Serum estradiol and progesterone in the mid-luteal phase predict clinical pregnancy outcome in IVF/ICSI cycles

Peri-implantation estradiol level has no effect on pregnancy outcome in vitro fertilization- embryo transfer

Background

The necessity of monitoring luteal endocrine functions in in vitro fertilization- embryo transfer (IVF-ET) remains uncertain. Specifically, the significance of luteal phase estradiol (E2) levels is a matter of debate in current literature.

Objective

To assess the impact of luteal phase (day 11 after HCG trigger) estradiol levels on IVF-ET outcomes.

Design

Twelve thousand five hundred and thirty-five (n = 12,535) IVF-ET cycles performed in our center between 2015 and 2021 were divided into 5 groups based on the middle and late luteal phase serum E2 (MllPSE2) level percentiles as follows: Group A < 50 pg/mL (N=500), group B 50 pg/mL≤E2<150 pg/mL (N=2545), group C 150 pg/mL≤E2<250 pg/mL (N=1327), group D 250 pg/mL≤E2<500 pg/mL (N=925), group E E2≥500 pg/mL (n=668). The clinical pregnancy rates, abortion rates, and live birth rates of each group were compared. Binary logistic regression analysis was carried out to assess the potential impact of MllPSE2 on the live birth rate (LBR).

Results

No significant differences were found in various parameters when comparing the five groups. The level of MllPSE2 showed no significant difference between the pregnant group and the non-pregnant group. The binary logistic regression analysis model demonstrated that MllPSE2 was not significantly related to LBR.

Conclusion

The influence of E2 during the peri-implantation period (day 11) on clinical outcome in IVF-ET is not affected, even if E2<50 pg/mL. It is speculated that ovarian-derived E2 in MllPSE2 is not deemed necessary for endometrial receptivity. Although caution is warranted due to the retrospective nature of the analysis and the potential for unmeasured confounding, it is argued that the need for luteal E2 monitoring in IVF-ET may be of questionable value.

The pathophysiological role of estrogens in the initial stages of pregnancy: molecular mechanisms and clinical implications for pregnancy outcome from the periconceptional period to end of the first trimester

BACKGROUND

Estrogens regulate disparate female physiological processes, thus ensuring reproduction. Altered estrogen levels and signaling have been associated with increased risks of pregnancy failure and complications, including hypertensive disorders and low birthweight babies. However, the role of estrogens in the periconceptional period and early pregnancy is still understudied.

OBJECTIVE AND RATIONALE

This review aims to summarize the current evidence on the role of maternal estrogens during the periconceptional period and the first trimester of pregnancies conceived naturally and following ART. Detailed molecular mechanisms and related clinical impacts are extensively described.

SEARCH METHODS

Data for this narrative review were independently identified by seven researchers on Pubmed and Embase databases. The following keywords were selected: 'estrogens' OR 'estrogen level(s)' OR 'serum estradiol' OR 'estradiol/estrogen concentration', AND 'early pregnancy' OR 'first trimester of pregnancy' OR 'preconceptional period' OR 'ART' OR 'In Vitro Fertilization (IVF)' OR 'Embryo Transfer' OR 'Frozen Embryo Transfer' OR 'oocyte donation' OR 'egg donation' OR 'miscarriage' OR 'pregnancy outcome' OR 'endometrium'.

OUTCOMES

During the periconceptional period (defined here as the critical time window starting 1 month before conception), estrogens play a crucial role in endometrial receptivity, through the activation of paracrine/autocrine signaling. A derailed estrogenic milieu within this period seems to be detrimental both in natural and ART-conceived pregnancies. Low estrogen levels are associated with non-conception cycles in natural pregnancies. On the other hand, excessive supraphysiologic estrogen concentrations at time of the LH peak correlate with lower live birth rates and higher risks of pregnancy complications. In early pregnancy, estrogen plays a massive role in placentation mainly by modulating angiogenic factor expression-and in the development of an immune-tolerant uterine micro-environment by remodeling the function of uterine natural killer and T-helper cells. Lower estrogen levels are thought to trigger abnormal placentation in naturally conceived pregnancies, whereas an estrogen excess seems to worsen pregnancy development and outcomes.

WIDER IMPLICATIONS

Most current evidence available endorses a relation between periconceptional and first trimester estrogen levels and pregnancy outcomes, further depicting an optimal concentration range to optimize pregnancy success. However, how estrogens co-operate with other factors in order to maintain a fine balance between local tolerance towards the developing fetus and immune responses to pathogens remains elusive. Further studies are highly warranted, also aiming to identify the determinants of estrogen response and biomarkers for personalized estrogen administration regimens in ART.

Population Pharmacokinetic Analysis of Follicle-Stimulating Hormone During Ovarian Stimulation: Relation with Weight, Prolactin and Gene Polymorphism in THADA and ADIPOQ

BACKGROUND

Personalisation strategies of ovarian stimulation for in vitro fertilisation (IVF)/ intracytoplasmic sperm injection (ICSI) treatments using exogenous follicle-stimulating hormone (FSH) have been extensively studied over the past 20 years. This research aimed to develop a FSH population pharmacokinetic (PPK) model taking into account the contribution of gene polymorphisms in Chinese reproductive-age women.

METHODS

Data from 173 patients undergoing GnRH agonist down-regulation long protocols of IVF/ICSI treatment were collected. PPK analysis was subsequently conducted using the nonlinear mixed-effect model (NONMEM) software. Several covariates, including 18 single nucleotide polymorphisms, demographic factors and biological characteristics, were evaluated. The final PPK model was extensively validated using bootstrapping and normalised prediction error distribution, as well as external validation on an independent group of 35 patients.

RESULTS

FSH PPK was accurately described by a one-compartment model with first-order absorption. The typical population value of apparent clearance was estimated to be 0.81 L/h [relative standard errors (RSE) 5.3%] with an inter-individual variability (IIV) of 16.0%. The typical apparent distribution volume was 8.36 L (RSE 9.7%, 59.7% IIV), and the absorption rate constant was estimated to be 0.0444 h-1 (RSE 9.1%). Body weight, basal prolactin concentration and the gene ADIPOQ (rs1501299) showed a significant covariate effect on the FSH clearance rate and exposure concentration. Genotypes of THADA (rs12478601) significantly influenced the distribution volume. Simulation results indicated that patients with the TT genotype of THADA (rs12478601) required a longer time to reach steady state and had less fluctuation in FSH levels. Model evaluations showed that the final model accurately and precisely described the observed data and demonstrated effective prediction performance.

CONCLUSION

PPK models of FSH have been developed, which could potentially be used for FSH dosage individualisation in the clinical setting.

CLINICAL TRIAL REGISTRATION

This study has been registered with the Chinese Clinical Trials Registry (ChiCTR2100049142).

Does mid-luteal progesterone predict pregnancy in intrauterine insemination cycles following sequential clomiphene citrate and gonadotropin treatment?

This study aimed to determine whether serum mid-luteal progesterone (MLP) levels measured in the current treatment cycles of infertile women undergoing controlled ovarian hyperstimulation and intrauterine insemination following the sequential use of clomiphene citrate and gonadotropin may predict pregnancy. A total of 107 consecutive anovulatory women were included in this prospective cohort study. Patients with other causes of infertility were also excluded from the study. None of the patients received progesterone treatment for luteal phase support. The data recorded for each woman included age, body mass index, infertility type and duration, basal hormone levels, and previous and current cycle characteristics with MLP levels. Ovulation was confirmed using MLP and sonographic evaluation in all patients. An MLP level of > 3 ng/mL was regarded as a sign of ovulation. After treatment, the patients were divided into 2 groups according to the presence or absence of pregnancy, and the obtained data were compared between the groups. There were no significant differences in age, body mass index, or basal hormone levels between the 2 groups (all P > .05). However, the duration of infertility was significantly shorter in the pregnancy group (P = .003). The anovulation rate in this cohort was 18.7% (n = 20). A total of 15 (14%) were examined. MLP levels were 25.1 ± 13.8 ng/mL and 18.3 ± 14.5 ng/mL in the pregnant and nonpregnant groups, respectively (P:.089). Based on the receiver operating characteristic curve analysis, it was determined that there was no predictive value of the mid-luteal phase progesterone level for pregnancy in patients in whom ovulation was detected. Mid-luteal serum progesterone levels did not predict pregnancy in infertile women who underwent controlled ovarian hyperstimulation with sequential clomiphene citrate plus gonadotropin treatment and intrauterine insemination.

A drop in serum progesterone from oocyte pick-up +3 days to +5 days in fresh blastocyst transfer, using hCG-trigger and standard luteal support, is associated with lower ongoing pregnancy rates

STUDY QUESTION

Do early- and mid-luteal serum progesterone (P4) levels impact ongoing pregnancy rates (OPRs) in fresh blastocyst transfer cycles using standard luteal phase support (LPS)?

SUMMARY ANSWER

A drop in serum P4 level from oocyte pick-up (OPU) + 3 days to OPU + 5 days (negative ΔP4) is associated with a ∼2-fold decrease in OPRs.

WHAT IS KNOWN ALREADY

In fresh embryo transfer cycles, significant inter-individual variation occurs in serum P4 levels during the luteal phase, possibly due to differences in endogenous P4 production after hCG trigger and/or differences in bioavailability of exogenously administered progesterone (P) via different routes. Although exogenous P may alleviate this drop in serum P4 in fresh transfer cycles, there is a paucity of data exploring the possible impact on reproductive outcomes of a reduction in serum P4 levels.

STUDY DESIGN, SIZE, DURATION

Using a prospective cohort study design, following the initial enrollment of 558 consecutive patients, 340 fulfilled the inclusion and exclusion criteria and were included in the final analysis. The inclusion criteria were: (i) female age ≤40 years, (ii) BMI ≤35 kg/m2, (iii) retrieval of ≥3 oocytes irrespective of ovarian reserve, (iv) the use of a GnRH-agonist or GnRH-antagonist protocol with recombinant hCG triggering (6500 IU), (v) standard LPS and (vi) fresh blastocyst transfer. The exclusion criteria were: (i) triggering with GnRH-agonist or GnRH-agonist plus recombinant hCG (dual trigger), (ii) circulating P4 >1.5 ng/ml on the day of trigger and (iii) cleavage stage embryo transfer. Each patient was included only once. The primary outcome was ongoing pregnancy (OP), as defined by pregnancy ≥12 weeks of gestational age.

PARTICIPANTS/MATERIALS, SETTING, METHODS

A GnRH-agonist (n = 53) or GnRH-antagonist (n = 287) protocol was used for ovarian stimulation. Vaginal progesterone gel (Crinone, 90 mg, 8%, Merck) once daily was used for LPS. Serum P4 levels were measured in all patients on five occasions: on the day of ovulation trigger, the day of OPU, OPU + 3 days, OPU + 5 days and OPU + 14 days; timing of blood sampling was standardized to be 3-5 h after the morning administration of vaginal progesterone gel. The delta P4 (ΔP4) level was calculated by subtracting the P4 level on the OPU + 3 days from the P4 level on the OPU + 5 days, resulting in either a positive or negative ΔP4.

MAIN RESULTS AND THE ROLE OF CHANCE

The median P4 (min-max) on the day of triggering, day of OPU, OPU + 3 days, OPU + 5 days and OPU + 14 days were 0.83 ng/ml (0.18-1.42), 5.81 ng/ml (0.80-22.72), 80.00 ng/ml (22.91-161.05), 85.91 ng/ml (15.66-171.78) and 13.46 ng/ml (0.18-185.00), respectively. Serum P4 levels uniformly increased from the day of OPU to OPU + 3 days in all patients; however, from OPU + 3 days to OPU + 5 days, some patients had a decrease (negative ΔP4; n = 116; 34.1%), whereas others had an increase (positive ΔP4; n = 220; 64.7%), in circulating P4 levels. Although the median (min-max) P4 levels on the day of triggering, the day of OPU, and OPU + 3 days were comparable between the negative ΔP4 and positive ΔP4 groups, patients in the former group had significantly lower P4 levels on OPU + 5 days [69.67 ng/ml (15.66-150.02) versus 100.51 ng/ml (26.41-171.78); P < 0.001] and OPU + 14 days [8.28 ng/ml (0.28-157.00) versus 19.01 ng/ml (0.18-185.00), respectively; P < 0.001]. A drop in P4 level from OPU + 3 days to OPU + 5 days (negative ΔP4) was seen in approximately one-third of patients and was associated with a significantly lower OPR when compared with positive ΔP4 counterparts [33.6% versus 49.1%, odds ratio (OR); 0.53, 95% CI; 0.33-0.84; P = 0.008]; this decrease in OPR was due to lower initial pregnancy rates rather than increased overall pregnancy loss rates. For negative ΔP4 patients, the magnitude of ΔP4 was a significant predictor of OP (adjusted AUC = 0.65; 95% CI; 0.59-0.71), with an optimum threshold of -8.73 ng/ml, sensitivity and specificity were 48.7% and 79.2%, respectively. BMI (OR; 1.128, 95% CI; 1.064-1.197) was the only significant predictor of having a negative ΔP4; the higher the BMI, the higher the risk of having a negative ΔP4. Among positive ΔP4 patients, the magnitude of ΔP4 was a weak predictor of OP (AUC = 0.56, 95% CI; 0.48-0.64). Logistic regression analysis showed that blastocyst morphology (OR; 5.686, 95% CI; 1.433-22.565; P = 0.013) and ΔP4 (OR; 1.013, 95% CI; 0.1001-1.024; P = 0.031), but not the serum P4 level on OPU + 5 days, were the independent predictors of OP.

LIMITATIONS, REASONS FOR CAUTION

The physiological circadian pulsatile secretion of P4 during the mid-luteal phase is a limitation; however, blood sampling was standardized to reduce the impact of timing.

WIDER IMPLICATIONS OF THE FINDINGS

Two measurements (OPU + 3 days and OPU + 5 days) of serum P4 may identify those patients with a drop in P4 (approximately one-third of patients) associated with ∼2-fold lower OPRs. Rescuing these IVF cycles with additional P supplementation or adopting a blastocyst freeze-all policy should be tested in future randomized controlled trials.

STUDY FUNDING/COMPETING INTEREST(S)

None. S.C.E. declares receipt of unrestricted research grants from Merck and lecture fees from Merck and Med.E.A. P.H. has received unrestricted research grants from MSD and Merck, as well as honoraria for lectures from MSD, Merck, Gedeon-Richter, Theramex, and IBSA. H.Y. declares receipt of honorarium for lectures from Merck, IBSA and research grants from Merck and Ferring. The remaining authors declare that they have no conflict of interest.

TRIAL REGISTRATION NUMBER

The study was registered at clinical trials.gov (NCT04128436).

Low Luteal Serum Progesterone Levels Are Associated With Lower Ongoing Pregnancy and Live Birth Rates in ART: Systematic Review and Meta-Analyses

Progesterone plays a key role in implantation. Several studies reported that lower luteal progesterone levels might be related to decreased chances of pregnancy. This systematic review was conducted using appropriate key words, on MEDLINE, EMBASE, and the Cochrane Library, from 1990 up to March 2021 to assess if luteal serum progesterone levels are associated with ongoing pregnancy (OP) and live birth (LB) rates (primary outcomes) and miscarriage rate (secondary outcome), according to the number of corpora lutea (CLs). Overall 2,632 non-duplicate records were identified, of which 32 relevant studies were available for quantitative analysis. In artificial cycles with no CL, OP and LB rates were significantly decreased when the luteal progesterone level falls below a certain threshold (risk ratio [RR] 0.72; 95% confidence interval [CI] 0.62-0.84 and 0.73; 95% CI 0.59-0.90, respectively), while the miscarriage rate was increased (RR 1.48; 95% CI 1.17-1.86). In stimulated cycles with several CLs, the mean luteal progesterone level in the no OP and no LB groups was significantly lower than in the OP and LB groups [difference in means 68.8 (95% CI 45.6-92.0) and 272.4 (95% CI 10.8-533.9), ng/ml, respectively]. Monitoring luteal serum progesterone levels could help in individualizing progesterone administration to enhance OP and LB rates, especially in cycles without corpus luteum.

SYSTEMATIC REVIEW REGISTRATION

https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=139019, identifier 139019.

A 10-year follow-up on the practice of luteal phase support using worldwide web-based surveys

BACKGROUND

It has been demonstrated that luteal phase support (LPS) is crucial in filling the gap between the disappearance of exogenously administered hCG for ovulation triggering and the initiation of secretion of endogenous hCG from the implanting conceptus. LPS has a pivotal role of in establishing and maintaining in vitro fertilization (IVF) pregnancies. Over the last decade, a plethora of studies bringing new information on many aspects of LPS have been published. Due to lack of consent between researchers and a dearth of robust evidence-based guidelines, we wanted to make the leap from the bench to the bedside, what are the common LPS practices in fresh IVF cycles compared to current evidence and guidelines? How has expert opinion changed over 10 years in light of recent literature?

METHODS

Over a decade (2009-2019), we conducted 4 web-based surveys on a large IVF-specialist website on common LPS practices and controversies. The self-report, multiple-choice surveys quantified results by annual IVF cycles.

RESULTS

On average, 303 IVF units responded to each survey, representing, on average, 231,000 annual IVF cycles. Most respondents in 2019 initiated LPS on the day of, or the day after egg collection (48.7 % and 36.3 %, respectively). In 2018, 72 % of respondents administered LPS for 8-10 gestational weeks, while in 2019, 65 % continued LPS until 10-12 weeks. Vaginal progesterone is the predominant delivery route; its utilization rose from 64 % of cycles in 2009 to 74.1 % in 2019. Oral P use has remained negligible; a slight increase to 2.9 % in 2019 likely reflects dydrogesterone's introduction into practice. E2 and GnRH agonists are rarely used for LPS, as is hCG alone, limited by its associated risk of ovarian hyperstimulation syndrome (OHSS).

CONCLUSIONS

Our Assisted reproductive technology (ART)-community survey series gave us insights into physician views on using progesterone for LPS. Despite extensive research and numerous publications, evidence quality and recommendation levels are surprisingly low for most topics. Clinical guidelines use mostly low-quality evidence. There is no single accepted LPS protocol. Our study highlights the gaps between science and practice and the need for further LPS research, with an emphasis on treatment individualization.

Luteal phase progesterone and oestradiol after ovarian stimulation: relation to response and prediction of pregnancy

Research has focused on optimizing luteal phase support and endometrial receptivity in ovarian stimulation cycles. In this study, serial endocrine measurements were taken in 600 patients after a gonadotrophin-releasing hormone antagonist stimulation protocol. On the day of blastocyst transfer, serum progesterone and oestradiol were similar irrespective of a subsequent positive or negative pregnancy test (median 99 ng/ml versus 103 ng/ml for progesterone, respectively) or a subsequent live birth or pregnancy loss. Serum progesterone was significantly correlated to each ovarian response parameter (total number of follicles, number of oocytes retrieved and oestradiol concentration; r = 0.45, 0.57 and 0.54 respectively, all P < 0.0001). These correlations were consistent irrespective of clinical outcome. On the day of the pregnancy test, these correlations had vanished except in the live birth subgroup showing a weaker correlation (r = 0.22, 0.27 and 0.32 respectively, all P < 0.005). The lowest HCG and progesterone levels associated with live birth were 59.3 IU/l and 12.3 ng/ml, respectively. Fourteen out of 92 patients (15.2%) with pregnancy loss had normal HCG but low progesterone levels (above and below their respective 5th percentile), and miscarried before the end of the 7th week, when the luteal-placental shift occurs.

Predictive value of repeated measurements of luteal progesterone and estradiol levels in patients with intrauterine insemination and controlled ovarian stimulation

OBJECTIVE

The objective of this study is to assess if the difference of repeated measurements of estradiol and progesterone during luteal phase predict the outcome of intrauterine insemination.

DESIGN

Prospective study.

SETTING

Reproductive clinic.

PATIENTS

126 patients with infertility.

INTERVENTION(S)

Patients underwent controlled ovarian stimulation with recombinant FSH (50-150 IU/d). The day of IUI patients were given p.o natural micronized progesterone in a dose of 100 mg/tds.

RESULTS

The area under the receiver characteristic operating curve (ROC curve) in predicting clinical pregnancy for % change of estradiol level on days 6 and 10 was 0.892 with 95% CI: 0.82-0.94. A cutoff value of change > -29.5% had a sensitivity of 85.7 with a specificity of 90.2. The corresponding ROC curve for % change of progesterone level was 0.839 with 95% CI: 0.76-0.90. A cutoff value of change > -33% had a sensitivity of 85 with a specificity of 75.

CONCLUSIONS

The % change of estradiol and progesterone between days 6 and 10 has a predictive ability of pregnancy after IUI with COS of 89.2% and 83.4%, respectively. The addition of % of progesterone to % change of estradiol does not improve the predictive ability of % estradiol and should not be used.

Can we alter pregnancy outcome by adjusting progesterone treatment at mid-luteal phase: a randomized controlled trial

Our study aimed to determine whether mid-luteal serum P concentrations can serve as a predictive factor for in vitro fertilization (IVF) outcomes and whether increasing P dosage for patients with low levels at mid-luteal phase may improve pregnancy rates. It was a prospective, randomized controlled study. A total of 146 patients undergoing IVF treatment were prospectively enrolled and received routine luteal phase support (LPS) regimen of Endometrin® (progesterone) 200 mg/day. Serum P levels were measured 7 days after embryo transfer (ET). Considering a cutoff level of 15 ng/ml on this day, patients with higher levels continued the same dosage until pregnancy test (control group). Patients with lower levels were randomly allocated to continue Endometrin® 200 mg/day (Group A) or to increase Endometrin® dosage to 300 mg/day (Group B). The Main Outcome Measures were pregnancy rates. Both biochemical and clinical pregnancy and live birth rates were comparable between all groups regardless of P level on day 7 of luteal phase and regardless of dose adjustment. ROC analysis determined that mid-luteal P levels of 17 ng/ml can be a better predictor of cycle outcome. In conclusion raising the P dose at mid-luteal phase to 300 mg daily did not improve cycle outcomes.

Mid-luteal estradiol levels of poor/good responders and intracytoplasmic sperm injection

Objective:

To assess mid-luteal estradiol (E2) levels in poor and good responders and determine its effect on the outcome after intracytoplasmic sperm injection (ICSI).

Methods:

The current study was carried out in females who underwent ICSI from June 2011 to September 2013 in “Islamabad Clinic Serving Infertile Couples”. They were categorized into good and poor responders on the basis of female age ≤40 years, basal follicle stimulating hormone ≤12 mIU/ml, and antral follicle count >5, respectively. Their mid-luteal E2 measured on the day of embryo transfer was stratified into groups (A-E) on the basis of 20th, 40th, 60th and 80th percentile values. The outcome was categorized into non-pregnant with beta human chorionic Gonadotrophin (hCG) 5-25 m IU/ml, and clinical pregnancy with beta hCG>25 m IU/ml.

Results:

The conception rate was 12% (63/513) in poor responders and 72% (237/329) in good responders respectively. The mid-luteal E2 levels were higher in conception as compared to non-conception cycles (p<0.001) in good and poor responders.

Conclusion:

Maximum pregnancies in poor and good responders (53% and 98% respectively) with mid-luteal E2 levels above 80^th percentiles confirm the role of the increase in mid-luteal E2 for augmentation in conception rate of females after ICSI.

Serum human chorionic gonadotropin measured 7 days following day 3 embryo transfer might predict pregnancy outcome in IVF

This prospective study investigated the predictive value of pregnancy outcomes with serum human chorionic gonadotropin (hCG) 7 days after day 3 embryo transfer (D3 ET), and whether estradiol (E2) and progesterone (P) improved the diagnostic efficiency. The study comprised 280 in vitro fertilization and embryo transfer (IVF-ET) cycles. Serum samples were obtained 7 days after D3 ET to measure hCG, E2, and P concentrations. Statistical analyses were conducted to evaluate the predictive value for pregnancy outcomes. We found significant differences in hCG level between pregnancy and non-pregnancy, viable and non-viable pregnancy, biochemical and viable pregnancy, as well as singleton and multiple pregnancy. An hCG cutoff value of 2.5 mIU/mL is predictive of pregnancy with a positive predictive value (PPV) of 95.9% and a negative predictive value (NPV) of 92.4%. An hCG value of 10.8 mIU/mL is predictive of a multiple pregnancy with an NPV of 98.1%. The area under the hCG curve between pregnancy and non-pregnancy was not improved by adding E2, P, or combined E2/P. Our results suggest a predictive value of pregnancy outcome with serum hCG drawn 7 days after D3 ET in IVF, and the diagnostic accuracy is not improved by adding measurements of E2/P.

How low is too low? Cycle day 28 estradiol levels and pregnancy outcomes

OBJECTIVE

To determine the utility of cycle day 28 estradiol (E2) levels in predicting pregnancy outcomes after IVF.

DESIGN

Retrospective, cohort study.

SETTING

Academic medical center.

PATIENT(S)

All IVF cycles resulting in a positive pregnancy test result at our center between January 2007 and December 2012 were included.

INTERVENTION(S)

In vitro fertilization with fresh embryo transfer.

MAIN OUTCOME MEASURE(S)

A total of 5,471 IVF cycles were identified. Cycles were stratified by day-28 E2 level (pg/mL) into three groups: A: ≤50; B: 51-100; and C: >100. Outcomes measured were live birth, clinical pregnancy, biochemical, ectopic, and spontaneous abortion rates.

RESULT(S)

There were 806, 588, and 4,077 IVF pregnancies in groups A, B, and C, respectively. Live birth rates were lower in groups A (15.4%) and B (41.2%) compared with group C (77.4%), representing decreased odds of live birth in patients with E2 levels of ≤50 pg/mL (odd ratio 0.05, 95% confidence interval 0.04-0.07) and in patients with levels of 51-100 pg/mL (odds ratio 0.20, 95% confidence interval 0.17-0.25) compared with patients with levels >100 pg/mL. Rates of biochemical and ectopic pregnancies were higher in groups A (66.5%, 6.20%) and B (30.7%, 3.57%) compared with group C (7.31%, 0.66%). An hCG level <50 mIU/mL was associated with increased odds of a biochemical pregnancy and decreased odds of a live birth.

CONCLUSION(S)

Low E2 levels early in IVF pregnancies are associated with poorer pregnancy outcomes. Estradiol can be used alone or in conjunction with hCG levels to predict the odds of a live birth.