However, the survey also revealed that applications of PGD have broadened to include choosing the baby's sex, as well as screening for diseases that may not strike until late in life, if at all. Non-medical sex selection accounted for 9 per cent of PGD tests conducted in the US in 2005, while 28 per cent of the clinics surveyed are now offering PGD to detect a propensity for adult-onset diseases such as Alzheimer's, Huntington's, hereditary breast cancer, and many other cancers.
PGD was originally intended to overcome infertility and to ensure that all children are born with achance of a full and flourishing life, says Eric Cohen, head of the bioethics programme at the Ethics and Public Policy Center in Washington DC. "PGD clinics are performing sex selection just to satisfy parental preferences, and embryo screening is being used for late-onset diseases that have nothing to do with childhood sickness," Cohen claims. He sees these as troubling trends.
The changes are taking place because there is little consensus on which tests are appropriate for PGD, and no federal or professional oversight of the clinics, according to Susannah Baruch, who led the GPPCstudy . "There are more than 1000 genetic tests currently available and there is no reason why any of these couldn't be used for PDG," she says.
In the UK, PGD is regulated by the Human Fertilisation and Embryology Authority. Its spokesman John-Paul Maytum says sex selection for nonmedical purposes is not permitted in the UK because the majority of the public are against it.
The GPPC survey, which was presented at the National Genetic Policy Summit in Washington DC last week, is the first to focus specifically on the uses of PGD in IVF clinics. It revealed that almost one-quarter of US clinics use PGD to select an embryo that would be an immunological match for a sick sibling. It also showed that PGD is not error-free: 21 per cent of clinics interviewed for the study knew of cases where children had been born with diseases that PGD was supposed to eliminate.
One reason for this is that all eight cells of the embryo may not be genetically identical, due to a condition called embryo mosaicism, making it possible that the cell taken for testing may not describe the whole embryo's genetic status. Exactly how many misdiagnoses have occurred is not known, but Baruch says the survey underscores the need for better data collection and regulation to ensure the safety and efficacy of the procedure.