Loading...

Google

Monday, May 14, 2007

Aspirin for prevention of bowel cancer?

Aspirin for prevention of bowel cancer?

Analysis of data from two long-term controlled trials of aspirin indicates that taking regular aspirin may reduce the risk of colorectal cancer, however this was shown only for doses of 300mg daily and over.

Several controlled trials have shown that aspirin and other NSAID may reduce colorectal adenomas, potential precursor lesions to carcinomas, however these have generally had relatively short follow-up and have not been able to show whether the progression from adenoma to cancer has been interrupted. Analysis of data up to ten years in the Women's Health Study did not show any effect. The authors therefore used long-term follow-up data from two large studies of aspirin as prophylaxis for vascular events to determine whether these demonstrated any reduction in colorectal cancer in the aspirin compared to the control groups. They also carried out a systematic literature review for observational studies on the association between aspirin or other NSAID use and colorectal cancer.

The British Doctor's Aspirin Study, published in 1988, investigated whether aspirin reduced the risk of death from stroke, MI, or other vascular conditions. It involved 5,139 male British doctors who were randomised to treatment with aspirin (normally 500mg daily, n=4,377) or no specific treatment (n=762). Study duration was five or six years, and all participants were flagged in national registers to collect all cancers and deaths in the group up to 2001 or emigration from the UK. The UK-TIA study investigated whether aspirin reduced the risk of recurrent stroke after a minor stroke or transient ischaemic attack (TIA). It recruited 2,449 participants who were randomised to aspirin 300mg or 1200mg daily (n=1,632) or placebo (n=817). Median follow-up for cancers and deaths was 23 years for both studies. Primary outcome for this study was colorectal cancer.

There were 216 colorectal cancers in the total study population (aspirin n=5,061, placebo n=2,527). Of these, 87 occurred in the placebo group (3.4%) and 129 in the combined aspirin groups. The difference was significant, with a hazard ratio of 0.74 (95% CI 0.56 to 0.97, p=0.02) overall. When just those who received aspirin for five years or more were included, the difference became greater (HR 063, 95% CI 0.47 to 0.85, p=0.002). The effect was seen only after at least ten years from randomisation and was dependent on duration of treatment and compliance. There were no significant effects on other cancers.

Analysis of data from observational studies indicated an association between regular use of NSAID and aspirin, and a reduction in risk of colorectal cancer. The effect for aspirin was only consistent in those taking 300mg daily or more. Based on their results, the authors conclude that use of aspirin at least 300mg daily for at least five years is associated with a reduction in risk of colorectal cancer. There is, however, a latency of about ten years, and this is consistent with the results of observational studies. Long-term follow-up of trials using lower doses of aspirin is important to clarify whether such doses have any benefit.

An accompanying Comment discusses the study. The author comments on the discrepancy between the results of this analysis and those of other studies. He suggests that a likely cause would be the different doses used, as there is substantial animal and laboratory evidence that higher doses are needed than those used for anti-platelet effects. He notes that the study has limitations, however concludes that in combination with other evidence it helps to confirm that aspirin at relevant doses can prevent colorectal cancer. More work is needed to identify those patients for whom the risks would outweigh the benefits, and to identify the mechanism so that other strategies for prevention can be developed.

Lancet 2007; 369: 1603-13 (link to abstract); Lancet 2007; 369: 1577-8 (Comment; link to full text, available to subscribers only).

No comments: