Colorectal cancer is the third most common cancer and the second most common cause of death from cancer in many industrialised countries. This disease will affect 5% of the US population, resulting in over 130 000 new cases and 57 000 deaths projected this year. The age adjusted death rates are highly variable in populations located in different parts of the world. For example, there are 3.4 cases per 100 000 in Nigeria compared with 35.8 cases per 100 000 each year in the state of Connecticut, USA. This indicates that, apart from the known genetic factors, certain environmental and dietary factors are probably involved in the aetiology of this disease. Epidemiological, clinical and genetic evidence indicates that a great number of colorectal adenocarcinomas develop from a benign adenomatous polyp progressing through a sequence of events which may take about 15–20 years.1 The progression of events leading to transformation of colonic epithelial cells includes a series of mutations in key genes which affect regulation of cell growth, cellular differentiation, DNA repair, apoptosis, and other biological processes.2

Some patients inherit a clear genetic predisposition to colorectal cancer and have been grouped into either polyposis or non-polyposis syndromes based on the phenotype they exhibit.3 Patients with familial adenomatous polyposis (FAP) typically develop hundreds to thousands of colonic polyps at an early age and possess a germline mutation in their adenomatous polyposis coli (APC) gene. Patients with hereditary non-polyposis colorectal cancer (HNPCC) typically have very few polyps and often develop carcinomas on the right side of the colon at an early age. Patients with HNPCC usually inherit a germline mutation in one of a set of genes required to repair mismatched bases in DNA.4

Patients with FAP have been of particular interest for studies designed to evaluate agents for their ability to induce polyp regression. Non-steroidal anti-inflammatory drugs (NSAIDs) have been evaluated in this patient population and results from clinical trials have shown that use of these drugs leads to a significant reduction in polyp size and number.5 Additionally, epidemiological studies have revealed that use of NSAIDs in the general population leads to a 40–50% reduction in the risk of developing colorectal cancer.6-10 These clinical studies have led researchers to evaluate potential mechanisms whereby NSAIDs could reduce the risk of developing colorectal cancer and induce polyp regression. One property shared by all NSAIDs is their ability to inhibit prostaglandin synthesis by inhibiting the activity of cyclooxygenase (COX) enzymes. Two COX isoforms have been identified, referred to as COX-1 and COX-2 as a result of the temporal order in which they were discovered.11 COX-1 is produced constitutively in many circumstances. However, COX-2 is induced at sites of inflammation and its synthesis is stimulated by a number of cytokines, growth factors and tumour promoters.12 ,13 Most NSAIDs currently available for clinical use inhibit both COX-1 and COX-2. Recently, a new class of NSAID has been developed comprising highly selective inhibitors of COX-2 and which seem to have greatly improved gastrointestinal safety profiles.14 ,15 Selective COX-2 inhibitors have been shown to be effective at preventing intestinal tumour formation and inhibiting colon carcinoma growth in animal models of intestinal tumorigenesis.16-18 Therefore, there is considerable interest in the potential role of COX-2 in the pathogenesis of colorectal cancer.

The relative levels of COX-2 and COX-1 expression in colorectal cancers have been evaluated by a number of different groups which have reported increased expression in colorectal carcinomas.19-21 Published reports indicate that roughly 85% of adenocarcinomas exhibit a 2–50-fold increase in COX-2 expression compared with the adjacent normal mucosa.19However, in the previous studies no attempt was made to compare the amount of COX-2 expression with the colonic location of the lesion. Dimberg et al (see page 730) report here for the first time that COX-2 is differentially expressed in colorectal cancers found in various regions of the colon. They found that COX-2 protein was greatly overexpressed in tumours located in the rectum compared with other regions of the colon. Interestingly, no association between increased COX-2 expression in tumour tissue and Dukes’ stage was found. These results highlight the need for future studies aimed at determining the precise location of COX-2 expression within tumours and the reason for the predominance of high expression in rectal carcinomas. More importantly, we need to answer the question whether these results portend a clinical outcome or predict any clinical significance with regard to treatment options.