The books I chose for my recent holiday turned out to be uniformly excellent, which doesn’t always happen. Another cracker was Siddartha Mukherjee’s [amazon_link id=”0007250924″ target=”_blank” ]The Emperor of All Maladies: A Biography of Cancer[/amazon_link]. Although detailed and authoritative on the medicine and science, and quite a chunky book, it reads like a thriller and is packed with fascinating stuff.

[amazon_image id=”0007250924″ link=”true” target=”_blank” size=”medium” ]The Emperor of All Maladies[/amazon_image]

One of the most interesting strands concerns the importance of randomized control trials in testing both causes of different cancers and potential treatments for it. Medical trials are expensive, fraught with ethical dilemmas, and slooooow. But failure to carry them out can maim and kill. For many years women underwent massively radical mastectomies that made no beneficial difference to their health and mortality. The surgeons were carried along on the crest of their own certainty and self-esteem. (In fact, the disconnect between cancer researchers and surgeons who treat patients is another interesting theme.) The Doll and Hill test of the link from smoking to lung cancer was so decisive because it assigned subjects to trial groups randomly, before they had developed the disease.

However, as Mukherjee notes, the campaigning by activists for people with HIV and AIDS to receive new drugs and combinations of drugs before trials had been completed – on the entirely understandable grounds that the patients were dying and so anything, even anything unproven, was worth trying – changed the environment for rigorous trials. We see this all the time now with new cancer drugs. Desperate patients learn very quickly from online forums that an experimental treatment is available, and each of them faces the choice between certain early death and a whisper of uncertainty. See this recent example described in The Guardian – although the article doesn’t state whether the pioneering treatment available in the US has actually been trialled there either).

Thank goodness the stakes in economic policy, increasingly looking at the methodology of randomized control trials, are not so high. I think Mukherjee makes a strong case for sticking with the methodology, and resisting the activism, although it would make sense to ask whether formal trials need to take so long or cost so much. After all, apart from slowing down results, they form a massive barrier to entry in the pharmaceuticals industry.

I could pick out lots of other terrific sections. The book describes Vannevar Bush’s view of research at the wartime Office of Scientific Research and Development – namely, that the government had to fund research directly serving the needs of war. Programmatic research was a much higher priority than basic research. And also his change of mind as soon as the war was over, writing to the President in 1945: “The striking advances in medicine during the war have been possible only because we had a large backlog of scientific data accumulated through basic research in many scientific fields in the years before the war.”

What else? The heterogeneity of cancers. The fact that humans and cancer are in an arms race because tumours develop resistance, and that a “war on cancer” is, like all metaphorical wars, unlikely to be won. That as early as 1910 a researcher (a Paul Ehrlich, although not that one) declared (wrongly) that he had found a ‘magic bullet’, a metaphorical weapon in the metaphorical war. A highly recommended book.

PS I’d no sooner posted this than I found this THES review of two new books on the tobacco industry’s (ab)use of areas of scientific doubt.