CAPP2: What the Colorectal Cancer Trial Found

Aspirin, tested properly, in the people most at risk

Plenty of observational studies had linked regular aspirin use to less bowel cancer. Observational studies cannot settle cause. CAPP2 was built to do that, in the group with the most to gain.

The trial enrolled people with Lynch syndrome. It ran for years, and it produced three different answers depending on when you looked.

What Lynch syndrome is

Lynch syndrome is an inherited condition, also called hereditary nonpolyposis colorectal cancer. NCI attributes it to harmful germline changes in the DNA mismatch repair genes MLH1, MSH2, MSH6 and PMS2, and in EPCAM.

Mismatch repair is the system that proofreads DNA when a cell copies it. When one of those genes is faulty from birth, errors build up much faster than normal in the lining of the bowel. The result is a sharply raised lifetime risk of colorectal cancer, along with raised risk of endometrial and several other cancers.

Because the risk is lifelong, prevention matters far more here than it does for the general population. NCI notes that colonoscopy every one to two years in people with Lynch syndrome has been shown to reduce both colorectal cancer incidence and deaths. Our page on colorectal cancer covers how the disease itself is diagnosed and staged.

What the trial did

CAPP2 was double-blind, placebo-controlled and randomized. It used a two-by-two design, testing two things at once: 600 mg of aspirin a day, and 30 g a day of resistant starch, each against its own placebo, for up to four years.

A total of 861 people with Lynch syndrome were randomly assigned to aspirin or aspirin placebo. Their mean age was 45. The registration number is ISRCTN59521990.

The main measure was simple: did the person develop colorectal cancer?

Three looks at the same trial

2008, at a mean of 29 months. The first report, covering 746 participants, found no significant preventive effect on colonic adenomas or carcinomas. The relative risk was 1.0. On this evidence, aspirin did nothing.

2011, at a mean of 55.7 months. Forty-eight people had developed 53 colorectal cancers: 18 of 427 on aspirin and 30 of 434 on placebo. The main intention-to-treat analysis gave a hazard ratio of 0.63, with a 95% confidence interval of 0.35 to 1.13. Because that interval crosses 1, the result did not reach statistical significance.

Among the 508 people who actually completed at least two years of tablets, the hazard ratio was 0.41, with an interval of 0.19 to 0.86.

2020, past ten years of follow-up. Now both analyses were significant. Intention to treat gave a hazard ratio of 0.65, with an interval of 0.43 to 0.97.

Why the answer changed

NCI's summary explains the pattern. The protective effect of aspirin did not appear until roughly five years after people started taking it, and it persisted for at least twenty years, even though the tablets were taken for a fixed period.

That is unusual, and it explains the early null result. The 2008 report was not wrong. It looked too soon.

NCI reports that in the long-term data, 24 people with Lynch syndrome needed to take this finite course of aspirin to prevent one colorectal cancer. That figure, the number needed to treat, is the most useful single number here.

Women in the aspirin group also developed fewer endometrial cancers, 7 against 17, though that difference did not reach statistical significance. Adverse events did not differ between the aspirin and placebo groups during the intervention.

When to get checked

For everyone at average risk, the U.S. Preventive Services Task Force recommends colorectal cancer screening from age 45 to 75, and a case-by-case decision from 76 to 85.

Lynch syndrome changes that entirely. Screening starts much earlier and repeats far more often, and NCI describes colonoscopy every one to two years as the approach shown to reduce incidence and deaths.

Ask a clinician about genetic counseling if any of these describe your family:

• Colorectal or endometrial cancer diagnosed under age 50 in a close relative.
• Two or more relatives with Lynch-related cancers, including bowel, endometrial, ovarian, stomach, small bowel, urinary tract, or pancreatic.
• A relative already known to carry a mismatch repair gene variant.
• Your own tumor reported as mismatch repair deficient or microsatellite instability high.

Symptoms deserve attention at any age. NCI lists bleeding from the rectum, a lasting change in bowel habits, belly pain, blocked bowels, loss of appetite, unplanned weight loss, and weakness. Iron-deficiency anemia on a blood test is another route in. Take any of these to a clinician if they last more than a couple of weeks.

What this trial cannot tell you

• It studied people with Lynch syndrome. It is not evidence that everyone should take aspirin to prevent bowel cancer.
• The dose was 600 mg a day, far higher than a low-dose aspirin tablet. Aspirin at that dose carries a real bleeding risk that this trial was not sized to measure fully.
• The 2011 intention-to-treat result was not statistically significant. The stronger figures came from later follow-up and from people who completed the course.
• NCI notes a limitation: how often participants had surveillance colonoscopy was not standardized across centers.
• Aspirin does not replace colonoscopy for anyone. It was tested in people who were already in surveillance programs.

Do not start daily aspirin on the strength of this page. That is a conversation with a clinician who knows your bleeding risk, your other medicines, and your genetic status. Clinical trial phases explains why a single trial rarely settles a prevention question on its own.

The wider picture

American Cancer Society projections for 2026, which SEER republishes, put new colorectal cancer diagnoses in the United States at about 158,850 and deaths at about 55,230. Survival is measured rather than forecast: NCI's SEER data give five-year relative survival across all stages of 65.4% for people diagnosed 2016 to 2022.

That is a group figure covering everyone diagnosed over several years, most of whom do not have Lynch syndrome. It describes a population and not a person.

Sources

• Burn J, Gerdes AM, Macrae F, et al. Long-term effect of aspirin on cancer risk in carriers of hereditary colorectal cancer: an analysis from the CAPP2 randomised controlled trial. Lancet 2011;378:2081-7
• NCI PDQ: Genetics of Colorectal Cancer (Health Professional Version)
• NCI PDQ: Rectal Cancer Treatment (Health Professional Version)
• SEER Cancer Stat Facts: Colorectal Cancer
• USPSTF: Colorectal Cancer Screening

How this page was made

An AI-assisted editorial system helped prepare this page. No named medical reviewer has reviewed it unless one is listed.

Cancer Explained is published by the National Cancer Information Foundation. It is not medical advice and does not suggest a test or treatment.

See an error, old source, or unclear wording? Tell us.