Little to show for 20 years' work

The success rate of genetic testing is poor, according to James Le Fanu. The promise was huge. All sorts of genetic disorders would be eliminated once we could pinpoint the genes involved, test a foetus to see if it was a carrier and then abort it. But look at the example of cystic fibrosis.

In Edinburgh, a ten-year trial has screened 25,000 people for the gene. They discovered that 22 prospective parents each had the fibrosis-linked gene, greatly increasing the risk that any child of theirs would develop the disease. In all, eight foetuses were found to have the lethal mutation and were aborted. That is a detection rate of 0.003 per cent, at a cost of Pounds 100,000 for each one detected. Meanwhile, the number of children born with cystic fibrosis has remained the same since the gene was discovered.

Gene therapy - the replacement of defective genes with normal ones - makes testing potentially even more worthwhile because it offers an alternative to abortion. But although all sorts of fancy effects have been achieved in the labs, since 1990, when a girl with a rare immunity problem was helped, there have been no other cures.

"Genes have turned out to be infinitely more complex and elusive then we ever imagined," Le Fanu says. "Sickle cell anaemia was the first disorder to be clearly linked to a gene but we now know that the way it works is almost unique in its simplicity. A change in just one letter of the genetic code, from GAG to GTG, means the wrong protein is made and so the blood cells are deformed. But in cystic fibrosis, for instance, 200 mutations can cause the disease but 200 more have no effect at all."

The effect of most genes "depends on what else is happening in the cell they are in and the genes linked to them", says Philip Gell, emeritus professor of genetics at the University of Birmingham. This degree of complexity creates a gap in our knowledge that is "unbridgeable".