Background: In developed countries, the lifetime risk of developing colorectal cancer is about 5 percent. Several studies with two to three years' follow-up have demonstrated that the use of aspirin and cyclo-oxygenase-2 (COX-2) enzyme inhibitors can reduce the recurrence of precancerous adenoma by about 40 percent. However, the only long-term study (the Women's Health Study) showed no protective effect against colorectal cancer over 10 years. Flossmann and Rothwell analyzed data on aspirin use, including two large randomized trials with more than 20 years' follow-up, to establish any protective effect of aspirin on colorectal cancer.
The Study: The study used data from two large British trials of daily aspirin use that began in about 1980: the British Doctors Aspirin Trial (5,139 male physicians) and the United Kingdom Transient Ischaemic Attack (UK-TIA) Aspirin Trial (2,449 patients after transient ischemic attack or minor ischemic stroke). These patients were followed regularly and their medical records were tagged to identify the cause of death. Using medical records and national cancer registries, researchers identified cases of colorectal cancer in aspirin users and nonusers. Patients who reported noncompliance with assigned treatment at more than one half of follow-up visits were excluded from the final analysis.
The researchers systematically reviewed data from published observational studies of aspirin or nonsteroidal anti-inflammatory drug (NSAID) use and the risk of colorectal cancer. For cohort studies, adjusted hazard rates were calculated based on the number of colorectal cancers per person-years of follow-up; for case-control studies, odds ratios were calculated for colorectal cancer in users versus nonusers.
Results: The median follow-up in both studies was 23 years. Diagnosis of colorectal cancer was reliably reported by cancer registries and death certificates. Based on intention-to-treat analysis, aspirin use was associated with a significant reduction in colorectal cancer incidence, but this effect was apparent only after 10 or more years of therapy. In the first nine years of therapy, the number of colorectal cancers was slightly less than expected (pooled hazard ratio [HR] = 0.92, P = .73). However, during years 10 through 19, the number of confirmed cancers (65) was significantly less than expected (106) based on the experience of the placebo group (pooled HR = 0.60, P = .007). No significant effect on the incidence of other cancers was identified.
The systematic search for observational studies yielded 19 case-control studies and 11 cohort studies. The studies varied greatly in populations, methodologies, dosages and types of aspirin or NSAID used, and in data collection and reporting of results. Nevertheless, regular aspirin use was consistently associated with a reduced risk of colorectal cancer (overall odds ratio [OR] = 0.80). The highest protective effect was found in patients with at least 10 years of aspirin or NSAID use and with dosages of at least 300 mg per day. In studies using the higher dosages of aspirin or NSAID, the pooled OR was 0.59, indicating a highly significant reduction in colorectal cancer cases.
Conclusion: The authors concluded that the use of at least 300 mg of aspirin daily for about five years provides effective protection against colorectal cancer. Consistent with the time required to progress from development of adenoma to neoplasm, the latency period for this protection is about 10 years. The authors stress that long-term follow-up from several trials is needed to establish the effectiveness of lower dosages of aspirin and to establish the net benefit to patients, such as reductions in cardiovascular and other adverse events, versus increased adverse effects, such as gastrointestinal bleeding.
editor's note: Because of the potential long-term benefit of their anti-inflammatory and antiplatelet properties, aspirin and other agents are being studied as preventive agents in several conditions. A recent study found no overall benefit of low-dosage aspirin (100 mg on alternate days) for the prevention or postponement of cognitive decline in women older than 65 years.1 Nevertheless, this study found marginally better performance on some cognitive measures after five to six years of follow-up. The relative risk of substantial decline in the global score was 0.92 (95% confidence interval [CI], 0.77 to 1.10); for decline in fluency, the relative risk was 0.80 (95% CI, 0.67 to 0.97). As the colorectal cancer study emphasized, dosage and duration of therapy are important; longer duration use and higher dosages of aspirin might result in significant changes in cognitive abilities between the treatment and control groups.
As front-line physicians, we must balance the optimism of these studies with the many unknowns (especially the risk of adverse effects, such as gastrointestinal bleeding) when advising patients. In particular, it is more important than ever to ask patients which nonprescription drugs and supplements they are taking, because many patients have started taking aspirin based on media reports that often focus on only the most newsworthy aspects of the study results and do not give the complete picture.—a.d.w.