Do serum androgens influence blastocysts ploidy in karyotypically normal women?

ABSTRACT

The aim of the study was to determine if serum testosterone (T) and dehydroepiandrosterone (DHEAS) levels are a factor in determining increased risk for embryonic aneuploidy in karyotypically normal women undergoing in vitro fertilization (IVF) and preimplantation genetic testing screening for aneuploidy (PGT-A). This is a retrospective cohort study of IVF cycles with PGT-A performed during 2015–2016. A total of 256 cycles with 725 embryos were initially considered for inclusion. A total of 208 cycles and 595 embryos determined to be either euploid or aneuploid were included in the analysis. The mean age of women was 37.4 ± 4.4 years. There were 193 (32.44%) euploid, and 338 (56.81%) aneuploid blastocysts. Sixty-four (10.76%) had ‘no diagnosis’ after PGT-A. The 32 embryos with ‘no diagnosis’ after first PGT-A were biopsied again and after the second analysis, 7 were found to be euploid and 3 aneuploid. The remaining 32 embryos were not reanalyzed due to the lack of patients’ consent for the second biopsy. The relationship between embryo ploidy and levels of serum testosterone and dehydroepiandrosterone sulfate was assessed using ordinal multivariable regression analysis. The model, adjusted for both anti-Mullerian hormone (AMH) and age, showed no association between ploidy status and serum levels of the two hormones. We concluded that the serum levels of testosterone and DHEAS do not influence embryo ploidy in karyotypically normal women undergoing IVF.

Abbreviations: T: testosterone; DHEAS: dehydroepiandrosterone; IVF: in vitro fertilization; PGT-A: preimplantation genetic testing screening for aneuploidy; AMH: anti-Mullerian hormone; FSH: follicle-stimulating hormone; LH: luteinizing hormone; E₂: oestradiol; P: progesterone

KEYWORDS:

• Testosterone
• dehydroepiandrosterone
• blastocysts ploidy
• androgens

Introduction

The successful outcome of in vitro fertilization (IVF) procedure depends on a variety of factors. Two of the most important ones, and independently influencing the results, are female age and her serum level of the anti-Mullerian hormone (AMH) (Broekmans et al. 2006; La Marca et al. 2011; Lehmann et al. 2014). Embryos obtained from the IVF procedure are transferred to the uterus in either fresh or frozen cycle. It has been established that ploidy of embryos influences both clinical pregnancy and live birth ratios (Franasiak et al. 2014; Minasi et al. 2016; La Marca et al. 2017). In older women (over 42 years old) aneuploidy rates reach close to 90% (Minasi et al. 2016). This age-dependent increase is mainly related to the lower quality of oocytes (Battaglia et al. 1996; Lim and Tsakok 1997; Pellestor et al. 2003). AMH level is also an independent factor related to aneuploidy where lower AMH concentrations have been shown to correlate with increased aneuploidy rates in embryos (Katz-Jaffe et al. 2013; La Marca et al. 2017).

It has been established that androgens play an important role in folliculogenesis (Vendola et al. 1998). They act via androgen receptor and influence the recruitment and growth of follicles during the early stages of the folliculogenesis (Ferrario et al. 2015). This action is in some extent achieved via increasing of IGF-1 (Weil et al. 1999). They also increase FSH receptor mRNA in granulosa cells and in that way exerts synergistic effect with FSH on folliculogenesis (González-Comadran et al. 2012). That leads to diminished amount of FSH during controlled ovarian hyperstimulation-COH.

The most potent androgens in serum are DHEAS and testosterone. They are predominantly synthetized in ovaries and suprarenal glands. The enzyme steroid sulphatase catalyzes the conversion of DHEAS to DHEA (Bonser et al. 2000). There are studies showing positive influence of serum androgen concentration on the IVF outcome (Dickerson et al. 2010; Ferrario et al. 2015; Sathyapalan et al. 2017). It has also been well established that supplementation of testosterone and DHEA in women with low ovarian reserve and inadequate androgen concentrations diminishes miscarriage rates and positively influences IVF outcome (González-Comadran et al. 2012). That effect has not been observed in women with normal response to stimulation (Yeung et al. 2016).

There are a limited number of studies that investigated the association between female serum androgens and embryo aneuploidy rates. The results come from either small number of women or from women supplemented with dehydroepiandrosterone (Gleicher et al. 2010, 2015) or women with polycystic ovary syndrome (Weghofer et al. 2007). Additionally, in some studies data regarding possible confounding factors are missing, i.e., information about parental karyotype or paternal age. The question regarding the influence of androgens on embryo ploidy remains open and thus we aimed to investigate it in karyotypically normal couples undergoing in vitro fertilization treatment.

Results

From the initial group of 207 patients (256 cycles) 168 patients with 208 cycles were included in the final analysis. The total number of blastocysts was 595. The number of euploid embryos was 193 (32.44%), aneuploid 338 (56.81%), and 64 (10.77%) had ‘no diagnosis’ after PGT-A. Thirty-two embryos with ‘no diagnosis’ after first PGT-A were biopsied again and after the next analysis, 7 were found to be euploid and 3 aneuploid. The remaining 32 embryos were not reanalyzed as patients did not consent to the second biopsy.

Baseline characteristics and ploidy of embryos in the subgroups, defined as below and over median for serum testosterone and DHEAS levels, are presented in Table 1.

Table 1. Baseline characteristics of the study population and differences between patients with low and high level of Testosterone and DHEA-S.

	Total N = 208	Testosterone <0.8 N = 98	Testosterone ≥ 0.8 N = 110	p	DHEA-S < 176 N = 104	DHEA-S ≥ 176 N = 104	p
Age of women (years)	37.4 ± 4.4	37.7 ± 3.3	37.0 ± 5.0	0.636	38.6 ± 3.5	36.2 ± 4.8	0.006
Age of men (years)	38 (35–41)	39 (37–42.5)	38.59 ± 4.7	0.4031	39 (36–42)	39 (34.5–41)	0.5566
BMI (kg/m^2)	22.9 ± 2.8	22.9 ± 2.5	23.0 ± 3.1	0.899	22.9 ± 2.2	22.9 ± 3.7	0.731
Primary infertility diagnosisSecondary infertility diagnosis	74% 26%	68% 32%	81% 19%	0.097	72% 28%	78% 22%	0.399
AMH (ng/mL)	2.1 (1.1–3.6)	1.6 (0.9–3)	2.8 (1.3–4.5)	0.002	2.0 (1.1–3.8)	2.2 (1.1–3.4)	0.854
FSH (mIU/L)	6.3 (4.8–7.7)	6.5 (5.4–8.9)	5.9 (4.1–7.1)	0.027	6.5 (5.4–7.9)	6.0 (4.3–7.5)	0.460
LH (mIU/L)	6.6 (5.1–7.9)	5.8 (4.4–7.3)	7.3 (5.6–8.4)	0.041	6.4 (5.1–7.8)	6.7 (5.4–8.8)	0.265
Testosterone (nmol/l)	0.8 (0.5–1.1)	-	-	-	0.6 (0.4–0.8)	1.1 (0.8–1.4)	<0.001
DHEA-S (mg/dL)	169 (126–259)	146 (108.5–170.5)	230 (144–288)	<0.001	-	-	-
E2 (pg/ml)	38 (29–55)	38 (29–54)	43 (33–100)	0.335	38 (33.5–55)	38 (24–82)	0.763
E2 on 8 ds. (pg/ml)	1653 (1148–2342)	1630 (1115–2249)	1590 (1148–2319)	0.942	1656 (1110–2189)	1621.5 (1226–2484)	0.667
Duration of infertility (years)	4 (2–7)	4 (3–7)	4 (2–8)	0.939	4 (2–7)	4 (2–8)	0.714
Stimulation length (days)	9.5 (8–11)	9 (8–10)	10 (9–11)	0.006	9 (8–10.5)	10 (9–11)	0.298
Antral follicle count (AFC)	11 (7–17)	10 (7–15)	12 (8–18)	0.152	10 (7–17)	11 (8–17)	0.604
Total gonadotropin dose (units)	2294.4 ± 556.3	2236.3 ± 471.1	2282.8 ± 617.3	0.653	2314.6 ± 461.3	2274.2 ± 640.7	0.693
Number of cumulus	9 (6–13)	8 (6–14)	9 (6–12.5)	0.734	9 (6–13.5)	9 (5.5–13)	0.918
No of mature oocytes	7 (4–9.5)	7 (4–10)	6 (4.5–8.5)	0.945	7 (5–10)	6 (4–9)	0.553
Number of available embryos	5 (3–7)	4 (3–6)	5 (2–6)	0.665	5 (3–6.5)	4 (3–7)	0.432
Blastocysts biopsied	2 (1–3)	2 (1–3)	2 (1–3.5)	0.933	2.5 (1.5–4)	2 (1–3)	0.082
Euploid blastocysts %	40 (0–66.7)	50 (0–66.7)	33.3 (0–66.7)	0.565	50 (0–66.7)	50 (0–100)	0.510
Aneuploid blastocysts %	60 (33.3–100)	50 (33.3–100)	66.7 (33.3–100)	0.565	50 (33.3–100)	50 (0–100)	0.510
No. of subjects with indicated cause of infertility History of miscarriage Advanced maternal age (set at 35) Male factor Anovulation Unexplained	36 70 70 2 30	19 33 31 1 14	17 37 39 1 16	0.952	19 32 41 2 10	17 38 29 0 20	0.091

BMI: body mass index; FSH: follicle-stimulating hormone; AMH: anti-Müllerian hormone; LH: luteinizing hormone; E₂: oestradiol; ds: day of stimulation

Data with not normal distribution are presented as: median (25th–75th)

Data with normal distribution as a mean ± standard deviation

P value for statistically significant results is shown in bold

In the subgroup with serum DHEAS below median the women were older (38.6 ± 3.5 vs. 36.2 ± 4.8; p = 0.006) and in the subgroup with testosterone over median the serum AMH and LH were significantly higher (2.8 vs. 1.6; p = 0.002 and 7.3 vs. 5.8; p = 0.041, respectively) and FSH significantly lower (5.9 vs. 6.5; p = 0.027).

Thus, it was necessary that our further analysis of embryo ploidy was conducted with adjustment for age and AMH. The data are presented in Table 2. However, we found that neither serum testosterone nor DHEA-S influences aneuploidy rates in embryos (p = 0.312 and p = 0.286, respectively). All analyzed correlations are presented in Figure 1.

Table 2. Ordinal multivariable logistic regression model for embryo ploidy.

		OR	95% CI for OR	p
Aneuploid blastocysts	Testosterone	0.396	0.066–2.387	0.312
	DHEA-S	1.004	0.996–1.013	0.286
All (euploid and aneuploid) blastocysts	Testosterone	0.277	0.042–1.838	0.184
	DHEA-S	0.647	0.993–1.011	0.647

Adjusted by AMH and age

Figure 1. Correlations among clinical and embryological variables.

Circles represent strength (size and color) and direction (color) of the correlation. Non statistically significant correlations are crossed out.

Discussion

The cumulative pregnancy rate after both fresh and frozen embryo transfers depends on the number of euploid blastocysts (Capalbo et al. 2014). Embryo ploidy is known to be dependent on female age, serum AMH, and the number of mature oocytes (Battaglia et al. 1996; Lim and Tsakok 1997; Pellestor et al. 2003; La Marca et al. 2017). There is less information available in the literature regarding the serum androgen concentrations and their relationships with the ploidy of embryos (Gleicher et al. 2009, 2010, 2015; La Marca et al. 2017).

There have been different techniques used in the embryo aneuploidy screening. The genomic hybridization arrays have been often used in many centers but have some limitations such as inability to detect changes in DNA sequences like: point mutations, intragenic insertions or deletions, triplet repeat expansion (Harper and Harton 2010). The next step in embryo analysis was the introduction of next-generation sequencing (NGS). This new technique gives more reliable and reproducible results in embryo ploidy testing which is important from the clinical point of view (Kuliev and Verlinsky 2004; Brezina et al. 2016).

In this study, we present the largest cohort of women (cycles) for which we investigate the association of androgens with embryo ploidy. We found that in the below median DHEA-S subgroup women were older and in the over the median testosterone subgroup serum AMH was significantly higher, which can be explained by the fact that DHEA-S levels in women decrease with age (Labrie et al. 1997).

We found that neither serum testosterone nor DHEA-S influences aneuploidy rates in embryos. The same conclusion was true when the model was adjusted for female age and serum AMH level. We have also attempted to create other models. First, we included each androgen variable separately. This model did not show the influence of androgens on embryo ploidy. In our patient cohort, over 40% of the women had no euploid embryos. Therefore, next we modeled the binary outcome defined as having at least one euploid embryo or no euploid embryos. This model failed to show any association of androgens with such defined outcome (data not shown). Finally, even adding the number of mature oocytes (MII) as a variable did not change our conclusions.

Our results are in agreement with few studies in which the relationship between serum androgens and ploidy was investigated (Weghofer et al. 2007; Gleicher et al. 2009, 2010, 2015). It is important to note that the previously published studies regarding androgens and ploidy of embryos included only patients who were supplemented with DHEA. Authors concluded that it could diminish aneuploidy and miscarriage rates (Gleicher et al. 2009, 2010).

In our study the number of women with extremely low AMH <0.4 (and supplemented with DHEA) was four (with five cycles) and we do not think it could have impact on the final results. Additionally, supplementation with DHEA was stopped before the beginning of ovarian stimulation.

In our study, the age of women correlated significantly with number of euploid embryos. This is in agreement with many other studies showing an increase in aneuploidy rates with woman’s age up to 90% for women older than 42 years (La Marca et al. 2017). In comparison, we did not show a significant correlation between embryo ploidy and serum AMH levels. This is a surprising finding worth to point out given conflicting reports on the influence of ovarian reserve on embryo ploidy (La Marca et al. 2017; Jiang et al. 2018).

It is important to note that we included in the analysis only couples with normal karyotypes (both partners). We aimed to include only couples with normal karyotype as abnormal parental karyotype increases the risk of embryonic aneuploidy. We have also excluded embryos that did not receive a PGT-A result (classified as ‘no diagnosis’ after PGT-A). Our goal was to obtain the study group that was as homogenous as possible.

The strength of this study is the strict inclusion criteria. We assessed karyotypes in all included couples (this is performed routinely in our clinic) and analyzed our data with respect to confounders such as age and AMH. We have also taken into consideration the men’s age as a factor which possibly has an impact on the blastocyst ploidy (García-Ferreyra et al. 2018). In our study, the men’s age was similar in all subgroups. Additionally, all IVF cycles and PGT were performed in a single center. This seems to be an important factor as having different laboratories that use various methods of aneuploidy screening could affect the results (Mateo et al. 2017; Munné et al. 2017; Grati et al. 2018; Spinella et al. 2018).

We have also excluded obese women as obesity can influence androgen levels. Our method of screening embryos was NGS which is routinely used in our clinic and currently considered to be the most reliable and accurate method of assessment of embryo ploidy.

The limitation could be the retrospective nature of the study, but we believe that the strict inclusion criteria outweighed this bias.

Finally, we have not classified embryos according to the current PGDIS criteria (PGDIS Position Statement on Chromosome Mosaicism and Preimplantation Aneuploidy Testing at the Blastocyst Stage 2016) as most of our data were collected before its publication. The issue of mosaicism is still under debate and various clinical, methodological and laboratory factors could influence the results (Scott et al. 2012; Treff et al. 2012; Wells et al. 2014; Maxwell and Grifo 2018; Munné 2018; Palmerola et al. 2019; Victor et al. 2019).

Conclusion

In conclusion, our data provide support for the assumption that the serum concentrations of testosterone and DHEA-S do not affect the risk for embryonic chromosomal abnormalities.

Materials and methods

We retrospectively analyzed data obtained from couples who underwent their cycle of IVF treatment with preimplantation genetic testing for aneuploidy at the Invicta Fertility Clinic, between January 2015 and December 2016.

During that period 207 patients had 256 cycles with the total number of 725 analyzed blastocysts. All procedures performed in this study were in accordance with 1964 Helsinki declaration, and its later amendments and the study protocol was approved by the appropriate Institutional Review Board and written informed consent was given by all participating patients.

The indications for PGT-A (Preimplantation genetic testing for aneuploidy – abnormal number of chromosomes) included: advanced maternal age, repeated implantation failure, recurrent miscarriage, and severe male infertility. Until November 2015 patients could also request PGT-A to check embryo ploidy even if they did not present with any of the abovementioned indications.

The inclusion criteria were: IVF cycle with PGT-A, normal karyotype of both partners, AMH level measured with either VIDAS AMH (bioMerieux) or Elecsys AMH (Roche Diagnostics) assay. It has been previously verified that results obtained with VIDAS AMH assay were well correlated and comparable with Elecsys assay (Pastuszek et al. 2017).

We excluded 44 cycles due to abnormal karyotype of one of the parents (female partner: 24 cycles, 22 patients, male partner: 20 cycles, 15 patients), 3 cycles due to polycystic ovary syndrome according to Rotterdam criteria (PCOS) (Rotterdam 2004) (1 patient), and 1 cycle due to obesity (1 patient).

Ultimately 208 cycles (168 women) were included in the analysis. The total number of analyzed blastocysts was 595.

This study was approved by the Institutional Review Board, decision number KB-29/17.

Hormone measurements

Venous blood samples were collected, between day 1 and 3 of the menstrual cycle, prior to stimulation. The serum levels of follicle-stimulating hormone (FSH), luteinizing hormone (LH), oestradiol (E₂), progesterone (P), testosterone (T), were measured with an electrochemiluminescence immunometric assays using commercial kits (Cobas 6000, e601). The intra- and inter-assay coefficients of variation (CV), respectively, were 7.4% and 4.7% for E₂, 2.5% and 2.3% for FSH, 2.2% and 2.4% for LH, 4.9% and 5% for T, and 3.3% and 3.2% for DHEAS.

The reference intervals were as follows: serum, T, 0.1–0.9 ng/ml; SHBG and DHEAS, 98.8–340 µg/dl. Serum AMH were measured by enzyme-linked immunosorbent assay (ELISA) using VIDAS AMH (bioMerieux) and Elecsys AMH assay (Roche Diagnostics).

Protocol

All patients began treatment in the early follicular phase between day 1 and 5 of the menstrual cycle. The long GnRH agonist protocol used in our center is started with oral contraceptive with Ovulastan (ADAMED, Poland). Starting on treatment day 16 or 17, triptorelin (Ferring, Germany) was administered every two days for 14 days. After pituitary suppression, confirmed by low serum levels of estradiol (<50 pg/ml) and the absence of ovarian cysts, human menopausal gonadotropins (HMGs) (Menopur, Ferring, Germany) were initiated for ovarian controlled hyperstimulation. It was continued until at least three follicles ≥17 mm were visualized. Follicular growth was monitored on treatment day 8 with an ultrasonographic scan and serum E2 measurement. Ovulation was then induced by administering 5,000 IU hCG (Choragon, Ferring, Germany). Oocyte retrieval was performed by transvaginal aspiration 36 h after the hCG administration (Lukaszuk et al. 2005).

PGT-A

In vitro fertilization, embryo culture and blastocyst biopsy techniques were performed as previously described (Lukaszuk et al. 2005). Intracytoplasmic sperm injection was performed routinely to remove potential sperm or cumulus cell DNA contamination. Trophectoderm biopsy was performed on all expanding or fully expanded blastocysts on postretrieval day 5 or 6, depending on the rate at which individual embryos reached this developmental stage. Following the biopsy embryos were frozen and their outcome is not discussed in the manuscript. Each trophectoderm biopsy sample underwent comprehensive chromosome screening analysis performed at the same laboratory using next-generation sequencing (NGS) technique (Kuliev and Verlinsky 2004). Embryos with over 50% of aneuploid cells for any chromosome were classified as aneuploid, those with up to 50% of aneuploid cells were reported as euploid. If we were to compare our classification criteria to those published by PGDIS (PGDIS Position Statement on Chromosome Mosaicism and Preimplantation Aneuploidy Testing at the Blastocyst Stage 2016) that suggest the ‘cut-off point for definition of mosaicism is >20%, so lower levels should be treated as normal (euploid), >80% abnormal (aneuploid), and the remaining ones between 20% and 80% mosaic (euploid-aneuploid mosaics)’, we would say that, in our clinic, embryos with mosaicism in the range of 20–50% were reported as euploid and embryos with mosaicism in the range of >50% to 80% were labeled as aneuploid (Liss et al. 2018). Segmental aberrations are reported as aneuploid when they cover over 50% of a chromosome.

Statistics

The results were presented separately in the low/high testosterone and DHEA-S groups. The cut-off point for both variables was the median value, which for testosterone was 0.8 mg/ml and for DHEA 176 µg/dl. Differences between continuous variables were calculated with the t-test for symmetrical distributions and with the Mann–Whitney test for asymmetrical distributions. Differences between qualitative variables were calculated using the chi-square test (Table 1). Correlations between quantitative variables were evaluated using the Spearman’s rank correlation coefficient (rS), along with its significance.

An ordinal multivariable logistic regression model was created to predict outcome defined as cycle resulting in all euploid, mixed euploid and aneuploid, and all aneuploid embryos. Variables included in the model were testosterone and DHEA levels as well as AMH and female age (Table 2). The statistical analysis was performed using the IBM SPSS version 23 statistical package. The significance level was set at α = 0.05. Due to the retrospective nature of the study, we did not have any missing data.

Authors’ contributions

Conception and design of the idea, data interpretation and preparation of manuscript: MK, PS, EP, GJ, RS, KŁ.

Disclosure statement

No potential conflict of interest was reported by the authors.