Why a Programmed Frozen Embryo Transfer May Be Cancelled

A programmed frozen embryo transfer uses oestrogen and progesterone to prepare the uterine lining. Even when the embryo is already in storage and medication is taken as scheduled, a clinic may pause the cycle before medication, before starting progesterone or shortly before transfer. Each stage involves different checks.

How a programmed cycle is timed

Oestrogen is usually started early in the menstrual cycle to build the endometrium. Around 10–14 days later, an ultrasound typically checks the lining, with hormone testing according to the protocol. If the findings are suitable, progesterone starts the secretory phase. The transfer is timed to the embryo’s developmental stage, the start of progesterone exposure and the clinic’s protocol.[1,2]

Checks before medication

A baseline ultrasound assesses the uterine cavity, lining and ovaries. It can identify findings such as polyps, adhesions, cavity fluid or an ovarian cyst that warrants assessment. The clinical team decides which other checks, such as an ECG, thyroid testing, blood pressure or blood glucose, should be completed in advance based on medical history and clinic requirements.

There is little direct evidence on the proportion of programmed cycles deferred before medication starts. In the TROPHY randomised trial in women with recurrent IVF failure, hysteroscopy identified abnormalities in 85 of 323 women (26%). Live birth was 29% in both arms. In the inSIGHT trial, women with a normal uterine cavity on ultrasound who were scheduled for their first IVF treatment had live-birth rates of 57% with hysteroscopy and 54% without it. The difference was not statistically significant. These trials support using hysteroscopy when there is a clinical indication.[3,4]

A small cross-sectional study reported chronic endometritis in 10.4% of infertile women preparing for frozen embryo transfer and 5.0% of fertile controls. The difference was not statistically significant. Detection rates for these findings describe the populations studied. They do not measure how often cycles are deferred before medication.[5]

Checks before progesterone

A thin lining

A retrospective analysis across 33 Canadian clinics included 18,942 cycles that proceeded to frozen embryo transfer. Around 2.9% had a lining thinner than 7 mm and 0.7% had a lining thinner than 6 mm. Among completed transfers, live birth was 23.7% at 6–6.9 mm and 28.4% at 8 mm or above. Clinics may use different thresholds when deciding whether to proceed. A separate single-centre case series reported 30 cases of treatment-resistant thin endometrium across 8,363 frozen embryo transfer cycles.[6,7]

An early progesterone rise or spontaneous ovulation

If endogenous progesterone rises before the scheduled medication, the secretory phase may start ahead of schedule. The team reviews blood tests, ultrasound findings and medication before deciding whether to proceed. Follicles can still develop or ovulate in programmed cycles without prior suppression. A review reported frequencies of approximately 1.9%–7.4%. In the COMPETE trial, 29 of 454 participants assigned to the programmed group (6.4%) switched to a natural cycle after spontaneous ovulation. Whether that option is available depends on confirming ovulation timing and meeting the other transfer criteria.[2,9]

In a single-centre retrospective cohort in Türkiye, 66 of 2,001 programmed cycles (3.3%) were cancelled for elevated progesterone. The centre used a threshold of ≥1.5 ng/mL on the scheduled day of progesterone initiation. Cancellation rates were 2.1%, 3.9% and 5.8% after 8–9, 10–11 and 12–14 days of oestradiol, respectively. The rates were 14.2% with diminished ovarian reserve, 2.6% with normal reserve and 0.5% with polycystic ovary syndrome. These figures reflect that centre’s patients and cancellation rule.[8]

Duration of oestrogen exposure

A retrospective study of 2,010 cycles resulting in singleton live birth reported that 37.1% used oestrogen for 13–15 days, 32.5% for 16–19 days and 11.9% for 20 days or more. Those figures describe cycles that had already resulted in a live birth. A separate observational study of 1,377 single-blastocyst cycles found live-birth rates of 29.7%, 27.8%, 21.8% and 17.2% across groups treated for ≤21, 22–28, 29–35 and 36–48 days, respectively. Longer exposure was associated with a lower live-birth rate. Decisions about extending a particular cycle depend on its follow-up findings.[10,11]

Cavity fluid or a thick lining near transfer

A retrospective study of 4,308 single-euploid frozen embryo transfer cycles reported live-birth rates of 58.5% without endometrial cavity fluid, 49.1% after fluid resolved spontaneously and 38.3% with persistent fluid. After adjustment, the resolved group did not differ significantly from the no-fluid group, while persistent fluid was associated with lower odds of live birth. In another 58 cycles transferred after a prior cancellation for fluid, the live-birth rate was 39.7%. The study did not randomise cancellation, so it could not establish which response gives a better outcome. With recurrent or persistent fluid, a team may evaluate causes such as hydrosalpinx or a caesarean scar niche.[14]

Evidence on a thick lining is mixed. A review cited an observational finding of lower implantation rates above 13 mm. A conference abstract covering 4,911 single-euploid transfers did not detect a significant decline above 15 mm after adjustment. When the lining looks thick, the team can consider ultrasound appearance and investigate findings such as polyps or hyperplasia.[2,15]

How often transfer was cancelled or did not take place

The service records covered 86 cycles in 61 women. Transfer took place in 31 cycles, and 33 women had no transfer during the observation period. Seventeen women had two or more cycles, of whom nine never reached transfer during that period. Recorded reasons included lining thickness, an early rise in progesterone and spontaneous ovulation. In a few cycles, an ECG abnormality was found after medication had started. The series is small, includes repeat cycles in the same women and may reflect different clinical criteria from the trials. It describes this group’s experience and does not estimate cancellation rates in the general population.

What to discuss after a cycle is paused

• At what stage did the cycle stop? What did the ultrasound, progesterone or other tests show, and what criteria does the clinic use?
• If a thin lining, an early progesterone rise or spontaneous ovulation recurs, could monitoring or medication timing change? Is a natural cycle an option?
• Which tests could be completed before medication starts? If fluid recurs, when should it be reassessed and what causes could be investigated?

A programmed cycle can simplify scheduling. For women who ovulate regularly, a natural cycle is also an option to discuss. A 2025 Cochrane review covered 32 randomised trials and 6,352 women, with limited certainty for several comparisons. A 2026 multicentre trial in ovulatory women observed similar rates of healthy live birth. Among those who achieved a clinical pregnancy, pre-eclampsia occurred in 2.9% in the natural ovulation group and 4.6% in the programmed group. Earlier observational research also found an association with hypertensive disorders of pregnancy. In a cohort of women with a prior caesarean delivery, early miscarriage was more common with programmed cycles, while the live-birth rate did not differ significantly.[1,16,17,18]

Medical information and data

This article is for education. The treating clinician assesses test results and decides whether to adjust medication, cancel or change the next protocol. The author’s aggregate service records have not been peer-reviewed. The artwork is conceptual.