A hurdle on the road to personalised cancer treatment

The new Nature is out, (since a couple of days, in fact) with a nice piece on the difficulties facing the development and approval of biomarkers-based cancer diagnostic assays.

Quickly, they work by taking a broad look at the proteins (or RNA, or else) expressed in your body at the moment, and compare them with similar sample in healthy and sick people. if the fingerprints (protein prints, or RNA-prints, or else-prints) match, then there's a good chance that you may share the same medical condition.

Trouble is, these tests do not yet seem able to differentiate between different kind of cancer enough to be useful in suggesting a treatment. To do so, large, long, expensive clinical trials are necessary. And the companies that produce those diagnostics do not have that kind of money. Pharmaceutical companies do, and they also have some interest in this: with clear diagnoses and treatment indication,s their drugs could be given only to patients likely to responds. Of course, this means that they would sell more or less drugs than they currently do. It can go both ways.

However, there are good chances. If it is true that the UK government will soon require certainty of effect on drugs before re-imbursing them to the company, then such a test would act as a shield in those cases where the drug were not to work nonetheless. I guess a middle ground will have to be found, with the government accepting a certain rate of failure in the prediction of the treatment and therefore shouldering the price of ineffective drugs rather than unloading it onto an already unstable pharmaceutical complex. Whether you like them (us) or not, the world needs new medicine and that's the most efficient way to create them.