Predictive role of aberrant crypt foci (ACF) as surrogate endpoint biomarkers of colorectal cancer.

A12

Colorectal cancer (CRC) chemoprevention seeks to target the earliest stages of tumorigenesis. Currently, adenomas serve as surrogate endpoint biomarkers (SEB) of CRC chemoprevention efficacy. Aberrant crypt foci (ACF) are possibly the earliest precursors of CRC. We sought to determine whether the natural history of ACF parallels that of adenomas. If so validated, these lesions would facilitate chemopreventative agent identification by shortening treatment time and obviating the need for colonoscopy. A cohort of patients from the Adenoma Prevention with Celecoxib (APC) Trial underwent baseline chromoendoscopy, and rectal ACF were counted and biopsied along with matched normal mucosa. Patients were randomly placed on celecoxib (200 mg bid, n=12 or 400 mg bid, n=11) or placebo (n=14) for 8 months. After treatments, patients were re-examined and the same tissue specimens obtained. Of 37 recruited patients, 35 completed baseline and 8-month follow-up. From 603 ACF counted in subjects, H&E staining of 70 biopsies revealed that all were nondysplastic. A mean of 9.6±10.4 ACF/patient was found at baseline and 7.1±8.0 ACF/patient at 8 months for all treatment groups. No difference in the degree of change in ACF number during the 8-month period was detected among the celecoxib or placebo groups (200 mg vs. placebo, P<0.62; 400 mg vs. placebo, P<0.37). The relative expression, localization, and cell of origin of β-catenin, Cox-2, CD31, and Ki-67 proteins were examined by immunohistochemistry (IHC), and a semi-quantitative scoring method assessed the area and intensity of positive staining for assays. All ACF showed normal β-catenin expression located at epithelial plasma membranes. Cox-2 and CD31 expression in stromal cells was unchanged between treatment and placebo groups. Ki-67 proliferative index (PI) was increased in ACF relative to matched normal mucosa (P<0.0002), but the decline in PI observed upon treatment was not significant. These data indicate that ACF are not a useful surrogate end-point biomarker (SEB) of adenoma risk or chemoprevention efficacy.