4.5.17

ECONOMETRIC ONCOLOGY?

A long time ago, I learned of the Ames test for mutagenic chemicals, in of all places, The Washington Monthly, and what struck me about the article were the similarities of that test to checking a time series for Granger causality, which was a popular thing in economics at the time, and it's still with us, if in more recondite forms.  The part that struck me was about the Ames test picking up changes in the history of mutations given the history of exposure to the possible carcinogen.  To quote the Wikipedia "intuition" of causality testing,  "We say that a variable X that evolves over time Granger-causes another evolving variable Y if predictions of the value of Y based on its own past values and on the past values of X are better than predictions of Y based only on its own past values."  That led me to speculate that something random was generating cancers, and exposure to some chemicals made the random variable more active.  At the time I mentioned this speculation to some people I knew who were doing cancer research, although we never had the opportunity to chat about the similarities, or not, face to face.

Since then, Granger causality has made its way into neuroscience.

More recently, cancer researchers have suggested that yes, there is a lot of randomness in cancer, and relatively few cancers respond to chemical exposures.
Random mutations may account for two-thirds of the risk of getting many types of cancer, leaving the usual suspects — heredity and environmental factors — to account for only one-third, say the authors, Cristian Tomasetti and Dr. Bert Vogelstein, of Johns Hopkins University School of Medicine. “We do think this is a fundamental mechanism, and this is the first time there’s been a measure of it,” said Dr. Tomasetti, an applied mathematician.
Smoking, or chewing tobacco, or working in a chemical plant, though, is rolling loaded dice.
Smoking greatly increases the risk of lung cancer, but for other cancers, the causes are not clear. And yet many patients wonder if they did something to bring the disease on themselves, or if they could have done something to prevent it.
Messrs. Tomasetti and Vogelstein have continued their research, to make their statements more precise, but the randomness inherent in cell division remains the prime suspect in many cancers.
In the case of pancreatic cancer, the researchers say that 77 percent of the mutations that cause tumours are due to random DNA copying mistakes, whereas 18 percent are down to environmental factors, with inherited genes accounting for the remaining 5 percent.

In other cancers – including prostate, brain, and bone cancer – the researchers found that 95 percent of cancer mutations are a result of cell division errors.

For some cancers, though, bad luck appears to play a smaller role.

In lung cancer, 65 percent of mutations are due to environment – such as smoking, or living in a polluted area – whereas DNA copying errors account for the other 35 percent, with heredity playing no role.

Overall, the team estimates that 66 percent of cancer mutations are due to random, unavoidable mistakes made during cell division, with 29 percent being attributable to environment, and 5 percent being inherited.
But don't roll those loaded dice: 40 percent of cancer cases can be prevented by avoiding exposure to mutagens.
Smoking was "far and away" the lifestyle factor most related to cancer, the study showed, contributing to 23 percent of cancers in men and nearly 16 percent in women. The rest of the factors differed between the genders.

Not eating enough fruits and vegetables contributed to 6 percent of cancers in men and occupational exposure to chemicals (e.g. asbestos) was tied to 4.9 percent of male cancers. Rounding out the top five risk factors were alcohol (4.6 percent) and obesity (4.1 percent).

Being overweight or obese on the other hand, was tied to 6.9 percent of cancers in women, followed by infections (3.7 percent), exposure to sunlight/tanning beds (3.6 percent) and lack of fruits and vegetables (3.4 percent)

"We didn't expect to find that eating fruit and vegetables would prove to be so important in protecting men against cancer," Parkin said in the statement. "And among women we didn't expect being overweight to have a greater effect than alcohol."

Overall, the researchers found 45 percent of all cancers in men could be prevented, along with 40 percent of all cancers in women.
That 66-29-5 stratification of primary causes? They're still primary causes. Interaction effects present opportunities for further research.

Drink your orange juice and eat your lettuce.

No comments: