GnRH agonists and antagonists for follicle stimulation in IVF

Document Sample
GnRH agonists and antagonists for follicle stimulation in IVF Powered By Docstoc

                  GnRH agonist dose and ovulation induction outcome
                     C.A.M. Jansen MD, PhD, K.E. Tucker PhD, HCLD
           Department of IVF, Reinier de Graafgroep, loc Diaconessenhuis Voorburg
                                       The Netherlands

“Off label use” is a term reserved for the prescription of medication for an unregistered
indication. There are numerous examples of “off label use” in medicine. Some of these occur
with the unspoken consent of the drug company [such as the use of GnRH agonists in
assisted reproductive technologies (ART)] and some without, (such as the use of the calcium
antagonist nifepidine - originally intended for treating hypertension - for alleviating premature
labor). Sometimes certain drugs are used directly against the directives of the company, such
as the use of almost homeopathic doses of human chorio gonadotropin (hCG) in the Simeons
Therapy to achieve weight loss.

Gonadotropin-releasing hormone (GnRH) and its analogues have been extensively used in
clinical medicine. Pulsatile secretion of native GnRH stimulates the anterior pituitary and
exogenous administration has been used for induction of ovulation. GnRH agonists initially
produce a stimulation of the gonadotrophs resulting in secretion of follicle stimulating
hormone (FSH) and luteinizing hormone (LH) and the expected gonadal response. Although
continuous administration initially has the same effect, it is subsequently followed by down-
regulation and inhibition of the pituitary-gonadal axis by clustering and internalization of the
specific receptors. In contrast, GnRH antagonists promptly suppress pituitary gonadotropin by
competitive GnRH-receptor binding, thereby avoiding the initial stimulatory phase of the
agonists. This binding occurs in a dose-dependent fashion and discontinuation of GnRH
antagonist treatment leads to a rapid and predictable recovery of the pituitary-gonadal axis.

The GnRH analogues are potent therapeutic agents that are considerably useful in a variety
of clinical indications. GnRH agonists can be used to obtain a state of hypophyseal
quiescence, which subsequently and concomitantly can be overruled by the exogenous
administration of gonadotrophins, or to produce an endogenous LH surge when follicle
stimulation is achieved with gonadotropins only or with the additional administration of an
GnRH antagonist. The appeal in ART is based on the fact that agonists prevent the
occurrence of a premature LH rise, thereby reducing its negative effect on the cohort of
growing follicles. This LH surge does not necessarily have to lead to an inadvertent ovulation
- when many follicles are present - but it does lead to follicle luteinization and diminished
oocyte fertilization, implantation and pregnancy rates.

The initial introduction of GnRH agonists in ART and ovulation induction (OI) was a classic
example of “off label use”. These medications had been registered previously for a totally
different indication, i.e. prostate carcinoma, for which it is necessary to achieve castrate
levels. It is surprising, however, that it has taken more than ten years before proper dose-
finding studies for this new indication have been performed.

In controlled ovarian Hyperstimulation (COH) there are a number of protocols with GnRH
agonists used, all meant to achieve down regulation, and many clinicians have also extended
the use to OI. This paper will deal with the results of the various dose-finding studies for
GnRH agonists, as well as the endocrinological consequences of variations in administration,
such as midfollicular cessation of agonists. In addition, we will discuss the benefits and
drawbacks of the endogenous LH peak produced by GnRH agonists with or without pre-
treatment of antagonists. The majority of studies stem from COH in ART cycles, but many of
the principles presented therein are also applicable to OI and stimulated IUI cycles.

It is extremely important that efforts be made to identify those patients that would be likely to
benefit from these strategies, as well as those that have no demonstrable advantages. At any
rate it should be clear that the present dosage schemes represent an “overkill‟ situation, and
should be replaced by more subtle schemes with individualized doses.



Gonadotrophin-releasing hormone (GnRH) and its analogues has been extensively used in
clinical medicine. Pulsatile secretion of native GnRH stimulates the anterior pituitary and
exogenous administration has been used for induction of ovulation. GnRH agonists initially
produce a stimulation of the gonadotrophs resulting in secretion of follicle stimulating
hormone (FSH) and luteinizing hormone (LH) and the expected gonadal response. Although
continuous administration gives the same initial effect, it is subsequently followed by down
regulation and inhibition of the pituitary-gonadal axis by clustering and internalization of the
specific receptor. Alternately, GnRH antagonists promptly suppress pituitary gonadotropin by
competitive GnRH-receptor binding, thereby avoiding the initial stimulatory phase of the
agonists. This binding occurs in a dose-dependent fashion and discontinuation of GnRH
antagonist treatment leads to a rapid and predictable recovery of the pituitary-gonadal axis.

GnRH agonists in reproductive medicine

Whenever a new medication is introduced into clinical practice, sequential, elaborate and
time-consuming investigations need to be performed. Phase I studies have been established
to describe drug mode of action, Phase II studies are designed for dose-finding and Phase III
studies for the determination of efficacy and efficiency. Phase III trials are also designed to
define the complication rate since it is often during this final investigative trial that any side
effects- occurring less frequently- can be reported. The seriousness of these is judged in
relation to the expected health benefits and to the significance of the disease for which the
medication is intended. For instance, even possible fatal side effects of the drug itself may be
acceptable when the condition itself is so serious that overall life expectancy in patients
having the disease is markedly improved by the medication.

This is usually the case for metastatic carcinoma and it should be clear that these conditions
couldn‟t be compared to non-life-threatening diseases or conditions, such as infertility. Once
the FDA or other regulating authorities have approved the medication for clinical use,
however, the prescribing physician has no obligation whatsoever to justify its use for
indications other than the intended one.

Originally intended for the treatment of prostate carcinoma by achieving castrate levels of the
sex steroids, the use of GnRH agonists in ART or infertility is a typical example of this “off
label use”. Since their introduction, however, GnRH agonists have gained widespread
notoriety for use in this new arena, even though there had been no dose finding or proper
health-gain-to cost studies designed to introduce these drugs into assisted reproductive

GnRH agonists were widely used since they could prevent premature follicle luteinization.
This peculiar type of drug introduction was only possible because these substances were
already available, as mentioned previously. As such it is surprising that it has taken longer
than 10 years after their entry into the field of ART before the first proper dose-finding studies
for COH in IVF were performed . In addition, surprisingly few articles deal with OI for
anovulatory infertility, as such. Some articles deal with what is euphemistically called
ovulation induction for intrauterine insemination (IUI), which usually refers to the stimulation of
multifollicular growth, which is the main basis for the added value of IUI for certain infertility
diagnoses, such as unexplained infertility.

Stimulation regimes

A meta-analysis of several prospective and randomized studies comparing down-regulated
and non-down-regulated treatments as first choice in IVF shows that a significant difference in
pregnancy rates exits between these two types of stimulation protocols . It should be noted,
however, that the differences are due to exceptionally low results in some studies in the non-
analogue group, much lower than previously published in comparable studies performed by
these same investigators. Nevertheless, the use of GnRH analogues has become ubiquitous
in an incredibly short time span, mainly because of the market availability and quite possibly,
for doctors‟ convenience. The use of GnRH analogues minimizes the need for careful follicle


monitoring, as even inaccurate measurements are not usually “punished” by premature LH
surges. The rapid acceptance of these drugs has occurred in spite of the increased risk of the
ovarian hyperstimulation syndrome (OHSS). The magnitude of this risk is more difficult to
quantify than pregnancy rate, mainly because of its lower incidence. Many prospective
randomized studies do not mention this; in addition there is a publication bias in reporting side
effects and almost a total publication deficit with regard to the mortality from this syndrome. A
Medline and Embase search revealed only one such case resulting from the complications of
the OHSS .

Several regimes have been proposed and are used throughout. The most well known are the
following four regimens. In the “ultra- short” protocol, GnRH agonists are given from day 1 to
3 of the stimulation cycle, after which the follicle stimulation is performed. The results from
this scheme have not fulfilled most clinicians‟ expectations and have therefore, been gradually
abandoned. During the “short” protocol, the agonist is given from day 1 of the cycle until the
day of hCG, concomitantly with the medication for ovarian stimulation. The disadvantage of
this regimen is that the stimulation lasts somewhat longer. The “long” protocol contains a
number of variants. With the long protocol the administration of GnRH agonists usually starts
during the preceding luteal phase. Follicular stimulation either hMG or FSH starts during the
follicular phase of the next cycle, with or without confirmation of down-regulation by
measuring circulating estradiol. Another variant to the protocol is to start on the day of
menstruation, or shortly thereafter (follicular phase onset), after which ovarian stimulation may
begin somewhere two to four weeks afterwards. A meta-analysis of randomized trials
between the short and ultra short protocols in comparison to the long protocol revealed a
significant difference in favor of the long protocol, with no difference between luteal or early
follicular phase start. Here again, however, not very much is known about the side effects, as
the incidence of these within individual articles is too low to perform any meaningful statistical
analysis and, so, they usually are not mentioned. Lastly, during the “ultra long” protocol
GnRH agonist is administrated for about 6 weeks : Often it begins with a depot injection that
lasts for at least about 4 weeks, after which the down regulation is maintained with daily
injections. Follicle stimulation with gonadotropins may start at any time after this, as long as
the agonists are continued. Due to the extreme long duration, coupled with the increasing
incidence of side effects such as vaginal dryness and hot flushes, this protocol has never
become popular, and hence the amount of data are insufficient.

Determination of down regulation

Often a serum estradiol level is used to confirm adequate down regulation and is usually
below 0.2 mol/l. It should be noted, however, that present Enzyme Immuno Assays are not
always appropriate, as there is a wide scatter in the low range. LH levels can be measured
as well, but many assays have a cut-off level of 1 IU/l and do not allow for the determination
of the minimum LH necessary to optimize follicle growth, as well as granulosa-cell estradiol
synthesis after stimulation with recombinant FSH.

It has been shown that standard suppression of LH may negatively affect results in terms of
ongoing pregnancy rate when stimulation is performed with recombinant FSH only. With an
initial standard dose of 0.5 mg of the agonist buserelin daily (s.c.), followed by a maintenance
dose of 0.2 mg daily, half of the patients had a LH below 0.5 IU/l on stimulation day 8 with rec
FSH. Although the pregnancy rate was similar in both groups, the early pregnancy loss was
45% in the low LH group versus 9% in the normal LH group . This suggests that indeed there
is a minimum concentration of LH necessary for adequate follicle growth. In the past, when
stimulation was performed with hMG this point was moot as exogenously administered LH
(and/or hCG) was present in the hMG formulation (mandatory 1:1 - FSH: LH ratio by the
pharmacopoeia) which was always sufficient to allow for proper folliculogenesis.

Pretreatment with oral contraceptives

Often oral contraceptives are prescribed for the cycle preceding the stimulation cycle. There
are two main advantages with this scheme. Firstly, there may be a diminished risk of cyst
formation , and secondly, it eliminates the risk of pregnancy. Although GnRH agonists
inadvertently administered during pregnancy do not lead to an increase in defects detectable


immediately after birth, there still is controversy with regard to their long-term effects. In fact, it
has been published that they may affect intellectual and motor development , especially with
regard to attention deficit. The possibility of long-term effects has already been suggested by
Swaab et al with the risk of what they called “functional neuroteratology” following in utero
                                                                10 11
exposure to small neuropeptides such as GnRH analogues , .

Modified protocols

Midfollicular cessation: Discontinuous long protocol

As GnRH agonists lead to desensitization due to a receptor internalization that only gradually
recovers over a time span of about over 2 weeks, some authors have advocated earlier
cessation, about halfway during the follicular phase. This does not seem to influence results
as long as exogenous LH (or hCG) is administered, but it negatively influences results once
FSH-only compounds are chosen as stimulants. It has been shown that when recombinant
FSH is given, cessation of the agonist on day 7 leads to extremely low LH levels in the days
following, much below what is generally considered the lower threshold necessary for
adequate follicular growth (which is between 0.5 and 1 IU/l) .

Continuation of agonists in the luteal phase
It has been suggested that inadvertent administration of agonists after ovulation might
improve the implantation potential; however this suggestion merely stems from the
observation that the patients concerned became pregnant in the cycle preceding the IVF
cycle. In view of this suggestion a prospective randomized study was performed in which
the agonist administration was continued until either a pregnancy was ruled out, or until the
occurrence of heartbeat on ultrasound . Although the pregnancy rate actually was higher in
the arm where the agonists were continued, the take home baby rate did not reach
significance (31/89 versus 21/86; P= 0.20) Especially in view of the recent reports of a
possible increased incidence of ADHD when agonists are administered during pregnancy,
extreme caution with this approach is warranted .

Studies with diminished doses

Some investigators have appropriately questioned the need for down regulation to castrate
levels and examined the effects of using lower doses of agonists . They have coined the
term “micro dose”, but one should question this as it may well be that what is considered
normal may actually be a “macro dose”. Results were similar in poor responders when 25
µg/day of triptorelin was administered instead of usual 100 µg dose.

The only real dose-finding studies were initiated by prof Schoemaker‟s group, performed at
the Free University in Amsterdam, as a prospective, double blind, randomized study. It was
clear that much lower doses, even up to 5 µg/day of triptorelin could suffice in obtaining
pituitary desensitization . In a further study there turned out to be no influence on pregnancy
rates between any of the dosages used (placebo, 15, 50 and 100 μg/ day), and even the
placebo led to a similar pregnancy rate , albeit a higher miscarriage rate. Table I shows the
details as deduced from this study.

Dose            ITT               Started           ET                Clin preg         Ongoing
Placebo         60                59                46                10                5
15 g/d         60                56                55                8                 8
50 g/d         60                56                55                10                9
100 g/d        60                58                55                13                7

Table I: Results of a dose-finding study with the GnRH agonist, triptorelin. ITT: intention to
treat; Started: cycles commenced; ET: Embryo transfer; Clin preg: Clinical pregnancy rate-
(erroneously called implantation rate in the article); Ong: Ongoing pregnancy rate .

This study is characterized by two peculiar aspects: one semantic, and one methodological.
Firstly, the implantation rate (IR) is defined as others would define the clinical pregnancy


rate/ET (i.e. the presence of at least one gestational sac per transfer). This means that
whenever the term IR is mentioned it should actually be clinical PR. Secondly, an LH peak is
defined in this study as a doubling from baseline with one later value higher than the doubled
one. This means, according to the authors, that if baseline level is 3 IU/l, a rise to 7IU/l, and
then to 8 IU/l the next day, depicts a LH peak. This, as such, cannot be considered to be a
full-blown LH surge, necessary to result in all changes prior to and after ovulation (i.e.
resumption of meiosis, adequate nuclear and cytoplasmic maturation and sufficient
luteinization) and it is known that during follicular stimulation such small LH peaks can be
present. In the placebo group, even after these small LH elevations, a follicle puncture was
performed, without additional hCG supplementation. It is, therefore, not surprising that there
were surprisingly few transfers in the placebo group, whilst this may also explain the high
miscarriage rate.

Depot preparations

Depot preparations are often used for benign gynecological disorders such as fibroids or
endometriosis. In some countries such as France and Israel, however, they are also used on
a large scale in IVF. Although they are guaranteed to act as long as 4 weeks, it has been
shown that their effects are still present and the analogues are still measurable about 8 weeks
after the last administration . In addition, the usual route of administration is subcutaneous
and this leads to a significantly longer course before cycle resumption takes place. For
ovulation induction this surely may pose the risk that the agonists are still present after
implantation, and that effects on pregnancy cannot be ruled out completely. Therefore,
caution should be warranted with depot preparations

Mimicking the LH peak

As GnRH agonists initially lead to a LH and FSH surge, these products can also be used to
                                                                                      19 20
mimic the endogenous LH surge in gonadotropin-only stimulation cycles for IVF or IUI , as
well as in clomiphene citrate cycles for IUI . As most centers resorted to down regulation with
agonists, however, these routes have not been explored sufficiently. The recent arrival of
GnRH antagonists may open up new avenues of ART treatment, especially in patients at risk
for OHSS. It has been shown that it is possible to mimic the LH surge after removal of GnRH
antagonist suppression , and it has been suggested that this approach may reduce the risk
of OHSS. More work in this area is needed to elucidate this point and to show if patients
really benefit from this approach.

Conclusion: The rehabilitation of minimal stimulation?

It is clear that GnRH agonists have become ubiquitous in IVF, but it is questionable whether
or not their use should be “first choice”, indiscriminate and leading to a maximal suppression.
The differences between patients treated with or without agonists can, at least in part, be
explained by the difference in cancellation rate. Agonists do not improve pregnancy rates in
centers with already good pregnancy rates, but they may normalize results in centers with
poor rates. They allow unrivalled flexibility, but with the same token, they may conceal
incompetence in follicle monitoring. They do not just prevent premature LH peaks, but also
mature LH peaks. There are clear signs that the complication rate, especially with regard to
the incidence of OHSS, is increased, and under these circumstances, the cost-benefit ratio is

It is clear from the first dose-finding studies that much lower doses than presently used will
suffice to prevent early luteinization. Especially now that recombinant FSH – devoid of LH
activity- is used, one should aim at levels of LH during suppression of at least between 1 and
3 IU/l. In view of the cost, side effects and increased risks, a number of centers are
abandoning the “maximal stimulation policy, regardless of cost and harm” of the past, and are
learning to accept the slightly decreased pregnancy rates . We would like to emphasize that
the „primum non nocere‟ (first do no harm) principle of Hippocrates should not be forgotten.


Future follicle stimulation schemes should become much more individualized. Instead of
completely and indiscriminately down-regulating all patients, clinicians should search for more
subtle approaches, and for selection criteria that will distinguish those patients who really
need GnRH agonists from those who need GnRH antagonists and those who need neither.


 Janssens RMJ, Lambalk CB, Vermeiden JPW, Schats R, Bernards JM, Rekers- Mombarg
LTM, Schoemaker J. Dose- finding study of triptorelin- acetate needed for prevention of a
premature LH surge in IVF: a prospective, randomised, double blind, placebo controlled
study. Hum Reprod 2000; 15: 2333-40.

  Hughes EG, Fedorkow DM, Daya S e.a. The routine use of gonadotropin releasing hormone
agonist prior to in vitro fertilization and gamete intrafallopian transfer: a meta-analysis of
randomized controlled trials, Fertil.Steril.1992; 58:888.
  Cluroe AD, Synek BA fatal case of ovarian hyperstimulation syndrome with cerebral
infarction. Pathology 1995; 27: 344- 6
  Daya S. Gonadotropin releasing hormone agonist protocols for pituitary desensitization in in
vitro fertilization and gamete intrafallopian transfer cycles. Cochrane Database Syst Rev.
2000: CD001299.
 Kim CH, Cho YK, Mok JE. Simplified ultralong protocol of gonadotrophin-releasing hormone
agonist for ovulation induction with intrauterine insemination in patients with endometriosis.
Hum Reprod. 1996; 11: 398-402.
 Tucker KE, Rietdijk A.M., Postma T, van de Water C, Jansen CAM,.
The accuracy of enzyme immunoassays of oestradiol for the determination of ovarian reserve
and the degree of down-regulation Hum. Reprod. Suppl 15; 2000; 15: 230.
 Westergaard LG, Laursen SB, Andersen CY. Increased risk of early pregnancy loss by
profound suppression of luteinizing hormone during ovarian stimulation in
normogonadotrophic women undergoing assisted reproduction. Hum. Reprod. 2000; 15:
 Biljan MM, Mahutte NG, Dean N, Hemmings R, Bissonnette F, Tan SL Pretreatment with an
oral contraceptive is effective in reducing the incidence of functional ovarian cyst formation
during pituitary suppression by gonadotropin-releasing hormone analogues. J Assist Reprod
Genet 1998; 15: 599-604.
 Lahat E, Raziel A, Friedler S, Schieber- Kazir M, Ron-El R. Long-term follow-up of children
born after inadvertent administration of a gonadotrophin-releasing hormone agonist in early
pregnancy. Hum Reprod 1999; 14: 2656-60.
  Swaab, D.F., G.J. Boer en Feenstra, M.F., Concept of functional neuroteratology and the
importance of neurochemistry. Prog. Brain Res, 1988; 73: 3.
  Boer, G.J., Sijdewint, G.M. en Swaab, D.F., Neuropeptides and functional neuroteratology.
Prog. Brain Res. 1988; 73: 245.

  Fujii S, Sagara M, Kudo H, Kagiya A, Sato S, Saito Y A prospective randomized
comparison between long and discontinuous-long protocols of gonadotropin-releasing
hormone agonist for in vitro fertilization. Fertil Steril 1997; 67:1166-8.
  Cedrin-Durnerin I, Bidart JM, Robert P, Wolf JP, Uzan M, Hugues JN. Consequences on
gonadotrophin secretion of an early discontinuation of gonadotrophin-releasing hormone


agonist administration in short-term protocol for in-vitro fertilization. Hum Reprod. 2000; 15:
  Balasch J, Martinez F, Jove I, Cabre L, Coroleu B, Barri PN, Vanrell JA.Inadvertent
gonadotrophin-releasing hormone agonist (GnRHa) administration in the luteal phase may
improve fecundity in in-vitro fertilization patients. Hum Reprod. 1993;8:1148-51.
     Pellicer A et al, 2000 (in preparation).
   Surrey ES, Bower J, Hill DM, Ramsey J, Surrey MW. Clinical and endocrine effects of a
microdose GnRH agonist flare regimen administered to poor responders who are undergoing
in vitro fertilization. Fertil Steril 1998; 69: 419-24.
   Janssens RMJ, Vermeiden JPW, Lambalk CB, Schats R, Schoemaker J Gonadotrophin-
releasing hormone agonist dose-dependency of pituitary desensitization during controlled
ovarian hyperstimulation in IVF. Hum Reprod 1998; 13: 2386-91.
  Filicori M, Cognigni GE, Arnone R, Pocognoli P, Tabarelli C, Ciampaglia W, Taraborelli S,
Casadio P. Subcutaneous administration of a depot gonadotropin-releasing hormone agonist
induces profound reproductive axis suppression in women. Fertil Steril 1998; 69: 443-9.
   Itskovitz J, Boldes R, Levron J, Erlik Y, Kahana L, Brandes JM Induction of preovulatory
luteinizing hormone surge and prevention of ovarian hyperstimulation syndrome by
gonadotropin-releasing hormone agonist. Fertil Steril 1991; 56: 213-20.
  Shalev E, Geslevich Y, Matilsky M, Ben-Ami M.Induction of pre-ovulatory gonadotrophin
surge with gonadotrophin-releasing hormone agonist compared to pre-ovulatory injection of
human chorionic gonadotrophins for ovulation induction in intrauterine insemination treatment
cycles. Hum Reprod. 1995; 10: 2244-7.

   Shalev E, Geslevich Y, Matilsky M, Ben-Ami M. Gonadotrophin-releasing hormone agonist
compared with human chorionic gonadotrophin for ovulation induction after clomiphene citrate
treatment. Hum Reprod. 1995; 10: 2541-4.

  Olivennes F, Fanchin R, Bouchard P, Taieb J, Frydman R.Triggering of ovulation by a
gonadotropin-releasing hormone (GnRH) agonist in patients pretreated with a GnRH
antagonist. Fertil Steril. 1996; 66: 151-3.
    Itskovitz-Eldor J, Kol S, Mannaerts B.Use of a single bolus of GnRH agonist triptorelin to
trigger ovulation after GnRH antagonist ganirelix treatment in women undergoing ovarian
stimulation for assisted reproduction, with special reference to the prevention of ovarian
hyperstimulation syndrome: preliminary report: short communication.
Hum Reprod. 2000; 15: 1965-8.
   DeVane GW, Gangrade BK, Wilson R, Loy RA.Optimal pregnancy outcome in a minimal-
stimulation in vitro fertilization program. Am J Obstet Gynecol. 2000; 183: 309-13.