Recombinant luteinizing hormone supplementation in women undergoing in vitro fertilization/ intracytoplasmic sperm injection with gonadotropin releasing hormone antagonist protocol: a systematic review and meta-analysis

Compared to GnRH agonists which dominate in the area of assisted reproductive technology(ART) accounting for its essential role in circumventing the problem of a premature luteinizing hormone (LH) surge since the mid-1980s, GnRH antagonists generate a prompt suppression of gonadotrophin release, but do not cause the flare-up effect, by specifically blocking the GnRH receptors and ultimately induce a decrease in serum LH levels and a less pronounced decrease in FSH secretion [1]. However, it is unpredictable whether or not GnRH antagonists cause a decline in serum oestradiol during follicular recruitment which would result in an adverse effect on the pregnancy outcome [2]. On the other hand, GnRH antagonists tend to oversuppress endogenous LH if the dosage or timing of use was not appropriately controlled. As it is reported that endogenous low level of LH influences detrimentally both on the development of normal healthy follicles, because growing follicles become increasingly sensitive to and finally dependent on LH for their development [3], and on the endometrium after ovulation because sufficient LH is indispensable for the resumption of meiosis and for the production of progesterone. Therefore, it seems urgent for clinical doctors to add exogenous LH while GnRH antagonist protocol is applied to pituitary down-regulation in case of adverse effect on the pregnancy outcomes.

Nevertheless, there still has been no ultimate conclusion about the effect of r-LH supplementation to r-FSH in GnRH antagonist protocol on the pregnancy outcomes according to the recent studies. The issue on LH supplementation in women undergoing IVF/ICSI with GnRH antagonist for pituitary down-regulation has caused a heated debate around the world [4]. The studies by Sauer et al. (2004),Griesinger et al. (2005), Levi-Setti et al. (2006) did not demonstrate any beneficial effect of LH supplementation on the oocytes quality and the pregnancy outcomes [5‐7], while two randomized trials have shown higher pregnancy rates among those receiving rLH with GnRH agonist protocol [8, 9]. It is noticeable that the meta-analysis published in 2007 and 2010 separately showed no advantage in combination of r-LH with r-FSH in women undergoing IVF/ICSI with GnRH antagonist protocol compared with r-FSH alone group [10, 11]. Given this background, the issue in this area warrants further research [12].

A recent systemic review and meta-analysis concluded that the combination of r-hLH with r-FSH stimulation enhanced the clinical pregnancy and implantation rates in patients aged ≥35 [13]. Similar results were reported in an open-label randomized controlled study by Bosch et al. (2011) which found that r-LH is beneficial in improving the implantation rate in women aged 36–39 years [14], although König et al. (2013) argued that the pretreatment with hormonal contraceptives before stimulation and the LH supplementation on stimulation day 1, while in his randomized controlled trial LH supplementation was given on stimulation day 6, might play an essential role in the discrepancy between two studies [15]. Until now, there has been no meta-analysis to review whether the LH supplementation benefits the advanced reproductive aged patients undergoing IVF/ICSI with GnRH antagonist protocol.

GnRH antagonist protocol depends on the occurrence of spontaneous menses, which is different from long GnRH agonist protocol in which ovarian stimulation can be initiated after pituitary desensitization has been achieved [16, 17]. Therefore, pretreatment with oral contraceptive pill (OCP) before stimulation was applied in order to prevent ovarian cysts, for the sake of synchronous follicular development and predictingtiming events in an IVF/ ICSI cycle regarding scheduling [18]. In the studies by Sauer et al. (2004), Levi-Setti et al. (2006), Bosch et al. (2011), the patients were pretreated with OCP and used the GnRH antagonist protocol for COH, but no special benefits was shown in r-LF + r-FSH group compared with the r-FSH only group [6, 7, 14]. Consequently, it is necessary to explore whether combination of r-LH with r-FSH for COH benefits the pregnancy outcomes in women undergoing IVF or ICSI-ET with GnRH antagonist protocol and oral contraceptive pills pretreatment by meta-analysis.

Based on the above considerations, the present meta-analysis was performed to answer the questions: (1) whether combination of r-LH with r-FSH for COH benefits the pregnancy outcomes in general women undergoing IVF/ICSI with GnRH antagonist protocol;(2) whether combination of r-LH with r-FSH for COH benefits the pregnancy outcomes in advanced reproductive aged women undergoing IVF/ICSI with GnRH antagonist protocol; (3) whether combination of r-LH with r-FSH for COH benefits the pregnancy outcomes in women undergoing IVF/ICSI with GnRH antagonist protocol and pretreated with oral contraceptive pills.

Our systematic review and meta-analysis addressed the issue on the comparison of the outcomes between the combination of r-LH with r-FSH and r-FSH alone for COH in women undergoing IVF/ICSI with GnRH antagonist protocol and the comparisons in the subgroups of advanced reproductive aged women and women pretreated with oral contraceptive pills were also carried out.

Based on the “two-cell, two-gonadotropin” theory, the LH and FSH play a critical role in stimulating the two cellular components of ovary, which are theca cell and granulosa cell, leading to the production of ovarian steroids [19, 20]. At the earlier stage of follicular development, FSH is indispensable for follicular growth and the formation of estrogen by inducing the aromatase enzyme converting androgen to estradiol [21], while the androgen production from cholesterol is dependent on the stimulation of the theca cells by LH and FSH together [22]. Although FSH can induce follicular growth even without LH, there was identified that the follicles would have developmental deficiencies, following hCG administration [23], which suggested that the effect of LH on follicular development was probably not only due to providing androgen substrate for aromatization, but also exerting a direct effect on the stimulation and modulation of folliculogenesis [24]. It is noticeable that both the theca cell and granulosa cell produce significant amount of progesterone, which was converted into androgens under the influence of LH. Therefore, the LH supplementation resulted in the lower serum progesterone level. As is well established, increased exposure to progesterone can advance the endometrium, leading to asynchrony of embryo development to endometrial development and the reduction of implantation. Under this context, the LH supplementation may be beneficial for the serum oestradiol and progesterone level on the day of HCG administration.

As is predicted, our results suggested a beneficial effect of r-LH supplementation on ovarian stimulation in serum oestradiol and progesterone level on the day of HCG administration in general population. However, there was no evidence of beneficial effect in ongoing pregnancy per ET; clinical pregnancy per ET; incidence of OHSS; r-FSH total dose used per treatment cycle; total days of stimulation per treatment cycle; number of retrieved oocytes per oocyte retrieval; number of mature oocytes (metaphase II) per oocyte retrieval; fertilization rate; implantation rate, which was in accordance with the result of the meta-analysis by Monique H Mochtar et al. (2010) [3].

With regard to LH supplementation for the advanced reproductive aged women undergoing IVF or ICSI with GnRH antagonist protocol, different trials showed different results. The study by Bosch et al. (2011) obtained a significantly better implantation rate and a clinically better ongoing pregnancy rate among those patients aged 36 to 39. However, the study by König et al. 2013 showed no benefit of LH supplementation in controlled ovarian stimulation for IVF/ICSI with GnRH antagonists on pregnancy rates in patients of 35 years or older. Then we pooled the data from the two trials, showing a significantly higher serum oestradiol level and significantly lower r-FSH total dose used per treatment cycle found in the combination of r-LH with r-FSH compared to r-FSH alone for COH in advanced reproductive aged women undergoing IVF/ICSI with GnRH antagonist protocol. Although a significantly lower number of retrieved oocytes per oocyte retrieval was also found in the combination of r-LH with r-FSH group and the data of ongoing pregnancy or clinical pregnancy per ET were not available, we can not make the conclusion that LH supplementation was not beneficial for advanced reproductive aged women, since according to Bosch et al., the combination of r-LH with r-FSH group showed similar metaphase II oocytes and a better fertilization rate, thus suggesting that the oocytes obtained were of better quality, which would, in turn, lead to a higher implantation [14]. More trials and meta-analyses are needed to explore the role of LH supplementation played in advanced reproductive aged women.

Since oral contraceptive pills pretreatment is a convenient way for clinics to schedule oocyte retrievals, it will be more often applied by US clinics, although it is reported that oral contraceptive pills pretreatment diminishes the advantages of a GnRH antagonist protocol by extending the duration of treatment and the amount of FSH required to get to the same criteria of hCG [25], especially when stimulation is started immediately after OC withdrawal. This meta-analysis was not designed to detect a clinical relevant difference in ongoing pregnancy rate between pretreatment with or without oral contraceptive pills. Our results suggest a good effect of r-LH supplementation in ovarian stimulation in serum progesterone level on the day of HCG administration (WMD -0.19, 95% CI -0.25 to -0.13). Significantly lower serum progesterone level was observed in the combination of r-LH with r-FSH group compared to r-FSH alone group for COH in women undergoing IVF or ICSI with GnRH antagonist protocol and oral contraceptive pills pretreatment. This might be due to the fact that oral contraceptive pills pretreatment could have had an influence on the endocrine environment in the follicular phase by means of endogenous gonadotropin control [26] and FSH acts on granulosa cells to facilitate the conversion of cholesterol into P, which is transferred to the thecal cells to be converted into androgens under the action of LH, therefore LH supplementation lowered serum progesterone level [27]. Moreover, lower progesterone level, subsequently, increased endometrium receptivity. Under this context, LH supplementation may be the optimal option by increasing the beneficial effect of LH administration in this particular population. Our results also show a significantly higher total days of stimulation per treatment cycle (WMD 0.49, 95% CI 0.12 to 0.85) while significantly lower r-FSH total dose used per treatment cycle (WMD -211.90, 95% CI -319.99 to -103.82).

However, it has to be acknowledged that there still are some limitations in our meta-analysis. Firstly, the combined sample size of the five studies is still too small to confidentially detect a clinically relevant difference with regard to pregnancy likelihood between the two treatment modalities, especially when taking the subgroups into consideration, only the data from two or three trials were available for meta-analysis. Secondly, inclusion criteria of the selected trials were not as strict as possible, meaning that the inclusion criteria bias existed in the paper. For instance, when analyzing the outcomes of combination of r-LH with r-FSH comparing with r-FSH alone for COH in women undergoing IVF or ICSI with GnRH antagonist protocol in general population general population, all trials should be limited into those in which all the patients were pretreated with oral contraceptive pills or not, the initiation of LH supplementation was on the same stimulation day, and also ended on another same day, moreover, the initial dosage of FSH and LH was kept accordance with each trials. The last but not the least, the data of the trials was not all available for the meta-analysis, although the author was contacted if necessary, which resulted into that less trials were analyzed in the subgroup.

To conclude, the present meta-analysis found no statistically significant differences in outcomes of pregnancy between the combination of r-LH with r-FSH group and r-FSH alone group for COH with GnRH antagonist protocol in general population, in advanced reproductive aged women and in women pretreated with oral contraceptive pills undergoing IVF/ICSI. More studies are necessary for more solid conclusions on pregnancy likelihood after combination of r-LH with r-FSH for COH in GnRH antagonist protocol to be drawn.