Recombinant luteinizing hormone supplementation to recombinant follicle stimulating hormone therapy in gonadotropin releasing hormone analogue cycles: what is the evidence?

Abstract

Objective: To look into current evidence exploring the added value of rLH supplementation to rFSH in GnRH analogues cycles, to identify groups of women that still have no evidence for adjuvant rLH therapy and to discuss ways that may advance research on this topic.

Methods: Eight systematic reviews and meta-analyses exploring the benefit for pregnancy achievement of rLH supplementation, excluding other LH activity preparations, to GnRH analogues cycles in the ART setting were thoroughly evaluated.

Results: Evidence exists to show that rLH supplementation seems to have added value for pregnancy achievement in women with poor ovarian response and in women ≥35 years of age employing the GnRH agonist protocol, while the evidence is still debatable when the GnRH antagonist is administered. In the general infertile population, rLH supplementation does not have added value in the GnRH-antagonist cycles while the evidence is still controversial when the GnRH agonist is employed. Whether rLH supplementation may have a benefit in some young, normo-gonadotropic women, who may develop LH deficiency following GnRH analogues, is still to be shown. The main task remains how to identify subgroups of women that may benefit from rLH supplementation.

Conclusion: An accurate definition of the LH threshold in GnRH analogue treated cycles may contribute to the discussion of which subgroups of women may benefit from adjuvant rLH therapy.

Keywords:

• GnRH analogues
• rLH
• LH threshold
• assisted reproduction

Introduction

Optimal follicular maturation is the result of the combined action of both follicle stimulating hormone (FSH) and luteinizing hormone (LH) stimulation on follicles progressing from the pre-gonadotropin-dependent to the gonadotropin-dependent phase. The essential roles of both FSH and LH have been well documented in natural cycles as well as in women with hypogonadotrophic hypogonadism. In cases of hypothalamic or pituitary deficiency, LH supplementation to FSH was shown to be an absolute necessity for adequate ovarian folliculogenesis, steroidogenesis and clinical outcome.

However, the added value of recombinant luteinizing hormone (rLH) to recombinant follicle stimulating hormone (rFSH) treatment during controlled ovarian hyperstimulation (COH) in assisted reproductive technologies (ART), employing gonadotropin releasing hormone (GnRH) analogues, continues to be actively discussed and debated. The main arguments remain: whether the employment of GnRH analogues could cause profound LH suppression, which may adversely affect folliculogenesis, steroidogenesis and oocyte competence, impinging on clinical outcome, and whether rLH supplementation would obviate these effects.

In this regard, it should be emphasized that the mechanisms behind the pharmacological action of the two GnRH analogues, agonists and antagonists, differ substantially. When employing the long GnRH agonist, it initially stimulates the release of gonadotropins (flare-up) and complete pituitary desensitization is only achieved after 2–3 weeks of pre-treatment when pituitary desensitization occurs due to GnRH receptor down-regulation. Conversely, when applying the GnRH antagonist protocol, serum LH level starts to drop following FSH superovulation and before GnRH antagonist initiation. This is attributed to the superovulation induction process itself via supra-physiological concentrations of estradiol. However, pituitary shut-down is achieved following GnRH antagonist initiation by an immediate GnRH receptor blockage^1,2.

It is interesting to note that early follicular LH priming following GnRH agonist down-regulation, but before FSH initiation, has been suggested to increase granulosa cell FSH sensitivity by increasing androgen synthesis during the early stage of folliculogenesis^3,4. This concept and its impact on clinical outcome are still under investigation and are outside the scope of the present study.

Why the delay?

Since the introduction of GnRH analogues, many studies have been published aiming to resolve this debate, yet the controversy continues. Many factors may have contributed to this delay and may have caused confusion among practitioners. Table 1 summarizes the confounding factors that have contributed to the complexity and confusion underlying this topic. These factors relate to the populations studied, LH activity and GnRH analogue preparations that were employed, as well as other intricate issues that may be difficult to control for in clinical randomized controlled trials (RCTs) or when comparing among different RCTs.

Table 1. Confounding factors that have contributed to the complexity and confusion surrounding the added value of rLH adjuvant therapy in GnRH analogue cycles.

Population targeted	LH activity preparation employed	GnRH analogue employed	Other factors that may have an impact on RCT results
General infertile population	rLH	GnRH agonist (long/short)	Only 1% of LH receptors should be occupied to stimulate steroidogenesis^41
Young/advanced reproductive age	hMG	GnRH antagonist	Immuno-reactive LH is not necessarily representative of LH bioactivity^42
Normal/high/low ovarian response	hCG	Which GnRH agonist and way of administration^40	LH and LH receptor (LHCGR) polymorphism^38^,^39
PCOS women			LH receptors have been identified in the human endometrium^43

LH: luteinizing hormone; GnRH: gonadotropin releasing hormone; rLH: recombinant luteinizing hormone; RCT: randomized controlled trial; hMG: human menopausal gonadotropin; hCG: human chorionic gonadotropin; LHCGR: luteinizing hormone choriogonadotropin receptor; PCOS: polycystic ovary syndrome.

What is the accumulated evidence for rLH adjuvant therapy?

Eight meta-analyses have been published in recent years examining specifically the topic of rLH supplementation to rFSH treatment employing GnRH analogues in the ART setting^5–12. These meta-analyses are summarized in Table 2 and they included four to 40 RCTs and 434 to 6443 pooled cases. All these studies have deliberately excluded other LH activity preparations, in order to control for rLH added effect. These reports focused on clinical endpoints including rFSH dosage requirement, serum E₂ level on the day of human chorionic gonadotropin (hCG) administration, number of oocytes, clinical pregnancy and live birth rates. While some of these reports demonstrated a clinical pregnancy rate advantage for rLH supplementation^9,10,12, others did not^5–8,11,12 contributing to the confusion among clinicians regarding the added value of rLH adjuvant therapy in ART.

Table 2. Summary of meta-analyses evaluating the advantage of rLH supplementation to rFSH in IVF/ICSI-ET cycles employing GnRH analogue protocols*.

Meta-analysis	Number of RCTs	Number of cases	Population	Number of RCTs employing GnRH-		Results
				Agonist	Antagonist	FSH dosage	Number oocytes	Max E2	Clinical pregnancy	Live birth
Oliveira et al.^5	4	1227	General	4	0	Y	Y	Y	N
Baruffi et al.^6	5	434	General	0	5	N	N	Y	N
Kolibianakis et al.^7	7	701	General	5	2	N	N	N		N
Mochtar et al.^8	14	2612	General Pooled low responders	11	3	N	N	N	NY	N
Hill et al.^9	7	902	≥35 years	6	1				Y
Lehert et al.^10	40	6443	General Pooled low responders	31	9		NY		YY	YN
Xiong et al.^11	5	1078	General Pooled ≥35 years	0	5	NY	NY	Y	N
Santi et al.^12	32	5840	General	22	10	Y	N		Y	N

* N, no advantage for rLH supplementation to rFSH treatment; Y, positive advantage for rLH supplementation to rFSH treatment; Blank, not addressed or insufficient data to reach conclusion.

rLH: recombinant luteinizing hormone; FSH: follicle stimulating hormone; IVF/ICSI-ET: in-vitro fertilzation / intracytoplasmic sperm injection - embryo transfer; GnRH: gonadotropin releasing hormone; RCT: randomized controlled trials; rFSH: recombinant follicle stimulating hormone.

The GnRH agonist protocol

Most studies included in the published meta-analyses analyzed RCTs that have employed the GnRH agonist, mainly the long protocol. The majority of these meta-analyses addressed the general infertile population1^7,8,10,12, while one targeted women ≥35 years of age⁹ and two performed subgroup analyses for women with poor ovarian response^8,10. The first three meta-analyses targeting adjuvant rLH to the GnRH agonist in the general infertile population did not find an added value^5,7,8 while the last two have found an advantage for pregnancy achievement^10,12, contributing to the confusion in the literature.

In women with poor ovarian response or ≥35 years of age, three meta-analyses agreed that rLH addition contributed favorably to pregnancy achievement^8–10. This suggests reduced LH bioactivity among advanced reproductive age women undergoing long GnRH agonist and rFSH mono-stimulation, which could be overcome by rLH adjuvant therapy.

None of the meta-analyses included a separate assessment of young, normo-gonadotropic women that responded sub-optimally to rFSH mono-therapy following long GnRH agonist down-regulation. Clinical evidence has shown that some women with apparently normal ovarian reserve may not respond optimally to rFSH and may demonstrate LH deficiency subsequent to GnRH agonist administration (hypo-responder women). It is estimated today that these women may represent 10–14% of young normo-gonadotropic women undergoing ART treatment employing the GnRH agonist protocol^13–15.

The GnRH antagonist protocol

Three meta-analyses targeted RCTs employing GnRH antagonist in the general infertile population, showing a positive effect on serum E₂ level (on the day of hCG administration) but not on clinical pregnancy rate^6,11,12. One meta-analysis performed a subgroup analysis for women ≥35 years of age but offered no definite conclusion regarding clinical pregnancy rate¹¹.

Only a few prospective RCTs have examined the impact of rLH supplementation to rFSH employing the GnRH antagonist protocol in women ≥35 years of age or those having poor ovarian response, with conflicting results^16–19. While one study found a favorable clinical pregnancy rate¹⁶ the other three did not find added value for the addition of rLH in this setting^17–19, contributing further to the controversy.

Similar to the long GnRH antagonist protocol, the added value of rLH adjuvant therapy in the subgroup of young, normo-gonadotropic women, which may show signs of LH deficiency following GnRH antagonist stimulation, has not been evaluated^15,20. On the theca-granulosa cellular level, the immediate fall in LH level following GnRH antagonist initiation may have an undesirable effect on follicular steroidogenesis and maturation, which may in turn adversely affect clinical results in these women.

Taken together, the evidence available to date suggests that rLH supplementation to rFSH seems to have a benefit, specifically on clinical pregnancy rate, when the long GnRH agonist protocol is employed in women with poor ovarian response and in women ≥35 years of age. In the general infertile population employing the GnRH agonist protocol, available evidence is still controversial. Conversely, a supplementary value for rLH adjuvant therapy in GnRH antagonist treated cycles in women ≥35 years of age or with low ovarian response has yet to be shown. Current evidence does not demonstrate a benefit for rLH supplementation for pregnancy rates in the general infertile population employing the GnRH antagonist protocol.

Additionally, whether adjuvant rLH therapy may be effective in some young, normo-gonadotropic women who may develop LH deficiency following GnRH agonist or antagonist treatment is still to be shown. The main task remains, therefore, to identify this subgroup of young women before GnRH analogue administration.

Methodology of LH threshold assessment

The residual LH level below which normal ovarian folliculogenesis and steroidogenesis and consequent clinical outcome might be impaired has been termed the LH “threshold” level^21,22. Previous investigators have employed single serum LH level, typically a single “snapshot” event, to investigate LH threshold, to examine its value and relation to clinical outcome.

Single sporadic determination

The long GnRH agonist protocol

Most reports employing the long GnRH agonist protocol have addressed mid-follicular LH threshold in normo-gonadotropic young infertile women, with conflicting results^23–28. Studies that defined an LH minimal threshold affecting clinical outcome and justifying LH supplementation to augment pregnancy rates did not agree on the residual LH level below which treatment outcome might be impaired. These studies indicated that the lower limit of the LH threshold ranged between 0.5 and 1.2 mIU/mL^23,24,26,28.

The GnRH antagonist protocol

A few studies have addressed the same question in normo-gonadotropic infertile women undergoing the GnRH antagonist protocol. Serum LH determination was explored 1–3 days following antagonist initiation as well as on the day of hCG administration, negating an association between serum LH level and clinical pregnancy attainment^29–31.

In addition, several studies investigating the GnRH antagonist protocol considered single serum LH levels on day 8^30–32 or on the day of hCG administration^31,32 employing Tukey’s hinges. In this methodology, LH threshold was stratified by quartiles: the lower, median and upper quartiles were compared. This was also performed in normo-gonadotropic young women; no significant difference in pregnancy rates was found among the LH quartiles.

Other methodologies

Single sample LH determination as an indicator for LH threshold has been challenged in studies employing both the GnRH agonist and antagonist protocols^14,28,33,34. In this context, individual area under the curve (AUC) of LH dynamics during COH, following GnRH antagonist initiation, has been suggested as a more appropriate way to evaluate LH threshold than single sample LH evaluation^33,34. The abrupt changes in serum LH level that may be encountered following GnRH antagonist initiation, the large differences in serum LH recovery following its employment and the wide variation in the number of days for its administration among patients may support this notion.

It has recently been suggested that it is not absolute LH serum levels, but rather the change in LH levels following GnRH analogue administration, that impacts outcomes. This has been shown in young, non-selected normo-gonadotropic women undergoing the long GnRH agonist protocol^14,28 as well as the GnRH antagonist protocol³³.

As an illustration of the LH level changes observed when employing the long GnRH agonist protocol, serum LH ratio of ≤0.5 during the mid to early follicular stage was suggested as an indicator of LH deficiency, causing a reduced live birth rate²⁸. A dynamic decrease of >50% in serum LH level was shown to have a detrimental effect on in-vitro fertilization - embryo transfer outcome. However, this concept should be further explored in prospective randomized trials.

The dynamic change in serum LH level may be more distinct in the GnRH antagonist cycles. In some women, the decline in LH level may be sharp and abrupt, if the theca-granulosa two-cell apparatus has insufficient time to adapt to this change^35,36. Accordingly, in a prospective single center study, pregnancy achievement was shown to correlate with low LH level change following GnRH-antagonist administration in young normo-gonadotropic women, while pregnancy rates significantly decreased when the changes in LH levels were too large³³. In such cases, serum LH level evaluation and comparison between post and pre GnRH antagonist initiation may dictate whether to supplement with rLH therapy. Based on considerations of the pharmacodynamics and pharmacokinetics of GnRH antagonist, a cut-off level of >50% decrease in LH level has been suggested as an indicator to supplement with rLH³⁷, but this should be further examined in prospective controlled trials.

Taken together, it seems that there is some discrepancy in LH threshold assessment between the long GnRH agonist and the GnRH antagonist protocols. While some studies support the use of mid-follicular serum LH level as an indicator for LH threshold, inconsistent results across studies have left the precise level in debate. However, all current data concurs that single serum LH evaluation following GnRH antagonist administration is not an adequate tool. The concept of LH dynamic changes following GnRH analogue employment as an indicator for LH threshold should be pursued, particularly among young normo-gonadotropic women. Furthermore, individual LH AUC level following GnRH antagonist initiation seems to be a more appropriate target for LH threshold evaluation than an isolated assessment, and should also be further explored.

Regardless of the methodology found suitable to define LH threshold, it must be remembered that women’s responses to similar serum LH levels, or comparable dynamic changes in their LH level following GnRH analogue administration, may differ. Individual variation in ovarian response has been attributed to different LH isoforms or to luteinizing hormone choriogonadotropic receptor (LHCGR) and follicle stimulating hormone receptor (FSHR) polymorphisms^38,39. Among the four distinctive LH isoforms identified, v-βLH was shown to be the most prevalent, with elevated bio-activity but short serum half-life³⁸. In addition, among many LHCGR and FSHR polymorphisms, LHCGR N312S and FSHR N680S seem to be the most prevalent and best studied³⁹. These LH, LHCGR and FSHR polymorphisms seem to be appropriate candidates for further exploration, to explain individual variation in ovarian response and the need for rLH adjuvant therapy before COH initiation. This may be especially true in young hypo-responder women following GnRH analogue administration.

All told, various methodologies have been applied to explore LH threshold following GnRH analogue treatment in ART with inconsistent results. Whereas the majority of studies have focused on various strategies of serum LH level evaluation, a few have explored LH, FSHR and LHCGR polymorphisms and their relationship with ovarian response. While the first approach directly targets the effect of GnRH analogue on serum LH dynamics, the second aims at individual LH in-vivo biological activity. The time may have come to find a way to combine assessment of both of these measures. In so doing, an accurate definition of LH threshold for adequate ovarian folliculogenesis and steroidogenesis in GnRH analogue treated cycles may be reached. An accurate definition of LH threshold may contribute to the discussion of which groups of women may benefit from adjuvant rLH therapy. Ascertaining a novel biomarker of LH deficiency prior to GnRH analogue initiation may provide an elegant route to individualized and more successful treatment.

Conclusions

Following the introduction of recombinant gonadotropin technology into routine clinical practice in ART, the way was paved to explore the added value of LH to FSH in GnRH analogue treated cycles, specifically their impact on pregnancy achievement. Many various factors have contributed to the complexity and confusion surrounding this topic. These factors are related to the populations studied, LH activity and GnRH analogue preparations employed, as well as to other intricate issues that may be difficult to control for in RCTs. Accumulating recent evidence supports the notion of an added value for adjuvant rLH therapy in women of advanced reproductive age undergoing the long GnRH agonist protocol, but is still debated in the general infertile population. Conversely, the added value of rLH therapy in GnRH antagonist cycles has yet to be demonstrated in the advanced reproductive age group while it seems to have no supplementary role in the general infertile population. A non-trivial portion of young, normo-gonadotropic women receiving the long GnRH agonist or the antagonist protocol, that may show signs of LH deficiency, may benefit from rLH adjuvant therapy. This is still to be demonstrated in prospective RCTs. An accurate definition of LH threshold in GnRH analogue treated cycles may contribute to the discussion of which subgroups of women may benefit from adjuvant rLH therapy.

Transparency

Declaration of funding

This manuscript was not funded.

Author contributions: J.S.Y. conceived the idea of this commentary, contributed to study design and construction of the study, executed the study, performed the analyses and interpretation of the data and drafted the manuscript. N.L. contributed to study design and execution, contributed to the analyses and the interpretation of the data and revised the manuscript for important intellectual content. Both authors approve the final version of the manuscript and accept responsibility for the paper.

Declaration of financial/other relationships

J.S.Y. and N.L. have disclosed that they have no significant relationships with or financial interests in any commercial companies related to this study or article.

CMRO peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Acknowledgment

The authors thank Ms. Sarah M. Cohen, MPH, for her editorial assistance and guidance.