Pharmacogenetics, uncertainty and ethics 9/12/03. By Adam Hedgecoe The ethics of pharmacogenetics are difficult to discern at present, argues Adam Hedgecoe.

Coming to the end of a three-year postdoctoral fellowship on pharmacogenetics funded by the Wellcome Trust's Biomedical Ethics section, one of the main themes coming out of this work is just how uncertain I am about the way in which pharmacogenetics will develop as a technology, and how in turn that technology will impact on healthcare, medical practice and society.

This lack of knowledge on my part is less to do with my own laziness or academic incompetence, and more to do with the inherent uncertainty that pervades pharmacogenetics and its newer sibling, pharmacogenomics.

Just take those two terms for example. Approach one scientist or pharmaceutical industry executive and ask them to define these two terms, and they will give you clear definitions, explaining how they differ with regard to technology used, the focus of research, and the ethical issues raised. Ask another expert and they may well tell you that there's no significant difference between them, that pharmacogenomics is merely a modern version of pharmacogenetics.

Such variations in definitions matter since ethical debate in this area often revolves around the idea that because pharmacogenomics is significantly different from traditional genetic approaches, it involves different (usually lesser) ethical problems. For example, a frequent claim is that pharmacogenomics is not the same thing as disease genetics. As pharmacogenomics focuses only on the way in which an individual may react to drugs, it tells us nothing about disease risk, and hence is free from all the ethical problems associated with disease genetics, the need for genetic counselling and restrictive genetic privacy.

While this is true for many cases of pharmacogenomics (or pharmacogenetics), there are a number of cases where the genes that indicate drug reaction are also implicated in disease risk. In these cases, what you tell a person about their genetic reactions to drugs will also tell them something about their disease risk, how the condition might progress and their relatives' chances of developing the same condition. The problem is that pharmacogenetics or pharmacogenomics are so broad, potentially covering such a range of conditions, drugs and genes, that it is hard to draw any ethical conclusions about the technology as a whole.

Another example that demonstrates the varied and broad nature of pharmacogenetics is the breast cancer drug Herceptin. Developed in the late-1990s, Herceptin is now in regular clinical use in the National Health Service. It is used to treat women whose breast cancers have too many copies of a gene called HER2, approximately one-third of all cases.

Although a form of genetic testing can be used to find out whether a tumour has too much HER2, doctors do not regard HER2 testing like other genetic tests; HER2 status tells us nothing about that person's genome, since it is just finding out something about a person's tumour. In this case, pharmacogenetics is far more like other treatments currently in use (like Tamoxifen) rather than a revolutionary approach to healthcare.

When we think about the ethics of a new technology like pharmacogenetics, we have to see it in context. It has far more in common with other, older, technologies than people might think, and can raise familiar ethical issues. Rather than vague generalisations about this diverse and varied technology, we need detailed explorations of individual examples, to try and highlight what problems really are new.

Dr Adam Hedgecoe is a lecturer in Sociology at the University of Sussex.

Image credit: Neil Leslie