A Multicenter Study of Prevalence and Risk Factors for Aberrant Crypt Foci

Abstract

Aberrant crypt foci (ACF) are the putative precursor of colorectal adenomas. However, there are limited data available on the prevalence and risk factors for ACF. Subjects from the Prostate, Lung, Colorectal and Ovarian cancer screening trial were recruited for an ACF study, with subjects with adenoma history being oversampled. By using a standardized protocol of magnified chromoendoscopy with methylene blue staining (up to the middle rectal fold), ACF were photo-documented and removed for histologic evaluation. A total of 505 (66% male; 55% > or =70 y) subjects from 4 institutions were examined; 42% had no adenoma, 32% had nonadvanced distal adenoma, and 25% had advanced distal adenoma at the baseline Prostate, Lung, Colorectal and Ovarian cancer screening trial examination (8.2 years before ACF examination on average). A total of 68% of this population had 1 or more ACF, 43% had 1 to 3, 19% had 4 to 6, and 5% had 7 or more. Baseline adenoma status was not associated with ACF prevalence (range, 66%-69%) or mean number of ACF (range, 3.1-3.5). Of 143 endoscopic ACF examined histologically, 68.5% were confirmed to be ACF. In a logistic model, current (odds ratio [OR], 2.6; 95% confidence interval [CI], 1.2-5.6) and former smoking (OR, 1.6; 95% CI, 1.1-2.5) were associated with higher ACF prevalence; a body mass index greater than 30 was associated with lower prevalence (OR, 0.53; 95% CI, 0.35-0.8). Age, sex, family history of colorectal cancer, and aspirin/nonsteroidal anti-inflammatory drug use were not associated significantly with ACF prevalence. ACF prevalence and number were not associated with adenoma history, and only 68.5% of endoscopic ACF were confirmed histologically. These results raise concern about the use of ACF as a surrogate marker of colorectal cancer risk.