Home / Resources / Why did my IVF cycle fail?

Patient guide · reviewed by Dr Sarafis

Why did my IVF cycle fail?

The questions your clinic may not have answered — and how to tell whether the embryo was the problem, or something nobody looked for.

By Dr Vasileios Sarafis MD, UK-trained fertility specialist, Thessaloniki and Athens. Last reviewed 16 August 2026.

A failed cycle rarely comes with a satisfying explanation. Most patients are told the embryo “didn’t take”, offered a slightly different protocol, and booked in again. Sometimes that genuinely is the right course. But the difference between a second cycle and a second cycle with a plan comes down to whether anyone has established which of two very different situations you are in.

1. The embryo is the most likely explanation — and the least well explained

The single largest cause of IVF failure is embryo aneuploidy: an embryo with the wrong number of chromosomes. These embryos frequently look normal down a microscope, grade well, and are transferred in good faith. They then fail to implant, or implant and stop.

The proportion of embryos affected increases substantially with maternal age. This is why a woman of 41 may need several transfers to achieve what a woman of 32 achieves on the first — not because her treatment is worse, but because more of her embryos are not chromosomally viable. Embryo grading, which describes appearance, is a weak proxy for chromosomal status.

Approximate figures pooled from published cohort data on chromosomally tested embryos. Individual results vary considerably, and these are averages rather than predictions for any one person.

The practical implication matters. If you are over 38 and transferred embryos that were never chromosomally tested, there is a substantial probability that the embryo itself was the reason — not your uterus, your lining, or your protocol. That does not make the loss easier. It does change what is worth doing next.

Should you simply add genetic testing?

Not automatically. Pre-implantation genetic testing for aneuploidy (PGT-A) identifies chromosomally abnormal embryos before transfer, and in selected groups it can reduce the number of failed transfers and miscarriages. What the randomised evidence has not reliably shown is an improvement in the chance of a live birth per patient who starts treatment. It changes which embryo is transferred first; it does not create embryos that were not there. Whether it helps you depends on your age, embryo numbers and history — which is precisely the kind of decision worth discussing properly rather than adopting as a default.

2. When the embryo is probably not the explanation

Once you have transferred two or more good-quality embryos — particularly if any were chromosomally tested and normal — without a pregnancy, the statistical case for “bad luck” weakens considerably. At that point the sensible question is no longer which protocol next? but what has never actually been examined?

3. What a standard work-up often does not include

These are the factors I most often find have never been assessed in patients arriving for a second opinion after repeated failure. Not all are relevant to every case, and several are only worth testing when the history points to them.

• The uterine cavity, properly examined. A standard ultrasound infers the cavity; hysteroscopy sees it. Small polyps, fibroids distorting the cavity, adhesions and residual septa are all correctable and all missable on routine scanning.
• Chronic endometritis. A persistent, low-grade inflammation of the lining. It is common in women with repeated implantation failure, almost always silent, diagnosed on endometrial biopsy (CD138 immunostaining) or targeted molecular testing — and usually treatable with a course of antibiotics.
• Adenomyosis. Endometrial tissue within the muscle of the uterus. It is frequently missed unless it is specifically looked for, and it changes both prognosis and the plan before transfer.
• Hydrosalpinx. A blocked, fluid-filled fallopian tube measurably lowers implantation rates. It is identifiable and surgically correctable, and it should not survive a thorough work-up unnoticed.
• Sperm DNA fragmentation. A normal semen analysis — count, motility, morphology — does not exclude damage to the DNA inside those sperm. It is worth measuring in recurrent implantation failure, recurrent loss, and unexplained poor embryo development.
• Thyroid, metabolic and nutritional status. Thyroid function, insulin resistance, ferritin and vitamin D are cheap to measure and, where genuinely abnormal, correctable before the next attempt.
• The endometrium at the time of transfer. Thickness alone is a crude measure. Pattern and blood flow on Doppler add information that a single millimetre reading does not.

4. What the evidence does not support — and why I tell you

A great deal is sold to patients after a failed cycle. An honest account has to include the things that were tested properly and did not help:

• Routine immune testing and treatment. Natural killer cell panels, steroids, intralipids and immunoglobulin infusions are widely offered for repeated failure. The evidence supporting routine use remains weak, and some of these treatments carry real risk. There are selected situations worth discussing; “everyone with two failures” is not one of them.
• Treating borderline thyroid results. The TABLET and T4LIFE randomised trials tested exactly this and found no improvement in live birth for most women. We test the thyroid to find genuine disease, not to medicate a borderline number.
• Routine endometrial receptivity timing (ERA-type testing). Randomised evidence has not demonstrated benefit when applied to unselected patients. It remains a case-by-case discussion, not a standard add-on.
• Multi-supplement protocols taken as a whole. No trial has ever tested a complete supplement stack as a package — including mine. Correcting a documented deficiency is a different proposition, and that is worth doing.

If a clinic offers you all of the above at once without explaining which of your findings justified each one, that is a reasonable moment to ask for the reasoning in writing.

5. “Unexplained” often describes the work-up, not you

When couples carrying a diagnosis of unexplained infertility were re-examined with laparoscopy in published series, identifiable pathology was found in roughly 58% — most often endometriosis, tubal disease or adhesions. The label is sometimes an accurate description of a difficult case. Just as often, it is a description of how far the investigation went.

6. Six questions worth asking at your follow-up appointment

These are collaboration questions, not confrontation. Good clinicians welcome them.

1. What did the embryology actually tell us — fertilisation rate, development pattern, grading at each stage?
2. Has my uterine cavity been examined directly, or only inferred from a scan?
3. Has chronic endometritis been excluded, and how?
4. Which factors on the standard list have never been tested in my case?
5. What specifically would change in the next protocol, and what is the reasoning?
6. What does the evidence say about that change — and what would we expect it to alter?

This page is educational and does not constitute medical advice. Which investigations are appropriate depends on your individual clinical history and previous treatment. For guidance specific to your situation, book a consultation.