Pre-implantation Genetic Screening (PGS): What are we really talking about?
The tenant behind pre-implantation genetic screening (PGS) is to biopsy one or few cells from each embryo after creation, analyze the chromosomes for each embryo and transfer the ones that has normal chromosomes back into the uterus to boost IVF success and increase the live birth rate.
Central to this idea is that abnormal chromosomes in the embryos is the main reason why an embryo does not yield a newborn. It is logic then that PGS should allow the selection for the best embryo (preferably one only) for transfer into the uterus ending into one singleton newborn.
If this premise is accepted then the following assumptions should also be generally accepted
a. All or the majority of embryos reached the appropriate stage of development and expansion to allow biopsy.
b. Biopsy of the embryo does not harm its ability to implant
c. The cell or few cells obtained represent the rest of the embryo (has identical chromosomes to all the other cells in the embryo)
d. The platform used to analyze the embryo chromosomes is close to 100% accurate (otherwise some embryos will be wasted because they are abnormal according to the test, while they are actually normal). The platform reports only the chromosomes of the embryo and is not accountable for other elements of implantation i.e. the endometrium.
e. The delay (one or more days) needed to finish the testing does not affect embryo implantation
f. Freezing and then thawing of a biopsied embryo does not affect its implantation potential
g. Patients and physicians have agreed on how to calculate success: how many live births one would obtain from all embryos resulting from a single IVF cycle (all fresh and frozen embryos) i.e. total potential of one IVF cycle versus fresh embryo transfer only.
h. The added cost of biopsy and testing of embryos, potential increases the delivery rate and reduces the incidence of multiple pregnancy and miscarriage is cost-effective from the viewpoint of individual and a modern society.
The initial attempt to perform Preimplantation genetic diagnosis using an old technology called FISH that tested 7 to 9 chromosomes proved harmful few years ago and that its wide adoption at that time was a form of medical illiteracy : because it depends on logic not actual well conducted study. When the studies were conducted, they all showed that women universally achieved lower pregnancy rates after PGS.
New platforms are now available to test for all the chromosomes (array cGH and SNP array) and using cells (trophoectoderm) obtained from more advanced stages of the embryo (blastocyst). The question in hand is should we adopt these techniques, not as a research tool, but as the standard of care that should be offered to the majority of women undergoing IVF?
How Effective is PGS? The case for Logic
Applying logical thinking to modern pre-implantation genetic screening (PGS) methods indicates:
a. Not all embryos will reach the blastocyst stage (day 5) to be suitable for biopsy. Not all physicians and patients push their embryos to the blastocyst stage especially if few embryos exist in culture on day 3. Moreover, some normal embryos may not survive extended culture to blastocyst.
b. There are no conclusive evidence that biopsy of the trophoectoderm (the part that makes the placenta) of an embryo does not harm the embryo.
c. Mosaicism ; when one or few cells are different in chromosomes than the rest of the cells, is known to take place in embryos. The cells in the trophoectoderm maybe abnormal while the cells in the embryo maybe normal. Interestingly the embryo can later get rid of the abnormal cells in the trophoectoderm. This can lead in misdiagnosis of the embryo as abnormal while the embryo itself has the potential to implant and yield a healthy baby.
d. The platform used to analyze the embryo chromosomes is not 100% accurate either because of the accuracy of the test itself or because of mosaicism. The accuracy reported by labs administering the test is 97%. This means some normal embryos will be discarded and some abnormal embryos will be transferred. Actually the accuracy was not validated by many labs, only very few worldwide. Clinically some physicians have experienced much lower accuracy (80 or 90%). The platform reports only the chromosomes of the embryo and is not accountable for other elements of implantation i.e. the endometrium. So it is possible that the lower accuracy is due to other elements on embryo gentics (other than the number of chromosomes) or the lining of the uterus.
e. Currently the transfer of embryos into the uterus has to be delayed for one day (day 6) or several weeks (embryo has to be frozen then thawed back after results are obtained). This delay may reduce implantation of the embryo because it will not match the window of implantation in the lining of the uterus. This is a controversial point as some researchers found no difference in implantation between day 5 and day 6. This research, however, is not widely replicated.
f. After PGS some ‘normal’ embryos will be frozen. The survival of thawed and biopsied embryos is maybe reduced, potentially leading to loss of normal embryos. No large studies on survival of biopsied embryos after thaw exist.
g. Patients and physicians have agreed on how to calculate success: if success is calculated based on how many live births one would obtain from all embryos resulting from a single IVF cycle (fresh and frozen) i.e. total potential of one IVF cycle, then PGD has no value as it will not make an abnormal embryo normal or vice versa. If the success is based on what happens in the fresh cycle only with no regard to frozen embryos then PGS may improve the success rate of IVF. All assuming an excellent embryo freezing program.
For exampe If you are a young woman <38, with a good number of available embryo on day 5, say 4 blastocysts that are suitable for biopsy, you may elect to
i. transfer one embryo in the fresh cycle and freeze 3 embryos. If you are not pregnant, then transfer one embryo in each subsequent frozen cycle. If you are destined to get pregnant you will do that within a maximum of 3 months after your initial IVF and the risk for multiple pregnancy is minimized to 1% or less. If you were not destined to get pregnant no testing would have helped you or
ii. Alternatively, you may elect to test all your embryos in the fresh cycle, transfer one normal embryo, if any and freeze any normal embryo remaining. The potential benefit is getting pregnant in the fresh cycle instead of getting pregnant 1-3 months later. Also you will reduce the risk of miscarriage because abnormal embryos will likely be eliminated. The potential risks are misdiagnosis by PGS (not 100% accurate), loss of a thawed embryo (did not survive biopsy and freeze) and lower implantation potential of a normal embryo due to biopsy and delayed transfer.
h. A cost-effective analysis for PGS is not available at this time. The added costs are biopsy and testing of embryos. The potential benefits are increase in the delivery rate and reduction in multiple pregnancy and miscarriage. In the scenario above you either pay for i. frozen embryo transfer(s) if you do not get pregnant in the fresh cycle or ii. pay for ICSI (required for PGS by the majority of programs), biopsy and testing in the fresh cycle and frozen embryo transfer(s) if you do not get pregnant in the fresh cycle. In terms of multiple pregnancy, it can be minimized in either pathways if your physician is transfers one embryo anyway, tested or not. Things are not that simple, the payer will also make a difference: PGS is completely out of a patient pocket as it is not covered by any insurance while frozen embryo transfer may or may not be covered.
How Effective is PGS? The case for Published Studies
In general decision making in biological sciences is not amenable to logic, but determined by well designed ad well conducted studies. So far, three studies were published using the new platforms for embryo chromosome analysis, aiming at increasing IVF success. The studies were criticized because of
1. Restricted to young women (median age 31 to 32) so results cannot be generalized to the general IVF population: 2 studies
2. Did not account for frozen embryos: all studies
3. The studies did not demonstrate superiority of PGS to transfer best embryos based on morphology (shape): one study. Specifically a transfer of a tested embryo in the fresh cycle was not inferior to transfer of two untested embryos. Non inferiority does not mean superiority. Noninferiority study design is not suitable for a PGS study as patients and physicians are only interested in such an expensive treatment that can harm their embryos only if it promises superior results for their infertility treatment. Moreover, treatment could actually be inferior because a limit is placed that will make the outcome non inferior, in that study 20%. So if the difference is less than 20% PGS is considered not inferior.
4. End point should be live birth or ongoing pregnancy. Surrogate or intermediate endpoints as pregnancy, implantation (short of a baby in hand or at least pregnancy beyond 20 weeks) are not ideal outcomes.
Randomized studies related to pre-implantation genetic testing using newer platforms were independently analyzed. So far no study showed that PGS is superior to the strategy of transferring the best embryo based on morphology (the standard of care). Moreover due to factors related to the biology of reproduction and that the accuracy of the test is unlikely to reach 100% accuracy soon, it is unlikely that PGS will prove beneficial to women undergoing IVF for fertility treatment. PGS may only shorten the time to pregnancy but will not be able to improve the pregnancy rate and due to inaccuracies may even reduce it.
Alternatives to PGS are being studied. One alternative is time lapse photography of the embryos to observe the cell division of the embryo cells and select the best embryo for transfer. It is noninvasive but further studies are required before its ready for general use. Another alternative is polar body biopsy of oocytes but results of ongoing studies are not available yet.
It is possible that factors in this article could be interpreted differently in a specific situation by patients and their physicians, in conjunction with the number of mature eggs produced, but it does not appear that PGS is ready for generalized application in the majority of IVF population.