Dr. Santiago Munne, co-founder of MedAnswers and renowned reproductive geneticist, addresses the misperceptions in the recent New York Magazine, which created a lot of anxiety in patients and their doctors by hyping a minority view and confusing the majority.

In keeping with our goal of educating and empowering patients, we are first going to clarify terms, many of which the author of the article confuses:

  • PGS – Preimplantation Genetic Screening – the lab test used to select the embryos with the correct number of chromosomes.
  • NGS – Next Generation Sequencing – the latest technology which biopsies on average five cells from an embryo.
  • Array CGH – Older, inferior, PGS technique to NGS.
  • AneuploidyALL cells tested by PGS are ABNORMAL
  • Mosaicism, SOME cells of the embryo are ABNORMAL
  • Euploidy, ALL cells TESTED are NORMAL  

Why does it matter if embryos are normal or abnormal?

Chromosomal abnormalities result in the embryo not implanting or in miscarriage. These abnormalities increase with the age of the woman from about 30% in young women to over 80% in women 42 and older. Once a normal embryo is selected, they implant equally well in women at any age (up to age 42).

How Does NGS for PGS work?

With the newest technology, Next Generation Sequencing (NGS), and biopsing on average 5 cells from an embryo. Those embryos with 0 normal cells are fully aneuploid and those with all normal cells are euploid (normal).

How do you Detect Mosaic Embryos?

The sequencing is so specialized now with NGS, we can also detect 1 in 5 (20%) to 4 in 5 (80%) abnormal cells in the biopsy. Those in between the 20 – 80% range, we call them mosaics. Mosaics have lower chances of implanting, higher chances of miscarrying. About 18% in young women have mosaic embryos, but the rate of mosaicism goes down as age goes up because they become fully abnormal for other chromosome defects. Women 40 and above have about 9% mosaic embryos.

Can Mosaic Embryos Make Babies?

Mosaic embryos (some abnormal cells) can make babies but at a lower chance, and a higher probability to miscarry. The quoted 24 babies from the Rome center or the series we recently published were not aneuploidy, they were mosaics. The seven by Gleicher were analyzed with an even older technique that could not differentiate aneuploid from mosaics.


What was the problem with Array CGH as it relates to mosaicism?

With the inferior, older technique, array CGH, 20% of previously classified mosaic embryos would have been classified as aneuploidy / abnormal (Ribustello et al. 2016) – 20% previously though to self-correct, would have instead failed. Of the 80% that would have been classified as euploidy or normal, we now know only 4% of them would have been true mosaics.

The best study so far, By Scott et al (2012) biopsied embryos but did not analyze until the transfer was done, so there was no selection, but by fingerprinting the resulting babies the embryos that implanted were known. In that study, 255 embryos were transferred, of those 133 were found to be normal and of those 41% implanted, while 99 were found to be abnormal and only 4% (4/99) implanted. In total 28% of unselected embryos implanted, which was significantly lower than for normal embryos (41%) and significantly higher than abnormal embryos (4%). This technique did not differentiate aneuploid from mosaic embryos, so it is possible that those 4% were mosaic.

How many embryos are discarded?

The article quotes Richard Paulson guessing from thin air that 40% of embryos are discarded while solid date points to a mere 2-4% are discarded. I hope my friend Richard, as was I, was also misquoted in the article. In IVF is about 50% of all embryos are abnormal (the full range is from 40% in young patients to 80% in older patients).

What if my lab is wrong?

It should not be a surprise to any patient, unless your clinic is not properly educating you that there is an error rate! PGS tests prior to Next Gen Sequencing (the most advanced technique) reported results as normal or abnormal, but each test reported (or should have) its error rate. For example, our array CGH test reported a 2% error (Gutierrez Mateo et al. 2011). The NGS failure rate is even less. I checked our database and found that of about 1,000 cycles of PGS with NGS that became pregnant, only five transfers of normal embryos resulted in a chromosomally abnormal fetus or 0.5%.

The critics are right that an elegant study such the Scott et al. (2012) has not been done, but the error rate of classifying a normal embryo as abnormal would need to be 10x higher than the observed error of classifying normal as abnormal embryos to be identical to the 2-4% error of prior techniques.

Do the number of mosaics increase with age like aneuploidy does?

No. The argument of Gleicher, Braveman and Vidale is that with advancing maternal age the likelihood of having normal embryos diminishes and the risk of throwing away mosaic embryos (which they mixed up with aneuploid) increases. However, only aneuploidy increases with maternal age, not mosaicism. In fact, because mosaicism is constant (and lab dependent, with some centers producing more than others), such that with advancing age more embryos are not only mosaic for one chromosome but also aneuploid for another, so all their cells are abnormal. Because of this, mosaic embryos with at least some normal cells decrease with age from 18% in young patients to 8% in patients 43 and older. That is, even with array CGH 20% x 8% = 1.6% of embryos at that age would have been classified as abnormal but be mosaic. Fewer than the 4% for the general population. Please understand, a mosaic embryo only has a 1.6% chance in an IVF clinic of success.

What are the success rates with the NGS technique?   

Now that we can identify mosaics better with NGS, we can prioritize transferring normal embryos, which have now lower chances of being mosaic. NGS tested normal embryos also are reported by two groups to implant better than normal embryos from array CGH. If there are no normal embryos left, transfer abnormal mosaic embryos, knowing they have a very low success rate. The remaining abnormal embryos (abnormal mosaic not aneuploidy) are now expected to have even fewer chances than the 2-4% previously reported, to produce a viable pregnancy. Certainly if a patient can afford the price of IVF and does not mind such low success rates, then it is their choice to transfer mosaic embryos.

Should IVF cycles include PGS?

Using genetic testing for embryo selection allows patients to more safely approach IVF by selecting one embryo for single-embryo transfer. This is another point the article failed to acknowledge; single embryo transfer is widely accepted as the safest approach given the reduced chance of multiples and the associated costs and complications of twin, triplet, or higher order pregnancies. [Cost on the order of 4 – 10x increase with multiples].

Without PGS, IVF cycles result in more miscarriages, especially in patients >40-years-old, when it can be as much as 40% of pregnancies. Patients need to be counseled that the risk of misdiagnosis is less than 4% (I would say 1.6%) but the risk of miscarriage is 40% at older ages. The article fails to mention that having no normal embryos and no mosaic embryos leads usually to the patient choosing egg donation, a procedure that has very high success rates.

PGS is also an important tool for those who suffer recurrent pregnancy loss – and the incredible emotional and physical toll their experience causes. A test which can substantially decrease (from 40% in women >40-years-old to 10%) the risk of miscarriage for these women holds tremendous value. This is a fact glossed over by our critics and ignored as a problem. Nor it talks about the patients that by using only morphological selection become pregnant with a trisomy 21 baby.


Perhaps the most critical correction to the article is this:

There is no solid evidence that ANEUPLOIDY embryos detected by the most advanced techniques (Next Generation Sequencing – NGS) can self-correct. Your doctor has not thrown out potential babies when they discarded or donated to science aneuploidy / abnormal embryos. (75% of embryos in nature are aneuploidy!)

Genetic testing labs do not make transfer recommendations nor do they discard embryos. Rather, PGS IS designed to help patients and providers better understand the options they face, and prioritize embryos with the highest chance of success for transfer.  The article strongly implied that PGS is designed to ‘eliminate’ embryos, which is never the goal or intent, of any PGS lab worldwide.

The article fails to quote 4 randomized clinical trials (those of the highest quality of evidence) all supporting PGS as improving ongoing pregnancy rates when normal embryos are found.

The Gleicher paper reporting 7 babies born after replacing abnormal embryos did not follow the standard procedure that we have for a discrepant result. In those cases, we retest left-over DNA from the biopsy to ascertain that there were no clerical errors. We test the same DNA with another technique, for technical errors, and then compare DNA from the fetus to DNA of the embryo (to determine the embryo replaced was the correct one) and DNA from the parents. To our knowledge this was not done. We have several instances of embryos being conceived in vivo, embryos switched within or between cycles and patients, and other errors.

The ASRM has produced guidelines which further detail and recommend the PGS approach. It’s very important patients and clinics have all the facts and are not misled.

If you have additional questions regarding PGS, please consider downloading our iOS app and asking your own questions, anonymously, on MedAnswers platform.


Gutiérrez-Mateo C, Colls P, Sánchez-García J, Escudero T, Prates R, Wells D, Munné S (2011) Validation of microarray comparative genomic hybridization for comprehensive chromosome analysis of embryos. Fertil Steril 95:953-958

Gleicher N, Vidali A, Braverman J, Kushnir VA, Barad DH, Hudson C, Wu YG, Wang Q, Zhang L, Albertini DF, International PGS Consortium Study Group (2016) Accuracy of preimplantation genetic screening (PGS) is compromised by degree of mosaicism of human embryos. Reprod Biol Endocrinol Sep 5;14(1):54.

Scott RT Jr., Ferry K, Su J, Tao X, Scott K, Treff NRT. (2012) Comprehensive chromosome screening is highly predictive of the reproductive potential of human embryos: a prospective, blinded, nonselection study. Fertil Steril, 97:870-875

Ribustello L, Escudero T, Lanzas J, Liu E, Munne S (2016) Mosaic embryos, detected by NGS, are mostly classified as euploid by aCGH. Reprod Biomed online 33:258-259

This blog response was upvoted by the following MedAnswers experts:

David Ball, PhD – Seattle Reproductive Medicine
Wendy Chang, MD – Southern California Reproductive Medicine
Catherine Welch, MBA, TS(ABB) – Fertility Tomorrow
Antonio Martinez Lara, Biologist and biotechnologist – Pronacera Therapeutics S.L.
Iman Abdelhadi – Senior Clinical Embryologist, Al Ahli Hospital
James Stachecki, PhD – Innovative Cryo Enterprises
Esther Vellila, Senior Embryologist and PGD Specialist

The comments below were contributed by MedAnswers Expert: Esther Vellila, Reproductive Biology Service Director, Senior Embryologist and PGD Specialist from Barcelona, Spain

  • Old Techniques cannot differentiate mosaic embryos from abnormal embryos. NGS permits the lab to have a deeper look into the embryo and we are able to detect mosaic embryos and even the degree of mosaicism. The degree of mosaicism provides important information for the physician to properly counsel patients in the cases in which the patient has no euploid embryos at all. It is a physician’s responsibility to counsel on the risks and benefits.
  • When we talk about aneuploidy detected with old technology we did not know if the abnormal embryo were fully abnormal or mosaic. This point is important because in the text they refer all time to abnormal embryos giving a healthy baby. This message is completely wrong.
  • There is also a difference between transferring an abnormal mosaic and an aneuploidy mosaic….in the article they mix all these concepts up (they talk about PGS mixing technologies and efficiencies) and is giving the general idea to the population that an actual ‘abnormal embryo’ can give a healthy child, which is not possible… those embryos they are referencing that grew into a healthy child were undetected, mosaics.
  • The success rates of transferring normal embryos and mosaic embryos depend on multiple factors in an IVF clinic (stimulation protocol, IVF culture conditions, micromanipulation proficiency, doctor skills etc…). In fact, abnormal IVF conditions with high levels of oxidative stress could lead to mosaicism.
  • A healthy baby born after transferring what they have called an ‘abnormal embryo’  (mosaic one) does not mean that is completely normal, they can show euploid cells in blood test but I want to know what will happen in germinal lines or other important lines in the body. I do not know, but no one is telling us that abnormal cells can be there and express as a higher risk in cancer, for example… is that all explained in the consent form for transferring this abnormal embryo? That we do not know what the full risks are? I certainly hope we do not have patients demanding we transfer mosaic embryos when we do not know the risks.