Skip to content

Cancer and Luck (2/2/15)

February 2, 2015

Feeling Lucky?

Is getting cancer mostly a matter of luck? That was the spin on a a recent study by Vogelstein and his team at Johns Hopkins. Media stories led to furious counterarguments among scientists – before running back into media self-commentary. Yet the real lessons of this interesting, assumption filled paper have been widely neglected. They include 1. Take out lifestyle events like smoking, drinking and obesity – and elements of genetics – and most of the reasons for cancer risk remain unknown. 2. Cancer may well be a disease of incorrect regeneration. 3. The randomness of much cancer incidence argues for blanket treatment coverage. If we treat everyone experiencing “random” accidents, why not cancer?

So What Did the Study Really Show?

The study by well known cancer researchers looked mainly at statistical variability – how much a disease appears in a population based on factors you know – and the ones you don’t.

The authors focused on stem cells – the relatively undifferentiated stuff used to make tissues new. As most of your body gets replaced quickly, stem cells matter.

The plan was to look at estimates of stem cell regeneration rates – how fast different tissues are remade – to understand cancer rates. Data on breast and prostate cancer were left out. The authors felt the data on their stem cell regeneration rates were too lousy to use. Other researchers thought their guesses about other tissues not great, either.

But when looking at how fast stem cells regenerate – how quickly tissues remake themselves – the correlation with cancer rate variability came out a high .81. This was interpreted by the authors to mean using the statistical rule of multiplying correlation by itself – that approximately two thirds (.81 times .81) of the rates of tumor variability were “explained” by how fast stem cells remake tissues.

In other words – the more completely new cells you make, the more errors. The more errors – the more tumors. Tissues that are replaced in days – like gut lining cells – have a lot more tumors found that tissues that replace themselves slowly, over years.

This is like saying that a group of drivers traveling a hundred thousand miles is more likely to experience accidents than if they drove five thousand miles. It does not tell you about cause – like bad weather or faulty brakes. It only says statistical variability goes up – in a fairly predictable way – if you have to remake stuff more often.

What Does This Mean For Cancer Biology?

That despite over 40 years since the beginning of Nixon’s “War on Cancer,” we really have not majorly changed cancers rates, nor, for most tumors, ultimate survival statistics. We certainly know a lot more about cancers – including that the tissue types we’ve used to diagnose disease do not include the dozens of different tumors glimpsed on the molecular level.  We also know tumors rapidly mutate – effectively becoming different animals. That partly explains why most chemotherapies work for limited periods – the tumors change before our eyes, giving them a hydra headed capacity to block effective treatment.

One remedy that has been suggested – including by the Johns Hopkins authors – to look for earlier tumor detection. For now, that’s a pipe dream. As H. Gilbert Welch pointed out in “Overdiagnosis,” a lot of present day cancer screening is not supported by evidence. If for every prostate cancer identified and cured via the prostate serum antigen test – the PSA – you end up treating 30 to 100 people unnecessarily, you have not accomplished your goal.

With the growth of health monitoring equipment and “nanomedicine” expect to see widespread, sometimes uncontrolled screening. Beware. The American IT foray into medical care- with hundreds of different incompatible electronic health record systems that have nurses and doctors treat charts instead of people, and security systems sufficiently robust to allow your most intimate secrets to be hacked by an intelligent teenager – represents a small foretaste of what may come. The process can be done right – and generally is done far better in rationalized, nationalized, European health systems. But “personalized medicine” will only be ready for prime time when it improves the health of populations – not just corporate ledgers.

Lastly, we have to face two curious facts: that cancer is much related to how accurately our tissues regenerate themselves – and that many of these errors are indeed random.

Though not all. The biggest impacts on human health are still those of lifestyle, not medicine. Lifestyle adds far more to lifespans than medical care does.   To prevent cancer you don’t send people first for screening tests – you get them to walk, eat whole foods, drink less and  smoke not at all.

And if cancer, the most feared and soon to be greatest overall killer of humanity, is mainly random in who it destroys – whether it’s a two year old with acute lymphocytic leukemia or an accounting professor with brain cancer – how can we withhold treatment from people on the basis of poverty or employer? If someone is in a car crash we treat them. Why treat tumors differently?

Ultimately, the data on cancer tells us what we’ve known all along – that we’re all in the same boat. It’s incumbent on us to help everybody else in the boat, too – if only to save ourselves.

No comments yet

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: