Miscarriage After a Euploid Embryo: What to Check Next

Endometriosis Surgical Specialists International — Reviewed September 2026

In one paragraph: A euploid result is a filter, not a finish line, and roughly one in eight of these transfers still ends in loss. The story that endometriosis made your embryo abnormal does not survive the evidence. This post sets out what PGT-A actually tested, what the 2026 evidence rules out, what is still worth looking for — and the one question about your miscarriage tissue that almost nobody is asked.

You Did the Thing That Was Supposed to Remove the Doubt

You biopsied the embryos. You waited for the report. You transferred the one that came back normal, and the beta rose the way it was supposed to, and there was a heartbeat, or there was very nearly a heartbeat. And then there wasn’t.

The cruelty here is specific. Every other loss comes with a ready explanation — the embryo, the age, the odds. This one arrived with that explanation already crossed out. The one variable everyone said was the problem had been tested and cleared. So when it happened anyway, there was nothing left to say, and what you were given instead was some version of bad luck.

That is not an answer. It is the sound a clinic makes when its testing menu runs out.

What “Euploid” Actually Tells You — and What It Does Not

PGT-A does not test your baby.

It tests a few cells pinched from the trophectoderm — the outer shell of the blastocyst, the part that becomes placenta. The inner cell mass, which becomes the fetus, is not sampled at all. The result is an inference drawn from neighbouring tissue at one moment in time.

It is a good inference. It is not a guarantee.

• In 1,051 single euploid frozen blastocyst transfers, the miscarriage rate was 14.5 in 100.

• In a separate series of 270, pregnancy loss was 13.3 in 100.

• In 459 frozen euploid transfers, loss ran between 8.0 and 13.5 in 100 depending on the group.

So the honest figure is that roughly 11 to 15 euploid transfers in every 100 end in loss, and a large share of those losses have no identified cause.

Nobody told you that before the transfer. “Euploid” was presented to you as a finish line. It is a filter.

The Finding That Should Change This Conversation

If you have endometriosis, or suspect it, you have been handed the egg-quality story: inflammation in the pelvis damages the eggs, damaged eggs make abnormal embryos, abnormal embryos miscarry.

It is tidy. It does not survive contact with the data.

First. When only chromosomally normal embryos are transferred, endometriosis makes no measurable difference to loss. In a study of frozen euploid blastocyst transfers, miscarriage was 11.1 in 100 with endometriosis versus 13.5 in 100 in male-factor controls — not significant, and numerically lower in the endometriosis group. Live birth was 61.1% versus 49.6%, also not significant. Control for embryo quality and the gap closes.

Second, and more decisive. Investigators examined the chromosomes of 543 miscarried pregnancies after IVF, comparing women with and without endometriosis. The rate of chromosomal abnormality was 35.29% versus 43.54%, and after propensity-score matching the odds ratio was 1.03 — no difference.

If endometriosis were making embryos chromosomally abnormal, the miscarried tissue would be more abnormal. It was not. It was, if anything, slightly less.

The same pattern holds at scale: a 2026 meta-analysis of 29 studies covering 93,071 women with endometriosis having assisted reproduction found miscarriage not significantly increased (RR 1.17, 95% CI 0.71–1.94).

What this rules out: The claim that your endometriosis made your embryos chromosomally abnormal and that is why you lost a normal pregnancy. The precise statement is quantity, not quality: endometriosis more consistently affects the number of eggs and ovarian reserve than the chromosomal normality of the embryos those eggs make. Anyone still selling you the damaged-egg explanation is working from a story rather than from evidence.

What it does not rule out: Everything else.

So What Is Left?

Some of it is findable. Some of it is not.

• Chromosomal Error the Biopsy Missed: Mosaicism, a segmental change, a discordance between trophectoderm and fetus. The biopsy sampled the placental lineage. It can be right about those cells and wrong about the pregnancy.

• Antiphospholipid Syndrome — The Real Thing: Lupus anticoagulant, anticardiolipin or anti-β2-glycoprotein antibodies, at diagnostic level, persisting on a repeat sample 12 weeks later, in someone who meets clinical criteria. One positive result on one draw is not a diagnosis, and it is not a reason to start treatment.

• Uterine Structure: A septum, intrauterine adhesions, a submucosal fibroid distorting the cavity.

• Adenomyosis: A larger miscarriage signal than endometriosis: 31.9 in 100 versus 14.1 in 100 across nine IVF studies, roughly double, with wide uncertainty (RR 2.12, 95% CI 1.20–3.75). Two small studies have tried to take the embryo out of the equation. In a retrospective study of 171 pregnancies after a single euploid frozen embryo transfer, miscarriage was 53% in the 34 women with adenomyosis against 19.7% in those without, mostly at the very early, biochemical stage. In a prospective study of 228 women receiving donor eggs, miscarriage was 35.4% with adenomyosis against 18.1% without, although live birth did not differ significantly (36.8% against 43.9%). These are small studies, and not all agree. A prospective study of 648 women having a single euploid transfer found no link between adenomyosis on 3D ultrasound and live birth, although the scan readers often disagreed about who had it, and a 2024 conference abstract of 282 single euploid transfers found that mild adenomyosis made no significant difference to miscarriage or live birth. Mild disease may simply not matter.

  Our reading, and it is our hypothesis rather than a finding: When a normal embryo is lost anyway, the uterus becomes the leading remaining candidate to map. That does not prove the uterus caused the loss. It means that if nobody has assessed your junctional zone, that is a gap.

• Chronic Endometritis: Reported in anywhere from about 10 to 30 in 100 women with recurrent loss, depending on how it is defined and counted. The plasma-cell finding usually clears on a repeat biopsy after antibiotics. But the only large randomised trial, CERM, which gave 438 women with recurrent miscarriage and chronic endometritis doxycycline or placebo, found no improvement in live birth: live births or ongoing pregnancies reached 52 in 100 with doxycycline and 50 in 100 with placebo (adjusted RR 1.02). Clearing the biopsy has not been shown to change the outcome that matters. So we test selectively, not routinely: a CD138-stained endometrial biopsy belongs in the work-up after a euploid loss or an otherwise unexplained picture. If it is positive, we still treat it, because the plasma cells usually clear. We do so without promising that it will change what happens in your next pregnancy.

• Endocrine Factors: Uncontrolled thyroid disease or diabetes.

• Factors Rarely Revisited After a Euploid Loss: In the 270-transfer series, smoking and day-6 blastocyst transfer were each independently associated with loss, with odds ratios near 4.6. That is roughly four times the odds — which on this cohort’s numbers would put absolute risk somewhere near one in three. That is an estimate from one study, not a measured rate — but it is worth raising before the next transfer.

• Unidentified Etiology: And for a real proportion of women: no identified cause. Not “nothing wrong.” No cause identified with the tools that currently exist. Those are different sentences, and only one is honest.

The Single Question Almost Nobody Is Asked

If tissue from your miscarriage was tested and came back “normal female,” 46,XX, ask one question:

Was a maternal contamination screen run?

Miscarriage tissue is not pure fetal tissue. It is mixed with your own — decidua, maternal blood, maternal cells at the edges of the specimen. If the laboratory analysed maternal cells rather than fetal ones, the report will read 46,XX, normal female. It will be describing you.

A maternal contamination screen compares the specimen against a maternal sample to establish whose cells were actually tested. Without it, a normal-female result on products of conception tells you very little — and it may have sent the rest of your workup in the wrong direction, because everything after it proceeded on the belief that the pregnancy was chromosomally normal.

This is a question to ask, not a diagnosis. Ask the laboratory. The answer is in the report or it is not.

One label on that, since this page holds everyone else to sourcing: Maternal cell contamination is a well-recognised laboratory pitfall, and it is not rare: in one laboratory’s review of 1,222 miscarriage specimens tested by SNP microarray, 269 of 456 “46,XX” results — about 59 in 100 — turned out to be maternal cells only, with no fetal component. Even so, no guideline we are aware of specifically mandates a contamination screen on products of conception. Raising it is our clinical position, not a society recommendation, and we are flagging it as such.

What ASRM’s 2026 Committee Opinion Actually Recommends

The American Society for Reproductive Medicine updated its recurrent pregnancy loss committee opinion in 2026 — the first revision since 2012. Three parts of it matter directly to you:

1. The Definition Changed: Recurrent pregnancy loss is now two or more losses, and they need not be consecutive. Ultrasound confirmation is not required, which means biochemical losses count. Many women told they did not yet qualify for evaluation now do.

2. The First Step Is Not a Blood Panel: It is chromosomal analysis of the miscarriage tissue by array-based technology — aCGH, next-generation sequencing or SNP array — followed by targeted testing directed by what that shows. Not a broad immune panel. Not a pre-emptive protocol. The tissue first.

3. The Prognosis Figure Is an Average, and You Should Know What Kind: ASRM states that between 50 and 80 in 100 patients with recurrent pregnancy loss go on to have a successful pregnancy even without medical intervention. That is population context, not a probability calculated for you. Age and the number of previous losses move it, and our own commentary on the ASRM opinion (manuscript in preparation) points out that it fits least well after repeated losses of chromosomally normal pregnancies, which is the situation this page is about. It still matters for one reason: it is the background rate against which any unproven treatment has to be judged.

What the Same Document Says Not to Do — Including, in Part, What We Do

We are an endometriosis surgical practice. This is the paragraph where we have the most to lose by being straight with you, so hold us to it.

ASRM 2026 lists as not recommended in recurrent loss: natural killer cell testing, endometrial receptivity testing, and routine immune testing and treatment. IVIG, intralipids and prednisone appear among treatments proven ineffective or without evidence of benefit. That heading joins two different things, treatments tested and found not to work and questions that have simply not been answered, and it is fair to ask which one you are looking at. ESHRE reached the same conclusions in 2023. The biology is not fake — uterine NK cells do behave differently in recurrent miscarriage — but the largest review of it, covering 1,100 women, found no established normal range, no predictive value for whether a pregnancy continues, and no correlation between blood NK levels and uterine NK levels.

In immune care, too, we go beyond these recommendations, and you should know where. For a narrow group of patients who have already had a complete evaluation, we may treat a defined immune abnormality, one drug at a time and never a combination. We never treat on a blood NK result. That is phenotype-matched reasoning, not established treatment. What we test, what we may treat, what we never use and what the trials actually show are set out in NK cells in endometriosis: why the problem is weakness, not attack.

And ASRM 2026 goes one step further, into our own territory: it lists “treatment of endometriosis and adenomyosis” among interventions without evidence of benefit for recurrent pregnancy loss.

We are not going to work around that. It is correct that no benefit has been demonstrated: no trial has tested treating endometriosis or adenomyosis as a way of preventing loss. What the opinion does not address is diagnosis. It makes no diagnostic recommendation on adenomyosis at all, while ESHRE’s recurrent-loss guideline suggests ultrasound to look for it. ESHRE’s endometriosis guideline goes no further on treatment either: its only statement on the point asks clinicians to “be aware that there may be an increased risk of first trimester miscarriage and ectopic pregnancy” — a Strong recommendation with the hedge written into it, and not a statement that anything can be done about it. No randomised trial has ever shown that surgery for endometriosis reduces miscarriage; the Cochrane review of 14 trials in 1,563 women reports no live-birth data at all.

So here is our position, stated as a position rather than as a finding:

We think endometriosis and recurrent loss may frequently be two expressions of one underlying dysfunction rather than two diseases sharing a patient — the question we posed in Fertility and Sterility in 2023. That is a hypothesis about biology. It is not a reason to operate.

Separately, some patients’ illness predates, exceeds and outlasts the pelvic disease eventually found. For them we have proposed Pelvic Overlap Spectrum Syndrome (POSS): endometriosis or adenomyosis, confirmed or strongly suspected, alongside features in other systems — autonomic, vascular, immune, connective tissue, nerve sensitisation, gut motility, bladder and pelvic floor, and mood. Pregnancy outcomes are recorded but deliberately do not count toward the definition, and POSS is a proposal with provisional research criteria: not a validated diagnosis, and not a claim that these conditions share one biology. One consequence matters here: surgery is aimed at structural disease and is not offered as treatment of the whole illness.

Complete excision has honest indications: pain, disease burden, organ involvement. Those are real. It is not a treatment for your miscarriage. We would rather operate on the honest indication than manufacture a better-sounding one. There is one situation we weigh differently: recurrent loss in a woman with endometriosis, once the rest of the evaluation has come back empty. There we treat surgery before the next transfer as a strong consideration, decided case by case. That is phenotype-matched reasoning, not a proven benefit, and it is explained in Surgery before your next transfer: questions to ask before you consent.

What a Real Re-Evaluation Looks Like After a Euploid Loss

Not a longer test menu. A different sequence:

1. Retrieve the tissue result and read it properly: Array-based analysis, and whether a maternal contamination screen was run. If tissue was never sent, record that plainly — it changes what the rest of the workup can mean.

2. Re-read the PGT-A report: Including the platform, the biopsy day, and any mosaicism call reported and then set aside.

3. Assess the cavity and the uterine wall: Structure, adhesions, fibroids, and the junctional zone, where adenomyosis lives and where most reports are silent. Where it fits the picture, add an endometrial biopsy stained for CD138 to look for chronic endometritis.

4. Test for antiphospholipid syndrome correctly: On clinical criteria with the 12-week repeat, or not at all.

5. Evaluate the whole patient, not the transfer: Pain, bowel and bladder symptoms, cycle history, prior surgery, prior imaging read rather than skimmed.

6. Say out loud what is unexplained: And decide together what is worth doing and what is not.

You are allowed to want an answer. You are also allowed to be told, by someone who has actually looked, that one does not yet exist — rather than handed a treatment designed to end the conversation.

How ESSI Approaches This

After a euploid loss, most patients are offered one of two things: another transfer, or a long list of tests. We offer a structured re-evaluation, run by one physician who is trained in reproductive endocrinology, reproductive immunology and surgery.

• We start with the pregnancy tissue: We read the array result and check whether maternal cell contamination was screened. A “normal female” result without that check may be your own cells.

• We map the uterus ourselves: Dr. Vidali performs the ultrasound, with specific attention to the junctional zone and adenomyosis. Where adenomyosis is found and you want to keep the uterus, we can treat it with uterus-sparing microwave ablation. That is a reasoned option, not a proven way to prevent the next loss. Pregnancy data after it are still limited, and we plan the timing of the next transfer with you.

• We look at the cavity and the lining: With hysteroscopy and a CD138 biopsy where the history fits. We treat chronic endometritis if it is found, and we tell you plainly what the CERM trial showed.

• We test for immune causes selectively, not by default: We look for antiphospholipid syndrome on proper criteria, check thyroid, and add a limited immune profile only after the standard work-up has come back empty.

• We protect what you have: If surgery is being considered and your ovaries are at risk, the sequence is planned around your remaining embryos, not around the operating schedule.

A first consultation here is a diagnostic conversation, not a surgical booking. Your outside imaging, operative reports, PGT-A reports and pathology are read before you are seen — including the miscarriage tissue result, if there is one. You will be told what we think is happening and what nobody yet knows.

Keep Reading

• Does endometriosis cause miscarriage? A surgeon’s reading of the evidence — the full evidence review this page sits under.

• Can an ultrasound miss endometriosis? — why a normal workup is not the same finding as no disease.

• NK cells in endometriosis: why the problem is weakness, not attack — why the NK finding in endometriosis is weakness, not attack, and why it isn’t a test to buy.

References

1. Boynukalin FK, Abalı R, Gultomruk M, Yarkiner Z, Mutlu A, Bahceci M. Reprod Biomed Online. 2021;42(6):1187–1195.

2. Dong X, Shi J, Liu X, et al. Front Endocrinol. 2025;15:1461088.

3. Bishop LA, Gunn J, Jahandideh S, et al. Fertil Steril. 2021;115(2):416–422.

4. ASRM Practice Committee. Recurrent pregnancy loss: a committee opinion. Fertil Steril. 2026;125:1023–1041.

5. Kong H, Fan W, Ye T, Du L. Sci Rep. 2025;15:2193.

6. Wei Y, Xiao X, Wu X, Xie C, Li T. Front Med. 2026;13:1630529.

7. Vercellini P, Consonni D, Dridi D, Bracco B, Frattaruolo MP, Somigliana E. Hum Reprod. 2014;29(5):964–977.

8. Stanekova V, Woodman RJ, Tremellen K. The rate of euploid miscarriage is increased in the setting of adenomyosis. Hum Reprod Open. 2018;2018(3):hoy011. doi:10.1093/hropen/hoy011.

9. Cozzolino M, Cosentino M, Loiudice L, et al. Impact of adenomyosis on in vitro fertilization outcomes in women undergoing donor oocyte transfers: a prospective observational study. Fertil Steril. 2024;121(3):480–488. doi:10.1016/j.fertnstert.2023.11.034.

10. Neal S, Morin S, Werner M, et al. Three-dimensional ultrasound diagnosis of adenomyosis is not associated with adverse pregnancy outcome following single thawed euploid blastocyst transfer: prospective cohort study. Ultrasound Obstet Gynecol. 2020;56(4):611–617. doi:10.1002/uog.22065.

11. Messelt A, Dunn TN, Peregrin Alvarez I, Powe Dillon JC, Detti L. Mild adenomyosis does not impact pregnancy rate or pregnancy outcomes in single euploid frozen embryo transfers [conference abstract]. Fertil Steril. 2024;122(4):e303–e304. doi:10.1016/j.fertnstert.2024.07.929.

12. Pirtea P, Cicinelli E, De Nola R, de Ziegler D, Ayoubi JM. Fertil Steril. 2021;115(3):546–560.

13. Vomstein K, Feil K, Strobel L, et al. Immunological risk factors in recurrent pregnancy loss: guidelines versus current state of the art. J Clin Med. 2021;10(4):869.

14. Odendaal J, Black N, Fishwick K, et al. (senior author Quenby S). A double-blind multi-centre randomised placebo-controlled trial of doxycycline in women with recurrent miscarriage and chronic endometritis defined by CD138 positive stromal cells (CERM). Hum Reprod. 2026; advance article. doi:10.1093/humrep/deag133.

15. Lathi RB, Gustin SLF, Keller J, et al. Reliability of 46,XX results on miscarriage specimens: a review of 1,222 first-trimester miscarriage specimens. Fertil Steril. 2014;101(1):178–182.

16. ESHRE Guideline Group on RPL. Hum Reprod Open. 2023;2023(1):hoad002.

17. Woon EV, Greer O, Shah N, et al. Hum Reprod Update. 2022;28(4):548–582.

18. ESHRE Endometriosis Guideline Group. Hum Reprod Open. 2022;2022(2):hoac009.

19. Bafort C, Beebeejaun Y, Tomassetti C, Bosteels J, Duffy JMN. Laparoscopic surgery for endometriosis. Cochrane Database Syst Rev. 2020;10:CD011031.pub3.

20. Vidali A, Riccio LGC, Abrao MS. Endometriosis and recurrent pregnancy loss: two manifestations of the same underlying dysfunction? Fertil Steril. 2023;119(5):836–837.

Endometriosis Surgical Specialists International (ESSI)