Scat CAT! Get the picture?

A story on the Dec 15th NPR site (Radiation From CT Scans May Raise Cancer Risk) provides a discussion of the risks of cancer induced by the proliferating use of CT scans in medicine. The estimate is that as many as 29,000 future cancers may be caused by the 72 million CT scans done in 2007 alone, and as many new cases induced each coming year if practices don't change. The story also presents a graph of these possible cases (and about 15,000 of them estimated to be fatal).

It's sometimes said, reassuringly, that the amount of radiation a patient receives from one chest scan is equivalent to the natural radiation exposure from one transcontinental flight, but the amount of radiation can vary so much per scan that some are equivalent to 500 transcontinental flights. Or even the dose received by survivors of the atomic bombs dropped on Hiroshima or Nagasaki. And those caused, and are still causing, increases in cancer rates in those survivors (more or less so, depending on distance from the epicenter). This is no joke, if you get the picture.

We've said many things before about competing risks, and this is an excellent example. It is directly comparable to the recent controversies about mammography, though (as we've discussed in previous posts) the recent controversy has been about over-detection of tumors that would regress on their own, rather than about the risk of x-ray induced cancer, but that was a major reason why the recommendation has been not to start until menopausal or post-menopausal years.

A major part of our book is devoted to describing how life is characterized by partially sequestered environments -- in this case, cells -- that respond to each other and behave as a result of signaling and related processes. Radiation damages DNA and can cause a gene to be mutated and or misexpressed, altering the cell's function. Our cells have DNA-repair mechanisms that can detect changes under some conditions (e.g., when the two strands no longer match and fit closely together because one of the corresponding nucleotides has been altered).

But if the cell doesn't detect and repair the change, the cell permanently is on a misbehavior track. This can be especially dangerous when the mutation is in the genes responsible for DNA repair itself.

Since a tumor cell, unlike a bacterial cell, is part of you, your immune system may not be able to tell that it's not doing its proper job. It or other tissue-order-retaining mechanisms may thus fail to maintain the proper order. The tumor cell divides, and divides, and divides: it is an internal engine of life that is no longer receptive to external signals, and the external police can't detect it.

But the whole picture is more than a CT inset. If CT is useful, it's presumably because it detects serious disorders. We know now that a fraction of those would not actually need to be treated, as in the case of mammography and other diagnostic tests. That means needless cost and morbidity. On the other hand, if the CT scans do lead to effective treatment of serious disease, then people will live longer. This means that some of them will get cancer (unrelated to the CT radiation exposure) or other diseases of aging, as well as the iatrogenic (doctor-induced) cancers. So there's no simple winning or losing here.