research

Lim YH, Sim PK, Nadkarni P. Are All Frozen Embryo Transfer Protocols Equal: A Retrospective Cohort Study Comparing Single Day 3, Double Day 3, Single Blastocyst, and Double Blastocyst Transfers with Natural Cycle, Modified Natural Cycle, and Hormonal Replacement Therapy Cycles for Endometrial Preparation. Int J Infertil Fetal Med 2025;16(1):11-17. https://doi.org/10.5005/jp-journals-10016-1353

Received: 12 July 2024 | Accepted: 01 October 2024 | Published: 09 June 2025 | Last reviewed: July 2026

This page summarises original clinical research conducted at KL Fertility & Gynaecology Centre and published in a peer-reviewed journal. It is presented here for the clinical and scientific community and is not intended as a substitute for individualised patient counselling.

In This Research

Background and Rationale

The number of frozen embryo transfer (FET) cycles performed worldwide has increased substantially over the past decade. In several countries, FET cycles exceeded fresh transfers as early as 2015–2016. Contributing factors include improved stimulation protocols, advances in culture media and incubator technology, the shift from slow-freezing to vitrification, strategies to avoid ovarian hyperstimulation syndrome, and the growing use of preimplantation genetic testing. During fresh transfer cycles, supraphysiological hormone levels associated with controlled ovarian stimulation may impair embryo-endometrium synchrony and have been linked to adverse obstetric and perinatal outcomes, further driving the trend toward deferred frozen transfer.

Despite this increase, the optimal endometrial preparation protocol for FET remains debated. The three most common approaches are the natural cycle (NC), modified natural cycle (m-NC), and hormone replacement therapy (HRT) cycle. NC and m-NC protocols require the patient to ovulate and are most appropriate for women with regular menstrual cycles. HRT cycles use exogenous oestrogen and progesterone to prepare the endometrium and are typically reserved for women with anovulatory or irregular cycles, offering scheduling flexibility but requiring exogenous hormonal support throughout early pregnancy. Emerging data have raised concerns about increased adverse maternal and obstetric outcomes associated with HRT cycles, including hypertensive disorders of pregnancy, preterm delivery, premature rupture of membranes, postpartum haemorrhage, and macrosomia, plausibly linked to the absence of a corpus luteum in HRT cycles.

A recent Cochrane systematic review found insufficient evidence to favour any specific FET protocol in terms of clinical pregnancy rate or live birth rate. This study aimed to contribute additional real-world data by comparing NC, m-NC, and HRT protocols in a Malaysian single-centre cohort, across four embryo transfer strategies: single day 3 (SET-D3), double day 3 (DET-D3), single blastocyst (SBT), and double blastocyst (DBT).

Study Design and Methods: NC, m-NC, and HRT Protocols for FET

  • Design: Retrospective cohort study.
  • Subjects: 2,205 patients under 38 years of age who underwent FET at KL Fertility & Gynaecology Centre, Kuala Lumpur, Malaysia, from June 2017 to June 2021. The patient population was primarily Malaysian, with Malay, Chinese, Indian, and indigenous ethnicities represented. Inclusion: single and double autologous embryo transfers (day 3 cleavage-stage or blastocyst) using ICSI fertilisation; embryos vitrified using the Cryotop method.
  • Exclusion criteria: FET cycles involving assisted hatching, sperm donation, oocyte donation, and preimplantation genetic testing. Cases with incomplete information were also excluded.
  • Embryo grading: Day 3 embryos: Grades I–II (Grade I: 6–10 cells, symmetrical blastomeres, no fragmentation; Grade II: 6–8 cells, slightly irregular, <10% fragmentation) selected for vitrification and transfer. Blastocysts: ICM and trophectoderm of grade A, B, or C selected for vitrification and transfer.
  • Endometrial preparation groups: NC (n=826): ultrasound monitoring from days 2–3, spontaneous LH surge detected by urine testing; FET 4 or 6 days post-LH surge. m-NC (n=628): same monitoring; ovulation triggered with 5,000 IU hCG when dominant follicle reached 16 mm; FET 4 or 6 days post-trigger. HRT (n=751): oral oestradiol valerate 6 mg/day (Progynova 2 mg × 3) for 10–12 days; vaginal micronized progesterone (Utrogestan 200 mg × 3/day) commenced when endometrial thickness reached ≥7 mm; FET 4 or 6 days later. NC and m-NC groups received oral dydrogesterone 10 mg twice daily as luteal phase support.
  • Outcome measures: Clinical pregnancy rate (CPR): intrauterine sac at 6 weeks per embryo transfer. Implantation rate (IR): gestational sacs at 6 weeks per embryo transferred. Ongoing pregnancy rate (OG-PR): pregnancies passing 12 weeks per embryo transfer (note: defined in Methods as per clinical pregnancy, but reported values throughout consistent with per-ET denominator).
  • Statistical analysis: IBM SPSS Statistics 26 for Windows. Student’s t-test or Pearson’s correlation for continuous variables; chi-square for categorical data; one-way ANOVA for continuous variables. p < 0.05 considered significant.
  • Baseline age: NC 33.1 ± 2.9 years; m-NC 32.9 ± 3.0 years; HRT 32.7 ± 2.9 years. Post hoc testing showed a statistically significant difference between NC and HRT groups (p = 0.01) but no significant difference between NC and m-NC, or m-NC and HRT.

Results: CPR, Implantation, and Ongoing Pregnancy Rates by Protocol

Table 1. Overall outcomes by endometrial preparation protocol

Outcome NC (n=826) m-NC (n=628) HRT (n=751) p value
CPR/ET 56.4% 56.7% 55.7% 0.921
Implantation rate 47.3% 45.8% 46.7% 0.798
OG-PR/ET 49.3% 46.3% 46.2% 0.063

No statistically significant differences in CPR, implantation rate, or ongoing pregnancy rate were observed between NC, m-NC, and HRT cycles. Non-significant trends slightly favoured NC across all three outcomes. The mean age of patients in the HRT group was significantly lower than in the NC group (p = 0.01); no other significant age differences were observed between groups.

Table 2. CPR by embryo transfer strategy and endometrial preparation protocol

Transfer strategy NC m-NC HRT p value
SET-D3 (n=159) 41.9% 44.7% 26.0% 0.113
DET-D3 (n=342) 61.7% 54.4% 57.4% 0.526
SBT (n=1,028) 53.5% 51.4% 51.2% 0.786
DBT (n=676) 62.4% 69.4% 66.9% 0.295

No significant differences in CPR, implantation rate, or ongoing pregnancy rate were found between NC, m-NC, and HRT cycles within any of the four embryo transfer strategy subgroups. The direction of effect across subgroups was inconsistent, reflecting chance variation in small subgroups (particularly SET-D3) rather than a systematic pattern.

Discussion: How These Findings Compare With Published Evidence

The finding that NC, m-NC, and HRT protocols produce statistically comparable CPR, implantation, and ongoing pregnancy rates is consistent with the Cochrane review by Ghobara and Vandekerckhove and the systematic review by Groenewoud et al., both of which found no significant benefit of any specific endometrial preparation method for FET with respect to CPR or live birth rate.

However, an important distinction must be drawn between pregnancy rate equivalence and obstetric safety equivalence. The present study measured outcomes only through 12 weeks’ gestation and was not designed to detect differences in obstetric or perinatal outcomes. A growing body of evidence — including studies by Saito et al., Ginstrom Ernstad et al., and others — has documented higher rates of hypertensive disorders of pregnancy, preterm delivery, and other adverse obstetric outcomes following HRT (programmed) cycles compared with NC cycles, an effect attributed to the absence of corpus luteum-derived relaxin and vascular endothelial growth factor in HRT cycles. The authors of this study explicitly acknowledge this literature and its implications.

The study also briefly documents its ethnicity breakdown, noting that Chinese patients constituted the majority of those seeking IVF at this centre despite being the second-largest ethnic group in Malaysia by population. The authors suggest that studying the relationship between ethnicity and FET clinical outcomes in the Malaysian context would be a valuable direction for future research.

Mild ovarian stimulation (mild-OS) is noted by the study’s authors as an emerging alternative for patients unable to use true NC or m-NC protocols, circumventing the thromboembolism and cardiovascular risks associated with exogenous oestrogen and progesterone in HRT cycles.

Clinical Practice Implications

The absence of a significant difference in pregnancy outcomes between NC, m-NC, and HRT protocols supports clinical flexibility in endometrial preparation strategy selection. For women with regular ovulatory cycles, either NC or m-NC is appropriate and avoids the exogenous hormone burden and emerging obstetric risk profile associated with HRT cycles. HRT cycles remain indicated for patients with anovulatory or irregular cycles who cannot rely on natural ovulation for cycle timing. Where HRT is required but its obstetric risk profile is a concern, mild-OS protocols may represent a preferable alternative where clinically feasible.

Limitations of This Retrospective Cohort Study

  • Retrospective design: protocol allocation was not randomised; clinicians preferentially assigned HRT to patients with anovulatory or irregular cycles, introducing inherent selection bias that complicates direct comparison.
  • The study measured outcomes through 12 weeks’ gestation only and did not capture live birth rates, obstetric outcomes, or perinatal outcomes.
  • The statistically significant age difference between NC and HRT groups (0.4 years; p = 0.01) is unlikely to be clinically meaningful but represents a baseline difference.
  • SET-D3 subgroup numbers were small (NC: 62; m-NC: 47; HRT: 50), limiting the statistical power of subgroup analyses for this transfer type.
  • The OG-PR definition in the Methods section (per clinical pregnancy) is inconsistent with the reported values throughout (which match a per-ET denominator). Values as reported in the paper are used in this summary.
  • Single-centre study; findings may not generalise across settings with different patient populations or clinical protocols.
  • The cohort was restricted to patients under 38 years; findings may not apply to older patients.

Key Findings

  • In 2,205 patients undergoing FET, no statistically significant differences in CPR, implantation rate, or ongoing pregnancy rate were observed between NC, m-NC, and HRT endometrial preparation protocols.
  • The absence of a protocol effect held across all four embryo transfer strategy subgroups (SET-D3, DET-D3, SBT, DBT).
  • Non-significant trends slightly favoured NC across all three main outcomes.
  • These findings support clinical flexibility in protocol selection based on individual patient characteristics, particularly ovulatory function.
  • The comparable pregnancy outcomes between protocols should be interpreted alongside published evidence of adverse obstetric outcomes associated specifically with HRT cycles, which this study was not designed to assess.

About the Researchers

  • Dr. Yun-Hsuen (Helena) Lim is a Fertility Specialist and Consultant in Obstetrics & Gynaecology at KL Fertility & Gynaecology Centre (Department of Clinical Fertility). She was the corresponding author and was responsible for primary research and manuscript preparation. (ORCID: 0000-0003-4153-3033)
  • Poo Keen Sim is a member of the Scientific Team and the Chief Embryologist at KL Fertility & Gynaecology Centre, with over 20 years of hands-on experience in clinical embryology. She holds a Master’s degree in Clinical Embryology and was responsible for data collection for this study.
  • Dato’ Dr. Prashant Nadkarni is a Fertility Specialist in Reproductive Medicine at KL Fertility & Gynaecology Centre (Department of Clinical Fertility) and was responsible for reviewing and revising the manuscript.

Download the Original Research

Download PDF Now

References and Further Readings