If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Obstetrics and Gynecology Department, Universitat Autònoma de Barcelona, Campus Universitario UAB, Bellaterra Cerdanyola del Vallès 08193, SpainClínica de la Mujer Medicina Reproductiva, Alejandro Navarrete 2606, Viña del Mar, Chile
Obstetrics and Gynecology Department, Universitat Autònoma de Barcelona, Campus Universitario UAB, Bellaterra Cerdanyola del Vallès 08193, SpainFertty International, Carrer d'Ausiàs Marc, 25, Barcelona 08010, SpainDepartment of Obstetrics and Gynaecology, Hospital de la Santa Creu i Sant Pau, Carrer de Sant Quintí, 89 Barcelona 08041, Spain
Obstetrics and Gynecology Department, Universitat Autònoma de Barcelona, Campus Universitario UAB, Bellaterra Cerdanyola del Vallès 08193, SpainFertty International, Carrer d'Ausiàs Marc, 25, Barcelona 08010, SpainGRI-BCN, Barcelona Infertility Research Group, IMIM, Institut Hospital del Mar d'Investigacions Mèdiques, Carrer del Dr. Aiguader, 88, Barcelona 08003, Spain
NC-FET is associated with a lower risk of hypertensive disorders of pregnancy
•
The risk of LGA and macrosomia may be reduced in newborns after NC-FET
•
Postpartum hemorrhage and placenta accreta are other outcomes improved after NC-FET
•
Cesarean section rates are high in FET, both in natural and artificial cycles
Abstract
This systematic review of literature and meta-analysis of observational studies reports on perinatal outcomes after frozen embryo transfer (FET). The aim was to determine whether natural cycle frozen embryo transfer (NC-FET) in singleton pregnancies conceived after IVF decreased the risk of adverse perinatal outcomes compared with artificial cycle frozen embryo transfer (AC-FET). Thirteen cohort studies, including 93,201 cycles, met the inclusion criteria. NC-FET was associated with a lower risk of hypertensive disorders in pregnancy (HDP) (RR 0.61, 95% CI 0.50 to 0.73), preeclampsia (RR 0.47, 95% CI 0.42 to 0.53), large for gestational age (LGA) (RR 0.93, 95% CI 0.90 to 0.96) and macrosomia (RR 0.82, 95% CI 0.69 to 0.97) compared with AC-FET. No significant difference was found in the risk of gestational hypertension and small for gestational age. Secondary outcomes assessed were the risk of preterm birth (RR 0.83, 95% CI 0.79 to 0.88); post-term birth (RR 0.48, 95% CI 0.29 to 0.80); low birth weight (RR 0.84, 95% CI 0.80 to 0.89); caesarean section (RR 0.84, 95% CI 0.77 to 0.91); postpartum haemorrhage (RR 0.39, 95% CI 0.35 to 0.45); placental abruption (RR 0.61, 95% CI 0.38 to 0.98); and placenta accreta (RR 0.18, 95% CI 0.10 to 0.33). All were significantly lower with NC-FET compared with AC-FET. In assessing safety, NC-FET significantly decreased the risk of HDP, preeclampsia, LGA, macrosomia, preterm birth, post-term birth, low birth weight, caesarean section, postpartum haemorrhage, placental abruption and placenta accreta. Further randomized controlled trials addressing the effect of NC-FET and AC-FET on maternal and perinatal outcomes are warranted. Clinicians should carefully monitor pregnancies achieved by FET in artificial cycles prenatally, during labour and postnatally.
European IVF-monitoring Consortium (EIM) for the European Society of Human Reproduction and Embryology (ESHRE) ART in Europe, 2015: results generated from European registries by ESHRE.
European IVF-monitoring Consortium (EIM) for the European Society of Human Reproduction and Embryology (ESHRE) ART in Europe, 2015: results generated from European registries by ESHRE.
). The freeze-all strategy has reported successful results with higher live birth rates and lower incidence of ovarian hyperstimulation syndrome compared with fresh embryo transfer (
). Some studies have suggested that singletons born after FET also have better neonatal outcomes compared with singletons born after fresh embryo transfer, e.g. low birth weight (LBW), small for gestational age (SGA) and preterm birth (PTB) (
Compared with fresh embryo transfer, FET seems to carry a greater risk of hypertensive disorders in pregnancy (HDP), including gestational hypertension and preeclampsia (Opdahl et al., 2015;
Increased incidence of obstetric and perinatal complications in pregnancies achieved using donor oocytes and single embryo transfer in young and healthy women. A prospective hospital-based matched cohort study.
). The reasons behind these findings are not clearly understood, but it has been suggested that some cryoprotectants or the freeze–thawing process per se could develop some metabolic or epigenetic changes related to abnormal placentation and eventually preeclampsia (
The role of the endometrium has also been a focus of attention. Different options for preparing the endometrium for FET have been described, including a natural cycle (NC-FET) based on the detection of the endogenous LH surge in the blood, a modified natural cycle using HCG for final oocyte maturation and artificial cycle (AC-FET) based on an hormonal replacement treatment with or without co-treatment with a gonadotrophin releasing hormone analogue, and a stimulated cycle with anti-oestrogens, aromatase inhibitors or gonadotrophins (
), in which oestrogens are administered with the aim of mimicking the changes generated by steroids of ovarian origin that occur in natural cycles. Endometrial thickness and pattern are monitored by vaginal ultrasound, and progesterone administration is usually started from an endometrial thickness of about 7–8 mm (
). In clinical practice, AC-FET is popular because it involves less monitoring and the embryo transfer can be scheduled on a convenient day for the patient and the clinic (
In recent years, multiple studies have investigated the effectiveness of different endometrial preparation schemes in relation to implantation rates, clinical pregnancy rates or live birth rates (
). Studies evaluating perinatal and maternal outcomes, however, are scarce and have reported contradictory results. Some observational data have suggested higher rates of HDP, LGA and macrosomia after programmed FET cycles compared with NC-FET (
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
The reason behind the high rates of adverse obstetric outcomes is not known, but it has been suggested that some endometrial changes mediated by the altered levels of oestradiol and progesterone, possibly reached during the AC-FET, could be associated with the development of impaired decidualization and placentation, leading to placenta-related complications, such as HDP, placenta accreta, placenta previa and placental abruption (
In addition, the hypothalamic–pituitary–gonadal axis is inhibited by exogenous oestradiol. This can lead to a lack of corpus luteum. The corpus luteum produces reproductive hormones, including the vasoactive hormone relaxin. Recently published studies have found that scheduled FET cycles, in which no corpus luteum is present, are associated with higher rates of preeclampsia compared with natural FET cycles, in which one or more corpus luteum occur (
Given the increasing use of FET, it is critical to determine whether specific FET protocols could be related to the development of adverse obstetric and maternal outcomes, and if elements of the treatment could be modified to optimize outcomes. Therefore, the aim of this systematic review was to determine whether NC-FET decreased the risk of adverse perinatal outcomes compared with AC-FET.
Materials and methods
Search strategy and selection criteria
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were used for the present systematic review and meta-analysis (
). The study protocol can be accessed at https://inplasy.com/ (registration number INPLASY202060088).
The selection criteria were described according to Patients, Intervention, Comparison and Outcomes (PICO) statements. Studies that compared the reproductive outcomes of deliveries from autologous oocytes between NC-FET and AC-FET were included (Table 1).
Modified natural cycles: ovulation was stimulated with 10,000 IU of HCG. Artificial cycles: 4–6 mg/day of oral oestradiol starting on day 2–4 of the natural menstrual cycle. The oestradiol dosage was adjusted based on the endometrial thickness and level of serum oestradiol. Intramuscular progesterone administration was started after adequate endometrial proliferation (diameter ≥8 mm) and serum oestradiol concentration (200–300 ng/l) were documented. Not included in the meta-analysis: GnRH agonist artificial cycles; ovulation induction cycles.
Unstimulated cycles normal ovulatory cycle. Programmed FET protocol: GnRH agonist suppression in the luteal phase and then oral oestroadiol initiated at a dose of 2 mg daily and titrated to 6 mg daily over 12 days. Intramuscular progesterone was started at 50 mg when appropriate parameters were met.
Clinical pregnancies after frozen embryo transfer cycles. AC-FET, artificial cycle frozen embryo transfer; FET, frozen embryo transfer; GnRH, gonadotrophin releasing hormone; NC-FET, natural cycle frozen embryo transfer; NE, not explained; NOS, Newcastle–Ottawa Scale; PGT-A, preimplantation genetic testing for aneuploidy.
NC-FET: 31 (28–35) AC-FET: 31 (28–35)
Natural cycles: normal ovulatory cycle. Artificial cycle: oestrogen valerate 2 mg was administered orally: one pill on days 1, 2, 3 and 4; two pills on days 5, 6 and 7; three pills on days 8, 9, 10 and 11; four pills on days 12, 13, 14, 15 and 16; and two pills on days 17 to 31. Dydrogesterone was administered orally (10 mg per 12 h) when the endometrial thickness reached at least 8 mm.
Blastocyst and cleavage stage
Vitrification
NE
Hypertensive disorder of pregnancy; large for gestational age; macrosomia; pre-term birth; post-term birth; low birth weight; very low birth weight; small for gestational age; stillbirth; neonatal mortality; caesarean section; gestational diabetes.
Natural cycles: normal ovulatory cycle (one corpus luteum). Programmed cycles: included oestrogen and progesterone with or without suppression with a GnRH agonist or antagonist (no corpus luteum). Stimulated cycles were not included in the meta-analysis.
Blastocyst and cleavage stage
Vitrification and slow freezing
NE
Hypertensive disorder of pregnancy; preeclampsia; gestational hypertension; large for gestational age; macrosomia; pre-term birth; post-term birth; low birth weight; very low birth weight; small for gestational age; stillbirth; birth defects; neonatal mortality; caesarean section; postpartum haemorrhage; placental abruption; placenta previa.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Natural cycles: spontaneous conceptions, intrauterine insemination, or FETs in modified natural cycles with one corpus luteum. Programmed cycle with no corpus luteum. Patients with ovulation induction with timed intercourse, intrauterine insemination or IVF with fresh embryo transfer were not included in the meta-analysis.
Blastocyst and cleavage stage
NE
NE
Hypertensive disorder of pregnancy; preeclampsia; gestational hypertension.
Natural cycles (normal ovulatory cycle). Programmed cycle: downregulation with GnRH agonist in the luteal phase followed by increasing doses of oral or transdermal oestradiol after menses. Intramuscular progesterone was started when endometrial thickness measured 8 mm.
Natural cycle: the dominant follicle was monitored by transvaginal ultrasound until ovulation was achieved with or without the trigger of HCG. Programmed cycle regimen: oestrogen, at a dose of 4–6 mg daily, was initiated on the second or third day of the menstrual cycle and lasted for 10–14 days until the endometrial thickness reached at least 8 mm.
Blastocyst
Vitrification
NE
Preeclampsia; large for gestational age; macrosomia; pre-term birth; post-term birth; low birth weight; very low birth weight; small for gestational age; caesarean section; postpartum haemorrhage; gestational diabetes; placental abruption; placenta previa.
Natural cycle: when the dominant follicle was detected, the investigators decided whether to use HCG for ovulation triggering according to their clinical routine. Hormone replacement cycle: the endometrium was prepared with oral oestradiol valerate at a dose of 4–8 mg daily, starting on days 1–3 of the menstrual cycle. Vaginal progesterone gel 90 mg/day and oral dydrogesterone 10 mg twice daily were added when the endometrial thickness reached 7 mm or more.
Blastocyst
Vitrification
NE
Hypertensive disorder of pregnancy; preeclampsia; gestational hypertension; large for gestational age; preterm birth; small for gestational age; postpartum haemorrhage; gestational diabetes; placenta previa.
Natural cycle: when the dominant follicle was detected the investigators decided whether to use HCG for ovulation triggering according to their clinical routine. Hormone replacement cycle: oral oestradiol valerate was given daily at a dose of 4–8 mg started on the 1–3 day of the period. When the endometrial thickness reached 7 mm or more, twice-daily oral dydrogesterone (10 mg) and vaginal progesterone gel (90 mg/day) were added.
Cleavage stage
Vitrification
NE
Hypertensive disorder of pregnancy; large for gestational age; macrosomia; preterm birth; post-term birth; low birth weight; small for gestational age; caesarean section; gestational diabetes.
Natural cycle: cycles without ovarian stimulation and with physiological oestradiol levels during the follicular phase. HCG was used to trigger ovulation. Artificial cycle: oral 17b-oestradiol 2 mg, three times daily was commenced on the second or third day of menstrual cycle. When the endometrial thickness was 8 mm, vaginal progesterone suppositories were initiated. Stimulated FET cycles were excluded from the meta-analysis.
Cleavage stage and blastocyst
Vitrification
No
Hypertensive disorder of pregnancy; large for gestational age; macrosomia; preterm birth; low birth weight; very low birth weight; small for gestational age; gestational diabetes; placenta previa.
Natural cycles: HCG was administrated when the diameter of dominant follicle reaching 18 mm or more. Hormonal replacement cycles: patients were prescribed with 4 mg oral oestradiol valerate from day 2–4 of menstruation for 5–6 days, and then 6 mg for the following 5–6 days. The dose of oestradiol valerate was modulated according to the endometrial thickness and the oestradiol levels. When the endometrial thickness reached at least 7 mm, FET was scheduled in 5 days. Cycles with ovulation induction were not included in the meta-analysis.
Blastocyst
Vitrification
No
Hypertensive disorder of pregnancy; large for gestational age; pre-term birth; low birth weight; small for gestational age; gestational diabetes; placenta previa
9/9
a Frozen embryo transfer cycles.
b Live births after frozen embryo transfer cycles.
c Clinical pregnancies after frozen embryo transfer cycles.AC-FET, artificial cycle frozen embryo transfer; FET, frozen embryo transfer; GnRH, gonadotrophin releasing hormone; NC-FET, natural cycle frozen embryo transfer; NE, not explained; NOS, Newcastle–Ottawa Scale; PGT-A, preimplantation genetic testing for aneuploidy.
An electronic search of databases was conducted from 1982 to March 2020. These included PubMed, Scopus and the Cochrane Database. Reference lists of relevant articles were also searched for any additional studies not covered by the literature search. The search combined terms and descriptors related to variants for the interventions, population study and outcomes: IVF with or without intracytoplasmic sperm injection; frozen–thawed embryo transfer; endometrial preparation; natural cycle; hormone replacement cycle; hypertensive disorders in pregnancy; preeclampsia; large for gestational age; macrosomia; preterm birth; post-term birth; low birth weight, very low birth weight, small for gestational age, stillbirth, neonatal mortality, gestational diabetes; caesarean section; placenta previa; placenta accrete; placental abruptio; and postpartum haemorrhage. The search strategy was modified to fit with the syntaxes used in each database consulted.
Study selection and data extraction
All the abstracts retrieved from the search in a first screening were assessed by two researchers (JM and MC) independently. The full text of citations that met the inclusion criteria were screened. Both authors judged study eligibility, assessed quality and extracted data, and any discrepancies were resolved by agreement; if required, a consensus was reached with the involvement of a third author (JE). The extracted data were summarized for each outcome (Table 2 and Table 3). The authors referred to the Grading of Recommendations Assessment, Development and Evaluation (GRADE) to evaluate the quality of evidence for each outcome (
Study quality was assigned by two reviewers (JM and MC) following the guidelines described in the Newcastle–Ottawa Scale (NOS) for assessing the quality of included studies (
). The studies were assessed for bias in the selection process, comparability of cohorts and ascertainment of outcomes. The quality assessment and risk of bias of the included studies are presented in Table 1.
Outcome measures
The primary outcome measure was the rate of adverse pregnancy outcomes, including HDP (blood pressure of 140/90 mmHg or above on two or more occasions, at least 6 h apart and more than 20 weeks of gestation); gestational hypertension (hypertension arising de novo after 20 weeks’ gestation in the absence of proteinuria and without biochemical or haematological abnormalities); preeclampsia (the presence of de-novo hypertension after 20 weeks’ gestation accompanied by proteinuria and any of the following: evidence of maternal acute kidney injury; liver dysfunction; neurological features; haemolysis or thrombocytopaenia; fetal growth restriction (
International Society for the Study of Hypertension in Pregnancy (ISSHP) Hypertensive Disorders of Pregnancy: ISSHP Classification, Diagnosis, and Management Recommendations for International Practice.
); large for gestational age (LGA) (birth weight above the 90th percentile); and macrosomia (birth weight above 4000 g).
Secondary outcome measures were PTB (live birth before 37 weeks); post-term birth (a live birth after 42 weeks); low birth weight (LBW) (a birth weight below 2500 g); very low birth weight (a birth weight below 1500 g); small for gestational age (SGA) (birth weight under two SD or below the 10th percentile); stillbirth (the death of a fetus before the complete expulsion or extraction from its mother after 22 completed weeks of gestational age); neonatal mortality (death before 28 days postpartum); gestational diabetes; caesarean section; placenta previa; placenta accrete; placental abruption; and postpartum haemorrhage.
The definitions used were in line with the definitions prescribed by the International Society for the Study of Hypertension in Pregnancy (
International Society for the Study of Hypertension in Pregnancy (ISSHP) Hypertensive Disorders of Pregnancy: ISSHP Classification, Diagnosis, and Management Recommendations for International Practice.
) and according to ICD-10 codes gathered from the maternal hospital discharge data in the different articles included in this study.
Quantitative analysis
To ascertain the pooled effect of different variables, a Mantel–Haenszel model and fixed-effects model were used. The risk ratios were calculated for dichotomous data, along with the 95% confidence intervals. The extent of dissimilarity between studies attributable to heterogeneity with the I squared statistic (I2) was assessed. The random-effects model (
) was used in cases in which the heterogeneity was greater than 50% (I2 >50%). The Review Manager (RevMan Version 5.3 Software, Copenhagen, Denmark) was used for statistical analysis.
Results
The search yielded a total of 1903 records but 1868 were excluded after the titles and abstracts of these manuscripts were screened. Of the remaining 35 studies that were considered eligible by one or both reviewers, 22 were excluded, five of which were reviews and 17 did not report maternal and neonatal outcomes. The remaining 13 studies were included in the present systematic review and meta-analysis (
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Thirteen studies investigating perinatal and maternal outcomes in pregnancies after NC-FET compared with AC-FET met the inclusion criteria. Details of the included studies are presented in Table 1. All the studies had NOS scores above 7, which is considered to be of high quality.
Primary outcomes
Hypertensive disorders in pregnancy
Ten studies reported HDP, including 26,661 patients in the NC-FET group and 23,558 patients in the AC-FET group. The overall relative risk for HDP was 0.61 (95% CI 0.50 to 0.73; I2 = 65%) (Figure 2a), favouring the NC-FET group. The quality of evidence was low according to GRADE. A sub-analysis considering preeclampsia or gestational hypertension separately was conducted.
Figure 2Forest-plots comparing primary outcomes after natural cycle (NC) frozen embryo transfer and artificial cycle (AC) frozen embryo transfer. (A) Hypertensive disorders in pregnancy; (B) preeclampsia; and (C) gestational hypertension. AC-FET, artificial cycle frozen embryo transfer; NC-FET, natural cycle frozen embryo transfer.
Women who gave birth after NC-FET were at a decreased risk of preeclampsia compared with those after AC-FET (four studies, 22,856 patients, RR 0.47, 95% CI 0.42 to 0.53, I2 = 4%; low quality of evidence) (Figure 2b). No differences were found between the NC-FET and AC-FET groups in risk of gestational hypertension (three studies, 8383 patients, RR 0.72, 95% CI 0.51 to 1.02, I2 = 0%; low quality of evidence) (Figure 2c).
Large for gestational age
Nine studies (n = 77,629 patients) related to the complication of LGA. A lower risk of LGA (RR = 0.93, 95% CI 0.90 to 0.96, I2 = 0%) was found in pregnancies after NC-FET cycles than after AC-FET (low quality of evidence) (Figure 3a).
Figure 3Forest-plots comparing primary outcomes after natural cycle (NC) frozen embryo transfer and artificial cycle (AC) frozen embryo transfer. (A) Large for gestational age; and (B) macrosomia. AC-FET, artificial cycle frozen embryo transfer; NC-FET, natural cycle frozen embryo transfer.
Nine studies (n = 72,502 patients) provided the outcome of macrosomia. A lower risk of macrosomia was found in pregnancies after NC-FET cycles than after AC-FET cycles (RR = 0.82, 95% CI 0.69 to 0.97, I2 = 72%; low quality of evidence) (Figure 3b).
The results for the primary outcomes are presented in Table 2.
Secondary outcomes
Preterm birth
Eleven studies were involved in the meta-analysis (n = 78,200 patients). Analysis showed that the overall risk of PTB was significantly lower among pregnancies resulting from the NC-FET cycles (RR = 0.83, 95% CI 0.79 to 0.88, I2 = 41%; low quality of evidence) (Supplementary Figure 1A).
Post-term birth
Seven studies (n = 67,262 patients) reported results comparing the post-term rates for each group. The overall relative risk for post-term birth was 0.48 (95% CI 0.29 to 0.80, I2 = 81%; very low quality of evidence) (Supplementary Figure 1B) favouring the NC-FET group.
Low birth weight
Nine studies reported the prevalence of LBW (n = 78,272 patients). The risk of LBW was lower after NC-FET compared with AC-FET (RR = 0.84, 95% CI 0.80 to 0.89, I2 = 22%; low quality of evidence) (Supplementary Figure 1C).
Very low birth weight
Four studies (n = 29,069 patients) were analysed. The overall risk of very low birth weight was not significantly different among pregnancies resulting from the NC-FET and AC-FET cycles (RR = 0.91, 95% CI 0.65 to 1.27, I2 = 28%; low quality of evidence) (Supplementary Figure 1D).
Small for gestational age
Nine studies (n = 77,629 patients) reported SGA data. The overall risk of SGA observed after NC-FET compared with AC-FET was not significantly different (RR = 0.94, 95% CI 0.88 to 1.00, I2 = 0%; low quality of evidence) (Supplementary Figure 2A).
Stillbirth
Seven studies (n = 65,311 patients) were used in this analysis. The overall risk of stillbirth was not significantly different among the pregnancies resulting from the NC-FET and AC-FET cycles (RR = 0.88, 95% CI 0.66 to 1.18, I2 = 15%; very low quality of evidence) (Supplementary Figure 2B).
Neonatal mortality
Three studies (n = 11,931 patients) were analysed. Overall, no difference in the risk of neonatal mortality was found among pregnancies resulting from NC-FET and AC-FET cycles (RR = 0.55, 95% CI 0.22 to 1.37, I2 = 0%; very low quality of evidence) (Supplementary Figure 2C).
Birth defects
Three studies (n = 9117 patients) reported birth defects data. The risk of birth defects was not significantly different in pregnancies resulting from the NC-FET than from AC-FET (RR 0.86, 95% CI 0.67 to 1.11, I2 = 0%; very low quality of evidence) (Supplementary Figure 2D).
Caesarean section
Seven studies (n = 70,020 patients) were pooled in this analysis. A significantly lower risk in caesarean section was observed after NC-FET group compared with the AC-FET group (RR = 0.84, 95% CI 0.77 to 0.91, I2 = 95%; very low quality of evidence) (Supplementary Figure 3A) but heterogeneity was substantial.
Postpartum haemorrhage
Four studies (n = 23,343 patients) were involved in the meta-analysis. A significantly lower risk of postpartum haemorrhage was observed in the NC-FET group compared with the AC-FET group (RR = 0.39, 95% CI 0.35 to 0.45, I2 =30%; low quality of evidence) (Supplementary Figure 3B).
Gestational diabetes
Nine studies (n = 65,628 patients) were analysed. The overall risk of gestational diabetes was not significantly different among pregnancies resulting from the NC-FET and AC-FET cycles (RR = 1.08, 95% CI 0.80 to 1.47). Marked heterogeneity was observed among the studies (I2 = 90%; very low quality of evidence) (Supplementary Figure 3C).
Placental abruption
Four studies (n = 46,986 patients) were involved in the meta-analysis. Analysis showed that the overall risk of placental abruption was significantly lower among pregnancies resulting from the NC-FET cycles (RR = 0.61, 95% CI 0.38 to 0.98; I2 = 0%; low quality of evidence) (Supplementary Figure 3D).
Placenta previa
Seven studies, including 57,356 patients, evaluated rates of placenta previa. No difference was found in the placenta previa rates between NC-FET and AC-FET cycles (RR = 0.85, 95% CI 0.59 to 1.23, I2 = 69%; low quality of evidence) (Supplementary Figure 3E).
Placenta accreta
Two studies (n = 35,755 patients) were pooled in this meta-analysis. Overall, the risk of placenta accreta was significantly lower in the NC-FET cycles compared with the AC-FET cycles (RR = 0.18, 95% CI 0.10 to 0.33, I2 = 0%; low quality of evidence) (Supplementary Figure 3). The results for the secondary outcomes are presented in Table 3.
Sensitivity analysis
Sensitivity analyses were conducted for the primary outcomes to examine the influence of variation among studies on the overall risk estimates. The pooled effect size was not significant (Supplementary Figure 4 to Supplementary 6 and Supplementary Table).
Discussion
This systematic review and meta-analysis show a decrease in the risk of HDP, preeclampsia, LGA and macrosomia with the use of NC-FET in preference to AC-FET in the overall population undergoing FET. Low-quality evidence also shows that the use of NC-FET does not lead to any differences in the chances of gestational hypertension.
The GRADE quality of evidence was low mainly because this is a review based on observational studies and because of the substantial inter-study heterogeneity obtained, which was assumed to be caused by the variation between study populations.
Strengths
The large sample size is a major strength of this study; the selected articles included two population-based national registries from Sweden and Japan. This large sample allowed the investigation of relatively infrequent events, i.e. minor obstetrical complications.
Studies that used suitable research designs that matched for potential confounders, i.e. multiple births and maternal age, were considered. Furthermore, the meta-analysis indicates acceptable values through low I2 values and narrow confidence levels for primary outcomes, i.e. preeclampsia, gestational hypertension and LGA, and secondary outcomes, i.e. PTB, LBW, postpartum haemorrhage, placental abruption and placenta accreta. This implies that the precision of the meta-analysis is of good quality and that the estimated value is comparatively stable for these variables. The present systematic review and meta-analysis was carried out in accordance with the PRISMA statement. This ensured that the methodological quality was high. Moreover, the quality of evidence was rated according to GRADE. The validity of our results is notably improved owing to these factors.
Limitations
All published articles in this review comprised observational studies. A limitation of registry-based studies is the inherent lack of data. Hence, we were unable to adjust for potential confounding variables, i.e. parity, smoking status, alcohol intake, socioeconomic status, duration of infertility, women's ovulatory status, body mass index, embryo quality, previous caesarean section, the freezing protocol, the embryo stage of development for transfer and preimplantation genetic testing. The reason for the use of AC-FET deserves attention because it may be associated with a higher risk of perinatal complications, possibly distorting the outcomes of the analyses. In this study, this point could not be analysed. In addition, unpublished data as full-text articles and in languages other than English were excluded from the meta-analysis.
Comparison of the meta-analysis with individual studies
The present meta-analysis included a population-based registry study in Sweden, which reported on all IVF singleton deliveries from autologous oocytes from 2005 to 2015 grouped into FET in programmed, stimulated or natural cycles and fresh embryo transfers (
, AC-FET cycles were associated with a higher risk of hypertensive disorders in pregnancy compared with NC-FET (adjusted OR, 1.78, 95% CI 1.43 to 2.21) and postpartum haemorrhage (adjusted OR, 2.63; 95% CI 2.20 to 3.13). These results are also in accordance with a recent study suggesting an increased rate of preeclampsia in programmed FET cycles in which no corpus luteum is present (
). In the same study, no differences were observed in the risk of gestational hypertension comparing programmed FET, natural FET, fresh embryo transfer cycles and natural conception.
A retrospective cohort study of patients who conceived after AC-FET and those who conceived after NC-FET was conducted in 2014 based on the Japanese assisted reproductive technology (ART) registry and is included in the present meta-analysis. Multiple logistic regression analyses were carried out to investigate potential confounding factors. They reported that pregnancies achieved after AC-FET had increased odds of HDP (adjusted OR 1.43, 95% CI 1.14 to 1.80) and placenta accreta (adjusted OR 6.91, 95% CI 2.87 to 16.66) compared with pregnancies after NC-FET (
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
reported that a higher risk of macrosomia was seen after programmed FET with an adjusted OR of 1.62 (CI 95% 1.26 to 2.09). These findings are in accordance with the result of this meta-analysis, but heterogeneity was substantial.
The results of the meta-analysis are consistent with the results of some of the individual studies on the risk of caesarean section (
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
found an elevated risk of gestational diabetes in patients undergoing AC-FET; however, a significant difference was not found in the meta-analysis. A possible cause for this could be that the meta-analysis included different recent studies that expanded the number of cases. These results must be assessed cautiously, however, bearing in mind the possible variations in how this condition has been defined by the different studies (
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
Hypertensive disorders in pregnancy, preeclampsia and placental disease
The results indicate a possible link between endometrial preparation and adverse perinatal and maternal outcomes, mainly placenta-related diseases. The success of pregnancy is dependent on proper implantation and placentation. If any problem arises during this process, the production of vasculogenic and angiogenic factors could be affected, leading to changes that are linked with the appearance of major placental syndromes (
). The term ‘placental insufficiency’ refers to the atypical movement of uteroplacental nutrients, which can cause damage to the placenta as well as known pregnancy complications, such as intrauterine fetal growth restriction and preeclampsia. The severity of placental disease could influence the common underlying cause for many cases of preeclampsia and PTB (
The adverse perinatal and maternal outcomes associated with cryopreservation might be caused by supraphysiological steroid hormones levels during early trophoblast invasion, which lead to an anomalous placental development (
Placental volume and other first-trimester outcomes: are there differences between fresh embryo transfer, frozen-thawed embryo transfer and natural conception?.
). The AC-FET cycle may be less ‘physiological’ than a NC-FET cycle owing to the preparation of the endometrium with hormonal replacement requiring medication (
). In the physiologic implantation process, progestin is important in the decidualization of oestradiol-primed human endometrial stromal cells, assisting with the extravillous trophoblast (EVT) invasion and vascular remodelling (
). Uterine spiral arteries and arterioles are converted to low resistance, high-capacity vessels that deliver increased blood flow required by the developing fetal–placental unit. A shallow EVT invasion and impaired spiral artery remodelling are linked to preeclampsia, placental abruption, stillbirth, fetal growth restriction and many cases of spontaneous PTB (
Failure of physiologic transformation of spiral arteries, endothelial and trophoblast cell activation, and acute atherosis in the basal plate of the placenta.
Sex steroids are critical modulators of a wide range of maternal and placental processes during pregnancy, which regulate the uteroplacental vasculature (
). Exogenous hormonal supplementation leads to an increased oestrogenic and progestogenic milieu during implantation and early pregnancy. Even after discontinuing hormone supplementation and ovarian hormone production declines, oestradiol and progesterone levels remain persistently elevated in women with pregnancies conceived by IVF compared with those who conceived naturally, at a time when the placenta becomes the source of hormone production (
Maternal 17-β-estradiol and progesterone remain elevated into the late first trimester in pregnancies conceived with in-vitro fertilization (IVF) despite discontinuation of supplemental therapy.
). Metabolomic studies suggest that the supraphysiologic hormonal state during hormone supplementation may cause reprogramming of the trophoblast after implantation, leading to a dysfunctional hormone production from syncytiotrophoblasts (
). Previous studies in animals have assessed the effect of elevated oestradiol levels in the first trimester, reporting decreased extravillous trophoblast invasion of the uterine spiral arteries (
). Recent studies have suggested that the impairment in placentation can occur via oestradiol-induced differential expression of the GATA3 transcription factor (
). Elevated progesterone, like elevated oestradiol, also affects placentation, which, in conjunction with oestrogen, induces first-trimester trophoblast tubulogenesis through the lysophosphatidic acid pathway (
). As the elevation in steroids leads to dysfunction in trophoblast cells, these increased hormonal states may be contributors to adverse obstetric and perinatal outcomes, including low birth weight, fetal growth restriction, preeclampsia and abnormal placentation (
). Further studies are necessary to determine the exact mechanisms leading to these outcomes.
On the other hand, corpus luteum has become the focus of new research related to this topic. A link exists between the absence of a corpus luteum owing to the pituitary–ovarian axis suppression by oestradiol replacement in the context of a programmed cycle and the absence of products of the corpus luteum that are not administered, such as relaxin. This hormone has a significant role for maternal cardiovascular adaptation to pregnancy (
Maternal endothelial function, circulating endothelial cells, and endothelial progenitor cells in pregnancies conceived with or without in vitro fertilization.
Am. J. Physiol. Regul. Integr. Comp. Physiol.2020; 318: R1091-R1102
). Deficient circulatory adaptations during the first trimester in women conceiving after AC-FET (with the absence of the corpus luteum), is also linked to adverse pregnancy outcomes, including preeclampsia (
Potential influence of the corpus luteum on circulating reproductive and volume regulatory hormones, angiogenic and immunoregulatory factors in pregnant women.
Am. J. Physiol. Endocrinol. Metab.2019; 317: E677-E685
). Frozen embryo transfer with a natural cycle does not have hormonal substitution and enables the more physiological development of a corpus luteum. The results of a randomized trial in which 75% of the FETs were carried out in a natural, ovulatory cycle indicated no increased risk of preeclampsia compared with those with fresh embryo transfer (
). Results from another study found a link between programmed FET cycles and higher rates of preeclampsia (12.8% versus 3.9%; P = 0.02) and preeclampsia with severe features (9.6% versus 0.8%; P = 0.002) compared with modified natural FET cycles (
The causes of increased risk of high birth weight, LGA and macrosomia after FET are still unknown. Epigenetic disturbances during the early embryonic stages as a result of the freezing– thawing and vitrification–warming procedures might affect the developmental programming of fetal and placental tissues in FET offspring (
). Asynchrony might occur between the embryo and the endometrium in FET cycles, influencing fetal growth and development resulting in increased birth weight (
Are There Differences in Placental Volume and Uterine Artery Doppler in Pregnancies Resulting From the Transfer of Fresh Versus Frozen-Thawed Embryos Through In Vitro Fertilization.
Placental volume and other first-trimester outcomes: are there differences between fresh embryo transfer, frozen-thawed embryo transfer and natural conception?.
), and could explain the increased risk of LGA and macrosomia in FET offspring. In addition, a higher birth weight could be the result of epigenetic changes after IVF with altered methylation of genes involved in the metabolism of insulin growth factor 1 and 2 (
In the NC-FET group, the intrauterine environment may be more favourable to embryo growth because it is not affected by hormonal replacement compared with AC-FET. As mentioned above, higher levels of sex steroids could negatively affect the peri-implantation uterine environment. One study reported an association between excessive fetal growth and preeclampsia, especially late-onset (
). Such a high caesarean section rate, however, is associated with other risks, mainly maternal risks such as postpartum haemorrhage. Endometrial preparation using AC-FET itself or pregnancy complications may be responsible for the higher rate of caesarean section in AC-FET patients compared with NC-FET. In addition, sociocultural differences between the population of the different studies included in the meta-analysis could influence the caesarean section rates, i.e. Asian versus European population. As adverse maternal outcomes, such as HDP, are a significant risk factor for caesarean section, adjusting for confounding factors, including the incidence of hypertension, is required in future studies to clarify whether the higher incidence of caesarean section may be caused by the AC-FET itself.
In this review, we also observed a lower risk of placenta accreta in the NC-FET group.
reported that FET constitutes an independent risk factor for placenta accreta, after controlling for patient age, before caesarean section, placenta previa and uterine factor infertility. This risk may be directly related to factors associated with cryopreservation, including the freeze–thawing process itself and the mode of uterine preparation. During the AC-FET, abnormal oestradiol levels may modulate the degree of trophoblastic invasion and extent of vascular remodelling at the time of implantation, resulting in a later exuberant trophoblastic growth (
). In this review, we found no differences in the risk of placenta previa between NC-FET and AC-FET groups; therefore, the increased incidence of placenta accreta in the AC-FET population does not seem to have been influenced by this factor.
Clinical considerations
It is critically important to evaluate the safety of an intervention when balancing benefits against harm. Therefore, maternal and perinatal outcomes should be considered when making treatment decisions; however, these are rarely reported. In the FET population, we should consider maternal and perinatal risks when the endometrial preparation method is decided, as patients who undergo NC-FET are at decreased risks for obstetric complications, as we found in this study.
The results of the present study encourage the use of NC-FET cycles, whereas AC-FET cycles ought to be used only when ovulation fails. The high rates of HDP show similarities with oocyte donation pregnancies, a known high-risk population for these and other maternal adverse outcomes in ART (
From the data obtained in our meta-analysis, practitioners can increase the safety of their interventions by identifying those patients who potentially require additional care. Patients undergoing AC-FET will benefit from single embryo transfer, avoiding the risks of multiple pregnancies. Also, their obstetricians should implement adequate monitoring strategies during prenatal, labour and postnatal care.
Future research
A large, multisite randomized controlled trial comparing pregnancy outcomes between NC-FET and AC-FET would be the optimal method to validate our findings. Further research should, if confirming these results, clarify the role of corpus luteum and its compounds on perturbed maternal cardiovascular function and placentation to replace these in patients undergoing FET cycles in which NC-FET is not possible to carry out, such as non-ovulatory women. In addition, we must be aware of epigenetic reprogramming in very early development and its relation to ART. Follow-up studies on children born after ART should be carried out throughout their life to monitor the development of adult diseases. Future meta-analyses should involve uncommon neonatal and childhood outcomes to provide more reliable data on the effect of cryopreservation.
In conclusion, pregnancies after NC-FET have a more favourable outcome compared with AC-FET, with lower rates of HDP, preeclampsia, LGA and macrosomia. The development of gestational hypertension in FET cycles seems not to be influenced by the mode of endometrial preparation. This is valuable information, as the number of FET cycles has increased, including the ‘freeze-all’ strategy. Future studies are required to clarify the underlying biologic mechanisms of our findings, and further randomized controlled trials are needed to improve the quality of evidence.
Placental volume and other first-trimester outcomes: are there differences between fresh embryo transfer, frozen-thawed embryo transfer and natural conception?.
Potential influence of the corpus luteum on circulating reproductive and volume regulatory hormones, angiogenic and immunoregulatory factors in pregnant women.
Am. J. Physiol. Endocrinol. Metab.2019; 317: E677-E685
Maternal endothelial function, circulating endothelial cells, and endothelial progenitor cells in pregnancies conceived with or without in vitro fertilization.
Am. J. Physiol. Regul. Integr. Comp. Physiol.2020; 318: R1091-R1102
Failure of physiologic transformation of spiral arteries, endothelial and trophoblast cell activation, and acute atherosis in the basal plate of the placenta.
Maternal 17-β-estradiol and progesterone remain elevated into the late first trimester in pregnancies conceived with in-vitro fertilization (IVF) despite discontinuation of supplemental therapy.
Are There Differences in Placental Volume and Uterine Artery Doppler in Pregnancies Resulting From the Transfer of Fresh Versus Frozen-Thawed Embryos Through In Vitro Fertilization.
Increased incidence of obstetric and perinatal complications in pregnancies achieved using donor oocytes and single embryo transfer in young and healthy women. A prospective hospital-based matched cohort study.
Endometrial preparation methods for frozen-thawed embryo transfer are associated with altered risks of hypertensive disorders of pregnancy, placenta accreta, and gestational diabetes mellitus.
José Moreno-Sepúlveda obtained his medical degree from Valparaiso School of Medicine, 2009, and his PhD from the Department of Obstetrics and Gynecology, Universitat Autonoma de Barcelona, 2019. Currently, his main research project is on safety in assisted reproductive techniques, focusing on maternal and perinatal outcomes after IVF.
Key message
The risk of adverse perinatal outcomes, including hypertensive disorders of pregnancy, preeclampsia, large-for-gestational-age infants and macrosomia is significantly lower for frozen–thawed embryo transfers (FET) carried out in natural cycles compared with artificial cycles. Clinicians should carefully monitor pregnancies achieved by FET in artificial cycles prenatally, during labour and postnatally.
Article info
Publication history
Published online: March 10, 2021
Accepted:
March 1,
2021
Received in revised form:
February 23,
2021
Received:
November 16,
2020
Declaration: The authors report no financial or commercial conflicts of interest.
00103-6&id=gr1.jpg){kind=link}
00103-6&id=gr2.jpg){kind=link}
00103-6&id=gr3.jpg){kind=link}