undergoing assisted reproductive technology. It has been suggested that
transvaginal ovarian drilling has been effective in improving IVF results
in dif´¬ücult cases, and is less invasive and expensive when compared with
laparoscopic ovarian drilling (Ferraretti et al., 2001).
How Does Ovarian Drilling Work?
Stein and Leventhal (1935) were the ´¬ürst to perform ovarian wedge resection
for the treatment of anovulation and infertility associated with PCOS. Many
LAPAROSCOPIC OVARIAN DRILLING IN PCOS PATIENTS
Fig. VII.8: Laparoscopic ovarian drilling in PCOS patients
Reproduced with permission of Goldberg and Falcone (2000). Atlas of Endoscopic Techniques
in Gynecology. Edinburgh: Saunders, Chapter 13, p. 61
investigators postulated different mechanisms for resumption of ovulation
(Rizk and Nawar, 2001). Stein and Leventhal (1935) proposed that wedge
resection decreased the crowding of the ovarian cortex by cysts, allowing the
progression of normal follicles to the surface. Gjonnaess (1984) pioneered
the treatment of electrocautery by laparoscopy, and suggested that cautery
of the ovary destroyed a substance that inhibits ovulation. Gjonnaess and
Norman (1987) studied the endocrine effects of electrocautery in PCOS
patients and observed decreased androgen production with a subsequent
decline in LH secretion. Daniell and Miller (1989) performed laparoscopic laser
vaporization of the ovaries and postulated that removal of androgenic ´¬‚uid is
the underlying mechanism by which ovulation is enhanced. Kovacs et al. (1991)
observed a transient production of inhibin resulting in increased FSH. Balen
and Jacobs (1991) suggested that production of putative gonadotrophin
surge attenuating factor from the ovary might be restored after electrocautery.
A postoperative surge of LH and FSH followed by a normal pattern
of gonadotrophin secretion and gonadotrophin surge attenuating factor was
observed following surgery (Balen and Jacobs, 1991).
How Can You Perform Ovarian Drilling?
Stein and Leventhal (1935) pioneered wedge resection for the treatment of
infertility associated with amenorrhea, hyperandrogenism and polycystic
ovaries. Following the introduction of laparoscopy, a wide variety of techniques
have been used (Rizk and Nawar, 2001). Laparoscopic ovarian electrocautery
was pioneered by Gjonnaess in 1984. Since then, unipolar and bipolar
diathermy, as well as carbon dioxide (CO2), potassium titanyl phosphate
140 PREVENTION OF OVARIAN HYPERSTIMULATION SYNDROME
(KTP), neodymiumÔÇ“yttriumÔÇ“aluminumÔÇ“garnet (Nd:Yag) and argon lasers
have been utilized. The principles of microsurgery should be adhered to
during the use of any technique of ovarian drilling. The introduction of
microlaparoscopy has made the procedure less invasive, and therefore safer
(Rizk, 2001, 2002; Risquez et al., 1993; Almeida and Rizk, 1998; Rizk and
The technique is relatively straightforward: a triple puncture laparoscopy
is usually performed and the ovary is grasped by the ovarian ligament. A
variety of instruments have been used for many years with tremendous
variation in the literature regarding the three variables involved, which are
the number of ovarian punctures, the duration of diathermy and the wattage
used (Rizk and Nawar, 2001). Gjonnaess increased the number of punctures
to the ovary from three, initially, to four then ´¬üve then ´¬ünally eight. The
length of each diathermy was 3 s and the wattage used was 200ÔÇ“300 W
(Gjonnaess, 1984, 1994; Gjonnaess and Norman, 1987). Armar and Lachelin
(1993) performed 3ÔÇ“4 cauteries for each ovary, Kovacs (1998) cauterized the
ovary 10 or more times, Balen and Jacobs (1994) suggested that unilateral
cautery of one ovary at four sites, at 40 W for 4 s, resulted in ovulation in
three of four patients. Interestingly, ovulation occurred in both ovaries.
Naether and Fischer (1993) recommended adjusting the number of cauteries
according to the ovarian size and other clinical factors. Microlaparoscopic
scissors (2 mm) have been used in an attempt to minimize adhesion formation,
and the cautery size was 3ÔÇ“4 mm in diameter and 5 mm in depth (Almeida
and Rizk, 1998).
Daniell and Miller (1989) recommended 25 W continuous mode laser
to vaporize all the subcapsule follicles. They vaporized 20 to 40 sites in each
ovary. Ostrzenski (1992) performed laparoscopic CO2 laser ovarian wedge
resection with excellent results and minimal adhesion formation. Feste (1990)
used a sapphire tip at 25 W to perforate the ovarian stroma 15 to 20 times.
Classic vaporization of the ovarian capsule using an argon laser beam was
compared with simple perforation and subcapsule destruction in a Belgian
study by Verhelst et al. in 1993.
How Aggressive Should Ovarian Drilling Be?
Different authors have differing opinions on the extent of destruction necessary
to desensitize PCOS to exogenous gonadotrophins. While Balen (1999) sugges-
ted that minimal destruction is necessary to sensitize the ovaries, Rimington
(1997) found that, in order to avoid OHSS, a considerable amount of healthy
ovarian tissue destruction is required.
Complications of Ovarian Drilling
The risk of ovarian adhesions and, rarely, ovarian atrophy are of concern
(Adashi et al., 1981). This is particularly true if a macrosurgical rather than a
microsurgical approach is used. The need for general anesthesia for standard
LAPAROSCOPIC OVARIAN DRILLING IN PCOS PATIENTS
laparoscopy is another drawback. Most recently, Almeida and Rizk (1998)
reported the ´¬ürst case of microlaparoscopic ovarian drilling under local
anesthesia in a patient with PCOS. This minimally invasive procedure might be
the way to proceed in the future.
Laparoscopic Ovarian Drilling in Non-IVF
Rizk and Nawar (2001) reviewed the treatment of PCOS by ovarian drilling
using electrosurgery (Table VII.4) and laser treatment (Table VII.5). Similar
ovulation and pregnancy rates were observed. The pregnancy rates varied from
20% to 60%.
Fukaya et al. (1995) studied 26 infertile patients with PCOS who previously
had OHSS after ovarian stimulation with hMG and who failed to conceive.
All patients were treated by KTP and Nd:Yag laser. After laser vaporization,
spontaneous ovulation was con´¬ürmed in six patients. In the remaining
20 patients not ovulating spontaneously after laser vaporization, ovulation
was successfully induced by using clomiphene citrate (three patients) or hMG
(17 patients), in contrast to the dif´¬ücult ovulation induction prior to laser
treatment. Nineteen clinical pregnancies (73%) were con´¬ürmed after laser
treatment. Of the patients treated with hMG, mild OHSS occurred in only
three patients. Fukaya et al. (1995) concluded that laser vaporization of the
ovaries is promising for the prevention of OHSS in patients who previously had
Table VII.4 Ovulation and pregnancy rates following electrocautery
treatment of PCOS
Reproduced with permission from Rizk and Nawar (2001).
Laparoscopic ovarian drilling for surgical induction of ovulation in
polycystic ovarian syndrome. In (Allahbadia G, Ed.), Manual of
Ovulation Induction. Rotunda Medical Technologies, Mumbai, India,
Chapter 18, pp. 140ÔÇ“4
Author Year No. of women treated Ovulation Pregnancy
Farhi et al. 1995 22 41%
Gjoannaess 1994 252 201 (92%) 145 (58%)
Gjoannaess 1984 65 57 (88%) 24 (37%)
Naether et al. 1993 133 73 (55%)
Naether and Fisher 1993 199
Tutinen et al. 1993 10 3 (30%) 2 (20%)
Kovacs et al. 1991 10 7 (70%) 3 (30%)
Abdel Gadir et al. 1990a 29 71% 10 (34%)
Rizk and Abdalla 2000 7 86% 58%
142 PREVENTION OF OVARIAN HYPERSTIMULATION SYNDROME
Table VII.5 Ovulation and pregnancy rates following laser treatment
Reproduced with permission from Rizk and Nawar (2001). Laparoscopic
ovarian drilling for surgical induction of ovulation in polycystic ovarian
syndrome. In (Allahbadia G, Ed.), Manual of Ovulation Induction.
Mumbai, India, Rotunda Medical Technologies, Chapter 18, pp. 140ÔÇ“4
Author Year No. of women treated Ovulation Pregnancy
Heylen et al. 1994 44 35 (80%) 24 (55%)
Verhelst et al. 1993 17 14 (82%) 11 (65%)
Ostrzenski 1992 12 75%
Gurgan et al. 1992 40 20 (50%)
Keckstein et al. 1990 19 8 (42%)
11 60 (70%) 3 (27%)
Daniell and Miller 1989 85 48 (56%)
Abdel Gadir et al. (1990a, b) compared gonadotrophin therapy with
ovarian drilling in PCOS patients. The pregnancy rates were similar but the
miscarriage rate was lower in PCOS patients.
Farquhar et al. (2001) reviewed 15 clinical trials addressing laparoscopic
drilling by diathermy or laser for ovulation induction in anovulatory PCOS.
The main outcomes of the meta-analysis were ovulation and pregnancy rates.
The miscarriage rate, multiple pregnancy rate and OHSS were secondary out-
comes. The pregnancy rate following ovarian drilling compared with gonado-
trophins differed according to the length of follow-up. Overall the pooled odds
ratio (OR) was not statistically signi´¬ücant (OR ┬╝ 1.27; 95% CI, 0.77ÔÇ“1.98).
Multiple pregnancy rates were reduced in the ovarian drilling arms of the four
trials where there was a direct comparison to gonadotrophins (OR ┬╝ 0.16; 95%
CI, 0.03ÔÇ“0.98). There was no difference in the miscarriage rates in the drilling
group when compared to gonadotrophins in these trials (OR ┬╝ 0.61; 95% CI,
0.17ÔÇ“2.16). Farquhar et al. (2001) concluded that there is insuf´¬ücient evidence
that there is a difference in cumulative pregnancy rates between laparoscopic
ovarian drilling after 6ÔÇ“12 months follow-up and 3ÔÇ“6 cycles of gonadotrophins
as a primary treatment for subfertile patients with anovulation and PCOS.
Multiple pregnancy rates are considerably reduced in women who conceive
following laparoscopic ovarian drilling.
Laparoscopic Ovarian Drilling in PCOS Patients Before IVF
Laparoscopic ovarian drilling has been used for treating patients suffering from
PCOS before starting their IVF stimulation (Rimington et al., 1997; Egbase
GnRH ANTAGONIST AS AN ALTERNATIVE TO THE LONG AGONIST PROTOCOL
et al., 1998; Tozer et al., 2001). The laparoscopic drilling in three trials was
performed on both ovaries and on only one ovary in one trial. Electrocautery
and laser vaporization were used.
Rimington et al. (1997) performed a prospective randomized study
on 50 patients undergoing IVF. Patients who failed to become pregnant in
the previous trial, or whose cycle had been cancelled because of high OHSS
risk, were randomized in two groups. In the ´¬ürst group, 25 women were
stimulated by long protocol pituitary desensitization followed by gonadotro-
phin stimulation and IVF. In the second group, 25 patients underwent
laparoscopic cautery after pituitary desensitization followed by gonadotrophin
stimulation and IVF. The pregnancy and miscarriage rates were identical;
however, ´¬üve cycles were cancelled because of OHSS risk compared to none
in the laparoscopic-electrocautery group. Moderate OHSS occurred in four
patients in the ´¬ürst group compared to one patient in the second group.
Egbase et al. (1998) performed unilateral ovarian diathermy on three
patients with clinical and sonographic features of PCOS who previously had
severe OHSS requiring hospitalization. Two of the patients conceived and none
Tozer et al. (2001) performed a retrospective comparative study on
31 women with clomiphene-resistant PCOS. In 15 women (Group A), 22 cycles
of IVF were performed preceded by laparoscopic ovarian diathermy. In
16 patients (Group B), 25 cycles of IVF were performed. The incidence of severe
OHSS was higher in Group B than Group A (4.2% vs. 0%). There was also a
trend for a lower miscarriage rate and higher pregnancy rate in the laparascopic
ovarian diathermy group.
GnRH ANTAGONIST AS AN ALTERNATIVE
TO THE LONG AGONIST PROTOCOL
Modern GnRH antagonists such as Cetrorelix and Ganirelix reliably prevent
premature LH surges in controlled ovarian hyperstimulation for assisted
reproduction (European Orgalutran Study Group, 2000; EuropeanÔÇ“Middle
East Orgalutron Study Group, 2001; Fluker et al., 2001; Felberbaum and
Diedrich, 2003). GnRH antagonists have helped to overcome some major
disadvantages of GnRH agonists, especially of the long protocol, which is
currently the standard for ovarian stimulation. However, several studies have
indicated a slight reduction in pregnancy rate compared to the GnRH agonist.
If the GnRH antagonist is associated with a lower risk of OHSS, that could be
a signi´¬ücant advantage (Howles, 2002). In three clinical trials, the incidence
of OHSS was reduced using a GnRH antagonist compared with a GnRH
agonist long luteal protocol (Albano et al., 2000; Ludwig et al., 2000; Olivennes
et al., 2000). In a Cochrane review (Table VII.6), the ef´¬ücacy of GnRH
antagonists was compared to the long agonist protocol in assisted conception
(Al-Inany and Aboulghar, 2002). At the time of the review, ´¬üve published
randomized controlled trials ful´¬ülled the inclusion criteria. In four studies,
144 PREVENTION OF OVARIAN HYPERSTIMULATION SYNDROME
Table VII.6 Meta-analysis comparing GnRH anatagonist vs.
GnRH agonist co-treatment for ovarian hyperstimulation for IVF
Reproduced with permission from Al-Inany and Aboulghar (2002).
Cochrane Review. Hum Reprod 17:874ÔÇ“85
Odds ratio 95% con´¬üdence interval
Prevention of premature LH surge 1.8 0.8ÔÇ“4.2
Clinical pregnancy rate/cycle 0.8 0.6ÔÇ“0.99*
Miscarriage rate 1.0 0.5ÔÇ“2.0
Duration of ovarian hyperstimulation 1.1 ├Ç1.5ÔÇ“├Ç0.8
Amount of gonadotropins used ├Ç3.3 ├Ç5.2ÔÇ“├Ç1.5*
OHSS 0.5 0.2ÔÇ“1.2
* Statistically signi´¬ücant difference
the multiple low-dose 0.25 mg antagonist regimen was used, and in one study a
single high-dose (3 mg) antagonist regimen was investigated. In all ´¬üve trials,
reference treatment included a long protocol of GnRH agonist (buserelin,
leuprorelin or triptorelin) starting in the mid-luteal phase of the preceding
cycle. In comparison with the long GnRH agonist protocol, the overall odds
ratio for the prevention of premature surge was 1.76 (95% CI, 0.75ÔÇ“4.16),
which was not statistically signi´¬ücant (Table VII.6). There were signi´¬ücantly
fewer pregnancies in the GnRH antagonist group (OR ┬╝ 0.79; 95%CI, 0.63ÔÇ“
0.99). There was no statistically signi´¬ücant reduction in the occurrence of severe
OHSS (relative risk (RR) ┬╝ 0.50; OR ┬╝ 0.79; 95% CI, 0.22ÔÇ“1.18). It is interes-
ting, however, that, except for the study by Ludwig et al. (2000), which was not
included in this meta-analysis, none of the ´¬üve studies was primarily designed
to determine the difference in OHSS occurrence between the two protocols.
In this prospective randomized study, there were 85 patients and 188 patients
for stimulation in the GnRH agonist and the GnRH antagonist groups
respectively. The incidence of OHSS was signi´¬ücantly lower in the Cetrorelix
group vs. the buserelin group: 1.1% vs. 6.5% respectively (Ludwig et al., 2000).
Furthermore, the patients studied were normo-responders and, in two studies,
PCOS patients were excluded (Albano et al., 2000; Olivennes et al., 2000).
Ludwig et al. (2001) performed a meta-analysis to evaluate whether there is
a reduction in OHSS and/or pregnancy rates associated with the use of two
GnRH antagonists, Cetrorelix and Ganirelix. A signi´¬ücant reduction of OHSS
was observed in the Cetrorelix studies, (OR ┬╝ 0.2; 95% CI, 0.10ÔÇ“0.54) but
no reduction for Ganirelix (OR ┬╝ 1.13; 95% CI, 0.24ÔÇ“5.31). The incidences of
OHSS degree III cases was reduced in the Cetrorelix protocols as compared with
the long protocol to a nearly signi´¬ücant degree (OR ┬╝ 0.26; 95% CI, 0.07ÔÇ“1.0).
Ganirelix did not reduce the incidence of OHSS degree III at all (OR ┬╝ 1.08;
95% CI, 0.27ÔÇ“4.38). The pregnancy rate in the Cetrorelix studies was not
signi´¬ücantly different from the long GnRH agonist protocol (OR ┬╝ 0.91; 95%
CI, 0.68ÔÇ“1.22). The pregnancy rate in the Ganirelix protocols was signi´¬ücantly
SINGLE EMBRYO TRANSFER
lower compared to the long GnRH agonist protocol (OR ┬╝ 0.76; 95% CI,
0.59ÔÇ“0.98). Ludwig et al. (2001) concluded that Cetrorelix and not Ganirelix will
reduce the incidence of OHSS, and that Cetrorelix not Ganirelix will result in
the same pregnancy rates as the long GnRH agonist protocol. At this point
I believe that the ´¬ünal word has not been said in relation to the development
of OHSS in GnRH antagonist cycles, and that further studies, some of which
have recently been published, will clarify this situation (Ragni et al., 2005).
NATURAL CYCLE IVF
IVF without ovarian stimulation avoids the potential complications of OHSS.
However, premature LH surge, cycle cancellation and low pregnancy rates have
decreased its popularity. The development of GnRH antagonists has revived
the natural cycle IVF by preventing premature LH surges (Figure VII.9;
SINGLE EMBRYO TRANSFER
The incidence of OHSS is directly related to hCG levels, which in turn is also
increased in multiple pregnancies. The severity of OHSS is enhanced by the
number of gestational sacs (Mathur et al., 2000). Avoiding multiple pregnancies
is a legitimate way to prevent the occurrence of late, severe OHSS. This could in
fact be achieved by a single-embryo transfer policy (Gerris and Van Royen,
2000; Gerris et al., 2002, 2004). Transferring a single zona-free day-5 blastocyst
allows further time for observation, and offers the patient with moderate OHSS
an optimal chance for a singleton pregnancy, while avoiding the complications
of OHSS (Kinget, 2002).
Fig. VII.9: Cetrorelix single dose in a natural cycle with minimal stimulation
Reproduced with permission from Olivennes F, Fanchin R, Bertrand C et al. (2001).
GnRH antagonist in single dose applications. Infertility and Reproductive Medicine Clinics
of North America 12:119ÔÇ“28. Philadelphia: W.B. Saunders Company
146 PREVENTION OF OVARIAN HYPERSTIMULATION SYNDROME
In-vitro Maturation of Oocytes
The ´¬ürst in-vitro maturation of human oocytes and the ´¬ürst fertilization of an
in-vitro matured (IVM) oocyte were reported by Edwards (1965) and Edwards
et al. (1969), respectively. Today, more than 300 babies have been reported after
IVM followed by assisted reproduction treatment, and more than 100 healthy
babies from unstimulated cycles in PCOS patients (Mikkelsen, 2004; LeDu
et al., 2005) (Table VII.7). Successful in vitro culture of immature oocytes could
possibly replace controlled ovarian hyperstimulation. Since ovarian stimula-
tion is not required for IVM, this approach is both safer and cheaper for
women with PCOS. The indications for IVM include women requiring IVF
who have PCOS, those with poor-quality embryos in repeated IVF cycles for
no apparent reason and poor responders to high gonadotrophin stimulation
(Tan and Child, 2002). The pregnancy rate following IVM is correlated with the
number of antral follicles present, the peak ovarian stromal blood ´¬‚ow velocity
at the baseline ultrasound scan, the number of immature oocytes retrieved,
the absence of a dominant follicle and endometrial thickness at embryo transfer
(Tan and Child, 2002).
The ´¬ürst birth was achieved by Veeck et al. (1983), although in this case the
immature oocytes were retrieved during a cycle in which ovarian stimulation was
used. Other teams have reported live birth following IVM of oocytes retrieved
from non-stimulated cycles (Cha et al., 1991; Trounson et al., 1994; Barnes et al.,
1995; Chian and Tan, 2002). A triplet birth after donation of oocytes obtained
by aspiration of follicles recovered at the time of oophorectomy during a
Cesarean section was reported by Cha et al. (1991). The ´¬ürst pregnancy achieved
in PCOS patients from immature oocytes retrieved by puncture during
non-stimulated cycles was reported by Trounson et al. (1994). LeDu et al.
(2005) reported the French experience of IVM in which 45 cycles of IVM was
performed in 33 PCOS patients (Table VII.7). A total of 509 oocyte-cumulus
complexes were obtained; 276 oocytes matured in 24 h and 45 in 48 h. The
normal fertilization rate for oocytes matured in 24 h and 48 h was 69.5% and
73.3%, respectively. The clinical pregnancy rate was 20% per retrieval and 22.5%
per transfer. The role of IVM in the management of patients with PCOS is
increasing, and deserves further investigation in the context of randomized
Is There an Ideal Protocol for Prevention of OHSS?
Rizk (2002) proposed the following protocol as an ideal protocol for the
prevention of OHSS (Table VII.8). I suggest that, to minimize the risk of OHSS,
pretreatment with OCP and low-dose gonadotrophin is initiated, GnRH
antagonist used to prevent LH surge, GnRH agonist used to trigger ovulation