Even amazing studies can leave us stumped

There are days when it seems like nothing we doctors do actually works. Then there are days I feel much better. I’m not sure which kind today is:

Background: Recent evidence suggests that daily aspirin use decreases cancer risk, particularly for colorectal cancer, but evidence for alternate-day use is scant.

Objective: To examine the association between long-term, alternate-day, low-dose aspirin and cancer in healthy women.

Design: Observational follow-up of a randomized trial.

Setting: Female health professionals.

Participants: 39 876 women aged 45 years or older in the Women’s Health Study (ClinicalTrials.gov: NCT00000479), 33 682 of whom continued observational follow-up.

Intervention: 100 mg of alternate-day aspirin or placebo through March 2004, with a median 10-year follow-up. Posttrial follow-up continued through March 2012.

Measurements: Cancer incidence.

This is a long term follow-up of female health professionals in the Women’s Health Study, in which participants were randomized to receive either aspirin or placebo every other day from 1992 to 2004. This was a massive trial, not likely to be repeated again anytime soon. What did they find?

Overall, aspirin didn’t seem to prevent all cancer (hazard ratio (HR) 0.97), breast cancer (HR 0.98 ), or lung cancer (HR, 1.04). But colorectal cancer wassignificantly reduced in those taking aspirin (HR, 0.80; p = 0.02). This was especially true for proximal colon cancer (HR 0.73; p = 0.02). The difference was seen only after 10 years, with a posttrial reduction of 42%. Phenomenal, right?

Now some caveats. This was a new, and welcome finding, because the original 10 year follow-up study detected no differences. This would seem to point to the fact that the benefit is really long-term. But it’s also possible we have some biases here. There was more gastrointestinal bleeding and peptic ulcers in the aspirin group. this might have theoretically led to more screening in this group, and more prevention. However, the colonoscopy rates in the two groups were similar, so this seems less likely. But the biggest buzzkill is that overall mortality rates remained unchanged.

Research is hard. Following this many participants for this long seems like an impossible task. We should applaud the efforts to do rigorous studies like this for so long, and seriously consider the relative benefits and harms of aspirin moving forward. But while these results seem positive, it’s hard to tell whether they are significant enough for us to change our thoughts on the widespread use of aspirin in otherwise healthy women.


Hidden information below


Email Address*