Jan 05 2015

Cancer Risk Largely Bad Luck

This is one of those glass-half-full / glass-half-empty news items. Different headlines reporting on the same study present the results in opposite ways. The BBC, for example, writes, “Life choices ‘behind more than four in 10 cancers.'” Meanwhile the press release from Johns Hopkins states, “Bad Luck of Random Mutations Plays Predominant Role in Cancer, Study Shows.”

The Hopkins headline is more accurate. The BBC headline is not just focusing on life choices vs bad luck, it actually gets the data wrong.

What the researchers did was look at the replication rate of stem cells in different tissue types (they did not include breast cancer and prostate cancer as they could not find published replication rates). They then compared differences in these rates to differences in adult cancer rates in the same tissues. For example, colon cancer is more common than cancer of the small intestine, and colon cells replicate more frequently than small intestine tissue. (In mice, this is reversed, but the correlation holds.)

All cancers, the authors emphasize, are a result of random mutations, lifestyle factors, and inherited genes. The latter two can affect the rate of random mutations and perhaps also the probability of cancerous cells being destroyed by the immune system vs growing into a cancer. The mutations result in cancer by altering the cell’s growing characteristic, allowing them to grow without limit. Normally cells posses mechanisms that keep them from growing out of control, and mutations may disable one or more of those mechanisms.

The authors hypothesize that if getting unlucky random mutations were the sole cause of cancer, then 100% of the difference in cancer rates among various tissue types could be explained by variation in their stem cell division rates. Statistically this means there would be a complete correlation between these two variables (a correlation of 1.0). They found a correlation of 0.804. This translates into 65% of the difference in cancer rates being explainable by differences in stem cell mutation rates.

The correlation was not the same for all tissue types. Some tissues had a very tight correlation, meaning cancers of those tissues are almost entirely due to random mutations. Other tissue types had a low correlation, the obvious one being lung cancer, which is largely caused by smoking. Of the 31 tissue types they looked at, 22 had a strong correlation between cancer rates and stem cell division rates. Those that did not were basal cell skin cancer (sun exposure is a major risk factor), head and neck and lung cancer (smoking is a risk factor), hepatocellular (liver) cancer, thyroid cancer, and several types of colon cancer.

The study, however, did not look directly at the contributions of inherited genetics and lifestyle factors, which is why the BBC headline is wrong. They assumed that not bad luck equaled lifestyle, but it equaled lifestyle plus genetics, which were not separated by this study.

The two cancer types not included, breast and prostate, are not largely lifestyle cancers, and so once they are included I suspect the percentage of overall cancers that are due to random mutations will increase.


This is an interesting study and it will be interesting to look at replications and other methods, if they are available, of making the same sort of estimation. What this study suggests is that at least 2/3 of all all cancers are due to random mutations – bad luck. The figure may be higher once breast and prostate cancers are included. Of the remaining third it is not clear how much is due to inherited genes vs lifestyle factors.

The logic of the study is sound, in my opinion. The authors assume that lifestyle and genetic factors affect the risk of tissue-specific cancers, but not cancer in general. This study would miss, however, lifestyle or genetic factors that affected the risk of all cancers (regardless of tissue type) equally. One might argue, therefore, that it overestimates the role of random mutations, but that is only if you accept that there are universal risk factors out there.

Of course we want to focus on the lifestyle factors, because that is the one thing we can control. If you avoid smoking, avoid excess sun exposure, avoid excess alcohol, and have a generally healthful diet, you are probably covering the vast majority of lifestyle factors that affect cancer risk. That is some comfort.

But still the majority of cancer risk is something we cannot do anything about (our own genetics, and just being alive). The researchers emphasize from this that we therefore need to focus our attention on early detection and treatment of cancer.

The good news is we are making progress. The latest figures show that death from cancer has declined by 22% since its peak in 1991. This translates into 1.5 million cancer deaths avoided if rates had remained at their peak level.

Also, 27% of cancer deaths are due to lung cancer, which is strongly linked to smoking. Therefore, if you avoid smoking then you are avoiding most of the lifestyle-associated cancer risk. Breast, colon, and prostate are also among the most common cancers, with these four representing half of all cancers. Breast cancer deaths have decreased by 35% and colon and prostate by 47%.

Hopefully these trends will continue, with improved detection and treatment of various types of cancer. Meanwhile, if you want to reduce your own risk, then don’t smoke. Avoiding excess sun is also important, but not nearly as large of a risk factor. There are many reasons to have a generally healthy diet – well rounded with plenty of plants. Just as important, however, is to follow screening guidelines. See your doctor and follow their recommendations.


David Gorski has written a much more detailed analysis at Science-Based Medicine that I highly recommend. He correctly points out that the study is not actually a measure of what causes cancer, but makes various assumptions to estimate the correlation of variability in stem cell replication rates and cancer rates by tissue type. The authors then assume that the proportion of cancer types that correlate with stem cell division rates are largely due to random mutations in the replication process.

While this is not literally the same thing, it is not an unreasonable assumption, and David points out that their result is in range with other estimates of the percentage of cancers that are preventable.

65 responses so far