 cm rectal adenoma, not a cancer. All 34 patients in the study were from different families and none was from an FAP family or had evidence of FAP.Group 2 (sporadic colorectal neoplasia, n = 36) Patients in the sporadic colorectal neoplasia group were recruited from the same colorectal cancer clinic in the Department of Surgery. Thirty-four pat...