Background and aims: Discovery of an isoform of Cyclo-oxygenase (COX) 1, the inducible COX-2, has made it possible to avoid some side effects of non-specific COX inhibitors. The COX-2 gene is over-expressed in reflux oesophagitis, Barrett's oesophagus, gastric and colon cancer, familial adenomatous polyposis, pancreatic cancer, hepatocellular carcinoma, hepatotoxicity, cirrhosis, and inflammatory bowel disease, and specific COX-2 inhibitors have been tried experimentally and clinically and found effective.
Methods: A Medline search was performed of English-language experimental studies and controlled clinical trials from January 1980 to January 2002, and relevant citations were noted.
Results: Review of available literature shows that sulindac and COX-2 inhibitors are effective in preventing as well as regressing familial adenomatous polyposis. However, they have not been shown to prevent cancer in these patients. Studies evaluating NSAIDs and COX-2 inhibitors in carcinogen-induced and genetically manipulated animal models of various cancers have been promising especially in conditions such as Barrett's oesophagus, oesophageal and hepatocellular carcinoma and pancreatic cancer. COX-2 inhibitors may be of value in the treatment of reflux oesophagitis, pancreatitis and hepatitis, although carefully planned randomized controlled clinical trials demonstrating their efficacy need to be conducted. At present NSAIDs and COX-2 inhibitors cannot be recommended for average-risk individuals or for those with sporadic colorectal neoplasia (or other forms of cancers) as chemo-preventive agents.
Conclusion: COX-2 inhibitors may open up a new therapeutic era in which these drugs can be used for chemo-prophylaxis. However, COX-2 selective inhibitors retain renal adverse effects of the non-selective inhibitors and the concern regarding the pro-thrombotic potential of COX-2 inhibitors will limit their value as chemo-preventive agents.