<< . .

( 30)

. . >>

Urofollitropin 1982 urine 75 <0.1 none no im
Urofollitropin HP 1992 urine ¼ 9000 75 <0.001 no sc
Follotropin 1995 CHO cells ¼ 12 000 75 yes sc
À þþþ

FSH, follicle stimulating hormone; LH, luteinizing hormone; QC, quality control; HP, highly puri¬ed; CHO, Chinese
hamster ovary; im, intramuscular; sc, subcutaneous

60 cases of hyperstimulation, including two deaths, were reported when using
PMSG (Figueroa-Casas, 1958; Muller, 1963; Schenker and Weinstein, 1978).

Human Pituitary Gonadotrophins
Gemzell et al. (1958) were the ¬rst to describe the ¬rst successful pregnancy
after ovulation induction and pregnancy in humans utilizing follicle stimulating
hormone (FSH) derived from human pituitary glands removed at autopsy.
Gemzell et al. (1963) reported OHSS in 4 out of 22 cycles, but in 1970 only one
case of OHSS was observed by monitoring estradiol levels. Human pituitary
gonadotrophins are no longer used. There have been very few reported cases of
CJD in Australia in patients treated with pituitary gonadotrophin preparations
(Cochius et al., 1990; Brown et al., 1992).

Human Menopausal Gonadotrophin
Human menopausal gonadotrophin (hMG) was developed by extraction of
urine based on a process developed in 1947 by Pietro Donini of Serono in
Rome (Loumaye and Howles, 1999). This was ¬rst successfully used in
hypogonatropic hypogonadal women for inducing pregnancy by Bruno
Lunenfeld (Lunenfeld et al., 1962; Lunenfeld, 1963).
In 1967, Rabau et al. in their classic article (Rabau et al., 1967), found no
relationship between the incidence of OHSS and the dose of gonadotrophin
administered. Schenker and Weinstein (1978) found a difference between
experimental animal and human studies regarding the impact of hMG dosage
and the occurrence of OHSS. In experimental animals, a direct relationship was
observed between the hMG dose and the development of OHSS. The change in
ovarian size, the degree of capillary permeability, and severity of ascites and

Table II.5 Comparison of the stimulation characteristics of OHSS cycles to
normo-ovulatory women who became pregnant after treatment with the same
GnRH-a/hMG protocol
Reproduced with permission from Smitz et al. (1990). Hum Reprod 5:933À7

OHSS Normal cycles

(n ¼ 10) (n ¼ 40) Signi¬cance

30.2 + 6.0 21.0 + 7.0
Number of days before desensitization p < 0.01
9.6 + 1.7 12.3 + 2.5
Days of hMG stimulation p < 0.01
21.9 + 6.9 39.2 + 14.2
Number of ampoules of hMG used p < 0.001
3735.0 + 1603 1634 + 492
Preovulatory estradiol concentration (ng/l) p < 0.001
7.5 + 4.2
Number of oocytes retrieved 19.1+ 10.3 p < 0.001

pleural effusion were all related to the dose of gonadotrophin. In humans,
OHSS in the individual patient could be a consequence of overdose of
gonadotrophin, whereas, in groups of patients, no correlation between the dose
of hMG and the occurrence of OHSS was observed (Schenker and Weinstein,
OHSS occurred in 0.008À23% of hMG/hCG cycles (Table II.2), compared
with 0.6À14% in GnRH-a/hMG/hCG cycles (Table II.1). Comparison of the
endocrine patterns (Table II.5) in patients who developed OHSS, and in
normo-ovulatory patients who became pregnant after treatment with the same
ovarian stimulation protocol, showed that the former required less hMG to
achieve a higher preovulatory serum estradiol concentration and a higher
number of mature oocytes (Smitz et al., 1990). Rizk and Smitz (1992) found
that, in the groups reporting the lowest frequency of severe OHSS, ovarian
stimulation was started with 2 ‚ 75 IU hMG, whereas most other groups used
3 ‚ 75 IU hMG. The amount of FSH injected at the start could possibly induce
the growth of a larger number of follicles, which could develop suf¬ciently to
acquire receptors for luteinizing hormone (LH), and, as a result, luteinize

Purified Urinary Follicle Stimulating Hormone
Urofollitropin (FSH) has been available since the 1980s. It is devoid of LH but
is still contaminated with urinary proteins. Highly puri¬ed urofollitropin
(FSH-hp) has been available since the 1990s and contains very small amounts of
urinary proteins. Lack of urinary proteins diminishes adverse reactions such as
local sensitivity (Albano, 1996) while the absence of LH has no negative effects
on the stimulation of PCOS patients (Hayden, 1999). Raj et al. (1977) have
suggested that the use of FSH in anovulatory patients with PCOS offers a safe
treatment compared with hMG, resulting in higher pregnancy rates and lower

hyperstimulation rates. These authors suggested that endogenous LH levels in
patients with polycystic ovaries are quite adequate for follicular development,
and that the administration of exogenous LH is therefore unwarranted.
However, it was quickly apparent that the pFSH did not suppress the risks of
hyperstimulalation and multiple births when used in the conventional protocol
(Check et al., 1985; Garcea et al., 1985, Buvat and Buvat-Herbaut, 1986; Buvat
et al., 1989). Check et al. (1985) found a 23.7% incidence of OHSS in
18 women treated with 38 FSH cycles. Severe OHSS occurred in 5.3%,
indicating that puri¬ed FSH is no safer than hMG.
Seibel et al. (1984) reported a new protocol consisting of chronic, low-
dose pFSH administration, starting with 40 IU/day, without any hCG injection.
The rate of hyperstimulation was signi¬cantly decreased by using a very low
dose, and further re¬nements in the regimen resulted in a signi¬cant
improvement in the pregnancy rate without a concomitant increase in the
OHSS rate.

Recombinant Follicle Stimulating Hormone
Recombinant FSH is made from Chinese hamster ovarian cells, which are the
host cells in the production of glycoproteins. The Chinese hamster ovary cells
express FSH activity biologically in amounts that are suf¬cient to make the
production process viable. A genomic clone that contains the complete
sequence of the FSH b-subunit alone, or together with the gene of the
a-subunit, is transferred to the Chinese hamster ovary cells (Figure II.6). The
polypeptide chain of the recombinant FSH is identical to the natural one.
However, the carbohydrated structures can be identical or closely related. The
computer model of the FSH glycoprotein hormone (Gonal-F) is presented
in Figure II.7. Furthermore, chimeric molecules have been synthesized with
a longer half-life by modifying the carboxy peptide end (Risquez, 2003). The
speci¬c bioactivity of bioFSH is ¸ 10 000 IU FSH/mg of proteins (Loumaye and
Howles, 1999). The advantages of recombinant FSH compared to urinary
products are the consistency of the ¬nal product, high biological purity (which
allows its subcutaneous injection) as well as chemical characterization for better
quality control (Recombinant Human FSH Product Development Group,
1998). The complete lack of LH allows precise studies on ovarian folliculo-
genesis and, ¬nally, the production of new molecules with short or long activity
(Devroey et al., 1994; Tarlatzis and Billi, 1998; Risquez, 2003). The ¬rst
pregnancies after the use of recombinant FSH for ovulation induction in
anovulatory infertility (Donderwinkel et al., 1992) and for ovarian stimulation
in IVF (Devroey et al., 1992) were reported more than a decade ago. Today,
more than a million babies have been born worldwide. More recently, similar
ef¬cacy, tolerability and safety was observed in a randomized comparative IVF/
ICSI trial of highly puri¬ed menotropin (MENOPUR, Ferring, Copenhagen,
Denmark) vs. recombinant follicle stimulating hormone (Gonal-F, Serono,
Switzerland) (European Israeli Study Group on highly puri¬ed menotropin vs.
recombinant FSH, 2002). The OHSS rates were 7% vs. 5.1% respectively and,

Fig. II.7: Computer model of the glycoprotein hormone Gonal-F (r-hFSH)
Reproduced with permission from Howles (1996). Hum Reprod Update 2:172À91

interestingly, an advantage of exogenous LH in IVF but not ICSI cycles was
observed (Platteau et al., 2004).

What about Recombinant FSH and OHSS?
In a prospective randomized study of low-dose step-up protocols in PCOS
patients resistant to clomiphene citrate, Rizk and Thorneycroft (1996) found
that the use of recombinant FSH did not abolish the risk of OHSS. In fact, the
incidence was comparable to that of puri¬ed urinary FSH in a similar low-dose
protocol. Aboulghar et al. (1998b) found no difference between recombinant
FSH and hMG. In a Cochrane database systematic review, Bayram et al. (2001)
studied the safety and effectiveness of recombinant and urinary FSH in terms of
ovulation, pregnancy, miscarriage, multiple pregnancy rate and OHSS. Only
four randomized clinical trials of rFSH vs. uFSH have been identi¬ed.
Gonadotrophins used in these studies were Follitropin-beta (Puregon) vs.
urofollitropin (Metrodin), Follitropin-alpha (Gonal-F) vs. urofollitropin
(Metrodin). No signi¬cant differences were demonstrated for the relevant
outcomes. The odds ratio (OR) for ovulation was OR ¼ 1.19 (95% CI,
0.78À1.80), pregnancy rate ¼ 0.95 (95% con¬dence interval (CI), 0.64À1.41);
multiple pregnancy rate, OR ¼ 0.44 (95% CI, 0.16À1.21); miscarriage rate,
OR ¼ 1.26 (95% CI, 0.59À2.70) and OHSS, OR ¼ 1.55 (95% CI, 0.50À4.84);
ovulation rate. A systematic review and meta-analysis of 18 randomized
controlled trials comparing recombinant and urinary FSH con¬rmed that there
was no difference in the incidence of OHSS (Daya, 2002).

Long-acting Recombinant Follicle Stimulating Hormone
The relatively short half-life of FSH preparations (32+12 h)(Mannaerts et al.,
1993) requires daily injections, which cause considerable discomfort to the
patient. In an attempt to create a long-acting FSH preparation, chimeric genes
containing the sequence encoding the carboxy terminal peptide (CTP) of beta-
hCG fused with beta-FSH were constructed (Fares et al., 1992). The ¬rst human
trials showed that recombinant FSHÀCTP could be administered in
hypogonadal males (Bouloux et al., 2001) and showed an extended half-life
of 95 h (Duijkers et al., 2002). The pharmacodynamics of a single low dose of
long-acting recombinant FSH (Corifollitropin-alpha) has been studied in
women in WHO Group II anovulatory infertility (Balen et al., 2004). Following
a single dose of long-acting remcombinant FSH, serum FSHÀCTP initially rises
to peak levels at one to two days and thereafter, serum FSHÀCTP slowly
decreases. This overall pro¬le mimics the FSH ¬‚are-up induced by clomiphene
citrate treatment and a step-down approach in classical ovulation induction.
The objective of the study was to determine whether a single low dose could
replace ¬rst- and second-line treatment of anovulatory women, assuming that
both clomiphene citrate responders and clomiphene citrate resistors could be
treated by this long-acting FSH. At this point, having a single starting dose for
all patients is not feasible and additional research is required to achieve
monofollicular ovulation. Beckers et al. (2003) reported the ¬rst live birth after
ovarian stimulation using a chimeric, long-acting human recombinant follicle
stimulating hormone agonist (rFSHÀCTP) for IVF (Figure II.8).

Fixed-dose Gonadotrophins
A ¬xed gonadotropin regimen to increase the ef¬ciency of IVF cycles was ¬rst
used in France without a concomitant increase in the incidence of OHSS
(Rainhorn et al., 1987).

Fig. II.8: First live birth after long-acting recombinant FSH
Reproduced with permission from Beckers et al. (2003). Fertil Steril 79:621À3

Table II.6 GnRH agonists available worldwide
Reproduced with permission from Edwards RG, Rizquez F (Eds) (2003). Modern
Assisted Conception. Cambridge, UK: Reproductive Biomedicine Online:
Reproductive Healthcare, Ltd, p. 64

Agonist Structure

Leuprolide (Lupron• ) pGlu-His-Trp-Ser-Tyr-DLeu-Leu-Arg-Pro-EtNH2
Triptorelin (Decapeptyl• ) pGlu-His-Trp-Ser-Tyr-DTrp-Leu-Arg-Pro-Gly-NH2
pGlu-His-Trp-Ser-Tyr-DSer (OtBU)-Leu-Arg-Pro-EtNH2
Buserelin (Suprefact• )
Histrelin (Supprelin• ) pGlu-His-Trp-Ser-Tyr-DHis (Bzl)-Leu-Arg-Pro-AzaglyNH2
Nafarelin (Synarel• ) pGlu-His-Trp-Ser-Tyr-DNal(2)-Leu-Arg-Pro-Gly-NH2
pGlu-His-Trp-Ser-Tyr-DSer (OtBu)-Leu-Arg-Pro-AzaglyNH2
Goserelin (Zoladez• )

Interestingly, Rizk et al. (1991b) found that a ¬xed-regimen protocol with
a predetermined date of retrieval has a similar incidence of OHSS of approxi-
mately 1%. This study was performed at Norwich in the United Kingdom
between 1988 and 1991, where the National Health Service would allocate
a ¬xed operative session to perform Gamete Intrafallopian Transfer (GIFT).
There was no observed difference between short and long protocols (Rizk et al.,


Greenblatt and Bar¬eld (1961) introduced clomiphene citrate for ovulation
induction. Severe OHSS with clomiphene citrate (CC) is rare. Southan and
Janovsky (1962) reported a patient with polycystic ovaries who developed
massive ovarian enlargement, ascites and hydrothorax after the administration
of 100 mg of CC for 14 days. Scommegna and Lash (1969) reported a case of
ovarian hyperstimulation associated with conception after treatment with CC.


The development of gonadotrophin-releasing hormone agonist has had a
tremendous impact on the practice of reproductive endocrinology. Several
GnRH agonists are used worldwide (Table II.6) and a wide variety of protocols
have been implemented in clinical practice (Figure II.9).

Gonadotrophin-releasing Hormone Agonist without
OHSS has rarely been reported following the administration of GnRH agonist
without gonadotrophins (Campo et al., 2000; Weissman et al., 1998).

Fig. II.9: GnRH agonists and gonadotrophins: short and ultra-short and long and ultra-
long protocols
Reproduced with permission from Edwards RG, Risquez F (Eds) (2003). Modern
Assisted Conception. Cambridge, UK: Reprod Biomed Online, Reproductive Health Care
Ltd, p. 107

Gonadotrophin-releasing Hormone Agonist Long Protocol
It had been hoped that the use of GnRH-a/hMG protocols would decrease
the incidence of OHSS. With the use of GnRH-a, the practice of IVF has
been simpli¬ed, the blocking of the LH surge permits further stimulation of
the ovaries, increasing the number of oocytes (Belaish-Allart et al., 1989; Rizk,
1992, 1993a). Rizk and Smitz (1992) have summarized the major reports of
OHSS after the use of GnRH and gonadotrophins for IVF (Table II.1).
Golan et al. (1988) observed a high incidence of OHSS, 8.4%, after
combined GnRH-a and hMG for superovulation. The French report of IVF
results (Bilan FIVNAT, 1989) showed that the use of GnRH-a led to signi¬-
cantly higher preovulatory estradiol concentrations, and to more frequent
severe hyperstimulation (4.6% versus 0.6% for non-GnRH-a/hMG cycles). In a
Cochrane database review, Hughes et al. (2000) reported an odds ratio of
1.4 (0.5À3.92) for moderate and severe OHSS in cycles using GnRH agonist as
an adjunct to gonadotrophins compared with gonadotrophins alone.


Third generation GnRH antagonists became available for assisted reproduc-
tive technology (ART) in the 1990s (Table II.7). They suppress gonado-
trophin release by competitive receptor binding, resulting in immediate
suppression and blockage of gonadotrophin secretion (Rizk and Nawar,
2004). Several protocols for GnRH antagonists in ART have been used

Table II.7 GnRH antagonists available worldwide
Reproduced with permission from Edwards RG, Rizquez F (Eds) (2003). Modern
Assisted Conception. Cambridge, UK: Reproductive Biomedicine Online:
Reproductive Healthcare, Ltd, p. 64

Name Structure

[N-Ac-DNAL1, DpCl-Phe2, DPal3, Arg5, DGlu6, (AA), DAla10]GnRH
N-Ac-DNAL1, DpCl-Phe2, DTrp3, DhArg(ET2)6, DAla10]GnRH
N-Ac-DNAL1, DpCl-Phe2, DPal3, DCit6, DAla10]GnRH
N-Ac-DNAL1, DpCl-Phe2, DPal3, DhArg(ET2)6, DhArg(ET2)8, DAla10]GnRH
[N-Ac-DNAL1, DpCl-Phe2, DPal3, Lys(Nic)5, DLys(Nic)6, Lys(iPr)8, DAla10]GnRH
[N-Ac-DNAL1, DpCl-Phe2, DPal3, D(HcI)6, Lys(iPr)8, DAla10]GnRH
-Ac-DNAL1, DpCl-Phe2, DPal3, Daph(atz)6, D-Aph(atz)6, Lys(iPr)8, DAla10]GnRH
Azaline B

Fig. II.10: GnRH antagonist protocol for assisted reproduction technology
Reproduced with permission from European Middle East Orgalutron Study Group (2001).
Hum Reprod. 16:644À51

(Figure II.10). The ef¬cacy and safety of GnRH antagonists in IVF and ICSI
cycles were reported to be similar to those of the GnRH agonists (Albano et
al., 2000; Ludwig et al., 2000; Olivennes et al., 2000; European Orgalutran
Study Group, 2000; European-Middle East Orgalutran Study Group, 2001;

Fluker et al., 2001). No difference was observed in a Cochrane review between
the agonist and antagonist protocols and the incidence of OHSS (Al-Inany
and Aboulghar, 2002). As more studies are published, this conclusion could
change in favor of the GnRH antagonist, as will be discussed in detail in
Chapter VII.


Edwards et al. (1980), in the ¬rst extended report on IVF, stated that the
luteal phase of virtually all patients was shortened considerably after
treatment with gonadotrophins, and it was suggested that high follicular
phase estrogen levels due to ovarian hyperstimulation might be involved. In
the United States, initial studies in 1983 concerning hMG-stimulated IVF
cycles also con¬rmed the occurrence of an abnormal luteal phase in IVF cycle,
with characteristic features of elevated progesterone levels and signi¬cantly
reduced luteal phase length (Jones, 1996). Seppala (1985) compiled The
World Collaborative Report on IVF, which pointed out the non-uniform
approach to luteal phase support. Many centers used progesterone and its
derivatives, others used hCG and some avoided any luteal support. The
introduction of GnRH-a offers advantages in terms of pituitary desensitiza-
tion and prevention of a premature LH surge, thereby resulting in lower
cancellation rates, and increased numbers of preovulatory follicles. However,
because of its effect on corpus luteum function, treatment is required if a
satisfactory luteal phase endometrium is to be maintained (Rizk, 1992,
1993a, b; Rizk et al., 1997). Smitz et al. (1988) were the ¬rst to document
luteal phase insuf¬ciency in 23 IVF cycles using GnRH-a when luteal phase
support with progesterone was omitted. Golan et al. (1988) observed a high
incidence of OHSS (8.4%) after the use of hCG for luteal support. Herman et
al. (1990) in a prospective randomized trial, compared the pregnancy rate and
incidence of OHSS after luteal hCG in IVF cycles stimulated with GnRH-a
and hMG. Nine of the 18 patients who received hCG conceived, compared
with 3 out of 18 patients in the placebo group. OHSS occurred in 5 out of 18
patients treated with hCG. None of the 18 patients without luteal hCG
support developed OHSS, including those patients who became pregnant.
Because the number of patients who conceived in the placebo group was
small (3 out of 18), the authors concluded that the question of whether
repeated early luteal hCG injections are more important in the pathogenesis
of OHSS than endogenous hCG secreted later by the developing conceptus,
remains unresolved. A meta-analysis of 18 studies reported a signi¬cant
reduction in the occurrence of OHSS when progesterone is used for luteal
phase support instead of hCG (Soliman et al., 1994) as will be discussed in

<< . .

( 30)

. . >>

Copyright Design by: Sunlight webdesign