Aspirin and gastrointestinal cancer.

In rodents, nonsteroidal anti-inflammatory drugs (NSAIDs) including aspirin inhibit chemically induced adenomas and early carcinomas of the colon. The NSAID Sulindac inhibits the growth of polyps of the colon and rectum in two randomized trials of patients with familial adenomatous polyposis (FAP), although the inhibition is not complete. Eight epidemiologic studies have found a 40-50% reduction in polyps or colorectal cancer among persons who regularly use aspirin or other NSAIDs compared to those who do not. Two epidemiologic studies show a slight increase in risk. Interpretation of the epidemiologic studies is complicated, because bleeding induced by aspirin may enhance the diagnosis and early treatment of cancer, and at least in theory, the symptoms of cancer could cause patients to avoid aspirin. Clinical trials designed specifically to investigate the aspirin hypothesis in humans at high risk of colorectal polyps or cancer are needed to establish causality, and to define the optimal dose and drug. Experimental studies should further define the mechanism of tumor inhibition in animals.