Author: Gary Harton, PhD Market Development Leader, Genetics, PerkinElmer
I’ve been involved in the Preimplantation Genetic Testing field since my first case in 1992. I’ve seen it grow and change over time in response to the latest publications and I’ve seen the tests evolve from polymerase chain reaction (PCR) through fluorescence in situ hybridization (FISH), microarrays, and next generation sequencing (NGS). For me, this blog gets to the one question that continues to nag the field when discussing preimplantation genetics…where is the definitive data to support preimplantation genetic testing for aneuploidy (PGT-A)?
Preimplantation genetic testing was first introduced in 1990 at the Hammersmith Hospital in London, England1. At the time, PGT was used to diagnose the sex of an embryo for couples that carried an X-linked disease. In these couples, the transfer of a male embryo would come with a 50% chance of being affected by an X-linked disease carried by the female partner in a couple. Selective transfer of female embryos avoided having affected male children born.
The selection of embryos unaffected by an inherited genetic condition is called preimplantation genetic testing for monogenic disorders, or PGT-M. We know that there are other whole chromosome aneuploidies that are associated with well described genetic conditions including Down syndrome which is caused by an extra copy of chromosome 21. Soon after the introduction of PGT-M, screening embryos for Down syndrome and other potential chromosome errors that were well described at the time including trisomy 13, trisomy 18, Turner and Klinefelter syndromes, and trisomy 16 was performed using FISH. Fast forward to 2020 and we have a lot more information at our disposal, including the knowledge that every chromosome can be found as extra (trisomy) or missing (monosomy) in early human embryos, not just the chromosome aneuploidies that we see in prenatal or postnatal screening programs. PGT-A aims to test embryos in an IVF cycle for aneuploidy across all 24 chromosomes and to enable selective transfer of euploid embryos, or those carrying the correct number of chromosomes.
FISH, microarrays, and NGS have been used for years, so surely by now there is plenty of data out there to answer the simple question posed above…does PGT-A work? It seems intuitive to me that embryos containing an abnormal set of chromosomes (aneuploidy) are not the best candidate for transfer during an IVF cycle, so being able to pick them out of the embryos available for transfer and choose an embryo with the correct number of chromosomes should clearly improve IVF outcomes. However, the story so far has not been so clear.
During the late 1990’s and early 2000’s, there was a lot of evidence building that was positive for PGT-A, then along came a number of randomized controlled trials (RCT) that showed little to no benefit to IVF success rates with PGT-A. Some of the studies that showed no impact were quite small2, but some were quite large3 as well. Studies with large numbers of samples are always more convincing than smaller studies. Then, the worst possible thing happened, two studies were published that not only didn’t show a positive impact of PGT-A, but actually showed a negative impact4,5. Imagine this conversation…we are going to offer couples a test that costs extra money on top of the IVF cycle, perform an invasive procedure to remove cells from the developing embryo, and that couple has less chance of getting pregnant than if they just transferred embryos blindly. Imagine that news for a moment!
I am often asked to give lectures on PGT and often I review the above information during the talk. When I get to this data, I’d say why these maybe weren’t the best studies (small sample numbers, wrong choice of study subjects, and inexperienced or not best in class technique in the IVF and/or genetics lab). And I’d also point out that the common denominator for all of these initial studies was the use of FISH to test for aneuploidy. While FISH is an excellent technology and still has a place in the cytogenetics lab today, it clearly wasn’t the best choice for PGT-A. The early PGT-A tests were performed on an embryo that had been in culture for 3 days when an embryo typically has 6-8 cells. At the time, the test was typically performed on a single cell which meant that we were removing a high proportion of the embryo’s mass (1/6-1/8). With the high error rate of FISH on a single cell, it was no surprise that it turned out that this technology was not ideal for PGT-A. FISH technology ideally screens hundreds of cells so you can get a consensus result. In addition, there are only 5 fluorescent colors that can be seen by the human eye. This meant that FISH could only look at a subset of the 24 chromosomes in the one cell from the embryo. So, an embryo diagnosed as euploid for the 8-12 chromosomes that could be analyzed by FISH could have been aneuploid for another chromosome that wasn’t analyzed.
Around 2009, the field moved on to more comprehensive and less subjective molecular methods of testing human embryos. This era has been called PGT-A 2.0 by some in the field. These methods include array comparative genomic hybridization (aCGH) and more recently, next generation sequencing (NGS). The main advantage of these methods over FISH is the ability to test all 24 chromosomes in a single test. Single nucleotide polymorphism (SNP) arrays, targeted amplification with next generation sequencing as well as targeted repetitive sequencing can also screen all 24 chromosomes but, I will leave these details for my future blog with the working title of “My Platform is Better Than Your Platform”. Now that the IVF field had the ability to screen all 24 chromosomes, the question on whether PGT-A is of benefit to IVF success rates was renewed. Publication rates and trials were on the rise again.
Why am I telling you all of this you might wonder? Well, I’m trying to get to the point of this post, which is how much data does the field need before it agrees that a technology works? This might seem a simple question but it hasn’t proven so simple to date within the IVF field. Whenever a paper comes out in support of PGT-A, it is followed by commentaries and letters to the editor with data showing why the paper is biased or wrong. The same thing happens when a paper showing no benefit from using PGT-A is published. So there are clearly two opposing sides to the debate and a significant proportion of IVF couples caught in the middle.
Recently, the data from the much awaited STAR trial6 was published with a bit of fanfare since it was the largest RCT on PGT-A ever performed, it tested all 24 chromosomes with the most up to date embryo biopsy and NGS methods and was meant to clearly show whether PGT-A impacted IVF outcomes (positively or negatively) or not. However, the STAR trial didn’t end the debate, and it might have fueled it further because the conclusions of the trial are not black and white. Overall this trial showed that PGT-A improved IVF outcomes for couples over 38 years old, however it also showed that PGT-A wasn’t helpful for couples between 35 and 38 years old. In addition, because there were so many couples in the younger age group, the cumulative data across the entire age range of 35 to 40 years old showed that PGT-A was not helpful…wanh wanh wanh waaaaanh!
I’ve been around this field for a very long time (feel free to insert your own old guy joke here) and can tell you that no other procedure or technology in the ART field has ever garnered this much data or this much debate! So how do we go about settling the debate once and for all…well, more data of course!
I have two pieces of data that I’m going to share with you here that are so convincing that I think they should stop the debate once and for all.
The first one was presented at the annual meeting of the American Society for Reproductive Medicine (ASRM) last year in Philadelphia, PA. Here Dr. Ashley Tiegs and colleagues presented their latest non-selection style study where a cohort of couples underwent IVF treatment with PGT-A, however the PGT-A result was not used to decide on the single best embryo for transfer7. Once all of the couples had completed their embryo transfers and all of the outcomes were known, the data on the genetic makeup of each embryo that was transferred was ‘unlocked’. This style of study allows you to truly understand the use of PGT-A, and also understand the false positive and false negative rates your test will return. The authors of this study presented the outcomes from 251 PGT-A tests with embryo transfer; 50 single embryo transfers where an aneuploid embryo was transferred and 201 single embryo transfers where a euploid embryo was transferred. Remember, no one knew the genetic makeup of the embryo until after all outcomes were available, transfer decisions were based on traditional morphologic criteria on the transfer day. Of the 50 single embryo transfers of aneuploid embryos, not one resulted in a live birth, either the embryo failed to implant or led to a miscarriage. Of the 201 single embryo transfers with euploid embryos, 137 resulted in a live birth. This is a 68% positive outcome from euploid embryo transfer and a 0% positive outcome for aneuploid embryo transfer! See figure 1 below.
Figure 1: Adapted from Tiegs AW et. al. Pregnancy rate, ongoing pregnancy rate and delivery rate of aneuploid embryos and euploid embryos in a double-blind non-selection study. Pos. bHCG = pregnancy seen by hormone testing, Fetal Heart + = ongoing pregnancy at 12+ weeks gestation, Delivery = child born.
The second bit of evidence I want to share with you comes from some work performed by my friend and colleague Dr. David McCulloh who is an embryologist and researcher at NYU Langone Fertility Center in Manhattan, New York. The Centers for Disease Control (CDC) and the Society for Assisted Reproductive Technology (SART) keep a publically available database of IVF cycles performed in the US that contains a registry of nearly every IVF cycle performed in the US each year. Over the last few years, the registry has collected data from every egg retrieved in an IVF cycle and follows that egg’s fate through transfer and outcome. In addition, for each egg there is a record of whether the embryo was biopsied or not, which is very important to us here. What Dr. McCulloh did was download the entire database of egg retrievals and sorted for ‘with biopsy’ and ‘without biopsy’ and compared the outcomes between the groups based on age of the female partner at the time of egg retrieval. Now this isn’t perfect because some of these couples would have had a biopsy for a reason other than to simply perform PGT-A (single gene defect testing [PGT-M] without PGT-A, or chromosome translocation testing [PGT-SR]), which combined make up only a minor proportion of the total cycles worldwide with biopsy8. The key to this data is that it contains 451,434 total embryos transferred across the four years of data collection (118,684 with biopsy and 332,750 without biopsy). Think about that for a minute, most RCTs contain a few hundred total cycles and therefore a few hundred embryos transferred, this data set contains almost 500,000 embryos transferred! What he found was a statistically significant difference in live birth rate for every age group (under 35, 35-37, 38-40, 41-42, over 42) with biopsy versus without biopsy. The data shows that cases with a biopsy, and therefore some sort of PGT, had a significantly higher chance of a live birth as compared to those cases that didn’t have a biopsy. This data cannot be manipulated during analysis, it is simply almost every embryo transferred during an IVF cycle that occurred in the US over 2014, 2015, 2016, and 2017.
Figure 2: Percentage of delivered babies/transfer with and without PGT. Age group bands listed on the X-Axis. Number in the bar is the number of embryos transferred in each category.
For me, this study unequivocally shows the power of PGT-A.
It does however come with one caveat that must be discussed here. This data is for couples that had an embryo transfer…some number of couples over the four years of data collection did not make it to transfer. That doesn’t mean that only couples in the PGT group didn’t make it to transfer, in each age group there were couples that did not have a transfer both with and without PGT, however the percentage of couples with no transfer was always higher in the PGT group as compared to the non-PGT group.
Figure 3: Percentage of retrievals with no transferrable embryos with and without PGT. Age group bands listed on the X-Axis. Number in the bar is the number of retrievals in each category.
The point of PGT-A is to identify euploid embryos for transfer assuming that these embryos have the highest chance of a successful outcome. Obviously there is a risk for some couples that there will be no euploid embryos and therefore no embryo transfer. This might sound like a bad thing but it isn’t necessarily the worst thing possible. By eliminating transfer of aneuploid embryos, that in the Tiegs paper had a 0% chance of live birth, the couple is shortening the time between IVF cycles and therefore hopefully getting to transfer of a euploid embryo with a very high chance of pregnancy success much faster. With the aneuploidy rate increasing every year, time between cycles is important.
For me the debate is over, PGT-A works, it’s a simple as that. It won’t help every couple achieve a pregnancy and it isn’t the right choice for every IVF cycle. However, I do think that every couple deserves a cogent and thorough counseling session on the pros and cons of PGT-A specific to their circumstances so they can make an informed decision on what to plan during their IVF cycle. It is very clear that placing embryos in a transfer order based on genetics is a benefit. Published data has shown that PGT-A improves time to pregnancy9, and reduces miscarriage10. These are maybe the most critical items to a couple undergoing fertility treatment because time is not on their side.
So, can we please stop arguing about whether PGT-A works or not and start thinking about how to make the technology better, agree on the best IVF populations to offer PGT-A to, find more effective ways to counsel couples and work together to achieve the highest delivery rates in IVF possible? We’ve come a long way as a community by increasing success rates from single digits to over 50 % in 20 years…let’s set our sights on 100% in the next 20 years.