Preimplantation genetic testing

I’m going to compare results from two embryo creation cycles, both from Orchid employees — those for myself (I lead software engineering at Orchid), and for Noor Siddiqui (Orchid’s CEO).  My embryos were tested using a well-known commercial PGT lab.  Noor’s embryos were tested using Orchid’s preimplantation genetic testing with whole-genome sequencing (PGT-WGS).

I want to walk through these results because my wife and I would have more embryos available today if I had known more about PGT when we underwent IVF.  The technology behind PGT is sophisticated, but the data is broadly interpretable to anyone who knows basic genetics.  Comparing these cycles helps show how high-depth sequencing can avoid unclear results that lead to viable embryos being discarded.

I'm hoping my experience will give parents a better understanding of the strengths and limitations of their PGT choices.

My wife and I underwent IVF in 2019 after a year of trying to conceive naturally.  At the time, I did not work at Orchid, knew very little about IVF, and knew essentially nothing about PGT.

‍

Our cycle began reasonably well for our age cohort (30 years old).  15 eggs were fertilized, and 9 embryos proceeded to biopsy. To analyze the genetics of an embryo, a few cells (~5) from the edge, called the "trophectoderm" of the embryo are sampled by an embryologist and sent to a lab.

These were reasonably promising numbers — historical statistics suggested we would end up with 6-7 viable embryos .  However, our PGT results couple weeks later were somewhat disappointing: out of the 9 embryos, 4 were deemed chromosomally normal, 3 were aneuploid, and 2 were “chaotic”:

‍

This was a little disappointing.  For our age cohort, we’d hoped for 6-7 euploid embryos.  4 embryos were probably not enough to give us the family size we’d aimed for (on average, each embryo has a ~60% chance of implanting).

I knew enough genetics to know that an embryo with an extra or missing chromosome was not an option for implantation.  I knew that mosaic embryos were sometimes implanted, but I didn’t know what a “Chaotic” embryo was; however, our fertility clinic assured us that those two were also not options for implantation, and we didn’t ask more questions.

We transferred one of the 4 euploid embryos, it successfully implanted, and 9 months later we gave birth to a healthy boy.  I didn’t think about our PGT results again for a couple years.

In late 2022 I joined Orchid. In the interim I had learned more about preimplantation testing, and when I learned that the technology behind single-cell sequencing had advanced enough that whole-genome sequencing of embryos was actually possible, preventing more genetic conditions than standard preimplantation genetic testing, I wanted to help bring this technology to as many people as possible.

After a crash course in bioinformatics, I knew what to expect from the high-depth sequencing performed by Orchid, and also understood that interpreting PGT results is more difficult in some cases than others (for example, the viability of mosaic embryos is a topic of open discussion).  

In the US, HIPAA gives patients the right to request the data underlying their clinical test results.  I was curious to go back and see the data underlying our PGT results, so I requested the raw data from our PGT vendor.  After a few weeks and several consent forms, I got the raw copy number plots for the embryos in our cycle. There were 4 euploid embryos; here's the plot for KC-6, now our toddler (the other 3 are below^[1]):

‍

A quick guide to interpreting copy number (CN) plots: preimplantation genetic testing for aneuploidy (PGT-A) measures the number of copies of each section of your genome.  Chromosomally normal humans have 2 copies of chromosomes 1-22 (one from the father, one from the mother). Male embryos have a single X chromosome and a single Y chromosome, while female embryos have 2 X chromosomes and 0 Y.

On the PGT-A copy-number plot, this means that on chromosomes 1-22 should have points centered on the “2” line of the CN plot.  Males will have a signal of “1” for the X chromosome and “1” for the Y chromosome, while females will have an X at “2” and little to no signal for the Y chromosome.  Small variations above and below are due to the low depth of sequencing and difficulty in sequencing DNA from a small number of cells, since only 3-5 cells are sampled per embryo.

The 4 euploid embryos met expectations for a standard commercial PGT test, since none of the embryos had large extra or missing chromosome segments. 

Next, I looked at the 3 Aneuploid embryos; here are KC-12, KC-10, and KC-13: 

‍

The CN plots showed clear extra chromosomes on KC-12 and KC-10, as expected from the test results.  KC-13, however, did not show an obvious extra copy of chromosome 6.  I wanted to make sure I wasn’t missing any subtlety here, so I requested an updated interpretation from the original PGT lab.  I’ll quote the response below:

"As you know, this analysis was originally completed in 2019 and mosaic reporting was not the standard.  Therefore, after a multistep review process, this embryo was initially reported as aneuploid with trisomy 6.  As you may imagine, much has changed in the landscape of PGT-A and mosaic reporting, which has become increasingly routine.  After reviewing this with an LD, rather than there being an error with the call, under our current algorithm this embryo would be reported as low mosaic for trisomy 6.  An embryo is reported as “low mosaic” if the copy number variation in the biopsy is between 30% and 50%. "

Mosaic embryos were known in 2019 to have moderately lower viability than euploid embryos, with one study reporting around 47% of transferred mosaic embryos resulting in a live birth, as opposed to 59% for euploid embryos. This was exciting, and offered another opportunity to transfer should we run out of euploid embryos.

This left two embryos reported as “Chaotic”.  “Chaotic” is not a classification Orchid uses for PGT reporting, so I had to reference my report to understand this classification: "An embryo is considered to be Abnormal: Chaotic when six or more aneuploidies are detected in the provided sample."

‍

‍

I was surprised to see these results.  At Orchid, we report embryos as “Indeterminate” when high levels of noise (including on the chromosomes marked as normal) suggest that errors were introduced during DNA amplification, or that the biopsy did not retrieve sufficient DNA — it’s very unlikely for an embryo to have 24 genuine aneuploidies.  I asked about these two embryos as well, and I’ll quote the response:

Our understanding of embryos with chaotic biopsy results has evolved and we now recognize this particular result has reduced predictive value compared to other PGT-A result types. You will see that the interpretation section of our new reports has changed, and we now suggest consideration of a re-biopsy on these embryos. This decision should be made in partnership with the fertility clinic. There is a hypothesis that the high number of aneuploidies in a chaotic biopsy may be a result of biological, technical or procedural artefact. We have seen concordance of aneuploidies in second independent biopsies from the same embryo, so true aneuploidy is also possible.  

While there are no-peer reviewed studies on chaotic embryos, this is an area of interest for our team at [REDACTED].  [REDACTED] recently completed a pilot study of 58 embryos which had been initially reported as chaotic and were subsequently re-biopsied for re-testing. 22 of the 58 (38%) re-biopsies returned a euploid result. Some re-biopsies returned a chaotic result while others were reported as aneuploid in which at least some of the chromosomes found to be aneuploid in the original biopsy were again noted in the second biopsy.  If you are interested, I have attached an ASRM abstract with this information, which can be found on the right column of page e350.  

The idea that we had at least one, and potentially three additional embryos available for transfer was briefly exciting. However, after asking our fertility clinic whether a re-biopsy was an option, our clinic reminded us that they had not preserved any embryos deemed “abnormal” in the initial PGT results, and KC-13, KC-7, and KC-8 had been discarded:

The five abnormal embryos were discarded in accordance with our standard protocol and signed consent forms.

As it turned out, one other Orchid employee had PGT results available for comparison — Noor Siddiqui, the CEO of Orchid.  Unlike me, Noor had the opportunity to use an early version of Orchid’s 30x PGT-WGS (Whole-Genome Sequencing) to test her own embryos.

Orchid uses a proprietary whole-genome amplification technique and much deeper sequencing than standard PGT-A.  At Orchid we frequently talk about how whole-genome sequencing enables broad screening for monogenic disorders, but it also enables more precise PGT-A results as well as the detection of smaller chromosomal abnormalities.

Here are copy-number visualizations for all the euploid embryos:

‍

To contrast against the mosaic KC-13 plot from my cycle, here’s an embryo from the cycle with a low-level mosaicism on chromosome 21, indicating that some, but not all of the biopsied cells were missing a copy of chromosome 21 (here, roughly 30%):

‍

This cycle does not have an example of a full aneuploidy sequenced at high depth for comparison, but you’d see the same pattern as on my results — a clear missing or extra chromosome centered on the “3” or “1” line.

I'm just going to compare PGT-A results in this article, but since Noor's embryos were tested using Orchid, they also got the monogenic screening reports that would only be possible via PGT-WGS (you can find a guide to what you get in a PGT-WGS report here). So alongside the PGT-A results, Noor's report included monogenic screening panels I didn't get (I've included an example below^[2]).

This is not a comprehensive comparison of commercial PGT-A with PGT-WGS^[3]. But I wanted to walk through these results because my wife and I would have more embryos available today had I known in 2019 what I know now.  There are three important points I want to make sure patients and clinicians understand:

• As a fertility patient, you have the right to understand your test results, and have the right to request the underlying test data if you don’t feel comfortable with the results you see.
• Many clinics do not preserve abnormal embryos by default, which limits the ability to re-test embryos with unclear PGT results. Make sure you understand your clinic's preservation policies before your cycle.
• The technology behind PGT is sophisticated, but the data is broadly interpretable to anyone with an understanding of the underlying biology.  Patients and clinicians can observe the difference between high-fidelity and low-fidelity sequencing and appreciate why it affects the accuracy of the test results.

At Orchid we want everyone to have healthy children. Part of that is making sure that parents, as well as clinicians, understand what preimplantation genetic testing can, and cannot, detect.  If you’d like to understand more about Orchid’s PGT, you can speak to a genetic counselor today to see whether Orchid fits into your family’s plans.

[1] Here are the rest of the euploid embryos.

‍

[2] Orchid's standard PGT-WGS screening has four monogenic screening panels, plus targeted screening for microduplications and deletions. You'll see these panels on both the clinical report and Orchid's web portal.

‍

[3] However, these results weren't cherry-picked; Noor and I were the only two Orchid employees who had PGT results available to compare.