The principle of preimplantation testing for sporadic chromosome aneuploidy (PGS) (preimplantation genetic screening) to improve the reproductive efficiency of assisted conception for couples at increased risk is a sound one. It is a reasonable expectation that testing eggs and embryos and excluding embryos with chromosome aneuploidy from transfer should increase the proportion of transferred embryos that implant, decrease the risk of spontaneous abortion and increase the chance of taking home a healthy baby. However, we would contest the recent assertion by Handyside and Thornhill (BioNews 385, 20 November 2006) that the clinical application is generally accepted to be clinically sound. A comprehensive review of the evidence, by Shahine and Ceders (2006), concluded that randomised controlled trials with large patient groups by expert practitioners in PGD technology are still needed before testing can be routinely recommended for assisted conception patients presenting with advanced maternal age, recurrent pregnancy loss, and recurrent assisted conception failure.
We should be asking why, given that the principle is sound and after ten years and many thousands of cycles world-wide, the evidence in favour of testing is not overwhelming? We believe that the answer is unfortunate and simple; this expensive additional test for just one factor causative in reproductive failure is just not good enough even in the best hands, to make a large enough difference to the clinical outcome for conditions that are in their nature multifactorial. At present, the technology used to exclude chromosome aneuploidy is restricted to selection from a limited population of eggs or embryos. In some countries, national legislation may further limit this technology to testing polar bodies, which can only assess the female contribution to the zygote. Cleavage-stage blastomere biopsy assesses the genetic contribution of both reproductive partners but at a stage where the embryo's cells are likely to be at their chromosomally most non-uniform (mosaic), which confounds the principle that the cell sampled is representative of the whole embryo.
At present, FISH (fluorescence in situ hybridisation) is the technique of choice for testing; the limitations of the FISH technique make it inevitable that a significant proportion of healthy eggs or embryos will produce false abnormal test results and thereby be excluded from consideration for transfer and destroyed. It is also a perverse irony that by testing embryos it is unlikely, but possible, to 'cure' the embryo by removing one or more aneuploid cells, accurately diagnose the cell(s) and then discard the embryo; or by removing one or more euploid cells from a mosaic embryo, exacerbate the proportion of abnormal cells in an embryo that might be transferred. The capacity of embryos to sequestrate abnormal cells is unclear, but the frequency of finding 'normal' blastocysts or normal stem cell lines following exclusion as abnormal after aneuploidy screening is of concern.
Furthermore, diagnosing a substantial number of embryos as abnormal reflects on the constituents of that cycle only, and it would be irrational to assume that it reflects that all future embryos produced by that patient will be abnormal, even if heavily skewed in that direction. Many older patients, even with only one normal embryo still achieve pregnancies without PGS, and younger women who produce good numbers of embryos are also likely eventually to achieve a successful pregnancy. Similarly even those fertile women who have miscarried more than twice without obvious cause, stand a better chance of a livebirth without any intervention at all. Helping 'closure' based on specious aneuploidy testing results would seem unethical.
The future will no doubt offer yet more exciting technological and treatment prospects for assisted conception, but we agree with Handyside and Thornhill that we need to manage the expectations of physicians and those of our patients more carefully, rather than expect miracles from preimplantation genetic testing for sporadic chromosome aneuploidy using currently available techniques.
Shahine LK, Cedars MI. Preimplantation genetic diagnosis does not increase pregnancy rates in patients at risk for aneuploidy. Fertil Steril. 2006 Jan;85(1):51-6.