Serum and follicular fluid thyroid hormone levels and assisted reproductive technology outcomes

Thyroid hormones (THs) are related to infertility and multiple adverse neonatal and maternal consequences [1, 2]. Recently, many studies have evaluated the relationship between ART outcomes and thyroid function [3‐5]. A growing body of literature debates what constitutes “normal” gestational and preconceptional thyroid function and treatment cut-offs [4, 5]. The latest meta-analysis on preconception subclinical hypothyroidism (SCH), which included 14,846 participants, found no significant differences in ART-related outcomes between different thyroid-stimulating hormone (TSH) level groups when the TSH cut-off was set to 2.5 mIU/L. However, when a broader TSH cut-off range (3.5–5 mIU/L) was used, the miscarriage rate was higher in the preconception SCH group than in the normal group [4]. On the other hand, several studies of euthyroid infertile women undergoing intrauterine insemination (IUI) [6‐8], found no differences in TSH levels among different IUI outcome groups. In addition, most previous studies categorized women with no history of thyroid disease and normal TSH levels as euthyroid; TH levels and thyroid antibody status were not known.

Follicular fluid (FF) supports the acquisition of development competence in oocytes [9, 10], and provides the important microenvironment for oocyte maturation. Changes in the FF levels of hormonesand metabolites have been reported to influence oocyte quality, early embryo development, and subsequent pregnancy [11, 12]. Since biological effects of THs are regulated by deiodinase (DIO) in peripheral tissue [13, 14], serum TH levels do not always predict tissue-specific effects in target organs, and local THs may play a direct role in physiological functions. Although the presence of THs in human FF was verified in 1993 [15], the concentration of THs in FF has been analysed in only a small number of studies [16, 17]. One preliminary observational study found that fT4 in FF were higher in infertile patients than in the healthy population [18]. However, data from literature are not sufficiently clear to definitely state the relationship between serum and FF TH levels and the outcomes of assisted reproductive technology (ART),such as in vitro fertilisation (IVF).

Previous studies used only serum TSH as a biomarker to evaluate thyroid function. The aim of our study was to compare the relationship of serum and FFTHs withcycle parameters and ART outcomes…

This study was approved by the Ethics Committee of the First People’s Hospital of Yunnan Province and was carried out according to good clinical practices. Informed consent was obtained from each patient.

The present study verified the presence of THs and thyroid autoantibodies in the FF of women undergoing ART and assessed the impact of these factors on embryonic development and ART outcomes. We demonstrated the presence of TSH, T4, T3, fT4, fT3 and thyroid autoantibodies in FF, and estimated their impact on fertilization and embryo development during ART. Our study highlighted that the majority of patients have FF TH levels within the normal serum range [15, 16] We also observed a significant positive correlation between serum and FF TH levels, indicating that the majority of THs detected in FF seem to be derived from peripheral blood and enter follicles through theca interna cells. Moreover, we detected significantly higher concentrations of T3 in FF than in serum, whereas T4 concentrations were higher in serum than in FF, which is in line with the findings of previous studies [15, 17]. Our work also showedthat the T4/T3 ratio was much lower in FF than in blood, which supportsthe presence of an ovarian 5′-monodeiodinase system in FF capable of generating T3 (ovary- generated T3) by outer ring deiodination of T4 [17]. We also detected thyroid autoantibodies in FF, but the levels were much lower than in serum. The results suggest that the blood-follicle barrier is a permselective barrier for thyroid autoantibodies.

The pivotal role of THs in several aspects of female reproduction have been well documented by several investigators. THs may impact folliculogenesis [19‐21], ovarian steroidogenesis [22], and endometrial receptivity [23]. There is evidence that thyroid function is associated with pregnancy outcome, particularly in IVF. Despite treatment, women with hypothyroidism may have lower chances of pregnancy success after IVF [24]. SCH may also impact reproduction, and treatment of women with SCH has been shown to improve IVF outcomes [25, 26], including increasing the rates of implantation, clinical pregnancy, and delivery.

Because of the strong associations of SCH with subfertility, the definition of euthyroidism in subfertile women is currently a topic of debate. TSH is considered the most sensitive test for thyroid function, so currently, studies investigating the association between SCH and infertility are based on serum TSH levels. In the present study, we showed no difference in TSH levels in euthyroid women undergoing ART among different IVF outcome groups. Our results are in line with those recently reported by Karmon AE et al. [6‐8]. In the group of euthyroid women undergoing infertility treatment with ART, the authors observed no significant differences in clinical pregnancy or delivery rates among euthyroid infertile women with different preconceptional TSH levels after IUI.

However, evidence is lacking to support the use of T4 or fT4 to categorize euthyroid infertile women undergoing IVF and to predict pregnancy outcomes. It is unknown whether distinctions can be made among TH values within the normal range of TSH and the chances of certain fertility outcomes. In the current study, the major differences between women achieving pregnancy and those who experienced implantation failure were in serum T4, fT4 levels and the T4/T3 ratio, This discrepancy among TSH,T4 and fT4 levels might be explained by the physiology: COS-induced hyperestrogenism directly reduces serum fT4 levels by stimulating thyroid binding globulin (TBG) production, whereas the increase in serum TSH levels is achieved by a negative feedback loop [27, 28]. Taken together, these data suggest that when there is a rapid increase in TBG leading to high variability in fT4 levels, the changes in TSH lags. Our study results show that the determination of fT4 levels and the T4/T3 ratio in serum on the day of OPU, together with TSH, maybe have greater prognostic value than TSH alone.

A growing body of literature indicates that the hormonal follicular milieu, which includes anti-Mullerian hormone and inhibin B, is correlated with reproductive outcome after IVF [29]. However, few studies have examined the possible association between FF THs and IVF outcomes. Our results are the first to show a link among the FF T4/T3 ratio, embryonic developmental competence and successful pregnancy. Embryo quality is an important predictor of ART success. As expected,the successful pregnancy group had a significantly higher mean number of blastomeres, MII oocytes and retrieved oocytes than the implantation failure group. Significant correlations were found between the numbers of retrieved oocytes and fertilized oocytes and serum TSH levels and the FF T4/T3 ratio. Kamron’s study advocates that thyroid function in subfertile women could indicate an insufficient capacity for basic reproductive functions such as oocyte quality, ovulation, fertilization and implantation [30]. Free TH supplementation (50 ng/ml of T3 and T4) led to significant increases in blastocyst formation and hatching rates in bovine embryos [31], and in the expansion rate of the blastocoel cavity of cryopreserved bovine embryos. Although an exact mechanism cannot be assumed, it is noteworthy that thyroid dysfunction may impact IVF outcomes, possibly at the level of oocyte quality.

No statistically significant difference was found between groups in the levels of thyroid autoantibodies in serum and FF. There was no statistically significant correlation between the number of retrieved oocytes or fertilized oocytes and the levels of thyroid autoantibodies in serum and FF. These findings support the theory that thyroid autoantibodies in FF do not affect oocyte number, oocyte maturation, orquality or preimplantation embryos during ART [32, 33]. However, other studies which only focus on serum TAI have found a lower fertilization rate in the thyroid autoimmunity (TAI)-positive group in the TAI-negative group(63 vs 72% [34] and 64.3 vs 74.6% [35]). One of the limitations of our study and other study included FF TAI to analysis is the relatively small number of TAI-positive patients. Our study included 165 (15 TAI positive) patients, while Medenica’s study included 52 (26 TAI positive) patients [32]. In addition, a previous study showed that thyroid antibodies cancross-react with zona pellucida and granulosa cells due to molecular mimicry [36].

Our results revealed significant negative correlations between serum E2 and fT4 in serum and FF. Not surprisingly, we also found a significant positive correlation between serum E2 and serum TSH. which could be explained by the COS-specific changes in oestrogen levels. A previous study showed that COS could lead to hyperestrogenism, through a rise in TBG, subsequently leading to a decrease in fT4 [27]. Our results support the theory that the adaptation of maternal THmetabolism to hyperestrogenism during pregnancy may be important for embryo survival [30, 31]. Similarly, COS-specific changesmight put stress on hypothalamus-pituitary-thyroid axis in a very short time and manifest as gestational thyroid disease, which subsequently creates a suboptimal environment during the early phase of implantation, which may increase the risk for IVF failure.

The use of a population of infertile women undergoing ART has some limitations. First, our outcomes were limited to cycle parameters and clinical pregnancy. Therefore, no conclusions can be drawn on any other obstetric or fetal end points. Second, whether these findings are generalizable to women without known fertility problems is still under debate. Additionally, a strong association between TH levels and pregnancy rate could not be found.

In conclusion, thyroid autoantibodies and mostTHs present in FF are not generated in the FF, but rather enter from the blood. Thyroid autoantibodies do not directly impact oocytes and embryos during ART, but on the OPU day,the T4/T3 ratios in both serum and FF were significantly higher in the group of women who became pregnant than in those who did not conceive. Our results highlight the importance of the adaptation of THmetabolism during COS.

Moreover, one point of practical importance regarding our study results is the confirmation of a strong correlation between serum and FF TH levels,which suggests that the complicated determinations of FF TH levels may not be necessary because serum TH levels provide the same information. Further studies with more samples should be performed to confirm our findings and to elucidate the relationships among ovarian stimulation protocols, oocyte quality, clinical pregnancy rate and longitudinal changes of THs levels during COS.