Chemoprevention trials have utilized adenoma recurrence as the primary endpoint based upon multiple lines of evidence establishing adenomas as precursor lesions of colorectal cancers. These trials require large study populations followed for extended time periods and utilize extensive resources. Accordingly, there is a critical need to identify a surrogate endpoint biomarker (SEB) that can predict treatment efficacy and whose validation would enable smaller, shorter-term and less costly prevention trials. A promising candidate SEB is the aberrant crypt foci (ACF). ACF are putative precursors of adenomas and carcinomas based upon their age-related prevalence, left-sided predominance, frequency of dysplasia, and association with prior/current neoplasia. ACF can be detected and quantified in the human colorectum using magnification chromoendoscopy. Studies in animal models of colon cancer indicate that ACF can serve as a SEB for chemopreventive efficacy and the NSAID sulindac was shown to regress ACF in Japanese subjects. To date, limited data are available on ACF in the U.S. population and an important aim of this proposal is to identify, quantify, and characterize ACF in a high risk population. To this end, we propose a randomized, placebo controlled trial in patients with prior advanced colonic adenomas or carcinomas evaluating aspirin alone and in combination with calcium carbonate or difluoromethylornithine over a 12 month period. Agent selection is based upon compelling preclinical and human data. The primary study endpoint will be the percent change in ACF number in the left colon at 12 month compared to baseline using magnifying chromoendoscopy. ACF characteristics (size/morphology, histopathology) will be determined and tissue biomarkers will be evaluated in ACF or normal mucosa. These will include drug targets, i.e., prostaglandin E2 and mucosal polyamine levels, apoptotic and proliferative indices, downstream effectors of EGFR, i.e., cyclooxygenase-2 (COX-2) and cyclin D1, and COX-2-dependent apoptotic regulatory genes based upon preliminary data. Gene expression profiling of ACF will also be performed. In an effort to determine the natural history of ACF, adenoma recurrence will be determined at the 12 month colonoscopy and patients will be followed prospectively to record surveillance colonoscopy data over a 36 month period post study entry. ? ?
