Frozen embryo transfer: a review on the optimal endometrial preparation and timing

STUDY QUESTION

What is the optimal endometrial preparation protocol for a frozen embryo transfer (FET)?

SUMMARY ANSWER

Although the optimal endometrial preparation protocol for FET needs further research and is yet to be determined, we propose a standardized timing strategy based on the current available evidence which could assist in the harmonization and comparability of clinic practice and future trials.

WHAT IS KNOWN ALREADY

Amid a continuous increase in the number of FET cycles, determining the optimal endometrial preparation protocol has become paramount to maximize ART success. In current daily practice, different FET preparation methods and timing strategies are used.

STUDY DESIGN, SIZE, DURATION

This is a review of the current literature on FET preparation methods, with special attention to the timing of the embryo transfer.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Literature on the topic was retrieved in PubMed and references from relevant articles were investigated until June 2017.

MAIN RESULTS AND THE ROLE OF CHANCE

The number of high quality randomized controlled trials (RCTs) is scarce and, hence, the evidence for the best protocol for FET is poor. Future research should compare both the pregnancy and neonatal outcomes between HRT and true natural cycle (NC) FET. In terms of embryo transfer timing, we propose to start progesterone intake on the theoretical day of oocyte retrieval in HRT and to perform blastocyst transfer at hCG + 7 or LH + 6 in modified or true NC, respectively.

LIMITATIONS REASONS FOR CAUTION

As only a few high quality RCTs on the optimal preparation for FET are available in the existing literature, no definitive conclusion for benefit of one protocol over the other can be drawn so far.

WIDER IMPLICATIONS OF THE FINDINGS

Caution when using HRT for FET is warranted since the rate of early pregnancy loss is alarmingly high in some reports.

STUDY FUNDING/COMPETING INTEREST(S)

S.M. is funded by the Research Fund of Flanders (FWO). H.T. and C.B. report grants from Merck, Goodlife, Besins and Abbott during the conduct of the study.

TRIAL REGISTRATION NUMBER

Not applicable.

Identification and evaluation of the microbiome in the female and male reproductive tracts

BACKGROUND

The existence of an extensive microbiome in and on the human body has increasingly dominated the scientific literature during the last decade. A shift from culture-dependent to culture-independent identification of microbes has occurred since the emergence of next-generation sequencing (NGS) techniques, whole genome shotgun and metagenomic sequencing. These sequencing analyses have revealed the presence of a rich diversity of microbes in most exposed surfaces of the human body, such as throughout the reproductive tract. The results of microbiota analyses are influenced by the technical specifications of the applied methods of analyses. Therefore, it is difficult to correctly compare and interpret the results of different studies of the same anatomical niche.

OBJECTIVES AND RATIONALE

The aim of this narrative review is to provide an overview of the currently used techniques and the reported microbiota compositions in the different anatomical parts of the female and male reproductive tracts since the introduction of NGS in 2005. This is crucial to understand and determine the interactions and roles of the different microbes necessary for successful reproduction.

SEARCH METHODS

A search in Embase, Medline Ovid, Web of science, Cochrane and Google scholar was conducted. The search was limited to English language and studies published between January 2005 and April 2018. Included articles needed to be original microbiome research related to the reproductive tracts.

OUTCOMES

The review provides an extensive up-to-date overview of current microbiome research in the field of human reproductive medicine. The possibility of drawing general conclusions is limited due to diversity in the execution of analytical steps in microbiome research, such as local protocols, sampling methods, primers used, sequencing techniques and bioinformatic pipelines, making it difficult to compare and interpret results of the available studies. Although some microbiota are associated with reproductive success and a good pregnancy outcome, it is still unknown whether a causal link exists. More research is needed to further explore the possible clinical implications and therapeutic interventions.

WIDER IMPLICATIONS

For the field of reproductive medicine, determination of what is a favourable reproductive tract microbiome will provide insight into the mechanisms of both unsuccessful and successful human reproduction. To increase pregnancy chances with live birth and to reduce reproduction-related health costs, future research could focus on postponing treatment or conception in case of the presence of unfavourable microbiota and on the development of therapeutic interventions, such as microbial therapeutics and lifestyle adaptations.

Predicting suboptimal oocyte yield following GnRH agonist trigger by measuring serum LH at the start of ovarian stimulation

STUDY QUESTION

Are the LH levels at the start of ovarian stimulation predictive of suboptimal oocyte yield from GnRH agonist triggering in GnRH antagonist down-regulated cycles?

SUMMARY ANSWER

LH levels at the start of ovarian stimulation are an independent predictor of suboptimal oocyte yield following a GnRH agonist trigger.

WHAT IS KNOWN ALREADY

A GnRH agonist ovulation trigger may result in an inadequate oocyte yield in a small subset of patients. This failure can range from empty follicle syndrome to the retrieval of much fewer oocytes than expected. Suboptimal response to a GnRH agonist trigger has been defined as the presence of circulating LH levels <15 IU/l 12 h after triggering. It has been shown that patients with immeasurable LH levels on trigger day have an up to 25% risk of suboptimal response.

STUDY DESIGN, SIZE, DURATION

In this retrospective cohort study, all patients (n = 3334) who received GnRH agonist triggering (using Triptoreline 0.2 mg) for final oocyte maturation undergoing a GnRH antagonist cycle in our centre from 2011 to 2017 were included. The primary outcome of the study was oocyte yield, defined as the ratio between the total number of collected oocytes and the number of follicles with a mean diameter >10 mm prior to GnRH agonist trigger.

PARTICIPANTS/MATERIALS, SETTING, METHODS

The endocrine profile of all patients was studied at initiation as well as at the end of ovarian stimulation. In order to evaluate whether LH levels, not only at the end but also at the start, of ovarian stimulation predicted oocyte yield, we performed multivariable regression analysis adjusting for the following confounding factors: female age, body mass index, oral contraceptives before treatment, basal and trigger day estradiol levels, starting FSH levels, use of highly purified human menopausal gonadotrophin and total gonadotropin dose. Suboptimal response to GnRH agonist trigger was defined as <10th percentile of oocyte yield.

MAIN RESULTS AND THE ROLE OF CHANCE

The average age was 31.9 years, and the mean oocyte yield was 89%. The suboptimal response to GnRH agonist trigger cut-off (<10th percentile) was 45%, which was exhibited by 340 patients. Following confounder adjustment, multivariable regression analysis showed that LH levels at the initiation of ovarian stimulation remained an independent predictor of suboptimal response even in the multivariable model (adjusted OR 0.920, 95% CI 0.871-0.971). Patients with immeasurable LH levels at the start of stimulation (<0.1 IU/l) had a 45.2% risk of suboptimal response, while the risk decreased with increasing basal LH levels; baseline circulating LH <0.5 IU/L, <2 IU/L and <5 IU/L were associated with a 39.1%, 25.2% and 13.6% risk, respectively.

LIMITATIONS, REASONS FOR CAUTION

The main limitation of the study is its retrospective design.

WIDER IMPLICATIONS OF THE FINDINGS

This is the largest study of GnRH agonist trigger cycles only, since most of the previous research on the predictive value of basal LH levels was performed in dual trigger cycles. LH values should be measured prior to start of ovarian stimulation. In cases where they are immeasurable, suboptimal response to GnRH agonist trigger can be anticipated, and an individualized approach is warranted.

STUDY FUNDING/COMPETING INTEREST(S)

There was no funding and no competing interests.

TRIAL REGISTRATION NUMBER

Not applicable.

To trigger or not to trigger ovulation in a natural cycle for frozen embryo transfer: a randomized controlled trial

STUDY QUESTION

Is the clinical pregnancy rate (CPR) following a frozen embryo transfer (FET) in a natural cycle (NC) higher after spontaneous ovulation than after triggered ovulation [natural cycle frozen embryo transfer (NC-FET) versus modified NC-FET]?

SUMMARY ANSWER

The CPR did not vary significantly between the two FET preparation protocols.

WHAT IS KNOWN ALREADY

Although the use of FET is continuously increasing, the most optimal endometrial preparation protocol is still under debate. For transfer in the NC specifically, conflicting results have been reported in terms of the outcome following spontaneous or triggered ovulation.

STUDY DESIGN, SIZE, DURATION

In a tertiary hospital setting, subjects were randomized with a 1:1 allocation into two groups between January 2014 and January 2019. Patients in group A underwent an NC-FET, while in group B, a modified NC-FET was performed with a subcutaneous hCG injection to trigger ovulation. In neither group was additional luteal phase support administered. All embryos were vitrified-warmed on Day 3 and transferred on Day 4 of embryonic development. The primary outcome was CPR at 7 weeks. All patients were followed further until 10 weeks of gestation when the ongoing pregnancy rate (OPR) was defined by the observation of foetal cardiac activity on ultrasound scan. Other secondary outcomes included biochemical pregnancy rate, early pregnancy loss and the number of visits, blood samples and ultrasonographic examinations prior to FET.

PARTICIPANTS/MATERIALS, SETTING, METHODS

A total of 260 patients (130 per study arm) were randomized, of whom 12 withdrew consent after study arm allocation. A total of 3 women conceived spontaneously before initiating the study cycle and 16 did not start for personal or medical reasons. Of the 229 actually commencing monitoring for the study FET cycle, 7 patients needed to be switched to a hormonal replacement treatment protocol due to the absence of follicular development, 12 had no embryo available for transfer after warming and 37 had a spontaneous LH surge before the ovulation trigger could be administered, although they were allocated to group B. Given the above, an intention-to-treat (ITT) analysis was performed taking into account 248 patients (125 in group A and 123 in group B), as well as a per protocol (PP) analysis on a subset of 173 patients (110 in group A and 63 in group B).

MAIN RESULTS AND THE ROLE OF CHANCE

Demographic features were evenly distributed between the study groups, as were the relevant fresh and frozen ET cycle characteristics. According to the ITT analysis, the CPR and OPR in group A (33.6% and 27.2%, respectively) and group B (29.3% and 24.4%, respectively) did not vary significantly [relative risk (RR) 0.87, 95% CI (0.60;1.26), P = 0.46 and RR 0.90, 95% CI (0.59;1.37), P = 0.61, respectively]. Biochemical pregnancy rate and early pregnancy loss were also found to be not statistically significantly different between the groups. In contrast, more clinic visits and blood samplings for cycle monitoring were required in the NC-FET group (4.05 ± 1.39) compared with the modified NC-FET group (3.03 ± 1.16, P = <0.001), while the number of ultrasound scans performed were comparable (1.70 ± 0.88 in group A versus 1.62 ± 1.04 in group B). The additional PP analysis was in line with the ITT results: CPR in group A was 36.4% versus 38.1% in group B [RR 1.05, 95% CI (0.70;1.56), P = 0.82].

LIMITATIONS, REASONS FOR CAUTION

The results are limited by the high drop-out rate for the PP analysis in the modified NC-FET group as more than one-third of the subjects allocated to this group ovulated spontaneously before ovulation triggering. Nonetheless, this issue is inherent to routine clinical practice and is an important observation of an event that can only be avoided by performing a very extensive monitoring that limits the practical advantages associated with modified NC-FET. Furthermore, although this is the largest randomized controlled trial (RCT) investigating this specific research question so far, a higher sample size would allow smaller differences in clinical outcome to be detected, since currently they may be left undetected.

WIDER IMPLICATIONS OF THE FINDINGS

This RCT adds new high-quality evidence to the existing controversial literature concerning the performance of NC-FET versus modified NC-FET. Based on our results showing no statistically significant differences in clinical outcomes between the protocols, the treatment choice may be made according to the patient's and treating physician's preferences. However, the modified NC-FET strategy reduces the need for hormonal monitoring and may therefore be considered a more patient-friendly and potentially cost-effective approach.

STUDY FUNDING/COMPETING INTEREST(S)

No specific funding was available for this study. None of the authors have a conflict of interest to declare with regard to this study.

TRIAL REGISTRATION NUMBER

NCT02145819.

TRIAL REGISTRATION DATE

8 January 2014.

DATE OF FIRST PATIENT’S ENROLMENT

21 January 2014.

The freeze-all strategy versus agonist triggering with low-dose hCG for luteal phase support in IVF/ICSI for high responders: a randomized controlled trial

STUDY QUESTION

Does the freeze-all strategy in high-responders increase pregnancy rates and improve safety outcomes when compared with GnRH agonist triggering followed by low-dose hCG intensified luteal support with a fresh embryo transfer?

SUMMARY ANSWER

Pregnancy rates after either fresh embryo transfer with intensified luteal phase support using low-dose hCG or the freeze-all strategy did not vary significantly; however, moderate-to-severe ovarian hyperstimulation syndrome (OHSS) occurred more frequently in the women who attempted a fresh embryo transfer.

WHAT IS KNOWN ALREADY

Two strategies following GnRH agonist triggering (the freeze-all approach and a fresh embryo transfer attempt using a low-dose of hCG for intensified luteal phase support) are safer alternatives when compared with conventional hCG triggering with similar pregnancy outcomes. However, these two strategies have never been compared head-to-head in an unrestricted predicted hyper-responder population.

STUDY DESIGN, SIZE, DURATION

This study included women with an excessive response to ovarian stimulation (≥18 follicles measuring ≥11 mm) undergoing IVF/ICSI in a GnRH antagonist suppressed cycle between 2014 and 2017. Our primary outcome was clinical pregnancy at 7 weeks after the first embryo transfer. Secondary outcomes included live birth and the development of moderate-to-severe OHSS.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Following GnRH agonist triggering, women were randomized either to cryopreserve all good-quality embryos followed by a frozen embryo transfer in an subsequent artificial cycle or to perform a fresh embryo transfer with intensified luteal phase support (1500 IU hCG on the day of oocyte retrieval, plus oral estradiol 2 mg two times a day, plus 200 mg of micronized vaginal progesterone three times a day).

MAIN RESULTS AND THE ROLE OF CHANCE

A total of 212 patients (106 in each arm) were recruited in the study, with three patients (one in the fresh embryo transfer group and two in the freeze-all group) later withdrawing their consent to participate in the study. One patient in the freeze-all group became pregnant naturally (clinical pregnancy diagnosed 38 days after randomization) prior to the first frozen embryo transfer. The study arms did not vary significantly in terms of the number of oocytes retrieved and embryos produced/transferred. The intention to treat clinical pregnancy and live birth rates (with the latter excluding four cases lost to follow-up: one in the fresh transfer and three in the freeze-all arms, respectively) after the first embryo transfer did not vary significantly among the fresh embryo transfer and freeze-all study arms: 51/105 (48.6%) versus 57/104 (54.8%) and 41/104 (39.4%) versus 42/101 (41.6%), respectively (relative risk for clinical pregnancy 1.13, 95% CI 0.87-1.47; P = 0.41). However, moderate-to-severe OHSS occurred solely in the group that received low-dose hCG (9/105, 8.6%, 95% CI 3.2% to 13.9% vs 0/104, 95% CI 0 to 3.7, P < 0.01).

LIMITATIONS, REASONS FOR CAUTION

The sample size calculation was based on a 19% absolute difference in terms of clinical pregnancy rates, therefore smaller differences, as observed in the trial, cannot be reliably excluded as non-significant.

WIDER IMPLICATIONS OF THE FINDINGS

This study offers the first comparative analysis of two common strategies applied to women performing IVF/ICSI with a high risk to develop OHSS. While pregnancy rates did not vary significantly, a fresh embryo transfer with intensified luteal phase support may still not avoid the risk of moderate-to-severe OHSS and serious consideration should be made before recommending it as a routine first-line treatment. Future trials may allow us to confirm these findings.

STUDY FUNDING/COMPETING INTEREST(S)

The authors have no conflicts of interest to disclose. No external funding was obtained for this study.

TRIAL REGISTRATION NUMBER

ClinicalTrials.gov identifier NCT02148393.

TRIAL REGISTRATION DATE

28 May 2014.

DATE OF FIRST PATIENT’S ENROLMENT

30 May 2014.

Vitrified-warmed blastocyst transfer on the 5th or 7th day of progesterone supplementation in an artificial cycle: a randomised controlled trial

Prospective studies comparing different durations of progesterone supplementation before transfer of vitrified-warmed blastocysts in an artificial cycle are lacking. However, in oocyte donation programmes, the sporadic available evidence demonstrates considerable differences in clinical pregnancy rates according to the duration of progesterone administration. This randomised controlled trial (RCT), included 303 patients undergoing a frozen-thawed embryo transfer (FET) of one or two vitrified-warmed blastocyst(s) in an artificial cycle. Randomisation was performed when the endometrial thickness reached ≥7 mm after oestrogen supplementation. One hundred and fifty two patients in group A received 7 d of vaginal micronised progesterone tablets and 151 patients in group B received 5 d of micronised vaginal progesterone before FET. No differences were seen in clinical pregnancy rate between both groups: 42/152 (27.6%) in group A versus 49/151 (32.5%) in group B. Although no statistically significant difference was observed in clinical pregnancy rates, our study was powered to detect an absolute difference of 16%. In this regard, we cannot exclude that smaller, clinically relevant differences might exist and our study did not have the power to detect this. Patients were also not blinded for the intervention, causing a potential bias.

Impact of Serum Estradiol Levels Prior to Progesterone Administration in Artificially Prepared Frozen Embryo Transfer Cycles

Background: The need for endocrine monitoring in artificial cycles for frozen embryo transfer (FET) remains unclear and, more specifically, the value of the late-proliferative phase serum estradiol (E2) levels is with conflicting evidence in current literature. Objective: To investigate whether artificial FET cycles require endocrine monitoring for the serum E2 level prior to initiation of exogenous progesterone administration after an endometrial thickness of 6.5 mm has been reached. Design: One thousand two hundred and twenty-two (n = 1,222) artificial FETs performed in a tertiary center between 2010 and 2015 were subdivided into 3 groups according to the following late-proliferative serum E2 level percentiles: ≤p10 (E2 ≤144 pg/ml; n = 124), p11-p90 (E2 from 145 to 438 pg/ml; n = 977) and >p90 (E2 >439 pg/ml; n = 121). A mixed-effects multilevel multivariable regression analysis was performed to assess the potential effect of the late-proliferative E2 level on the live birth rate (LBR). Results: The level of late-proliferative circulating E2 showed no significant difference in terms of LBR after FET. Specifically, the multivariable regression model demonstrated a LBR of 19.5% for the p11-p90 reference group, compared to 24.4% for the ≤p10 (p = 0.251) and 19.5% for the >p90 group (p = 0.989). Conclusion: In this large retrospective dataset, no association was observed between late-proliferative phase serum E2 levels and LBR following FET in artificially prepared cycles. Although, caution is warranted due to the retrospective nature of the analysis and the potential for unmeasured confounding, we argue that monitoring of the late-proliferative serum E2 levels and using them to guide clinical decision-making (e.g., medication step-up, cycle prolongation or cancelation) may be of questionable value.

Frozen-warmed blastocyst transfer after 6 or 7 days of progesterone administration: impact on live birth rate in hormone replacement therapy cycles

OBJECTIVE

To study the difference in live birth rate (LBR) between frozen-warmed blastocyst transfer (FET) on the 6th or the 7th day of progesterone administration in artificially prepared cycles.

DESIGN

Retrospective cohort study.

SETTING

Tertiary university-based referral hospital.

PATIENT(S)

Patients who underwent FET between December 2015 and December 2017 in a hormone replacement therapy cycle (HRT).

INTERVENTION(S)

Group A included all eligible patients who underwent transfer of a vitrified-warmed blastocyst on the 6th day of progesterone administration; group B included patients who underwent blastocyst transfer on the 7th day of progesterone. The artificial HRT protocol in this study consisted of estrogen administration at a dose of 2 mg twice daily for 7 days followed by 2 mg three times daily for 6 days and micronized vaginal progesterone 200 mg three times daily from an adequately considered endometrial thickness onward.

MAIN OUTCOME MEASURE(S)

Live birth rate.

RESULTS

The study included 619 patients, 346 in group A and 273 in group B. The LBRs were comparable between both groups (36.6% for group A and group B), even after adjustment for confounding factors (adjusted odds ratio 1.073, 95% confidence interval 0.740-1.556). Subgroup analysis revealed significantly higher miscarriage rates for day 6 blastocysts transferred on the 6th day of progesterone supplementation compared with transfer on the 7th day of progesterone supplementation (50.0% versus 21.4%, respectively). Additionally, there was a tendency toward a higher LBR when the 7-day progesterone supplementation protocol was used for transfer of a day 6 blastocyst (21.5% and 35.5% for group A and group B, respectively).

CONCLUSION

Warmed blastocyst transfer on the 6th compared with the 7th day of progesterone administration in an HRT cycle results in similar LBR. Subgroup analysis of day 6 blastocysts showed significantly higher miscarriage rates when FET was performed on the 6th day of progesterone administration.

Outcome of in-vitro oocyte maturation in patients with PCOS: does phenotype have an impact?

STUDY QUESTION

Does the phenotype of patients with polycystic ovary syndrome (PCOS) affect clinical outcomes of ART following in-vitro oocyte maturation?

SUMMARY ANSWER

Cumulative live birth rates (CLBRs) after IVM were significantly different between distinct PCOS phenotypes, with the highest CLBR observed in patients with phenotype A/HOP (= hyperandrogenism + ovulatory disorder + polycystic ovaries), while IVM in patients with phenotype C/HP (hyperandrogenism + polycystic ovaries) or D/OP (ovulatory disorder + polycystic ovaries) resulted in lower CLBRs (OR 0.26 (CI 0.06-1.05) and OR 0.47 (CI 0.25-0.88), respectively, P = 0.03).

WHAT IS KNOWN ALREADY

CLBRs in women with hyperandrogenic PCOS phenotypes (A/HOP and C/HP) have been reported to be lower after ovarian stimulation (OS) and ART when compared to CLBR in women with a normo-androgenic PCOS phenotype (D/OP) and non-PCOS patients with a PCO-like ovarian morphology (PCOM). Whether there is an influence of the different PCOS phenotypes on success rates of IVM has been unknown.

STUDY DESIGN, SIZE, DURATION

This was a single-centre, retrospective cohort study including 320 unique PCOS patients performing their first IVM cycle between April 2014 and January 2018 in a tertiary referral hospital.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Baseline patient characteristics and IVM treatment cycle data were collected. The clinical outcomes following the first IVM embryo transfer were retrieved, including the CLBR defined as the number of deliveries with at least one live birth resulting from one IVM cycle and all appended cycles in which fresh or frozen embryos were transferred until a live birth occurred or until all embryos were used. The latter was considered as the primary outcome. A multivariate regression model was developed to identify prognostic factors for CLBR and test the impact of the patient's PCOS phenotype.

MAIN RESULTS AND THE ROLE OF CHANCE

Half of the patients presented with a hyperandrogenic PCOS phenotype (n = 140 A/HOP and n = 20 C/HP vs. n = 160 D/OP). BMI was significantly different between phenotype groups (27.4 ± 5.4 kg/m2 for A/HOP, 27.1 ± 5.4 kg/m2 for C/HP and 23.3 ± 4.4 kg/m2 for D/OP, P < 0.001). Metformin was used in 33.6% of patients with PCOS phenotype A/HOP, in 15.0% of C/HP patients and in 11.2% of D/OP patients (P < 0.001). Anti-müllerian hormone levels differed significantly between groups: 12.4 ± 8.3 µg/l in A/HOP, 7.7 ± 3.1 µg/l in C/HP and 10.4 ± 5.9 µg/l in D/OP patients (P = 0.01). The number of cumulus-oocyte complexes (COC) was significantly different between phenotype groups: 25.9 ± 19.1 COC in patients with phenotype A/HOP, 18.3 ± 9.0 COC in C/HP and 19.8 ± 13.5 COC in D/OP (P = 0.004). After IVM, patients with different phenotypes also had a significantly different number of mature oocytes (12.4 ± 9.3 for A/HOP vs. 6.5 ± 4.2 for C/HP vs. 9.1 ± 6.9 for D/OP, P < 0.001). The fertilisation rate, the number of usable embryos and the number of cycles with no embryo available for transfer were comparable between the three groups. Following the first embryo transfer, the positive hCG rate and LBR were comparable between the patient groups (44.7% (55/123) for A/HOP, 40.0% (6/15) for C/HP, 36.7% (47/128) for D/OP, P = 0.56 and 25.2% (31/123) for A/HOP, 6.2% (1/15) for C/HP, 26.6% (34/128) for D/OP, respectively, P = 0.22). However, the incidence of early pregnancy loss was significantly different across phenotype groups (19.5% (24/123) for A/HOP, 26.7% (4/15) for C/HP and 10.2% (13/128) for D/OP, P = 0.04). The CLBR was not significantly different following univariate analysis (40.0% (56/140) for A/HOP, 15% (3/20) for C/HP and 33.1% (53/160) for D/OP (P = 0.07)). When a multivariable logistic regression model was developed to account for confounding factors, the PCOS phenotype appeared to be significantly correlated with CLBR, with a more favourable CLBR in the A/HOP subgroup (OR 0.26 for phenotype C/HP (CI 0.06-1.05) and OR 0.47 for phenotype D/OP (CI 0.25-0.88), P = 0.03)).

LIMITATIONS, REASONS FOR CAUTION

These data should be interpreted with caution as the retrospective nature of the study holds the possibility of unmeasured confounding factors and misassignment of the PCOS phenotype. Moreover, the sample size for phenotype C/HP was too small to draw conclusions for this subgroup of patients.

WIDER IMPLICATIONS OF THE FINDINGS

Caucasian infertile patients with a PCOS phenotype A/HOP who undergo IVM achieved a higher CLBR than their counterparts with C/HP and D/OP. This is in strong contrast with previously reported outcomes following OS where women with PCOS and hyperandrogenism (A/HOP and C/HP) performed significantly worse. For PCOS patients who require ART, the strategy of OS followed by an elective freeze-all strategy remains to be compared with IVM in a prospective fashion; however, the current data provide support for IVM as a valid treatment option, especially in the most severe PCOS phenotypes (A/HOP). Our data suggest that proper patient selection is of utmost importance in an IVM programme.

STUDY FUNDING/COMPETING INTEREST(S)

The clinical IVM research has been supported by research grants from Cook Medical and Besins Healthcare. All authors declared no conflict of interest.

TRIAL REGISTRATION NUMBER

N/A.

Follicular-phase endometrial scratching: a truncated randomized controlled trial

STUDY QUESTION

Does intentional endometrial injury (scratching) during the follicular phase of ovarian stimulation (OS) increase the clinical pregnancy rate (CPR) in ART?

SUMMARY ANSWER

CPR did not vary between the endometrial injury and the control group, but the trial was underpowered due to early termination because of a higher clinical miscarriage rate observed in the endometrial injury arm after a prespecified interim analysis.

WHAT IS KNOWN ALREADY

Intentional endometrial injury has been put forward as an inexpensive clinical tool capable of enhancing endometrial receptivity. However, despite its widespread use, the benefit of endometrial scratching remains controversial, with several recent randomized controlled trials (RCTs) being unable to confirm its added value. So far, most research has focused on endometrial scratching during the luteal phase of the cycle preceding the one with embryo transfer (ET), while only a few studies investigated in-cycle injury during the follicular phase of OS. Also, the persistence of a scratch effect in subsequent treatment cycles remains unclear and possible harms have been insufficiently studied.

STUDY DESIGN, SIZE, DURATION

This RCT was performed in a tertiary hospital setting between 3 April 2014 and 8 October 2017. A total of 200 women (100 per study arm) undergoing IVF/ICSI in a GnRH antagonist suppressed cycle followed by fresh ET were included.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Participants were randomized with a 1:1 allocation ratio to either undergo a pipelle endometrial biopsy between Days 6 and 8 of OS or to be in the control group.The primary outcome was CPR. Secondary outcomes included biochemical pregnancy rate, live birth rate (LBR), early pregnancy loss (biochemical pregnancy losses and clinical miscarriages), excessive procedure pain/bleeding and cumulative reproductive outcomes within 6 months of the study cycle.

MAIN RESULTS AND THE ROLE OF CHANCE

The RCT was stopped prematurely by the trial team after the second prespecified interim analysis raised safety concerns, namely a higher clinical miscarriage rate in the intervention group. The intention-to-treat CPR was similar between the biopsy and the control arm (respectively, 44 versus 40%, P = 0.61, risk difference = 3.6 with 95% confidence interval = -10.1;17.3), as was the LBR (respectively, 32 versus 36%, P = 0.52). The incidence of a biochemical pregnancy loss was comparable between both groups (10% in the intervention group versus 15% in the control, P = 0.49), but clinical miscarriages occurred significantly more frequent in the biopsy group (25% versus 8%, P = 0.032). In the intervention group, 3% of the patients experienced excessive procedure pain and 5% bleeding. The cumulative LBR taking into account all conceptions (spontaneous or following ART) within 6 months of randomization was not significantly different between the biopsy and the control group (54% versus 60%, respectively, P = 0.43).

LIMITATIONS, REASONS FOR CAUTION

The trial was stopped prematurely due to safety concerns after the inclusion of 200 of the required 360 patients. Not reaching the predefined sample size implies that definite conclusions on the outcome parameters cannot be drawn. Furthermore, the pragmatic design of the study may have limited the detection of specific subgroups of women who may benefit from endometrial scratching.

WIDER IMPLICATIONS OF THE FINDINGS

Intentional endometrial injury during the follicular phase of OS warrants further attention in future research, as it may be harmful. These findings should be taken in consideration together with the growing evidence from other RCTs that scratching may not be beneficial.

STUDY FUNDING/COMPETING INTEREST(S)

This study was supported by 'Fonds Wetenschappelijk Onderzoek' (FWO, Flanders, Belgium, 11M9415N, 1524417N). None of the authors have a conflict of interest to declare with regard to this study.

Endometrial scratching in women undergoing IVF/ICSI: an individual participant data meta-analysis

BACKGROUND

In IVF/ICSI treatment, the process of embryo implantation is the success rate-limiting step. Endometrial scratching has been suggested to improve this process, but it is unclear if this procedure increases the chance of implantation and live birth (LB) and, if so, for whom, and how the scratch should be performed.

OBJECTIVE AND RATIONALE

This individual participant data meta-analysis (IPD-MA) aims to answer the question of whether endometrial scratching in women undergoing IVF/ICSI influences the chance of a LB, and whether this effect is different in specific subgroups of women. After its incidental discovery in 2000, endometrial scratching has been suggested to improve embryo implantation. Numerous randomized controlled trials (RCTs) have been conducted, showing contradicting results. Conventional meta-analyses were limited by high within- and between-study heterogeneity, small study samples, and a high risk of bias for many of the trials. Also, the data integrity of several trials have been questioned. Thus, despite numerous RCTs and a multitude of conventional meta-analyses, no conclusion on the clinical effectiveness of endometrial scratching could be drawn. An IPD-MA approach is able to overcome many of these problems because it allows for increased uniformity of outcome definitions, can filter out studies with data integrity concerns, enables a more precise estimation of the true treatment effect thanks to adjustment for participant characteristics and not having to make the assumptions necessary in conventional meta-analyses, and because it allows for subgroup analysis.

SEARCH METHODS

A systematic literature search identified RCTs on endometrial scratching in women undergoing IVF/ICSI. Authors of eligible studies were invited to share original data for this IPD-MA. Studies were assessed for risk of bias (RoB) and integrity checks were performed. The primary outcome was LB, with a one-stage intention to treat (ITT) as the primary analysis. Secondary analyses included as treated (AT), and the subset of women that underwent an embryo transfer (AT+ET). Treatment-covariate interaction for specific participant characteristics was analyzed in AT+ET.

OUTCOMES

Out of 37 published and 15 unpublished RCTs (7690 participants), 15 RCTs (14 published, one unpublished) shared data. After data integrity checks, we included 13 RCTs (12 published, one unpublished) representing 4112 participants. RoB was evaluated as 'low' for 10/13 RCTs. The one-stage ITT analysis for scratch versus no scratch/sham showed an improvement of LB rates (odds ratio (OR) 1.29 [95% CI 1.02-1.64]). AT, AT+ET, and low-RoB-sensitivity analyses yielded similar results (OR 1.22 [95% CI 0.96-1.54]; OR 1.25 [95% CI 0.99-1.57]; OR 1.26 [95% CI 1.03-1.55], respectively). Treatment-covariate interaction analysis showed no evidence of interaction with age, number of previous failed embryo transfers, treatment type, or infertility cause.

WIDER IMPLICATIONS

This is the first meta-analysis based on IPD of more than 4000 participants, and it demonstrates that endometrial scratching may improve LB rates in women undergoing IVF/ICSI. Subgroup analysis for age, number of previous failed embryo transfers, treatment type, and infertility cause could not identify subgroups in which endometrial scratching performed better or worse. The timing of endometrial scratching may play a role in its effectiveness. The use of endometrial scratching in clinical practice should be considered with caution, meaning that patients should be properly counseled on the level of evidence and the uncertainties.

Heterogeneity Among Poor Ovarian Responders According to Bologna Criteria Results in Diverging Cumulative Live Birth Rates

Research Question: Does reproductive outcome differ among the various subgroups of poor ovarian responders according to the Bologna criteria? Design: This was a retrospective, cohort study including poor ovarian responders according to Bologna criteria, undergoing an ICSI cycle from January 2011 until December 2017. Patients were divided into four groups: (1) age ≥ 40 years and abnormal ovarian response test, (2) age ≥ 40 years, abnormal ovarian reserve test and one previous poor response to stimulation, (3) age ≥ 40 years and one previous poor response, (4) abnormal ovarian reserve test and one previous poor response. Result(s): Overall, 846 cycles in 706 Bologna poor ovarian responders were included: 310 cycles in group 1, 169 in group 2, 52 in group 3, and 315 in group 4. There were significant differences in age, antral follicle count, antimüllerian hormone, cycle cancellation rates, and number of retrieved oocytes between the four groups. Live birth and cumulative live birth rate differed significantly between groups and were highest in Group 4 [Live birth rate: 7.4% (1) vs. 4.1% (2) vs. 5.8% (3) vs. 13.4% (4), p = 0.001 and Cumulative live birth rate: 8.3% (1) vs. 4.1 % (2) vs. 9.6% (3) vs. 16.8% (4) p < 0.001]. The multivariate GEE analysis revealed that the number of MIIs and the Bologna criteria pattern were the variables which were significantly associated with cumulative live birth rate. Conclusion(s): Poor ovarian responders represent a heterogeneous population. The young subpopulation has a better clinical prognosis in terms of fresh and cumulative live birth rate.

Evaluating the benefit of measuring serum progesterone prior to the administration of HCG: effect of the duration of late-follicular elevated progesterone following ovarian stimulation on fresh embryo transfer live birth rates

RESEARCH QUESTION

Progesterone overproduction during ovarian stimulation is associated with lower live birth rates (LBR) after fresh embryo transfer. Therefore, circulating P concentrations on the day of HCG administration are frequently measured in clinical practice and followed by an elective cryopreservation strategy whenever late-follicular elevated P (LFEP) occurs. A recent study concluded that the duration of LFEP >1.00 ng/mL prior to HCG administration may also affect clinical pregnancy rates. The objective of this current study was to assess whether this hypothesis was reproducible using LBR as the primary outcome.

DESIGN

Retrospective analysis including women undergoing IVF/ICSI between 2010-2015. LBR were compared among different P elevation duration subgroups (0, 1 or >1 day) using two LFEP thresholds (>1.00 ng/mL and >1.50 ng/mL).

RESULTS

The duration of LFEP >1.00 ng/mL was not associated with a significant decrease in LBR according to whether the patient had LFEP lasting for 0, 1 or >1 days (29.9%, 30.3% and 26.3%, respectively). Conversely, when using >1.50 ng/mL as the LFEP threshold, LBR decreased significantly (30.3% 20.4% and 20.5%, respectively). However, the relative frequency of having LFEP >1.50 ng/mL for >1 day was exceedingly rare (1.9%) and the additional benefit of evaluating LFEP beyond the day of HCG triggering no longer remained statistically significant after confounder-adjustment with multivariable regression analysis.

CONCLUSION

These results suggest a lack of benefit in measuring serum P in the days preceding HCG administration, since LBR in women with LFEP >1 day do not vary significantly from those with LFEP detected only on the day of HCG administration.

The proliferative phase endometrium in IVF/ICSI: an in-cycle molecular analysis predictive of the outcome following fresh embryo transfer

STUDY QUESTION

Does an early proliferative phase endometrial biopsy harvested during ovarian stimulation harbour information predictive of the outcome following fresh embryo transfer (ET) in that same cycle?

SUMMARY ANSWER

Transcriptome analysis of the whole-tissue endometrium did not reveal significant differential gene expression (DGE) in relation to the outcome; however, the secretome profile of isolated, cultured and in vitro decidualized endometrial stromal cells (EnSCs) varied significantly between patients who had a live birth compared to those with an implantation failure following fresh ET in the same cycle as the biopsy.

WHAT IS KNOWN ALREADY

In the majority of endometrial receptivity research protocols, biopsies are harvested during the window of implantation (WOI). This, however, precludes ET in that same cycle, which is preferable as the endometrium has been shown to adapt over time. Endometrial biopsies taken during ovarian stimulation have been reported not to harm the chances of implantation, and in such biopsies DGE has been observed between women who achieve pregnancy versus those who do not. The impact of the endometrial proliferative phase on human embryo implantation remains unclear, but deserves further attention, especially since in luteal phase endometrial biopsies, a transcriptional signature predictive for repeated implantation failure has been associated with reduced cell proliferation, possibly indicating proliferative phase involvement. Isolation, culture and in vitro decidualization (IVD) of EnSCs is a frequently applied basic research technique to assess endometrial functioning, and a disordered EnSC secretome has previously been linked with failed implantation.

STUDY DESIGN, SIZE, DURATION

This study was nested in a randomized controlled trial (RCT) investigating the effect of endometrial scratching during the early follicular phase of ovarian stimulation on clinical pregnancy rates after IVF/ICSI. Of the 96 endometrial biopsies available, after eliminating those without fresh ET and after extensive matching in order to minimize the risk of potential confounding, 18 samples were retained to study two clinical groups: nine biopsies of patients with a live birth versus nine biopsies of patients with an implantation failure, both following fresh ET performed in the same cycle as the biopsy. We studied the proliferative endometrium by analysing its transcriptome and by isolating, culturing and decidualizing EnSCs in vitro. We applied this latter technique for the first time on proliferative endometrial biopsies obtained during ovarian stimulation for in-cycle outcome prediction, in an attempt to overcome inter-cycle variability.

PARTICIPANTS/MATERIALS, SETTING, METHODS

RNA-sequencing was performed for 18 individual whole-tissue endometrial biopsies on an Illumina HiSeq1500 machine. DGE was analysed three times using different approaches (DESeq2, EdgeR and the Wilcoxon rank-sum test, all in R). EnSC isolation and IVD was performed (for 2 and 4 days) for a subset of nine samples, after which media from undifferentiated and decidualized cultures were harvested, stored at -80°C and later assayed for 45 cytokines using a multiplex suspension bead immunoassay. The analysis was performed by partial least squares regression modelling.

MAIN RESULTS AND THE ROLE OF CHANCE

After correction for multiple hypothesis testing, DGE analysis revealed no significant differences between endometrial samples from patients who had a live birth and those with an implantation failure following fresh ET. However secretome analysis after EnSC isolation and culture, showed two distinct clusters that clearly corresponded to the two clinical groups. Upon IVD, the secretome profiles shifted from that of undifferentiated cells but the difference between the two clinical groups remained yet were muted, suggesting convergence of cytokine profiles after decidualization.

LIMITATIONS, REASONS FOR CAUTION

Caution is warranted due to the limited sample size of the study and the in vitro nature of the EnSC experiment. Validation on a larger scale is necessary, however, hard to fulfil given the very limited availability of in-cycle proliferative endometrial biopsies outside a RCT setting.

WIDER IMPLICATIONS OF THE FINDINGS

These data support the hypothesis that the endometrium should be assessed not only during the WOI and that certain endometrial dysfunctionalities can probably be detected early in a cycle by making use of the proliferative phase. This insight opens new horizons for the development of endometrial tests, whether diagnostic or predictive of IVF/ICSI treatment outcome.

STUDY FUNDING/COMPETING INTEREST(S)

This study was supported by Fonds Wetenschappelijk Onderzoek (FWO, Flanders, Belgium, 11M9415N, 1 524 417N), Wetenschappelijk Fonds Willy Gepts (WFWG G160, Universitair Ziekenhuis Brussel, Belgium) and the National Medicine Research Council (NMRC/CG/M003/2017, Singapore). There are no conflicts of interests.

TRIAL REGISTRATION NUMBER

NCT02061228.

Culturomics to Investigate the Endometrial Microbiome: Proof-of-Concept

The microbiome of the reproductive tract has been associated with (sub)fertility and it has been suggested that dysbiosis reduces success rates and pregnancy outcomes. The endometrial microbiome is of particular interest given the potential impact on the embryo implantation. To date, all endometrial microbiome studies have applied a metagenomics approach. A sequencing-based technique, however, has its limitations, more specifically in adequately exploring low-biomass settings, such as intra-uterine/endometrial samples. In this proof-of-concept study, we demonstrate the applicability of culturomics, a high-throughput culturing approach, to investigate the endometrial microbiome. Ten subfertile women undergoing diagnostic hysteroscopy and endometrial biopsy, as part of their routine work-up at Brussels IVF, were included after their informed consent. Biopsies were used to culture microbiota for up to 30 days in multiple aerobic and anaerobic conditions. Subsequent WASPLab®-assisted culturomics enabled a standardized methodology. Matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS) or 16S rRNA sequencing was applied to identify all of bacterial and fungal isolates. Eighty-three bacterial and two fungal species were identified. The detected species were in concordance with previously published metagenomics-based endometrial microbiota analyses as 77 (91%) of them belonged to previously described genera. Nevertheless, highlighting the added value of culturomics to identify most isolates at the species level, 53 (62.4%) of the identified species were described in the endometrial microbiota for the first time. This study shows the applicability and added value of WASPLab®-assisted culturomics to investigate the low biomass endometrial microbiome at a species level.

Do we need to measure progesterone in oocyte donation cycles? A retrospective analysis evaluating cumulative live birth rates and embryo quality

STUDY QUESTION

Does late follicular-phase elevated serum progesterone (LFEP) during ovarian stimulation for oocyte donation have an impact on embryo quality (EQ) and cumulative live birth rate (CLBR)?

SUMMARY ANSWER

LFEP does not have an influence on EQ nor CLBR in oocyte donation cycles.

WHAT IS KNOWN ALREADY

Ovarian stimulation promotes the production of progesterone (P) which, when elevated during the follicular phase, has been demonstrated to have a deleterious effect in autologous fresh IVF outcomes. While there is robust evidence that this elevation results in impaired endometrial receptivity, the impact on EQ remains a matter of debate. The oocyte donation model is an excellent tool to assess the effects of LFEP on EQ from those on endometrium receptivity separately. Previous studies in oocyte donation cycles investigating the influence of elevated P on pregnancy outcomes in oocyte recipients showed conflicting results.

STUDY DESIGN, SIZE, DURATION

This is a retrospective analysis including all GnRH antagonist down-regulated cycles for fresh oocyte donation taking place in a tertiary referral university hospital between 2010 and 2017. A total of 397 fresh donor-recipient cycles were included. Each donor was included only once in the analysis and could be associated to a single recipient.

PARTICIPANTS/MATERIALS, SETTING, METHODS

The sample was stratified according to serum P levels of ≤1.5 and &gt;1.5 ng/mL on the day of ovulation triggering. The primary endpoint of the study was the top-quality embryo rate on Day 3, and the secondary outcome measure was CLBR defined as a live-born delivery beyond 24 weeks.

MAIN RESULTS AND THE ROLE OF CHANCE

Three hundred ninety-seven fresh oocyte donation cycles were included in the analysis, of which 314 (79%) had a serum P ≤ 1.5 ng/mL and 83 (20.9%) had a serum P &gt; 1.5 ng/mL. The average age of the oocyte donors was 31.4 ± 4.7 and 29.9 ± 4.5 years, respectively, for normal and elevated P (P = 0.017). The mean number of oocytes retrieved was significantly higher in the elevated P group with 16.6 ± 10.6 vs 11.5 ± 6.9 in the P ≤ 1.5 group (P &lt; 0.001).

In parallel, the total number of embryos on Day 3, as well as the number of good-quality embryos at this stage, was significantly higher in the elevated P group (6.6 ± 5.6 vs 4.15 ± 3.5 and 8.7 ± 6.3 vs 6.1 ± 4.4; respectively, P &lt; 0.001). However, maturation and fertilization rates did not vary significantly between the two study groups and neither did the top- and good-quality embryo rate and the embryo utilization rate, all evaluated on Day 3 (P = 0.384, P = 0.405 and P = 0.645, respectively). A multivariable regression analysis accounting for P groups, age of the donor, number of retrieved oocytes and top-quality embryo rate as potential confounders showed that LFEP negatively influenced neither the top-quality embryo rate nor the CLBR.

LIMITATIONS, REASONS FOR CAUTION

This is an observational study based on a retrospective data analysis. Better extrapolation of the results could be validated by performing a prospective trial. Furthermore, this study was focused on oocyte donation cycles and hence the results cannot be generalized to the entire infertile population.

WIDER IMPLICATIONS OF THE FINDINGS

This is the first study providing evidence that LFEP does not influence CLBR and is adding strong evidence to the existing literature that LFEP does not harm EQ in oocyte donation programs.

STUDY FUNDING/COMPETING INTERESTS

Not applicable.

Expression of adhesion and extracellular matrix genes in human blastocysts upon attachment in a 2D co-culture system

STUDY QUESTION

What are the changes in human embryos, in terms of morphology and gene expression, upon attachment to endometrial epithelial cells?

SUMMARY ANSWER

Apposition and adhesion of human blastocysts to endometrial epithelial cells are predominantly initiated at the embryonic pole and these steps are associated with changes in expression of adhesion and extracellular matrix (ECM) genes in the embryo.

WHAT IS KNOWN ALREADY

Both human and murine embryos have been co-cultured with Ishikawa cells, although embryonic gene expression associated with attachment has not yet been investigated in an in vitro implantation model.

STUDY DESIGN, SIZE, DURATION

Vitrified human blastocysts were warmed and co-cultured for up to 48 h with Ishikawa cells, a model cell line for receptive endometrial epithelium.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Six days post-fertilization (6dpf) human embryos were co-cultured with Ishikawa cells for 12, 24 (7dpf) or 48 h (8dpf) and attachment rate and morphological development investigated. Expression of 84 adhesion and ECM genes was analysed by quantitative PCR. Immunofluorescence microscopy was used to assess the expression of three informative genes at the protein level. Data are reported on 145 human embryos. Mann-Whitney U was used for statistical analysis between two groups, with P < 0.05 considered significant.

MAIN RESULTS AND THE ROLE OF CHANCE

The majority of embryos attached to Ishikawa cells at the level of the polar trophectoderm; 41% of co-cultured embryos were loosely attached after 12 h and 86% firmly attached after 24 h. Outgrowth of hCG-positive embryonic cells at 8dpf indicated differentiation of trophectoderm into invasive syncytiotrophoblast. Gene expression analysis was performed on loosely attached and unattached embryos co-cultured with Ishikawa cells for 12 h. In contrast to unattached embryos, loosely attached embryos expressed THBS1, TNC, COL12A1, CTNND2, ITGA3, ITGAV and LAMA3 and had significantly higher CD44 and TIMP1 transcript levels (P = 0.014 and P = 0.029, respectively). LAMA3, THBS1 and TNC expressions were validated at the protein level in firmly attached 7dpf embryos. Thrombospondin 1 (THBS1) resided in the cytoplasm of embryonic cells whereas laminin subunit alpha 3 (LAMA3) and tenascin C (TNC) were expressed on the cell surface of trophectoderm cells. Incubation with a neutralizing TNC antibody did not affect the rate of embryo attachment or hCG secretion.

LARGE SCALE DATA

None.

LIMITATIONS, REASONS FOR CAUTION

This in vitro study made use of an endometrial adenocarcinoma cell line to mimic receptive luminal epithelium. Also, the number of embryos was limited. Contamination of recovered embryos with Ishikawa cells was unlikely based on their differential gene expression profiles.

WIDER IMPLICATIONS OF THE FINDINGS

Taken together, we provide a 'proof of concept' that initiation of the implantation process coincides with the induction of specific embryonic genes. Genome-wide expression profiling of a larger sample set may provide insights into the molecular embryonic pathways underlying successful or failed implantation.

STUDY FUNDING AND COMPETING INTEREST(S)

A.A. was supported by a grant from the 'Instituut voor Innovatie door Wetenschap en Technologie' (IWT, 121716, Flanders, Belgium). This work was supported by the 'Wetenschappelijk Fonds Willy Gepts' (WFWG G142 and G170, Universitair Ziekenhuis Brussel). The authors declare no conflict of interest.

Comparing Vaginal and Endometrial Microbiota Using Culturomics: Proof of Concept

It is generally accepted that microorganisms can colonize a non-pathological endometrium. However, in a clinical setting, endometrial samples are always collected by passing through the vaginal-cervical route. As such, the vaginal and cervical microbiomes can easily cross-contaminate endometrial samples, resulting in a biased representation of the endometrial microbiome. This makes it difficult to demonstrate that the endometrial microbiome is not merely a reflection of contamination originating from sampling. Therefore, we investigated to what extent the endometrial microbiome corresponds to that of the vagina, applying culturomics on paired vaginal and endometrial samples. Culturomics could give novel insights into the microbiome of the female genital tract, as it overcomes sequencing-related bias. Ten subfertile women undergoing diagnostic hysteroscopy and endometrial biopsy were included. An additional vaginal swab was taken from each participant right before hysteroscopy. Both endometrial biopsies and vaginal swabs were analyzed using our previously described WASPLab-assisted culturomics protocol. In total, 101 bacterial and two fungal species were identified among these 10 patients. Fifty-six species were found in endometrial biopsies and 90 were found in vaginal swabs. On average, 28 % of species were found in both the endometrial biopsy and vaginal swab of a given patient. Of the 56 species found in the endometrial biopsies, 13 were not found in the vaginal swabs. Of the 90 species found in vaginal swabs, 47 were not found in the endometrium. Our culturomics-based approach sheds a different light on the current understanding of the endometrial microbiome. The data suggest the potential existence of a unique endometrial microbiome that is not merely a presentation of cross-contamination derived from sampling. However, we cannot exclude cross-contamination completely. In addition, we observe that the microbiome of the vagina is richer in species than that of the endometrium, which contradicts the current sequence-based literature.

Early pregnancy loss in patients with polycystic ovary syndrome after IVM versus standard ovarian stimulation for IVF/ICSI

STUDY QUESTION

Is the incidence of early pregnancy loss (EPL) in patients with polycystic ovary syndrome (PCOS) higher after IVM of oocytes than after ovarian stimulation (OS) for IVF/ICSI?

SUMMARY ANSWER

Women with PCOS who are pregnant after fresh embryo transfer have a higher probability of EPL following IVM, but after frozen embryo transfer (FET), no significant difference in the incidence of EPL was observed following IVM compared to OS.

WHAT IS KNOWN ALREADY

There is conflicting evidence in the current literature with regard to the risk of EPL after IVM of oocytes when compared with OS. Because of the limited sample size in previous studies, the use of different IVM systems and the possible bias introduced by patient characteristics and treatment type, firm conclusions cannot be drawn.

STUDY DESIGN, SIZE, DURATION

This was a retrospective cohort study evaluating 800 women, with a diagnosis of infertility and PCOS as defined by Rotterdam criteria, who had a first positive pregnancy test after fresh or FET following IVM or OS between January 2010 and December 2017 in a tertiary care academic medical centre.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Pregnancies after non-hCG triggered IVM following a short course of highly purified human menopausal gonadotropin were compared with those after conventional OS. The primary outcome was EPL, defined as a spontaneous pregnancy loss before 10 weeks of gestation.

MAIN RESULTS AND THE ROLE OF CHANCE

In total, 329 patients with a positive pregnancy test after IVM and 471 patients with a positive pregnancy test after OS were included. Women who were pregnant after IVM were younger (28.6 ± 3.4 years vs 29.3 ± 3.6 years, P = 0.005) and had higher serum anti-Mullerian hormone levels (11.5 ± 8.1 ng/ml vs 7.2 ± 4.1 ng/ml, P < 0.001) compared to those who were pregnant after OS. The distribution of PCOS phenotypes was significantly different among women in the IVM group compared to those in the OS group and women who were pregnant after OS had previously suffered EPL more often (28% vs 17.6%, P = 0.003). EPL was significantly higher after fresh embryo transfer following IVM compared to OS (57/122 (46.7%) vs 53/305 (17.4%), P < 0.001), while the results were comparable after FET (63/207 (30.4%) vs 60/166 (36.1%), respectively, P = 0.24). In the multivariate logistic regression analysis evaluating fresh embryo transfer cycles, IVM was the only independent factor (adjusted odds ratio (aOR) 4.24, 95% CI 2.44-7.37, P < 0.001)) significantly associated with increased odds of EPL. On the other hand, when the same model was applied to FET cycles, the type of treatment (IVM vs OS) was not significantly associated with EPL (aOR 0.73, 95% CI 0.43-1.25, P = 0.25).

LIMITATIONS, REASONS FOR CAUTION

The current data are limited by the retrospective nature of the study and the potential of bias due to unmeasured confounders.

WIDER IMPLICATIONS OF THE FINDINGS

The increased risk of EPL after fresh embryo transfer following IVM may point towards inadequate endometrial development in IVM cycles. Adopting a freeze-all strategy after IVM seems more appropriate. Future studies are needed to ascertain the underlying cause of this observation.

STUDY FUNDING/COMPETING INTEREST(S)

The Clinical IVM research has been supported by research grants from Cook Medical and Besins Healthcare. All authors declared no conflict of interest.

TRIAL REGISTRATION NUMBER

N/A.