We have studied aspirin as a potential chemopreventive for colorectal cancer, completing Phase I studies on aspirin pharmacology and potential biomarker assays (prostaglandins, PGE2 and PGF(2α) and cyclooxygenase modulation) in normal human subjects. These studies have determined the optimal dose of aspirin for future Phase IIa and IIb chemopreventive trials in high-risk cohorts of patients for colon cancer. Aspirin's effects on rectal prostaglandins are prolonged, detectable even after aspirin and its metabolite are removed from the plasma. Aspirin-mediated inhibition of prostaglandin production in the human rectal epithelium may be related to direct suppression of cyclooxygenase transcription and not to enzyme inactivation by acetylation. A systematic method to monitor adherence (self- report, telephone contact, pill count, and microelectronic monitoring) has been established for future trials. Strategies to improve recruitment of high-risk cohorts have been developed. Phase IIa non-randomized studies with aspirin at 81 mg in high-risk cohorts (resected Duke's A colon cancer, Duke's C colon cancer treated with adjuvant therapy and disease-free at 5 years, history of colon adenomas > 1 cm, two or more first-degree relatives with colon cancer, and familial adenomatous polyposis and hereditary non-polyposis colorectal cancer syndromes) are currently being conducted for surrogate end- point biomarker (prostaglandins, cyclooxygenase, cellular mucins, and proliferation) modulation.

Epidemiological, experimental, and clinical observations provide support for a colorectal cancer chemopreventive role for aspirin. We have evaluated the effects of aspirin on proliferation biomarkers in normal-risk and high-risk human subjects for colorectal cancer. Colorectal biopsies were obtained at baseline and at 24h after 28 daily doses of 81mg of aspirin from 13 high-risk and 15 normal-risk subjects for colorectal cancer. We evaluated aspirin's effects on proliferating cell nuclear antigen (PCNA) immunohistochemistry and epithelial mucin histochemistry using the lectin, Amaranthus caudatus agglutinin (ACA) in crypt sections from rectal biopsies. The baseline whole crypt PCNA LIs differed significantly between normal-risk and high-risk subjects. PCNA LIs are not affected by 28 days of aspirin at 81mg daily. ACA LIs are decreased by 28 days of aspirin at 81mg daily in both normal-risk and high-risk subjects. Aspirin's effects on ACA LIs may have mechanistic and biological implications that deserve further attention. PCNA and ACA LIs are not useful as proliferation biomarkers for aspirin's chemopreventive activity in morphologically normal human colorectal mucosa.