Preimplantation genetic testing for aneuploidy versus morphology as selection criteria for single frozen-thawed embryo transfer in good-prognosis patients: a multicenter randomized clinical trial

For women with at least two blastocysts able to be biopsied, NGS-based PGT-A did not improve pregnancy outcomes after the first embryo transfer in all women, but did for women aged 35–40 years.

Like Comment
Related Content

Volume 112, Issue 6, Pages 1071–1079.e7


Santiago Munné, Ph.D., Brian Kaplan, M.D., John L. Frattarelli, M.D., H.C.L.D., Tim Child, M.D., Gary Nakhuda, M.D., F. Nicholas Shamma, M.D., Kaylen Silverberg, M.D., Tasha Kalista, M.A., Alan H. Handyside, Ph.D., Mandy Katz-Jaffe, M.D., Dagan Wells, Ph.D., Tony Gordon, Ph.D., Sharyn Stock-Myer, Ph.D., Susan Willman, M.D., on behalf of the show STAR Study Group 



To evaluate the benefit of next-generation sequencing (NGS)–based preimplantation genetic testing for aneuploidy (PGT-A) for embryo selection in frozen-thawed embryo transfer.


Randomized controlled trial.


Not applicable.


Women aged 25–40 years undergoing IVF with at least two blastocysts that could be biopsied.


Randomization for single frozen-thawed embryo transfer with embryo selection based on PGT-A euploid status versus morphology.

Main Outcome Measure(s)

Ongoing pregnancy rate (OPR) at 20 weeks' gestation per embryo transfer.


A total of 661 women (average age 33.7 ± 3.6 years) were randomized to PGT-A (n = 330) or morphology alone (n = 331). The OPR was equivalent between the two arms, with no significant difference per embryo transfer (50% [137/274] vs. 46% [143/313]) or per intention to treat (ITT) at randomization (41.8% [138/330] vs. 43.5% [144/331]). Post hoc analysis of women aged 35–40 years showed a significant increase in OPR per embryo transfer (51% [62/122] vs. 37% [54/145]) but not per ITT.


PGT-A did not improve overall pregnancy outcomes in all women, as analyzed per embryo transfer or per ITT. There was a significant increase in OPR per embryo transfer with the use of PGT-A in the subgroup of women aged 35–40 years who had two or more embryos that could be biopsied, but this was not significant when analyzed by ITT.

Clinical Trial Registration Number


Read the full text here.

Fertility and Sterility

Editorial Office, American Society for Reproductive Medicine

Fertility and Sterility® is an international journal for obstetricians, gynecologists, reproductive endocrinologists, urologists, basic scientists and others who treat and investigate problems of infertility and human reproductive disorders. 


Go to the profile of M. Blake Evans
about 2 years ago

Very interesting work on a hot topic in our field. In light of the authors acknowledging the difficulties of large-scale multicenter RCTs and the importance of standardizing diagnostic tests: I'm interested to know what the cut off was amongst the labs when classifying embryos as segmental mosaic vs normal?

Go to the profile of Prof Alan Handyside
about 2 years ago

The genetic labs involved in the study set their own criteria for calling samples euploid/aneuploid, using the software as a guide. For intermediate ('mosaic') and segmental copy number changes, similarly, they used their own criteria. In 2014-16, there were few data on the clinical relevance of these changes and reanalysis of the study data shows there was variation between labs on the criteria they used. For example, at least one of the labs was clearly risk averse and called all segmental changes as aneuploid. Fortunately, in practice, samples with only segmental changes are relatively rare. Copy number estimation in segmentals is less accurate depending on the size of the segment involved. So referal to a genetic counsellor about the specific risks is advisable when considering the embryo for transfer.

Go to the profile of M. Blake Evans
about 2 years ago

Excellent, thank you for the elaboration.

Go to the profile of Luis Hoyos
about 2 years ago

This is an enlightening study with important implications in the way we practice. I would be interested in knowing the authors', and everybody else's, opinion about the role of PGT-A in our field moving forward.

Go to the profile of Prof Alan Handyside
about 2 years ago

In principle, there can be no clinical justification for transferring non-viable aneuploid embryos for infertility treatment. If we had a non-invasive, low cost and accurate method of identifying these embryos, I believe we would apply it to all IVF patients in the same way that we perform pronucleus checking in every cycle. However, we are not there yet and this study demonstrates that women in their late 30s are most likely to benefit from NGS based PGT-A.

Go to the profile of Samuel Santos-Ribeiro
about 2 years ago

Very interesting results from this very much anticipated study! I am curious to know if the authors are considering a post-hoc follow-up of cumulative live birth rates? Such could bring into light the true importance of mosaic embryos for overall ART success.

Go to the profile of Prof Alan Handyside
about 2 years ago

The problem with comparing cumulative live birth rates is that it requires a long term study and patients rarely transfer all of their stored embryos. (It would also be very difficult for this large multicentre study). So I agree that this would be the definitive way of studying the contribution of mosaic embryos. But ongoing efforts to report the outcomes following transfer of mosaic embryos is also an effective strategy as long as patients understand the risks and give appropriate consent.

Go to the profile of Pandiyan  Natarajan
about 2 years ago

Interesting study and paper. I have been following the development of Preimplantation genetic diagnosis from 1987. The technique which was introduced to avoid the birth of  an  abnormal baby was slowly extended for other unproven indications. PGT-A as it stands today is really on very shaky ground. The extrapolation from PGT-M and PGT-SR to PGT-A for all was an overkill with considerable damage to even the existing indication. The paper clearly proves the futility of PGT-A for all. Are there really any indication for PGT- A? With increasing trend towards Single Embryo Transfer, the technique would soon be phased out and avoiding known genetic defects would remain the only indication.

Go to the profile of Luis Hoyos
about 2 years ago

I understand your point but I don't entirely agree. I do agree that it was widely adopted too soon resulting in an overkill just like most new technologies in our field, nevertheless this study only proves that in "good prognosis" patients PGT-A offers no advantage, but how about patients with "poor prognosis"? 

Overall, I do think that PGT-A has a role but not the current one. We are using a screening test as a diagnostic test and that is the main problem I have with it. I believe that PGT-A (and may be we should change the name again to PGS to emphasize that after all it is just a screening test) should be used to prioritize embryo transfer in "poor prognosis" patients but embryos should not be discarded based on the results. PGT-A will not phase out but I think it will transform. I can imagine a future where PGT-A is done non-invasive and the result used along with other parameters to prioritize embryo transfer in "poor prognosis" patients. I do agree with you about PGT-M and PGT-SR continuing to play a crucial role in the future.

Go to the profile of Pandiyan  Natarajan
about 2 years ago

Thank you very much for your valuable feedback and the points raised.  

In poor prognosis patients also PGT- A is unlikely to improve pregnancy rates over the long term. Transfer of 2 embryos is more likely to improve pregnancy rates with the attendant risk of twin gestation. . Sequential transfer of available embryos- 1 in the first cycle and 2 in subsequent cycles is likely to be of greater benefit than PGT-A.

 PGT-A has many inconsistencies; there is confusion about the definition of aneuploidy and mosaicism ( current definitions are arbitrary) and the recommendation to discard aneuploid embryos is even more worrisome. Cells in the inner cell mass, blastocyst fluid and trophoblast are not the same and do not have identical karyotype. Mitotic errors may have crept in trophoblast cells which are not reflective of the cells of the inner cell mass.

As for the terminology, all was well when it was PGD. The test was designed to primarily diagnose certain genetic defects in embryos prior to implantation. The test was also used to exclude sex chromosome linked disorders. 

Preimplantation genetic diagnosis and HLA typing of the embryo and transfer of a HLA compatible normal embryo was a later development leading to the birth of a saviour sibling and the popular term Designer baby. The term was the invention of Journalists with a penchant for exaggeration as was the term Test tube baby. A true Designer baby is a far cry as science stands to day. All major human attributes, intelligence, musical talents, athleticism and sporting skills are not just inherited. They are 99% (nurtured) perspiration and 1% (genetic) inspiration. We now know that DNA is not our destiny. They are part of our destiny and much more important, they are our history.

Current PGT-A testing does not screen for genetic defects. It only screens for chromosomal aneuploidy. So strictly speaking, ipso facto, it should only be termed as Pre implantation Chromosome Testing for Aneuploidy PCT- A or Pre implantation Chromosome Screening- PCS as you would prefer to call.

Screening the embryo was done only for a few chromosomes earlier. Now with NGS, we are able to screen for all the 24 chromosomes. To expect this screening  to improve prognosis is a huge jump in faith with no substantial data to support this claim. Failure of implantation is due to several factors, chromosomal aneuploidy is just one of them.

Intervention of an early embryo was clearly harmful when it was cleavage stage embryo biopsy. The results were poor. Is  Trophectoderm biopsy safe ? Only time will tell, with more follow up studies of these babies in to adulthood. I have my own reservations about the safety, specificity  and sensitivity of these procedures.

The first principle of medicine is, 'Primum non Nocere.' 

Simple, non invasive Sequential embryo transfer should achieve this purpose rather safely. Non invasive testing of the embryo is ideal; morphology does that with all its attendant limitations. Proteomics, metabalomics to aid in embryo selection is an utopian idea with no clear light visible at the end of the tunnel.

Professor Dr Pandiyan Natarajan,

Professor and Head, Chief Consultant in Andrology and Reproductive Medicine,

Chettinad Super Speciality Hospital,

Kelambakkam, Tamil Nadu.


Go to the profile of Prof Alan Handyside
about 2 years ago

PGT-A is actually highly effective for embryo selection in the context of elective single embryo transfers. We ran a clinical programme in which all patients had single vitrified-warmed blastocyst transfer with optional PGT-A (Gorodeckja et al (2019) Human Fertility). The average implantation rate per embryo was 80% and estimates of clinical pregnancy/live birth rates far exceeded the national UK data for frozen embryo transfer with one or more embryos with women up to the age of about 40 years.

Go to the profile of Prof Alan Handyside
about 2 years ago

Chromosome aneuploidy is a gross genetic defect with phenotypes ranging from lethality before implantation, miscarriage, abnormal fetal development and rarely affected live births. The incidence of aneuploidy at conception following natural conception has been estimated to be 20-30% and NGS based testing at the blastocyst stage shows an average incidence of 50%. At birth, the incidence of aneuploidy is only approximately 0.3-0.5%. On the principle of 'primum non nocere', the choice is either to apply an arbitrary age limit for IVF treatment to avoid high risk women of advanced maternal age or to use PGT-A to avoid repeated failed transfers of aneuploid embryos and miscarriage.

Go to the profile of Hosam Zaki
about 2 years ago

This is a very interesting study for a very hot topics, however the study design was incomplete to give a definite answer for the following reasons:

1- defining exactly which embryos are eligible for biopsy, as there are many embryos were biopsied with less than grade 4 expansion. Also many embryos were of “C” grading. There is ample evidence that the time of biopsy and grade are crucial.
2- The percentage of good quality embryos were higher in the control group, and there was no comparison between equal quality embryos transferred.

3- there is no homogenousity  in-between centers with 85% of the centers done less than 30 samples. There was also wide variations with the percentage of euploid levels. That might reflect that “small contributing centers may be were still in the beginning of the learning curve at the time of study!

4- There was no unification in the definition what exactly considered as EUPLOID embryo, as some would consider less than 50% mosaicism as euploid, with some research showing significant difference between implantation rate of euploid embryo and mosaic with less than 50% aneuploidy cells.

Hence, I would not reach the same conclusion of the authors with the above issues.

Dr Hosam Zaki FRCOG
Ganin Fertility Center
Cairo Egypt