Of genes and genomes

Doctors often prescribe medicines that are not adequately tailored to the individual. There is not much we can do to avoid this with conventional treatments. But with the human genome now mapped, some are predicting a new era of “personalised medicine”, in which patients can have their genes matched to the best treatment.

It is a nice theory, but I don’t buy it. We may now know the ingredients of the genome, but it doesn’t mean we can cook up whatever we please. For example, we are able to identify genetic foetal disorders in pregnancy, such as Down’s syndrome, but the only “treatment” remains abortion. There remains a lack of true genetic “therapies” – where someone can have a treatment which manipulates their genetic material. Trials with conditions such as Parkinson’s disease or rheumatoid arthritis have offered some hope that genetic treatment could be provided – but to date human data has been obtained on a very small scale.

The other major problem is that most genetic information does not give us clear-cut answers. Genes will tell you what you may be at increased risk and decreased risk of. These risks do not operate in isolation. If a smoker has a particular genetic sequence putting him or her at lower than average risk of cancer, smoking is still going to be riskier than not smoking.

Other doctors have pronounced the genome as the answer to getting better cancer therapies to those most likely to benefit. But this is also far messier than we might expect. HER2 genes, on breast cancer cells, have been celebrated as something that science can pinpoint to save lives. Having HER2 means that a woman is more likely to respond to medication such as trastuzumab (Herceptin). However, the test for HER2 is itself suboptimal, with the amount of cells picking up a stain in a petri dish used to judge if this genetic protein is present. Various new sets of guidelines have tried to minimise error but it is not a test which always gives true results. Even then, only between 25 and 30 per cent of those with the HER2 gene will respond to the drug targeting it. It may be a useful treatment, but is hardly the revolution we were promised.

My worry is that the use of genetic risk assessments to direct treatment will produce more dilemmas rather than less. Many of these conundrums will rest on our ability to assess and judge our risk of disease – and our abilities to prevent it. If anything is meant by the term “personalisation”, it should be our own view on how to deal with the risks we know we are running, and far less about our genetic code.

Margaret McCartney is a GP in Glasgow

margaret.mccartney@ft.com

To follow Margaret’s blog go to: www.margaretmccartney.com/blog

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't cut articles from FT.com and redistribute by email or post to the web.