ART outcome after euploid frozen embryo transfer is not affected by previous Cesarean section delivery in the absence of intracavitary fluid

Abstract

Purpose

To evaluate the impact of a cesarean section (CS) on the chance of clinical pregnancy and live birth (LB) in frozen embryo transfer (FET) cycles in the setting of euploid embryos and the absence of intracavitary fluid (ICF) as causes of implantation failure were excluded.

Methods

Retrospective study, including patients with at least one previous CS or at least one previous vaginal delivery, who underwent a euploid FET cycle.

Results

A total of 412 euploid embryo transfer cycles had been included. Patients’ mean age was 34.5 years and 42.48% of patients have had at least one previous CS. A clinical pregnancy was seen in 69.42% and 60.19% of the patients had a LB. Positive pregnancy test, clinical pregnancy, and LB rate were not significantly different between the groups without/with a history of a previous CS (p = 0.6/0.45/0.94, respectively). LB rate was significantly reduced by the presence of mucus on the ET catheter (OR: 0.413; p = 0.010), the BMI (OR: 0.946; p = 0.006), the combined embryo quality (embryo quality fair: OR: 0.444; p = 0.001; embryo quality low: OR: 0.062; p < 0.001), and by the HRT endometrial preparation approach (OR: 0.609; p = 0.023).

Conclusion

The possible negative impact of a CS can be overcome when a euploid FET after exclusion of ICF is performed.

Introduction

Over the last decades, the number of cesarean sections (CS) increased dramatically and data from 154 countries, spanning the years from 2010 to 2018, estimate that approximately 21.1% of women gave birth by CS worldwide, with a range from 5% (sub Saharan Africa) and an average of 42.8% (Latin America) [1]. Unequivocally, a CS can be a life-saving procedure; however, it is often overused for non-medical indications. Reasons for the preference of a CS are—among others—the strong fear of pain and injuries to mother and child during delivery, uncertainty regarding vaginal birth, and positive views or perceived advantages of CS [2]. Furthermore, pregnant women, who have conceived through assisted reproductive technologies (ART) and their care provider, often consider ART pregnancies as “more precious” [3]. This also results in a higher incidence of CS in patients after invasive infertility treatment [4, 5], independently from a singleton or a multiple pregnancy [6, 7].

With the rising numbers of CS, the short- and long-term consequences for the health of both, mother and child, became increasingly obvious [8]. The impact of a previous CS on the future fertility in the general population is discussed controversially, but it seems that the clinical and social circumstances leading to the CS have a greater effect on future fertility than the surgical procedure itself [9–11]. However, this might be different in women who do not conceive spontaneously after a previous CS and need to undergo ART due to secondary infertility. Whereas some studies describe a reduction of implantation, ongoing pregnancy, and LB rates [12–15], others describe a negative impact of a previous CS only in the presence of an isthmocele [16] or no impact [17].

However, there is a considerable heterogeneity in the previously mentioned studies as they include different embryonic developmental stages as well as fresh and frozen embryo transfers (FET) cycles. Our group published previously that approximately 40% of patients with a previous CS, undergoing ovarian stimulation for IVF/ICSI, develop intracavitary fluid (ICF) [18]. The presence of ICF could impact the intracavitary environment and prevent implantation in case of a fresh ET, when going unnoticed. Further on, ploidy status of the transferred embryo(s) is a crucial factor for achieving a pregnancy and was neglected in almost all studies evaluating the impact of CS on ART outcome.

The last years of ART are characterized by a distinct shift from fresh ET to FET cycles due to various reasons [19]. Therefore, the aim of this retrospective analysis was to evaluate the impact of a previous CS on the chance of a clinical pregnancy and LB in a FET cycle, when aneuploidy of the embryo was excluded through preimplantation genetic testing for aneuploidy (PGT-A) as well as ICF during FET preparation as possible causes of implantation failure.

Material and methods

This retrospective study included patients with secondary infertility and at least one previous cesarean delivery (= previous CS) or at least one previous vaginal delivery (= no previous CS), who underwent a euploid single and double FET cycles after ovarian stimulation with a “freeze-all” approach due to PGT-A at the blastocyst stage. The study was conducted between March 2017 and October 2019 in a tertiary referral fertility center. Ethical approval was given by the Ethical Committee (REFA032 and REFA032a) in 2019 and 2020.

Ovarian stimulation

Patients underwent ovarian stimulation using standardized protocols with either rec-FSH (recombinant follicle-stimulating hormone) or HMG (human menopausal gonadotropin). Dosage was chosen according to the ovarian reserve parameters [20]. As soon as ≥ 3 follicles with a size of ≥ 17 mm were identified on ultrasound scan, final oocyte maturation was initiated, and oocyte retrieval procedure was carried out 34–36 h later under ultrasound guidance. Oocytes were either subjected to IVF or ICSI technique for insemination. Cycles in which the progesterone level was > 1.5 ng/ml on the day of final oocyte maturation were excluded from the analysis as progesterone elevation in the late follicular phase might affect embryo quality [21].

Embryo culture and blastocyst biopsy

Embryos were cultured either in SAGE Quinn’s Advantage Sequential medium or in Life Global single step medium. Fertilization was assessed 17–20 h post-insemination and embryo development was recorded until the day of blastocyst biopsy. Trophectoderm biopsy was performed on day 5–6 post-insemination according to blastocyst development (≥ Bl3CC according to Gardner and Schoolcraft) [22].

Determination of aneuploidy by NGS

A next-generation sequencing (NGS) platform was used (Resproseq, Life-Thermofisher, USA) [23] to assess ploidy status of TE samples.

Frozen embryo transfer (FET) cycle preparation

The choice of the endometrial preparation was at clinicians’ discretion and the patients’ preference, as no protocol was proven to be superior [24]. For both endometrial preparation approaches (hormone replacement therapy (HRT)-FET/natural cycle (NC)), baseline transvaginal ultrasound scans were performed on cycle day 2/3 to exclude uterine and ovarian pathology. Patients were monitored according to clinical standard protocol. No threshold of endometrial thickness was required. In case of the presence of ICF during the endometrial preparation course, the FET was canceled. Ultrasound scans were performed by three designated reproductive medicine specialists, following the same ultrasound technique to avoid bias.

Hormone replacement therapy cycle

Patients were started on E2 (estradiol valerate) 4 mg (2 × 2 mg) on cycle day 2/3 and dosage was increased to 6 mg on day 4 of E2 treatment. Depending on the EMT measured between day 8 and 10, E2 dose was either continued or increased at clinician’s discretion. When the EMT reached a triple line pattern after a minimum of 10 to a maximum of 16 days of E2 administration and ICF, follicle growth and spontaneous ovulation were excluded; an initial vaginal progesterone (P4) dose of 100 mg was started that evening. No minimum endometrial thickness for the embryo transfer was required. From the next day, P4 administration was increased to 100 mg vaginally three times daily and E2 was continued. Embryo transfer was scheduled after 120 h of P4 administration.

Natural cycle

Patients were seen intermittently throughout the cycle and ultrasound scans were performed to monitor follicular growth. In conjunction with ultrasound monitoring, patients underwent serial measurements of serum LH, E2, and P4 levels to determine the timing of ovulation. On day 1 after the LH rise, a decrease in E2 concentration and a rise of the P4 level ≥ 1.5 ng/ml confirmed ovulation (day 0) [25] and vaginal P4 100 mg was started that evening. The following day, patients increased P4 administration to 100 mg vaginally three times daily and embryo transfer was scheduled after 120 h of P4 exposure.

Embryo vitrification and warming

Blastocyst vitrification and warming was performed using the Cryotop method (Kitazato, Biopharma) [26]. A single/double euploid blastocysts were warmed and incubated for 2–4 h to allow blastocoele re-expansion prior to transfer. Embryos, who did not re-expand, were excluded from the analysis. Embryo quality was classified into categories good/fair/low, according to Asociación para el Estudio de la Biología de la Reproducción (ASEBIR) criteria [27]. In cases of DET, the best embryo quality grade was used for classification.

Embryo transfer procedure

The number of embryos (single-embryo transfer (SET) vs double-embryo transfer (DET)) to be transferred was based on medical facts (thickness of myometrium after CS in the CS group, quality of the embryo(s), patients’ individual characteristics) and patients’ preference, if medically justifiable. It has to be kept in mind that DET remains a common practice in Middle Eastern regions, predominantly driven by culture and patient request [28], and the herein included study population was predominantly composed of Arab couples, native to the United Arab Emirates. Embryo transfers were performed under abdominal ultrasound guidance. After the procedure, the embryologist checked the transfer catheter regarding the presence of blood and/or mucus on the outer and inner catheter. The embryo transfer procedure was noted as difficult when it required additional instrumentation such as an obturator (Guardia Obturator; Cook) and/or of a tenaculum. Embryo transfers were performed by three designated reproductive medicine specialists, following a common embryo transfer technique to avoid any bias through different techniques. Our internal audits showed similar pregnancy rates among the physicians, who conducted the ETs in this study.

Definitions of the outcome variables

A pregnancy was defined as an hCG level of ≥ 15 IU/ml 10 days after the embryo transfer procedure; clinical pregnancy and LB follow the ICMART definitions [29].

Statistical analysis

Patient characteristics are described using mean ± SD, median, minimum, and maximum values for continuous variables, frequencies, and percentages for categorical variables. Bivariate comparisons (Wilcoxon test and fisher exact test) and multivariate logistic regression analysis were done. For all statistical tests, a fixed significance level was indicated by p < 0.05. The association between outcomes of pregnancy rate, clinical pregnancy, live birth, patient characteristics, and procedural variables were tested by multiple logistic regression models. Generalized estimating equation (GEE) functions were used to account for the presence of multiple cycles per patient. A backward stepwise selection method was implemented to reach the final regression model by removing parameters with p > 0.1 and adding those with p < 0.05. Confidence intervals (95% CI) were calculated for each significant model. Data analysis was performed using STATA 16.1, StataCorp LLC.

Results

A total of 412 euploid embryo transfer cycles (227 SETs, 185 DETs) had been performed in 342 patients. The mean (± SD) age of the patients was 34.5 (± 4.87) years; mean AMH 2.55 (± 1.54) ng/ml and 175 patients (42.48%) have had at least one previous CS. Endometrial preparation was performed as an HRT cycle in 253 (61.41%) cases, and in the total group, a mean number of 1.45 (± 0.5) embryos was transferred. The ET procedure required some additional instrumentation in 11.65% (48 patients), and in 101 (24.51%)/43 (10.44%) patients, blood/mucus was found on the ET catheter after the procedure, respectively. Table 1 summarizes the data of the included patients. The pregnancy test was positive in 75% (309 patients), a clinical pregnancy was seen in 69.42% (286 patients), and 60.19% (248) of the patients had a LB.

Table 1.

Descriptive of the total group

Variable	n	Mean	Median	Std. Dev	Min	Max
Age (years)	412	34.50	35.00	4.87	20.00	45.00
AMH (ng/ml)	412	2.55	2.39	1.54	0.04	6.86
BMI (kg/m2)	412	27.69	27.39	4.90	13.41	44.06
Infertility duration (years)	412	3.42	2.00	3.10	0.00	25.00
EMT (mm)	412	7.80	7.60	1.40	3.00	13.00
Embryos transferred (n)	412	1.45	1.00	0.50	1.00	2.00
Categorical variables
	Freq	Percent
CS history
Previous CS no	237	57.52
Previous CS yes	175	42.48
Combined embryo quality
Good	110	26.70
Fair	269	65.29
Low	33	8.01
Endometrial preparation regimen
NC	159	38.59
HRT	253	61.41
Presence of blood on ET catheter
No	311	75.49
Yes	101	24.51
Presence of mucus on ET catheter
No	369	89.56
Yes	43	10.44
Difficult ET procedure
No	364	88.35
Yes	48	11.65
Positive pregnancy test
No	103	25.00
Yes	309	75.00
Clinical pregnancy
No	126	30.58
Yes	286	69.42
Live birth
No	164	39.81
Yes	248	60.19

Legend: CS: cesarean section; AMH: anti-Muellerian hormone; BMI: body mass index; EMT: endometrial thickness; NC: Natural Cycle; HRT: hormonal replacement therapy cycle; ET: embryo transfer

Comparison of patients with/without previous CS revealed statistically significant differences in the endometrial preparation regimen with more NC protocols (p = 0.003), a higher incidence of mucus on the ET catheter (p = 0.04), and a higher number of ET procedures, classified as difficult (p = 0.04) in the patients with a history of CS. No statistically significant differences were seen for the other parameters (such as age, AMH, BMI, infertility duration, endometrial thickness at the time of planning the embryo transfer, number of SETs/DETs, combined embryo quality, and the presence of the blood on the ET catheter).

Positive pregnancy test, clinical pregnancy, and LB rate were not statistically significant different between the groups without/with a history of a previous CS (p = 0.6/0.45/0.94, respectively) (Table 2).

Table 2.

Comparison of patients without previous cesarean section versus patients with previous cesarean section

Variables		Prev CS no	Prev CS yes	p-value
Continuous
		N = 237	N = 175
Age (years)		34.1 (5.0)	35.0 (4.6)	0.057
AMH (ng/ml)		2.5 (1.5)	2.6 (1.5)	0.850
BMI (kg/m2)		27.55 (0.3)	27.88 (0.4)	0.500
Infertility duration (years)		3.4 (3.3)	3.4 (2.8)	0.910
EMT (mm)		7.9 (1.4)	7.7 (1.3)	0.110
Categorical		Prev CS no	Prev CS yes	p-value
Embryos transferred (n)	SET	124 (52.3%)	103 (58.9%)	0.190
	DET	113 (47.7%)	72 (41.1%)
Combined embryo quality	Good (1)	67 (28.3%)	43 (24.6%)	0.590
	Fair (2)	153 (64.6%)	116 (66.3%)
	Low (3)	17 (7.2%)	16 (9.1%)
Endometrial preparation regimen	NC	77 (32.5%)	82 (46.9%)	0.003
	HRT	160 (67.5%)	93 (53.1%)
Presence of blood on ET catheter	No	187 (78.9%)	124 (70.9%)	0.061
	Yes	50 (21.1%)	51 (29.1%)
Presence of mucus on ET catheter	No	221 (93.2%)	148 (84.6%)	0.004
	Yes	16 (6.8%)	27 (15.4%)
Difficult ET procedure	No	216 (91.1%)	148 (84.6%)	0.04
	Yes	21 (8.9%)	27 (15.4%)
Positive pregnancy test	No	57 (24.1%)	46 (26.3%)	0.600
	Yes	180 (75.9%)	129 (73.7%)
Clinical pregnancy	No	69 (29.1%)	57 (32.6%)	0.450
	Yes	168 (70.9%)	118 (67.4%)
Live birth	No	94 (39.7%)	70 (40.0%)	0.940
	Yes	143 (60.3%)	105 (60.0%)

Legend: Prev CS, previous cesarean section; AMH, anti-Muellerian hormone; BMI, body mass index; EMT, endometrial thickness; SET, single-embryo transfer; DET, double-embryo transfer; NC, natural cycle; HRT, hormonal replacement therapy cycle; ET, embryo transfer

To evaluate the impact of the variables on the clinical pregnancy rate/LB rate, a multiple regression model (generalized estimating equations (GEE) as some patients contributed with more than one cycle) was applied. After elimination of the parameters, which did not contribute significantly, clinical pregnancy was significantly negatively influenced by the presence of mucus on the ET catheter (OR: 0.389, p = 0.005), by the BMI (OR: 0.953; p = 0.026) and the combined embryo quality (embryo quality fair: OR: 0.41; p = 0.002; embryo quality low: OR: 0.084; p < 0.001). LB rate was also significantly reduced by the same variables (mucus on the ET catheter (OR: 0.413, p = 0.010), BMI (OR: 0.946; p = 0.006), combined embryo quality (embryo quality fair: OR: 0.444; p = 0.001; embryo quality low: OR: 0.062; p < 0.001), and by the HRT endometrial preparation approach (OR: 0.609; p = 0.023) (Table 3).

Table 3.

Multivariate GEE (generalized estimating equations) regression model for the outcomes “clinical pregnancy” and “live birth”

Independent variables		Clinical pregnancy				Live birth
		OR	p	[95% Conf	Interval]	OR	p	[95% Conf	Interval]
Age		1.044	0.076	0.996	1.094	1.042	0.064	0.998	1.088
AMH		0.969	0.669	0.838	1.120	0.955	0.501	0.836	1.092
BMI		0.951	0.024	0.911	0.993	0.941	0.003	0.904	0.980
EMT		1.043	0.609	0.887	1.227	1.008	0.915	0.870	1.168
Previous CS (No = reference)
	Yes	0.896	0.628	0.574	1.398	0.975	0.904	0.647	1.469
Endometrial preparation regimen (HRT = reference)
	NC	0.777	0.293	0.485	1.244	0.597	0.022	0.384	0.929
Number of embryos transferred (SET = reference)
	DET	1.286	0.297	0.802	2.064	1.346	0.178	0.874	2.073
Combined embryo quality (Good = reference)
	Fair	0.387	0.001	0.217	0.689	0.446	0.001	0.271	0.734
	Low	0.085	0.000	0.033	0.218	0.071	0.000	0.025	0.198
Presence of blood on ET catheter (No = reference)
	Yes	1.252	0.422	0.723	2.168	1.338	0.268	0.799	2.238
Presence of mucus on ET catheter (No = reference)
	Yes	0.416	0.015	0.205	0.843	0.413	0.015	0.202	0.843
Difficult ET procedure (No = reference)
	Yes	0.620	0.172	0.312	1.232	0.674	0.242	0.348	1.306
Only with significant variables (using backward stepwise)		Clinical pregnancy				Live birth
		OR	p	[95% Conf	Interval]	OR	p	[95% Conf	Interval]
BMI		0.953	0.026	0.914	0.994	0.946	0.006	0.909	0.985
Combined embryo quality (Good = reference)
	Fair	0.410	0.002	0.234	0.718	0.444	0.001	0.272	0.727
	Low	0.084	0.000	0.034	0.208	0.062	0.000	0.023	0.171
Presence of mucus on ET catheter (no = reference)
	Yes	0.389	0.005	0.200	0.754	0.413	0.010	0.210	0.812
Endometrial preparation regimen (NC = reference)
	HRT	ns	ns	ns	ns	0.609	0.023	0.398	0.934

Legend: CS, cesarean section; AMH, anti-Muellerian hormone; BMI, body mass index; EMT, endometrial thickness; SET, single-embryo transfer; DET, double-embryo transfer; NC, natural cycle; HRT, hormonal replacement therapy cycle; ET, embryo transfer; ns, not significant (hence not included in the final model)

Discussion

Due to the constant increase in the number of CS and the number of FETs, reproductive medicine specialists will be treating more frequently patients with a history of CS. Our retrospective analysis did not demonstrate a negative impact of a previous CS on the clinical pregnancy and LB rate, when a euploid FET cycle after exclusion of ICF was performed, thereby supporting the findings of others [16, 17]. Through the exclusion of embryo aneuploidy as a factor of implantation failure, our data add important knowledge to the existing controversial data on the influence of a previous CS on the ART outcome.

Most of the studies, evaluating the impact of a CS on the ART outcome, are heterogeneous in the design as they include either only fresh [12, 16, 30] or a mix of fresh and frozen embryo transfer cycles [13, 15], at different embryo developmental stages [17] and without genetic testing of the embryos. The only other study, which also excluded embryo aneuploidy as a source of implantation failure, is the retrospective analysis of 551 SET-FET cycles by Friedenthal et al. [14], describing a decreased implantation rate and a reduction in the ongoing pregnancy and LB rate in patients after CS. Our analysis included SET as well as DET cycles; hence, this fact does not pose a bias as there was no significant difference in the number of SETs and DETs between patients with and without a previous CS.

A possible explanation for the different findings between Friedenthal et al. [14] and our study could lie in the fact that authors used only an HRT approach as endometrial preparation. Contrary to their approach, the herein presented study used HRT as well as NC regimen for the endometrial preparation. Hence, the multivariate analysis, involving the parameters which are considered important for ART success, demonstrated a significant negative impact of the HRT approach on LB rates. According to our results, it seems beneficial to choose a NC approach for the endometrial preparation; however, these data must be confirmed by larger studies.

A previous CS alters the uterine anatomy and might result in a CS defect of varying magnitude, also termed “isthmocele.” Through the availability and routine use of ultrasound machines with high-resolution vaginal ultrasound probes, an isthmocele is usually easily detectable [31]. It is defined as an anechoic area of different shapes at the site of the CS scar, with a depth of at least 1 mm [32]. Patients with an isthmocele are at a 40% risk of developing ICF during ovarian stimulation [18], a finding which could impact the chances of a pregnancy in a fresh embryo transfer cycle, if missed during the monitoring process [33]. The adverse impact of ICF was described previously [34]. The process of “apposition” and “attachment “ of the embryo to the endometrial surface is crucial for implantation and this process could potentially be severely disturbed in the presence of ICF. In patients with ICF, caused by an existing hydrosalpinx, the embryotoxic properties of the ICF seem to play an additional role for the detrimental outcome [35, 36]. Due to a lack of sufficient data of the impact of ICF, caused by an isthmocele, on the ART outcome in FET cycles, it can be only be speculated that this event would reduce the chance of a pregnancy. Further prospective studies are warranted, but might seem undue due to ethical concerns. Until today, there are no commonly accepted guidelines how to treat patients with persistent/recurrent ICF and different treatment options could be discussed, according to the individual case: alternative endometrial preparation approach, aspiration of the ICF before the ET procedure, and/or surgical correction of the isthmocele.

As a result of a disturbed contractility in the scar region with a reduced expulsion of secretions from the niche, the isthmocele is often filled with “old” menstrual blood from the previous period or with mucus. This explains a tendency for the prevalence of blood and a significantly higher prevalence of mucus on the ET catheter in patients after CS. In the multivariate analysis, the presence of mucus on the ET catheter revealed a significantly negative impact on the clinical pregnancy and LB rate. Interestingly, when Diao et al. [16] stratified their study population into cases with and without an isthmocele after CS, a significant negative impact on the LB rate was seen in patients with isthmocele, whereas the history of a CS without a defect did not impact the ART outcomes [16]. Despite the mechanism, in which way the presence of mucus reduces clinical pregnancy and LB rate, is not clear, it can be hypothesized that the presence of mucus in the niche and the possible spread of the mucus into the cavity by the ET procedure causes the ART failure.

The alteration of the uterine anatomy as a long-term consequence of the CS can increase the difficulty of the ET procedure [17]. Similar to other studies [14, 17], and in our study as well, the embryo transfer procedure itself was described significantly more often as “difficult” in the CS group, hence without a negative impact on the ART outcome. Therefore, it can be concluded that the transfer difficulties can be overcome when the procedure is performed in a careful and gentle manner under ultrasonographic guidance. Theoretically, a difficult ET procedure might present a cause for an implantation of the pregnancy in the CS scar itself. Data on the prevalence of CS pregnancies in IVF are scarce and are estimated to be around 1:1700 [37] in patients undergoing a fresh embryo transfer. So far, no data are published for patients undergoing a FET cycle and no CS scar ectopic pregnancy occurred in the herein included study population.

The conclusions of this study might be limited by the retrospective design and the non-random choice of the endometrial preparation approach, therefore resulting in the higher number of NC cycles in group with a previous CS. Still, the strength of the data is the exclusion of embryo aneuploidy as source of implantation failure, thereby adding important information on the controversial data on the influence of CS on ART outcome.

Conclusion

The number of CS is rising continuously, and it seems that a CS as mode of delivery is especially popular among physicians and patients who conceived after a fertility treatment. Undoubtedly, this development is unfavorable as some data point to the fact that secondary infertility might be more common in patients after CS. This might impact the outcome in patients who have to undergo an ART treatment. Despite the fact, that the negative impact of a CS can be overcome when a FET with a euploid embryo(s) after exclusion of ICF is performed in natural cycle, it is of utmost importance to prevent the abuse of CS as delivery mode.

Author contribution

Bayram A: conceptualization of study, data analysis, review of paper

Elkhatib I: performance of embryo transfers, review of paper

Abdala A: performance of embryo transfers, review of paper

Noguiera D: review of paper

Melado L: performance of embryo transfers, review of paper

Fatemi HM: review of paper

Lawrenz B: data analysis, performance of embryo transfers, drafting of paper

Declarations

Conflict of interest

The authors declare no competing interests.