**2. Potential mechanisms of exogenous LH benefit in ART**

54 Enhancing Success of Assisted Reproduction

Cholesterol

Progesterone

Cholesterol StAR mediated mitochondrial transport

CYP17

mediated by LH are indicated in red.

Pregnenalone

Androgens

CYP11a

Mitochondrial membrane

CYP17

human LH was later produced (18).

(9, 10). The specific importance of LH activity can be demonstrated in patients with LHβ or LH receptor gene mutations. Case reports of these male and female patients have demonstrated hypogonadism, infertility, pseudohermaphroditism, and amenorrhea (11-13).

**Luteinizing Hormone** 

**Thecal Cells Ovarian Follicle Luteinized Granulosa Cells** 

Metaphase I oocyte Metaphase II oocyte

Loss of gap junctions Uncoupling of granulosa cells from the oocyte Decreased granulosa cell proliferation genes

Resumption of meiosis

**Figure 1.** Key actions of LH within the ovary on the thecal cells, oocyte, and granulosa cells. Actions

In assisted reproduction technologies (ART), the importance of LH is demonstrated clearly in hypogonadotropic hypogonadic patients. Patients with a profound lack of endogenous LH fail to undergo complete follicular maturation in the absence of exogenous LH (14, 15). Such patients require the exogenous administration of both LH and FSH to optimize reproductive outcomes (4, 16, 17). Urinary human menopausal gonadotropins were initially utilized in assisted reproductive technologies. These preparations were islolated and purified from large pools of human urine. One of the early urinary hMG products was Pergonal 75. One ampule of Pergonal 75 contained 75 international units (IU) of FSH and 75 IU of LH, which became an industry standard for ampules (18). These urinary hMG preparations contained both FSH and LH, as well as some hCG, and therefore patients were stimulated with both gonadotropins. Later advancements in monoclonal antibody technology enabled the production of urinary purified FSH and a more purified hMG, which is still used today (19, 20). Recombinant DNA technology using a mammalian cell culture system (Chinese hamster ovary cells) was used to produce recombinant human FSH, which was first licensed in 1995, and quickly replaced urinary FSH products. Recombinant

Ovulation

cAMP

PKA

Epiregulin and Amphiregulin

Increased cell survival

Despite the clear biologic importance of LH outlined in the preceding paragraphs, numerous studies have demonstrated successful ART outcomes with the use of exogenous FSH only (21, 22). A likely explanation is that LH is a very potent hormone, activating the LH receptor for adequate ovarian steroidogenesis when only 1% of LH receptors are bound (23). Even after GnRH agonist or antagonist down-regulation, a majority of patients will There are theoretical benefits of the use of exogenous LH for the oocyte and the endometrium. The putative purpose of controlled ovarian stimulation in ART is to maximize the number of oocytes retrieved. However, the evidence is clear that the addition of LH is not associated with an increase in the number of oocytes or the number of mature metaphase II oocytes (MII) retrieved. Indeed, the use of hMG has been shown to decrease the number of follicles, oocytes, and metaphase II oocytes (MII) as compared to rFSH alone (21, 22, 24-28), presumably due to the action of LH contained in hMG. This is confirmed by similar data comparing rLH plus rFSH versus rFSH alone which has shown a decrease in developing follicles and oocytes retrieved with rLH (29, 30). In the majority of these trials, the decrease was in oocytes from small to intermediate follicles, and the number of oocytes retrieved from large follicles and the number of MIIs retrieved were not different. This suggests the possibility that the use of exogenous LH activity is associated with a decreased in the development of small follicles which may have been unlikely to yield a fertilized 2PN. There appears to be no negative effect on the development of larger follicles.

In a series of *in vivo* studies evaluating the effect of LH activity on follicle growth, Filicori and colleges confirmed the findings that LH activity can decrease the growth of small follicles without impacting the continued growth and maturity of larger follicles. First, they demonstrated that the number of follicles under 10mm in size during ART stimulation positively correlated with FSH dose (r=0.193, p<0.05) but negatively correlated to LH dose (r=0.648, p<0.0001) (31). In another study, it was demonstrated that incrementally decreasing the dose of FSH from day 7 of stimulation and increasing the dose of LH resulted in a decrease in the number of follicles <10mm in size, without affecting follicles over 14mm in size (32). To evaluate if this effect was due to the decreasing FSH dose or the increasing LH dose, they performed a similar experiment where FSH was held steady at 150IU per day and patients were placed into groups of incrementally increasing LH doses. In this experiment, increasing doses of LH (in the presence of a constant dose of FSH) was again associated with a decrease in number of small follicles while not affecting the larger follicles (33). When the experiments were repeated utilizing hMG, hMG was also associated with a decrease in small follicles (34). These experiments and the results of many randomized controlled trials demonstrate that any beneficial effect of LH activity is not the result of an increase in oocyte yield.

While the number of total oocytes, especially from small follicles, appears to be diminished in ART cycles utilizing LH, the quality of those oocytes may be increased. While direct

measures of oocyte quality are difficult to assess clinically, some studies have noted an increased fertilization rate in oocytes obtained from cycles stimulated with LH (24, 30). Numerous trials have also demonstrated that the addition of LH activity results in an increase in serum estradiol on the day of hCG (Figure 2), which may represent a higher quality cohort of developing follicles (22, 26, 30, 35-43). LH supplementation was demonstrated to result in lower levels of apoptosis in cumulous cells as compared to FSH stimulation only (44). Cumulous cell apoptosis has been used a marker of oocyte quality and the decrease in apoptosis with the addition of LH is consistent with its post-receptor effects through increased epiregulin and amphiregulin.

The Use of rLH, HMG and hCG in Controlled Ovarian Stimulation for Assisted Reproductive Technologies 57

progesterone in the donor is not associated with decreased implantation in the recipient (52). Progesterone is necessary for endometrial development and embryo implantation. However, premature rises in progesterone can advance the development of the endometrium and lead to asynchrony with the embryo development (46, 51, 53). FSH drives the conversion of cholesterol to progesterone but lacks CYP17 to further convert progesterone to androgens (54, 55). LH stimulates CYP17 in thecal cells to further convert the progesterone to androgens, which are subsequently aromatized in the granulosa cell (3). Under the two-cell two gonadotropin model, LH is protective of premature progesterone elevations prior to luteinization (24, 45) (Figure 3). Further investigation is needed to

determine if exogenous LH administration is protective for the endometrium.

**Figure 3.** Model demonstrating a possible mechanism by which the administration of exogenous LH decreases premature rises in serum progesterone. FSH stimulates granulosa cells to convert cholesterol to progesterone. Lacking CYP17, the granulosa cells cannot convert progesterone to androgens and thus progesterone is secreted from the cells. In the absence of adequate LH levels, this progesterone is secreted into the circulation where it can advance the endometrium prematurely. In the presence of adequate LH levels, the progesterone is converted into androgens by CYP17 in the thecal cells. The androgens are then taken up by the granulosa cells and converted to estrogens. In this model exogenous LH protects the endometrium from exposure to premature progesterone rises. Green arrows represent

increased action. Red arrows represent decreased action.

**Figure 2.** Randomized controlled trials demonstrating an increased estradiol level on the day of hCG with rLH (top) or hMG (bottom) as compared to rFSH alone (adapted from Hill *et al.*, 2012 (21)).

Another possible effect of LH is on the endometrium and embryo implantation. LH receptors are present in the endometrium during the window of implantation (9, 10), but whether these receptors play a direct role in embryo implantation needs further investigation. An indirect effect on the endometrium has been proposed via decreased premature progesterone secretion (24, 45). There is a growing body of evidence to suggest that prematurely elevated progesterone levels on the day of hCG have a negative impact on embryo implantation without affecting embryo quality (46-51). The evidence that this is an endometrial effect is supported by studies in oocyte donor cycles, where elevation of progesterone in the donor is not associated with decreased implantation in the recipient (52). Progesterone is necessary for endometrial development and embryo implantation. However, premature rises in progesterone can advance the development of the endometrium and lead to asynchrony with the embryo development (46, 51, 53). FSH drives the conversion of cholesterol to progesterone but lacks CYP17 to further convert progesterone to androgens (54, 55). LH stimulates CYP17 in thecal cells to further convert the progesterone to androgens, which are subsequently aromatized in the granulosa cell (3). Under the two-cell two gonadotropin model, LH is protective of premature progesterone elevations prior to luteinization (24, 45) (Figure 3). Further investigation is needed to determine if exogenous LH administration is protective for the endometrium.

56 Enhancing Success of Assisted Reproduction

through increased epiregulin and amphiregulin.

measures of oocyte quality are difficult to assess clinically, some studies have noted an increased fertilization rate in oocytes obtained from cycles stimulated with LH (24, 30). Numerous trials have also demonstrated that the addition of LH activity results in an increase in serum estradiol on the day of hCG (Figure 2), which may represent a higher quality cohort of developing follicles (22, 26, 30, 35-43). LH supplementation was demonstrated to result in lower levels of apoptosis in cumulous cells as compared to FSH stimulation only (44). Cumulous cell apoptosis has been used a marker of oocyte quality and the decrease in apoptosis with the addition of LH is consistent with its post-receptor effects

**Figure 2.** Randomized controlled trials demonstrating an increased estradiol level on the day of hCG with rLH (top) or hMG (bottom) as compared to rFSH alone (adapted from Hill *et al.*, 2012 (21)).

Another possible effect of LH is on the endometrium and embryo implantation. LH receptors are present in the endometrium during the window of implantation (9, 10), but whether these receptors play a direct role in embryo implantation needs further investigation. An indirect effect on the endometrium has been proposed via decreased premature progesterone secretion (24, 45). There is a growing body of evidence to suggest that prematurely elevated progesterone levels on the day of hCG have a negative impact on embryo implantation without affecting embryo quality (46-51). The evidence that this is an endometrial effect is supported by studies in oocyte donor cycles, where elevation of

**Figure 3.** Model demonstrating a possible mechanism by which the administration of exogenous LH decreases premature rises in serum progesterone. FSH stimulates granulosa cells to convert cholesterol to progesterone. Lacking CYP17, the granulosa cells cannot convert progesterone to androgens and thus progesterone is secreted from the cells. In the absence of adequate LH levels, this progesterone is secreted into the circulation where it can advance the endometrium prematurely. In the presence of adequate LH levels, the progesterone is converted into androgens by CYP17 in the thecal cells. The androgens are then taken up by the granulosa cells and converted to estrogens. In this model exogenous LH protects the endometrium from exposure to premature progesterone rises. Green arrows represent increased action. Red arrows represent decreased action.

There is evidence to suggest that suppressed LH levels in women during ART stimulation can have negative effects (Figure 4). Depending on the study, adverse outcomes have been demonstrated with LH below 0.5-1.2 IU/L. LH levels < 1.2 IU/L have been reported to be associated with decreased serum estradiol, poor follicular development, decreased oocyte yield, decreased high quality embryos, and lower pregnancy rates (14, 15). Below LH levels of 1IU/L, other researchers demonstrated slower follicular growth and decreased estradiol (56). Finally, LH levels < 0.5 IU/L have been associated with increased pregnancy loss, lower implantation rates, and lower live birth rates (57, 58).

The Use of rLH, HMG and hCG in Controlled Ovarian Stimulation for Assisted Reproductive Technologies 59

with GnRH analogues, most patients are not in danger of having elevated endogenous LH levels. Indeed, by day 6 of GnRH antagonist administration, endogenous LH levels are depressed to a mean level of 1.6 IU/L, a value much closer to the LH threshold than the ceiling (65). Similarly, long agonist protocols also suppress endogenous LH levels to a mean near 1 IU/L (66). The evidence would suggest the clinician should be more concerned with replacing an adequate LH level in patients under pituitary down-regulation and the threat

1. LH activity causes atresia of small follicles during ART stimulation 2. Indirect evidence suggests increased oocyte quality with LH

3. LH activity decreases oocyte yield due to a loss of these small follicle

5. LH activity may protect the patient from premature progesterone elevations

hMG is a urinary gonadotropin preparation consisting equal activity of both LH and FSH and some hCG. It is available in highly purified forms, minimizing earlier preparation disadvantages of protein contamination leading to the risk of allergic reactions. Evaluating studies with hMG have the advantage of homogeneity. Due to the nature of hMG containing equal FSH and LH activity, all patients in the hMG group receive equal amounts of LH and FSH activity and start LH activity on the same day as the FSH activity is started. As hMG has been available for longer than rFSH, there are more studies and total data available for

When looking at intermediate outcomes and surrogate markers for ART, hMG has not been demonstrably different than rFSH for ovarian stimulation. The results are similar in the proportion of MII oocytes, the number of high quality embryos, zona pellucida morphology, and polar body evaluation (42, 67-69). Studies have also shown no benefit in the number of oocytes retrieved with hMG and indeed numerous studies have shown a small decrease in the number of oocytes retrieved (typically around 1 oocyte less per retrieval) (22, 25-28, 41, 67). In the majority of studies, the decrease in oocytes did not translate into a decrease in the number of MIIs retrieved per cycle, indicating the loss was in smaller, immature oocytes. hMG administration has been associated with higher serum and follicular fluids androgens and estrogens and lower serum progesterone levels on the day of hCG (22, 25, 41-43, 56, 70- 72). It has been proposed that this more favorable endocrine milieu reflects a healthier cohort of developing follicles in hMG cycles. One study also demonstrated increasing implantation rates with increasing doses of LH supplementation (73). This dose dependent benefit of LH could be due to an increase in the quality of the oocytes retrieved or due to an endometrial effect on implantation. However, this study was small and we are not aware

4. LH activity increases estradiol production from follicles

**3. Human menopausal gonadotropin** 

that the findings have been confirmed.

of high LH levels is less prevalent.

**2.1. Summary points** 

analysis.

**Figure 4.** Demonstrates the concept of an LH window. Low LH levels have been associated with decreased poor pregnancy outcomes with levels below 0.5-1.2 IU/L, demonstrating a threshold below which low LH causes poor outcomes. High LH levels have also been associated with poor pregnancy outcomes with levels over 6.8-10 IU/L. This gives rise to the concept of a therapeutic LH window (in green) to maximize ART outcomes.

It has also been demonstrated that elevated LH levels are associated with negative ART cycle outcomes. Decreased pregnancy rates and increased spontaneous abortion were reported with LH levels above 10 IU/L (59). Increased follicular arrest, decreased fertilization, higher recurrent pregnancy loss, and lower implantation rates have all been reported in patients with higher LH levels that controls, although ceiling values were not established in these studies (56, 60-63). This evidence that too much or too little LH activity can have negative outcomes has led to the concept of an LH window (4, 61, 64). In reality, with GnRH analogues, most patients are not in danger of having elevated endogenous LH levels. Indeed, by day 6 of GnRH antagonist administration, endogenous LH levels are depressed to a mean level of 1.6 IU/L, a value much closer to the LH threshold than the ceiling (65). Similarly, long agonist protocols also suppress endogenous LH levels to a mean near 1 IU/L (66). The evidence would suggest the clinician should be more concerned with replacing an adequate LH level in patients under pituitary down-regulation and the threat of high LH levels is less prevalent.
