Writing about this study, the BBC reports:
For individual cancers the reduction was about 40% for bowel cancer, 30% for lung cancer, 10% for prostate cancer and 60% for oesophageal cancer.These results are from a meta-analysis of 3 large observational studies of approximately 25,000 adults in the UK, originally randomly assigned to daily aspirin or control (placebo, another antiplatelet agent or nothing) to assess the effects of aspirin on vascular events (heart attack or stroke) and followed for 4 - 20 years. Each study collected statistics on deaths from cancer as well as cancer incidence.
The number of some cancers was too small to allow estimation of risk reduction, if it indeed occurred. Although how cancer reduces risk is not actually known, it either reduces incidence or the rate of growth of the tumor, and animal studies showed this was "mediated at least in part by inhibition of the cyclo-oxygenase (COX) enzymes and reduced production of prostaglandins and other inflammatory mediators", according to the BBC.
Now, these results are interesting, not least because they were found serendipitously, as the result of looking for the possible effect of a daily dose of salicylic acid (aspirin) on vascular events, via reduction of blood clotting. And one or more of the investigators were following up on earlier findings (perhaps in the same study) that aspirin may be protective against colorectal cancer. Because environmental factors are correlated and highly variable, whether aspirin on its own would have persistent effects in changing lifestyle environments is an open question. People who have taken aspirin may be different in many unmeasured ways from others. But let's take this study as given.
The surprising thing is that this seems an out of the blue finding: why aspirin? One possible explanation suggested by the investigator was that cancerous cells respond to aspirin by undergoing apoptosis -- cellular suicide. Why normal cells would not do that as well is unclear, and whether there was any substantial evidence for the explanation (as opposed to just a guess), we can't say. However, by chance we were having dinner with a visiting cell biologist from South Africa, and he said this is likely the result of the much higher metabolic rate of tumor cells, so that the aspirin effect would affect normal cells as well, but a far lesser rate; that is, the aspirin effect isn't targeting the tumor cells specifically. Aspirin may affect vascular tissue (small blood vessels) in ways that deprive the 'hungrier' tumor cells.
A more important point to us is that we're in the age in which our NIH and other global research supporters seem determined to turn every disease into a genetic disease. Cancer does indeed seem to be a genetic disorder at the cell level--genes gone bad for some reason lead the cell to misbehave and grow out of normal control. Though many different means to alter gene function or expression in cells may be involved, they seem largely to occur somatically during life. That means that cancer isn't mainly 'genetic' in the sense of its causal mutations being inherited by children from their parents.
Some cancer susceptibility certainly is inherited. But the risk increases associated with most genes currently known to affect inherited cancer risk are far less than the 20-40% figures claimed in this aspirin study. This means that cheap aspirin prevention would be worth many times what we would be able to do using genotypes in expensive high-tech approaches. So why devote so much funding to search for 'cancer genes', rather than intensely figuring out what to do with the genetic variants that we know really do increase risk? We hasten to add as a case in point that such progress is certainly being made in this direction in at least some instances, including the notorious BRCA1/2 genes associated with breast/ovarian cancer.
More importantly, if we concentrated on those clear-cut cancer risk genes, while we let aspirin prevention take its course for many other common tumors, we would be left with the cancer cases that are more likely to be really genetic -- the ones that occur despite taking aspirin. Maybe that is where genetic research in cancer should go, rather than into huge, open ended black holes of long-term megagenomics projects.