<< . .

( 30)

. . >>

The results of Trial 21447 showed that a clinically signi¬cant OHSS (severe and
all cases) was signi¬cantly lower in the r-hLH compared to the u-HCG group
(p ¼ 0.001), however the pregnancy rates and the clinical pregnancy rates
were signi¬cantly lower in the r-hLH group compared to the u-hCG group
(p ¼ 0.018 and p ¼ 0.023, respectively). In order for the r-hLH to be as effective
as hCG, the dose should be increased. However, the cost/bene¬t ratio may
not be bene¬cial. The study has not yet been published and the manufacturer
of r-hLH decided not to register the high dose used for triggering ovulation
(Aboulghar and Al-Inany, 2005). Unfortunately the clear advantage of r-hLH
in reducing the incidence of OHSS which was demonstrated by the European
Recombinant LH Study Group (2001) cannot be yet utilized in clinical practice
because of the lower pregnancy rates after triggering ovulation with r-hLH
(Aboulghar and Al-Inany, 2005).
Emperaire (2005) did not accept the ¬ndings of this unpublished study
for at least four reasons. First, the debate upon the safety and ef¬ciency of
recombinant LH vs. hCG to trigger ovulation is important enough that it could
not be ignored because of the results of an unpublished trial; and he also
wondered why the trial was not published if the design was correct and
prospectively performed. Second, the results are restricted only to the triggering
of ovulation for IVF after GnRH agonist desensitization and cannot be extended
to other types of stimulation (Emperaire and Edwards, 2004), or it could be used
in IVF cycles where GnRH antagonists have been utilized. This second point has
led to a reply by Aboulghar and Al-Inany (2005), where they quoted a
signi¬cantly lower pregnancy rate when GnRH agonist was used instead of hCG
to trigger ovulation in GnRH antagonist cycles, leading to the discontuation
of clinical trials (Humaidan et al., 2005; Kolibianakis et al., 2005). A third
but important point that Emperaire (2005) discussed was the fact that OHSS
was signi¬cantly reduced when recombinant LH was used instead of hCG
(p < 0.001). On the other hand the signi¬cance of a lower pregnancy and clinical
pregnancy rate appears to be statistically much weaker (p < 0.018 and p < 0.023,
respectively). Fourth, and ¬nally, the dose of recombinant LH could be subject
to further testing to determine the most optimal regimen. He observed that, in
primates, two doses of 2500 IU rLH given 12 h apart were more ef¬cient than
a single double-dose (Zelinski-Wooten et al., 1991).

Native GnRH to Trigger Ovulation
Native GnRH has also been investigated for triggering ovulation instead
of hCG in cycles stimulated by hCG. Gerris et al. (1995) compared intra-
venously administered gonadotropin hormone (100 and 500 mg), GnRH
agonist (buserelin, 500 mg) and human chorionic gonadotrophin (10 000 IU).
Endogenous LH surges occurred in all cycles of patients treated with GnRH
or GnRH agonists. The rise was slowest but highest in the group of patients
treated with GnRH agonist (500 mg) compared with GnRH both at the 100
and 500 mg doses (p < 0.0001). Although the serum estradiol was similar in all
groups, day 8 estradiol was signi¬cantly higher in the hCG group. Luteal phase
insuf¬ciency, de¬ned as cycles with progesterone concentrations of <8.0 ng/ml,
occurred more frequently in patients whose ovulation was triggered by GnRH
or GnRH agonist despite progesterone supplementation. Clinically, day 8 luteal
scores showed more conspicuous OHSS in the hCG group (p ¼ 0.292). The
authors concluded that ovulation occurs and pregnancies can be achieved
following endogenous LH surge induced by GnRH and its agonist. A high
frequency of luteal insuf¬ciency occurs in such cycles even with luteal supple-
mentation. Finally, OHSS cannot be totally prevented by this approach although
cycles with an endogenous LH surge, in general, result in fewer subclinical
instances of ovarian hyperstimulation.

Intravenous Albumin Administration
Experience in subjects with different forms of third space ¬‚uid accumulation
have shown that albumin is effective in preventing and correcting hemodynamic
instability. Using a similar approach, in an effort to increase the oncotic pressure
and to reverse the leakage of ¬‚uids from the intravascular space, high-risk
subjects for severe OHSS were treated with albumin. Asch et al. (1993) were the
¬rst to suggest that intravenous albumin during follicular aspiration could be
potentially useful to prevent OHSS. The authors proposed that the role of
albumin in prevention of OHSS is multifactorial. First, it acts to sequester
vasoactive substances released from the corpora lutea. Albumin has a half-life
of 10“15 days. Patients generally develop OHSS symptoms 3“10 days after
hCG injection, regardless of the embryo transfer. Timely administration of
albumin, during oocyte retrieval and immediately following, may serve to bind
and inactivate this factor. Second, albumin also serves to sequester any
additional substances which may have been synthesized as a result of OHSS.
Finally, the oncotic properties of albumin also serve to maintain intravascular
volume and prevent the ensuing effects of hypovolemia, ascites and

Following the ¬rst clinical trial by Asch et al. (1991) a series of publications
for and against the ef¬cacy of albumin were published (Table VII.16). Several
investigators have conducted prospective trials and agreed with their con-
clusion (Shoham et al., 1994; Shalev et al., 1995a, b). Other authors have
disagreed with the conclusion (Ng et al., 1995; Mukherje et al., 1995; Shaker
et al., 1996; Lewitt et al., 1996).

Cochrane Database Review of Albumin for the
Prevention of OHSS
In a Cochrane review of the use of intravenous albumin to prevent severe OHSS,
seven randomized controlled trials were identi¬ed (Aboulghar et al., 2002).
Five studies (Shoham et al., 1994; Shalev et al., 1995c; Isik et al., 1996;
Gokman et al., 2001; Ben-Chetrit et al., 2001) met the inclusion criteria
(Aboulghar et al., 2002). All were single-center, randomized controlled trials
with parallel design. In these ¬ve clinical trials, 378 patients were enrolled, of
which 193 were in the albumin-treated group and 185 in the control group
(Table VII.17). The meta-analysis of the ¬ve trials showed a signi¬cant reduction
in severe OHSS with an OR of 0.28 and 95% CI, 0.11“0.73. The RR was 0.35
(0.14“0.87) and the absolute RR was 5.5. Albumin infusion would save one case
of OHSS for every 18 women at risk for severe OHSS. The authors concluded
that intravenous albumin shows a clear bene¬t for the prevention of OHSS;
however, whether the number needed to treat would justify the routine use of
albumin in high risk patients must be judged by the clinical decision-makers and
future large randomized trials (Aboulghar et al., 2002).
A recent study, published after the Cochrane review, represents the
largest randomized controlled trial to date of albumin infusion vs. no treat-
ment in IVF patients who are at risk of developing moderate“severe OHSS
(Belver et al., 2003). Between March 1999 and February 2002, women
undergoing IVF with more than 20 retrieved oocytes were included. A total
of 988 patients were involved. Immediately after retrieval, patients were
allocated to one of two groups based on computer randomization. The ¬rst
group received intravenous albumin (40 g) and the second group received no
treatment. In women who developed moderate“severe (n ¼ 66) or severe
(n ¼ 46) OHSS, there was no difference based on prior albumin administration
between blood parameters or body weight on the day of oocyte retrieval or
seven days later. Furthermore, the number of patients with paracentesis,
hospital admissions, complications and days of OHSS until resolution did
not differ. The authors concluded that albumin infusion on the day of
oocyte retrieval is not a useful means for preventing the development of
moderate“severe OHSS.
It is my impression that if a new meta-analysis is performed taking
into account the more recently published trials, it is possible that the role
of intravenous albumin at the time of follicular aspiration for the prevention
of OHSS would be nonsigni¬cant.
Table VII.16 Intravenous albumin for the prevention of OHSS
Reproduced with permission from Delvigne and Rozenberg (2002b). Hum Reprod Update 8:559“77

OHSS incidence
Study Study design (n) Control group Risk patients with albumin use OHSS incidence in controls Comment

Asch et al. (1993) not controlled (36) historical 0% 80% in 21 patients no
E2* 46000 pg/ml
high-risk patients and 430 oocytes transfer occurred
Shoham et al. (1994) prospective placebo of NaCl 0/16 4/15 severe no information
E2 41906 pg/ml
randomized and multiple OHSS (p < 0.05) about moderate forms
controlled (31) follicular development
Shahata et al. (1994) retrospective (200) historical whole 0/104 8/96 only 18% of controls
E2 42997 pg/ml
IVF population and 420 oocytes had E2 42997 pg/ml
or 430 follicles
Ng et al. (1995) prospective placebo of 2/49 10/158 NS* albumin blunted the
E2 42724 pg/ml
controlled (207) ringer™s solution and 415 follicles severity of OHSS
Mukherjee et al. (1995) case report (2) “ 2 severe OHSS “ “
E2 44500 pg/ml
and 420 oocytes (1 early and 1 late)
Orvieto et al. (1995) case report (1) “ early severe OHSS “ “
E2 2293 pg/ml
and 46 oocytes
Ben-Rafael et al. (1995b) case report (1) “ early severe OHSS “ “
E2 42293 pg/ml,
435 oocytes
Halme et al. (1995) case report (1) “ 1 oocyte donor, early severe OHSS “ “
E2 2400 pg/ml,
15 oocytes

Shalev et al. (1995b) prospective no treatment 0/22 4/18 no transfer in 5.5% of
E2 42500 pg/ml
randomized (40) and 420 follicles controls and 13.6% of
study group
Shaker et al. (1996) prospective cryopreservation 4/13 moderate 3/13 moderate pregnancy signi¬cantly
E2 43540 pg/ml
randomized or 42745 pg/ml OHSS, no severe OHSS, higher in controls
controlled (26) and 415 oocytes not severe NS
Isik et al (1996) prospective no treatment 0/27 1/28 severe and “
E2 43000 pg/ml
randomized 4/28 moderate,
controlled (55) p < 0.05
Lewit et al. (1996) retrospective cases “ previous OHSS, 2/5 early severe, “ the most severe received
review (5) 2/5 moderate 75 g and had no transfer
E2 43600 pg/ml
and large number
of follicles
Orvieto and retrospective “ “ 2/30 early “ “
Ben-Rafael (1996) review (30) severe OHSS
Chen et al. (1997) prospective (72) historical controls 4/30; 0/16 14/42; p ¼ 0.047, prevention is effective in
E2 43600 pg/ml
and 420 oocytes not pregnant, 5/23 not pregnant, non-pregnancies and
4/14 pregnant 9/19 pregnant singleton pregnancies
Egbase et al. (1997) uncontrolled (31) “ 9.7% severe “ early follicular aspiration
E2 43269 pg/ml
and 412 follicles before hCG was also
412 mm per ovary performed (n ¼ 16)
Ndukwe et al. (1997) retrospective (60) “ 5/60 severe “ no preventive effect,
E2 44086 pg/ml
and 420 follicles (1 early, 4 late); especially in pregnant
8/60 moderate patients
Koike et al. (1999) prospective no treatment 420 oocytes 11/43 early, 15/55 early, “
randomized 2/43 late severe 6/55 late severe
controlled (98) OHSS OHSS
Panay et al. (1999) prospective no treatment 2/37 mild, 4/49 mild Pregnancy rate per
E2 43541 pg/ml
randomized (86) or 420 follicles 2/37 moderate cycle signi¬cantly
higher in controls
Costabile et al. (2000) prospective 200 mg/day 4/42 moderate 0/54 moderate high progesterone dose
E2 42452 pg/ml
randomized progesterone and 420 follicles (no severe) (no severe) is effective in preventing
controlled (96) from 1 day OHSS and better for
post-retrieval pregnancy rate
Gokmen et al. (2001) prospective placebo 0/85 severe and 4/83 severe and
E2 43000 pg/ml
randomized or 420 follicles 4/85 moderate 12/83 moderate
placebo (168) p < 0.05

* E2 ¼ 17b-estradiol
NS ¼ not signi¬cant


Hydroxyethyl Starch Solution Administration
Rizk (2002) reviewed the role of synthetic macromolecules used to prevent
OHSS and avoid the potential risks from using human products such as
albumin. Hydroxyethyl starch (HES) is a synthetic colloid, glycogen-like
polysacharride which is derived from amylopectin. It has been used as an
effective volume expander and is available in several molecular weights with
different chemical properties. Graf et al. (1997) used 1000 ml of 6% HES at the
time of oocyte retrieval and an additional 500 ml 48 h later in 100 patients
considered at high risk of OHSS (estradiol 4 11 000 pm/l and 20 follicles or
more). They compared the outcome to an historical control group of 82
patients without any prophylactic measures. Seven cases of severe OHSS and 32
cases of moderate OHSS occurred in the control group compared with two
cases of severe and 10 cases of moderate OHSS in the HES group (p < 0.05 and
p < 0.001, respectively).
Konig et al. (1998) in a prospective, randomized trial evaluated HES
and placebo in patients with estradiol levels 41500 pg/ml and with 10 follicles
on the day of hCG. The dose of HES was 1000 ml after the oocyte retrieval.
In the HES group, there was one case of moderate OHSS and no severe cases
compared with six cases of moderate OHSS and one case of severe OHSS in the
control group.
Gokman et al. (2001) performed a prospective randomized placebo-
controlled trial on 253 patients considered at risk of OHSS. They compared the
ef¬cacy of 500 ml of 6% HES (n ¼ 85) and 50 ml 20% human albumin (n ¼ 85)
or placebo (n ¼ 83). In high-risk patients whose estradiol was 43000 pg/ml, or
ultrasound showed more than 20 follicles, all the treatments were administered

Table VII.17 Intravenous albumin for the prevention of OHSS included in the
Cochrane database review
Reproduced with permission from Aboulghar and Monsour (2003). Hum Reprod Update 9:275“89

No. patients E2 level
on day No. No.
Albumin Albumin of hCG OHSS OHSS
Study Type Total (IV) Control dose (pg/ml) (albumin) (control)

Shoham et al. (1994) prospective randomized 31 16 15 50 g 1906 0 4
Shalev et al. (1995c) prospective randomized 40 22 18 20 g 0 4
Isik et al. (1996) prospective randomized 55 27 28 10 mg 0 4
Ben Chetrit et al. prospective randomized 87 46 41 50 g 2724 4 1
Ng et al. (1995) cohort controlled 207 49 158 50 g 2724 2 10
according ¸3600
Chen et al. (1997) prospective 72 30 42 4 14
historical control to BMI*

Odds ratio ¼ 0.42; 95% CI 0.21“0.88 (p ¼ 0.012)
* BMI ¼ body mass index

during oocyte retrieval. No severe OHSS cases were observed in the albumin
and HES groups while four patients developed severe OHSS in the placebo
group. Moderate OHSS was observed in four patients in the albumin group,
¬ve patients in the HES group and 12 patients in the placebo group (p < 0.05).
Gokman (2001) recommended HES for the prevention of OHSS, since it is as
ef¬cient as but safer and cheaper than albumin.
These three studies provide consistent results suggesting a bene¬cial effect
of HES in decreasing OHSS. Although the patient cohort was too small to draw
de¬nitive conclusions, these studies suggest that HES rather than albumin
should be further investigated (Delvigne and Rozenberg, 2002b).

Glucocorticoid Administration
Rizk (1993) has found no protective effect of intravenous glucocorticoid. The
pathophysiology suggests the involvement of an in¬‚ammatory mechanism
during the development of the ¬‚uid leakage associated with the syndrome.
Therefore, investigators hypothesized that glucocorticoids could possibly
prevent OHSS in patients at high risk. Tan et al. (1992) in a prospective
randomized trial investigated the usefulness of glucocorticoids in the reduction
of the rate of OHSS. Thirty-one patients, who were stimulated with hMG and
who were desensitized with GnRH agonists and developed more than 20
follicles 4 12 mm and/or had serum estradiol of 410 000 pmol/l on the day of
hCG administration, were recruited. The patients were randomized into two
groups. Group A (n ¼ 17) were administered intravenous hydrocortisone
immediately after transvaginal ultrasound oocyte recovery. Prednisolone 10 mg
three times daily was given for ¬ve days, starting on the day of oocyte recovery,
followed by prednisolone 10 mg twice daily for three days and 10 mg once daily
for two days. Group B (n ¼ 14) did not have any intravenous or oral gluco-
corticoid treatment. Luteal phase support was given in the form of intra-
muscular Gestone 100 mg/day. Seven of the 17 patients (41.2%) who received
glucocorticoids developed OHSS compared with six of the 14 patients (42.9%)
who did not. The authors concluded that the administration of glucocorticoids
to high-risk patients did not diminish the risk of developing OHSS.
In contrast, Lainas (2002) reported that methylprednisolone 16 mg/day,
starting on the 6th day of controlled ovarian hyperstimulation and tapered to
day 13 after embryo transfer, was effective in reducing OHSS signi¬cantly to
10%, compared with 43.9% in the control group.

Intramuscular Progesterone vs. Intravenous Albumin
Costabile et al. (2000) compared the use of intramuscular progesterone versus
intravenous albumin for the prevention of OHSS. Progesterone was admini-
stered at the time of oocyte retrieval and continued throughout the luteal phase.
The OHSS rates in this interesting study do not permit a conclusion to be
drawn. The authors suggested that progesterone appears likely to be safer than
albumin, with a possible bene¬t to pregnancy rates.


Does follicular aspiration protect against OHSS? The ¬rst case of severe OHSS
in an IVF program was reported by Friedman et al. (1984). The incidence of
moderate and severe OHSS in the major IVF series varied from 0.6% to 14%.
Rabinowitz et al. (1987) postulated that multiple follicular aspirations, which
empty most of the follicles of follicular ¬‚uid granulosa cells, may have a
protective effect against OHSS. Four of the 81 patients in the series who
developed OHSS during a cycle where egg retrieval was cancelled did not
develop the syndrome in cycles when multiple follicular punctures and
aspiration were performed.

Follicular Aspiration at the Time of Oocyte Retrieval
In a retrospective study, the incidence of OHSS was compared in IVF to
intrauterine insemination (IUI) patients who had undergone similar controlled
ovarian hyperstimulation (Aboulghar et al., 1991). At that point, the incidence
of OHSS was thought to be lower in those who had undergone IVF. This
was attributed to the effect of follicular aspiration. However, when the results
were analyzed according to the cause of infertility in 182 patients who
underwent follicular aspiration compared with 137 patients who did not,
it became apparent that OHSS was signi¬cantly higher in patients with PCOS
regardless of whether they had undergone IVF or IUI procedures (Aboulghar
et al., 1992). There was no difference in OHSS whether follicular aspiration
occurred or not in each group independently.

Follicular Aspiration Prior to hCG
In a prospective randomized study, Egbase et al. (1997) performed unilateral
follicular aspiration prior to administration of hCG for the prevention of severe
OHSS. The authors recruited 31 patients who were considered to be at risk of
OHSS. This was de¬ned by a serum estradiol of 412 000 pmol/l and with
a total of 412 follicles, each larger than 12 mm in diameter per ovary. These
patients were randomized into two groups: unilateral ovarian follicular
aspiration was performed in Group I patients (n ¼ 16) 6“8 h prior to hCG
administration under conscious sedation using midazolam 5 mg IV and
fentanyl 100 mg IV. In 15 patients (Group II), no aspiration was performed
prior to hCG. All patients in both groups had transvaginal ultrasound-guided
oocyte recovery 35“36 h after the administration of 10 000 IU of hCG, pro-
phylactic human albumin was administered to patients in all groups in the
form of 100 ml of 20% human albumin during oocyte recovery, and this was
repeated every 12 h for 48 h. The mean number of oocytes retrieved from
patients in Group I was signi¬cantly lower than from Group II (14.9 vs. 22.6,
p < 0.001). The incidence of OHSS was 25% in Group I and 33.3% in Group II.
Two patients in Group I and one patient in Group II developed severe OHSS.
The clinical pregnancy rate was 37.5% in Group I and 46.6% in Group II.

They concluded that unilateral follicular aspiration 6“8 h prior to hCG fails to
protect against the incidence of severe OHSS in women at risk. They postulated
that the hCG induced mediators of OHSS in the contralateral ovary, and they
may have reached too high a level of activity to be modi¬ed by intervention
in the ipsilateral ovary. They also noted that prophylactic human albumin
administration does not protect against the development of severe OHSS.

Follicular Aspiration of One Ovary 12 h after hCG
Tomazevic and Meden-Vrtovec (1996) investigated whether follicular aspira-
tion 12 h after hCG administration would prevent OHSS. In the ¬rst group,
106 patients at high risk of developing OHSS underwent follicular aspiration
of one ovary, 12 h after hCG administration. In the control group, 92 patients,

<< . .

( 30)

. . >>

Copyright Design by: Sunlight webdesign