Background
In women, anti-Müllerian hormone (AMH) is synthesized by granulosa cells in preantral and early antral stage (5–8 mm diameter) follicles [
1,
2]. AMH regulates follicular development by inhibiting the initial recruitment from the primordial pool [
3,
4]. Since AMH produced in the ovary is secreted into the circulation, it can be measured in the serum. Some studies reported that the serum AMH level remained stable throughout the menstrual cycle [
5‐
8]. In addition, the serum AMH level strongly correlated with the resting pool of ovarian follicles; therefore, AMH measurement may be used to predict the reduction in the number of growing follicles and subsequent menopause [
9‐
11]. These AMH physiological characteristics can be applied to assisted reproductive technologies.
Previous studies reported that serum AMH level was an effective predictor of oocyte retrieval among women with infertility undergoing controlled ovarian stimulation (COS) for in vitro fertilization (IVF): that is, the serum AMH level positively correlated with the number of retrieved oocytes after the administration of exogenous gonadotropins [
12‐
16]. Furthermore, in COS cycles, women with higher serum AMH levels have higher ongoing pregnancy rates as well as live birth rates [
17,
18]. On the contrary, in the case of natural conception without any exogenous gonadotropins, a low serum AMH level is not associated with a reduced fecundability rate or a lower cumulative probability of conceiving, suggesting that the serum AMH level cannot be used as a predictor of pregnancy outcomes in spontaneous natural cycles without gonadotropin [
19].
Clomiphene citrate (CC) is often administered alone in the minimal ovarian stimulation protocol for IVF [
20‐
23]. CC binds hypothalamic estrogen receptors and induces the secretion of gonadotropin-releasing hormone (GnRH) by altering the negative feedback effect of estrogen on the hypothalamus, resulting in the promotion of endogenous gonadotropin secretion. Minimal ovarian stimulation protocols can decrease the dose and duration of exogenous gonadotropin administration, thereby reducing the influence on the ovaries and reducing patient distress and complications [
24‐
26]. In addition, IVF outcomes in minimal ovarian stimulation protocols are considered similar to those of COS cycles, especially in patients who are poor responders [
20,
27‐
29]. Although the serum AMH level is predictive of ovarian response and IVF outcomes after gonadotropin administration, the predictive ability of the serum AMH level in clomiphene-based minimal stimulation cycles is still unknown. Therefore, to reveal whether the serum AMH level can be used as a predictor for IVF success in minimal stimulation cycles, we aimed to retrospectively analyze the correlation of the serum AMH level with ovarian responsiveness, embryonic outcomes, and cumulative live birth rates in CC-based minimal ovarian stimulation.
Discussion
To the best of our knowledge, this is the first study demonstrating a correlation between the serum AMH level and the rates of blastocyst formation, blastocyst cryopreservation, and live birth per oocyte retrieval in CC-based minimal ovarian stimulation cycles without any exogenous gonadotropin administration. Furthermore, the present investigation showed the cumulative live birth rate and treatment period in CC-based minimal ovarian stimulation cycles to be strongly associated with the serum AMH level at fertility treatment initiation.
Some studies have investigated the correlation between serum AMH levels and IVF outcomes in COS cycles and reported that the serum AMH level can be predictive of the oocyte quality, embryo morphology, and subsequent pregnancy outcomes after exogenous gonadotropin administration [
14,
40,
41]. However, studies regarding the correlation of the serum AMH level with IVF outcomes in minimal ovarian stimulation cycles remain limited. We first assessed the association between the serum AMH level and IVF outcomes in CC-based minimal ovarian stimulation cycles. As mentioned in the Introduction, serum AMH level was not associated with a cumulative probability of conceiving in cases of natural conception [
18]. Normally, only one follicle is induced to develop into a Graafian follicle, and a COC is ovulated into the fallopian tube in natural conception; therefore, pregnancy outcomes in natural conception would be affected by the rest of the follicular pool as well as the serum AMH level in young women.
According to a previous study, we expected that the oocyte retrieval outcomes may not be associated with the serum AMH level in women who underwent CC-based minimal ovarian stimulation IVF as the CC-only protocol slightly promotes follicular development, and in most cases, only 1–3 COCs are ovulated. However, our data showed that the serum AMH level positively correlated with the number of retrieved oocytes even without exogenous gonadotropin administration. This AMH feature was similar to that in COS IVF cycle [
12‐
16]. Furthermore, our results demonstrated that the decline in the serum AMH level correlated with the decrease in blastocyst formation and cryopreservation rates, but not the fertilization and cleavage rates. In line with this finding, in women with low serum AMH levels (< 1.42 ng/mL), the live birth rate after cleaved embryo transfers was significantly lower than that in women with high serum AMH levels (≥ 1.42 ng/mL), suggesting that decreased serum AMH levels are also correlated with the impaired embryo developmental competence of the oocyte and cleaved embryos to the blastocyst stage, as in COS cycles [
14,
42‐
44].
In this study, the correlation of serum AMH levels with the live birth rates in the first cycle was also analyzed. A previous study reported that AMH is a better clinical predictor of cycle success in COS IVF cycles, although FSH, antral follicle count, and AMH are widely used to assess the ovarian reserve in women undergoing infertility evaluation [
45]. Contrarily, another study demonstrated that antral follicle count was more predictive of clinical pregnancy than the serum AMH level [
46]; therefore, in COS cycles, factors that best predict ovarian reserve and live birth are controversial. In CC-based minimal stimulation cycles, AMH appears to be superior to FSH and antral follicle count for the prediction of successful live birth in the first cycle. In this study, a multivariate logistic regression analysis demonstrated that the patient’s age and serum AMH level were significantly associated with the live birth rate in the first retrieval cycles, and that the FSH level and antral follicle count on day 3 were not (Table
3). Furthermore, we stratified the pregnancy outcomes by the patient’s age groups according to the Society for Assisted Reproductive Technology (SART) classification (age: young, ≤ 37; middle, 38–40; and advanced, ≥ 41 years) and the percentile of serum AMH levels (Supplemental Table
8). When the patient’s age was limited, pregnancy outcome was significantly correlated with the serum AMH level in all age groups. When the serum AMH level was limited, the patient’s aging resulted in poor pregnancy outcomes. These results confirm the previous reports that examined the association among the pregnancy outcomes, patient’s age, and serum AMH level in COS IVF cycles [
47,
48]. Our results also suggest that the clinical outcomes in CC-based minimal stimulation cycles can be predicted using the patient’s age and serum AMH level.
Furthermore, we investigated the association between the serum AMH level and cumulative live birth rate. Women with a high serum AMH level had a higher cumulative live birth rate and shorter treatment period than women with a low serum AMH level. These results were similar to those observed in COS IVF cycles [
49,
50]. However, some previous studies showed that basal AMH has modest predictive performance for the occurrence of cumulative live birth rate and may not give additional value to the patient age [
51,
52]; therefore, studies on the correlation between serum AMH and cumulative live birth rate have inconsistent findings even in COS cycles. To validate our findings, the correlation of the serum AMH level with cumulative live birth rate in CC-based minimal stimulation should be further evaluated.
The strength of this study is the assessment of a relatively rare cohort in the current IVF practice. Specifically, given that very few clinics use a CC-based minimal ovarian stimulation protocol, our findings provide useful insights for the improvement of clinical outcomes in CC-based minimal stimulation IVF. However, considering the single-center, retrospective design of our study, the possibility of selection biases cannot be excluded and therefore our results are limited. Further prospective, large-scaled studies are warranted to investigate the prediction value of AMH levels in CC-based minimal ovarian stimulation.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.