Background
Observational retrospective data suggest that an artificial cycle frozen embryo transfer may be associated with a higher risk of hypertensive disorder of pregnancy than a natural cycle frozen embryo transfer among women with regular ovulatory cycles. The corpus luteum, which is not present in the artificial frozen cycles, is at least partly responsible for this poor obstetrical outcome. However, an artificial cycle is the most frequently used regimen for women with polycystic ovary syndrome undergoing frozen embryo transfer. Whether the risk of hypertensive disorder of pregnancy could be mitigated by employing physiological frozen embryo transfer protocols that lead to the development of a corpus luteum in patients with polycystic ovary syndrome remains unknown.
Objective
This study aimed to investigate the impact of letrozole use during frozen embryo transfer cycles on obstetrical and perinatal outcomes of singleton and twin pregnancies compared with artificial frozen cycles among women with polycystic ovary syndrome.
Study Design
This retrospective cohort study involved women with polycystic ovary syndrome who had undergone artificial frozen cycles or letrozole-stimulated frozen cycles during the period from 2010 to 2018 at a tertiary care center. The primary outcome was the incidence of hypertensive disorder of pregnancy. A multivariable logistic regression analysis was performed to control for the relevant confounders.
Results
A total of 2427 women with polycystic ovary syndrome were included in the final analysis. Of these women, 1168 underwent artificial cycles and 1259 underwent letrozole treatment, of which 25% of women treated with letrozole alone and 75% of women receiving letrozole combined with gonadotropins. After controlling for maternal characteristics and treatment variables, no significant difference was noticed regarding gestational diabetes mellitus, abnormal placentation, and preterm premature rupture of membranes between groups in both singleton and twin pregnancies. For birth outcomes, the prevalence rates of preterm birth, perinatal death, and birthweight outcomes were all comparable between groups in both singletons and twins. However, singleton pregnancies resulting from letrozole-stimulated cycles had a lower risk of hypertensive disorder of pregnancy than those conceived by artificial cycles (adjusted odds ratio, 0.63; 95% confidence interval, 0.40–0.98). Furthermore, a decreased risk of hypertensive disorder of pregnancy was seen among women with twin deliveries resulting from letrozole-stimulated cycles vs artificial cycles (adjusted odds ratio, 0.52; 95% confidence interval, 0.30–0.87). In addition, the cesarean delivery rate was significantly lower for singletons but not for twins in the letrozole group compared with pregnancies from the artificial cycle group (adjusted odds ratio, 0.63; 95% confidence interval, 0.50–0.78, and adjusted odds ratio, 1.20; 95% confidence interval, 0.65–2.23, respectively).
Conclusion
In women with polycystic ovary syndrome undergoing frozen embryo transfer, letrozole use for endometrial preparation was associated with a lower risk of hypertensive disorder of pregnancy than artificial cycles for endometrial preparation. Our findings provided a foundation that the increased risk of hypertensive disorder of pregnancy associated with frozen embryo transfer might be mitigated by utilizing physiological endometrial preparation protocols that lead to the development of a corpus luteum, such as a mild ovarian stimulation cycle for oligo- or anovulatory women.
Why was this study conducted?
Recent evidence suggested that an artificial cycle frozen embryo transfer (FET) may be associated with a higher risk of hypertensive disorder of pregnancy than a natural cycle FET among women with regular ovulatory cycles. However, whether this increased risk could be mitigated by employing a mild ovarian stimulation protocol during FET for women with polycystic ovary syndrome (PCOS) remains unknown.
Key findings
Among women with PCOS, letrozole-induced FET cycles were associated with a lower risk of hypertensive disorder of pregnancy than artificial FET cycles in both singleton and twin pregnancies.
What does this add to what is known?
Our findings suggested that the increased risk of hypertensive disorder of pregnancy associated with FET might be mitigated by utilizing a physiological FET protocol for women with PCOS.
Introduction
More than 30 years have passed since the first baby conceived by frozen embryo transfer (FET) was born in 1984. , Since then, the use of FET has progressively increased, and embryo cryopreservation has become an indispensable component of assisted reproductive technology (ART). In 2014, the number of FET cycles for the first time exceeded that of the conventional in vitro fertilization (IVF) in Europe (192,017 vs 146,148). A similar trend was also found in the United States where the proportion of FET has doubled since 2005, comprising 30% of all transfers in 2015.
Compared with a fresh embryo transfer (ET), FET was associated with equivalent reproductive outcomes and with a lower risk of ovarian hyperstimulation syndrome. However, in recent years, concerns have been raised regarding whether FET has an adverse impact on the resulting offspring and their mothers. A recent meta-analysis concluded that singletons conceived by FET were at lower risks of preterm birth (PTB), low birthweight (LBW), and small for gestational age (SGA) but faced a higher risk of being born large for gestational age (LGA) and macrosomia compared with counterparts derived from fresh transfers. Moreover, higher rates of hypertensive disorder of pregnancy (HDP) were observed in FET cycles compared with both fresh ET and naturally conceived cycles.
The mechanism underlying the observed increased risk of HDP after FET remains unknown. It has been proposed that the type of endometrial preparation (EP) might play a role. Recent evidence found that artificial cycle FET (AC-FET) may be associated with an increased risk of HDP vs fresh transfers; nonetheless, this unfavorable outcome was not apparent when FET was performed in natural or stimulated cycles. , A large epidemiologic study indicated that compared with a natural cycle FET (NC-FET), pregnancies after AC-FET had an increased incidence of HDP. These findings supported the hypothesis that the corpus luteum (CL), which was not present in AC-FET, is at least partly responsible for the increased risk of HDP associated with FET. ,
Results from the aforementioned studies might encourage physicians to use NC-FET instead of AC-FET for women with regular ovulatory cycles. Nonetheless, artificial cycle is the most frequently used regimen for women with polycystic ovary syndrome (PCOS) undergoing FET. A multicenter randomized control trial (RCT) showed that among women with PCOS, AC-FET led to a higher prevalence of preeclampsia compared with fresh transfers. Interestingly, another study by the same authors found no difference in the HDP rates between frozen and fresh transfers when FET was performed in a natural cycle among ovulatory women. These contrasting results suggested that the risk of HDP might be mitigated by employing physiological FET protocols that lead to the development of a CL for oligo- or anovulatory women.
We recently investigated the live birth rates between 2 different regimens during FET and reported that letrozole use for EP might be a potentially better alternative to artificial cycles in women with PCOS. However, obstetrical and perinatal outcomes following letrozole-induced FET cycles (L-FET) were not previously addressed. With the increasing use of a freeze-only strategy globally, in particular to PCOS patients, exploration of this topic is of pivotal importance. Thereafter, the aim of this study was to clarify the pregnancy complications and birth outcomes among women with PCOS who underwent AC-FET and L-FET.
Materials and Methods
Study population and design
A retrospective study was conducted at the Center for Reproductive Medicine of the Ninth People’s Hospital of Shanghai Jiao Tong University School of Medicine. All FET cycles performed in women with PCOS from January 2010 to December 2018 were reviewed for potential inclusion. The diagnosis of PCOS was based on Rotterdam consensus, with women fulfilling at least 2 of the 3 criteria: (1) oligo- or anovulation, (2) clinical or biochemical signs of hyperandrogenism, and (3) polycystic ovarian morphology on ultrasound, as defined by at least 1 ovary with ≥12 follicles or volume ≥10 cm 3 . The inclusion criteria for women were age of ≤40 years and having pregnancies beyond 20 weeks of gestation. Patients were excluded if they had a history of smoking, recurrent spontaneous miscarriage (defined as ≥2 previous biochemical or clinical losses), congenital uterine malformations, preexisting diabetes and hypertension before pregnancy, and use of antidiabetic agents, such as metformin or myo-inositol, before or during pregnancy. Pregnancies from women with congenital adrenal hyperplasia, Cushing syndrome, and androgen-secreting tumors were also excluded. If women had more than 1 delivery in the database, only the first pregnancy was retained. Data analysis was restricted to singletons and twins. The study was approved by the institutional review board of the hospital.
Endometrial preparation
The EP methods for women with PCOS have been extensively described in our previous study. Briefly, in L-FET, letrozole was given orally for 5 days, beginning on day 3 of the menstrual cycle. Ultrasound and endocrine monitoring were performed from cycle day 10 onward. If the dominant follicle reached a mean diameter of ≥14 mm, no other drug was added until ovulation triggering. If the leading follicle was <14 mm on cycle day 10, a low dosage of human menopausal gonadotropin was injected to stimulate follicle growth. When the leading follicle reached a mean diameter of ≥17 mm and attained an endometrial thickness of ≥7 mm, with serum estradiol (E 2 ) levels of >150 pg/mL and progesterone levels of <1.0 ng/mL, 1 of 2 procedures was performed, depending on the luteinizing hormone (LH) value. If LH was ≥20 IU/L, human chorionic gonadotropin (hCG) was administered in the same afternoon, and day 3 of ET was scheduled 4 days later (6 days later for blastocysts). Exogenous progesterone (400 mg/d; Utrogestan; Besins Healthcare, Paris, France) was given vaginally starting 2 days after hCG administration. If LH was<20 U/L, hCG was injected at 9:00 PM, and ET was arranged 5 days later for 3-day-old embryos or 7 days later for blastocysts. Progesterone exposure was initiated 3 days after ovulatory trigger.
In AC-FET, oral estradiol was administered on the second or third day of the menstrual cycle. After 12 to 14 days, a transvaginal ultrasound examination was performed to confirm no dominant follicle emerged and to measure the endometrial thickness. When the endometrial thickness was ≥7 mm, exogenous progesterone supplementation (400 mg/d; Utrogestan; Besins Healthcare) was provided. ET was performed 3 days after progesterone administration for 3-day-old embryos or 5 days later for blastocysts. In both L-FET and AC-FET groups, luteal support was continued to 10 weeks of gestation if a pregnancy occurred.
The choice of EP regimen was determined by patients’ or physicians’ preference. In general, L-FET is preferred to be administered because our data showed superiority of this protocol over AC-FET. However, women who did not wish to be frequently monitored or who lived at considerable distance from our IVF center were allocated to the artificial cycles.
Embryo vitrification, thawing, and transfer
The vitrification and thawing procedures were used as described by Kuwayama. Briefly, embryo vitrification was performed via Cryotop carrier system, along with dimethylsulfoxide-ethylene glycol-sucrose as cryoprotectants. For thawing, embryos were transferred into dilution solution in a sequential manner (1, 0.5, and 0 mol/L sucrose).
Embryo scoring was performed on day 3, day 5, and day 6 based on the Istanbul consensus on embryo assessment. A maximum of 2 embryos could be transferred according to the Chinese legislation. Of note, in early years (before 2013) of our laboratory culture system, a commercially available sequential medium (Irvine Scientific) was used. From 2013 onward, embryos were cultured in a single-step culture medium (Irvine Scientific). Except for the culture medium, other laboratory conditions and IVF protocols remained unchanged throughout the study period.
Outcome measures
The primary outcome was the incidence of HDP. Secondary outcomes included other pregnancy complications and outcomes of offspring, such as PTB and neonatal birthweight. Obstetrical outcomes (International Statistical Classification of Diseases and Related Health Problems—Tenth Revision [ICD-10 code]) were as follows：HDP (including gestational hypertension [ICD-10 code O13], preeclampsia [ICD-10 code O14–O15]), gestational diabetes mellitus (GDM [ICD-10 code O24.4]), and preterm premature rupture of membranes (PPROM [ICD-10 code O42]). Abnormal placentation included placenta previa (ICD 10 code O44) and placental abruption (ICD 10 code O45).
The perinatal outcomes evaluated were infant’s sex, gestational age (calculated from the day of ET, defined as day 17 of the cycle for cleavage-stage ET and day 19 for blastocyst transfer), PTB (<37 weeks of gestation), LBW (<2500 g), very LBW (<1500 g), macrosomia (≥4000 g), SGA (defined as birthweight below the 10th percentile for gestational age), LGA (defined as birthweight above the 90th percentile for gestational age), stillbirth (intrauterine or intrapartum death of a child born at a gestational age of ≥20 weeks or weighing ≥500 g), perinatal death (stillborn and death of a live-born infant before days 7), and neonatal death (death of a live-born infant between days 7 and 28). The reference of birthweight for gestational age was based on the Chinese population–based neonatal birthweight curve of singletons or twins. , The follow-up system in our center has been described previously. , Briefly, obstetrical and neonatal data were extracted from medical records of women who had a delivery in our university hospital, whereas written reports were obtained from gynecologists and/or pediatricians for women who delivered elsewhere. In addition, any adverse outcomes and pregnancy complications were adjudicated by a trained staff.
Statistical analysis
Continuous data were presented as mean (standard deviation [SD]), and between-group differences were analyzed by the Student t test. Categorical variables were expressed as numbers and percentages, and these variables were compared using the chi-square test or the Fisher exact test, where appropriate. The impact of EP type on obstetrical complications was evaluated by multiple logistic regression analysis. Adjustments were made by maternal characteristics (age, body mass index [BMI], parity, infertility cause, and duration) and treatment variables (number of embryos transferred, embryo stage at transfer, insemination method, year of treatment, and FET cycle rank). Of note, concerning the model implemented, gestational diabetes for HDP was also added to the adjusted analysis. For birth outcomes, the adjustment was moreover made by pregnancy complications.
Results
A total of 2427 patients with PCOS who fulfilled the inclusion criteria were included in the final analysis, with no loss to follow-up. Of these women, 1168 underwent AC-FET and 1259 underwent letrozole treatment, consisting of 25% of women treated with letrozole alone and 75% of women receiving letrozole plus gonadotropins. Demographic and treatment characteristics are shown in Table 1 . For women with singleton pregnancies, all baseline data were similar to the exception of the distribution of embryo developmental stage at transfer and insemination method between groups. The baseline characteristics of the 2 cohorts were all comparable among women with twin pregnancies.
| Characteristic | Singletons | Twins | ||||
|---|---|---|---|---|---|---|
| AC-FET n=832 | L-FET n=891 | P value | AC-FET n=336 | L-FET n=368 | P value | |
| Age at OPU (y) | 29.42±3.31 | 29.36±3.31 | .706 a | 29.06±3.01 | 28.91±3.08 | .516 a |
| Age at ET (y) | 30.39±3.38 | 30.38±3.38 | .911 a | 29.99±3.07 | 29.80±3.14 | .441 a |
| Maternal BMI (kg/m 2 ) | 23.43±3.85 | 23.25±3.60 | .335 a | 23.33±3.77 | 23.15±3.51 | .513 a |
| Infertility duration (y) | 3.35±2.65 | 3.32±2.44 | .798 a | 3.30±2.23 | 3.27±2.42 | .844 a |
| Parity | .056 b | .278 b | ||||
| First | 812 (97.6) | 855 (96.0) | 328 (97.6) | 354 (96.2) | ||
| High order | 20 (2.4) | 36 (4.0) | 8 (2.4) | 14 (3.8) | ||
| FET cycle rank | .061 b | .353 b | ||||
| First | 468 (56.3) | 551 (61.8) | 209 (62.2) | 248 (67.4) | ||
| Second | 230 (27.6) | 217 (24.4) | 80 (23.8) | 75 (20.4) | ||
| High order | 134 (16.1) | 123 (13.8) | 47 (14.0) | 45 (12.2) | ||
| Infertility cause | .660 b | .948 b | ||||
| PCOS only | 177 (21.3) | 177 (19.9) | 73 (21.7) | 75 (20.4) | ||
| PCOS + male factor | 212 (25.5) | 250 (28.1) | 85 (25.3) | 95 (25.8) | ||
| PCOS + tubal factor | 391 (47.0) | 410 (46.0) | 158 (47.0) | 173 (47.0) | ||
| PCOS + other factors | 52 (6.3) | 54 (6.1) | 20 (6.0) | 25 (6.8) | ||
| Insemination method | .001 b | .185 b | ||||
| IVF | 525 (63.1) | 486 (54.5) | 209 (62.2) | 214 (58.2) | ||
| ICSI | 162 (19.5) | 205 (23.0) | 58 (17.3) | 84 (22.8) | ||
| IVF + ICSI | 145 (17.4) | 200 (22.4) | 69 (20.5) | 70 (19.0) | ||
| Developmental stage | .002 b | .583 b | ||||
| Day 3 | 686 (82.5) | 781 (87.7) | 296 (88.1) | 329 (89.4) | ||
| Day 5 or 6 | 146 (17.5) | 110 (12.3) | 40 (11.9) | 39 (10.6) | ||
| Number of embryos transferred | .339 b | .250 c | ||||
| 1 | 144 (17.3) | 139 (15.6) | 0 (0) | 3 (0.8) | ||
| 2 | 688 (82.7) | 752 (84.4) | 336 (100) | 365 (99.2) | ||
| Year of treatment | .137 b | .338 b | ||||
| 2010–2012 | 79 (9.5) | 63 (7.1) | 29 (8.6) | 37 (10.1) | ||
| 2013–2015 | 280 (33.7) | 324 (36.4) | 115 (34.2) | 141 (38.3) | ||
| 2016–2018 | 473 (56.9) | 504 (56.6) | 192 (57.1) | 190 (51.6) | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
