<< . .

( 30)

. . >>

and progesterone for luteal phase support. I emphasize that no comparison of
pregnancy rates has been performed between this protocol and more
conventional protocols. Simply put, you may see a drop in pregnancy rate if
you adopt any or all of the above. However, it would be estimated that the

Table VII.7 Biological data and birth reports after in-vitro maturation from unstimulated cycles
in polycystic ovarian syndrome patients
Reproduced with permission from Le Du et al. (2005). Hum Reprod 20:420“4.

Immature oocytes Maturation Fertilization Transfers Pregnancies

Cycles Total Mean Total % Total % (n) Embryos Biochemical Ongoing Births

Trounson et al. 9 308 13.4 169 60.4 41 13 17 1 1
Barnes et al. 3 57.6 62 2 4 1 1
Barnes et al. 9 165 16.5 102 60.0 27 26 0
Cha and Chan 832 499 60.0 364 73 64 306 16 16
Chian et al. 3 17 13 76.5 10 77 4 10 2 2 4
Chian et al. 25 249 209 84.0 10 5
Chian et al. 24 183 142 77.6 125 88 24 63 8 6 6
Cha et al. (2000) 94 1139 13.6 708 62.0 481 75 85 416 23 20
Mikkelsen and 12 81 36 44 25 69 30 53 7 0
Lindenberg (2001)
Child et al. (2001) 68 11.3 79 67 217 20 10
Chian et al. (2001) 1 63 22 15 1 3 2
Abdul-Jalil et al. 1 12 6 50 4 67 1 3 2
Child et al. (2002) 107 1102 10.3 834 75 652 78 107 28 23 17
Nagele et al. 1 16 11 7 64 1 3 1 1
Son et al. (2002) 1 61 40 65.6 38 95 1 3 2 2
Kyono et al. 1 12 12 1 3 2 2
Lin et al. (2003) 33 762 23.1 548 71.9 383 69.5 33 125 12 12
Le Du (2005) 45 509 11.4 321 63.0 70 40 103 11 6 5 (þ1)

Table VII.8 Is there an ideal protocol for the prevention of OHSS?

Pretreatment with oral contraceptive pills
Low-dose gonadotropins
GnRH antagonist to prevent LH surge
GnRH agonist to trigger ovulation
Progesterone for luteal phase support

incidence of severe OHSS would decrease from 1%“2% to 0.5%“0.8%. It has to
be mentioned that with any protocol there is still a 0.5% chance of OHSS unless
a more liberal policy for cancellation is adopted.

Preventive Measures During Controlled Ovarian Stimulation

Withholding hCG
Rizk and Aboulghar (1991) found that withholding hCG was the most
commonly used method of preventing OHSS in patients predicted to be at high
risk of developing the syndrome. Strict criteria for withholding hCG will lower
and possibly abolish the incidence of OHSS, but at the expense of canceling the
cycle. Today, cancelation of the IVF cycle is the least-favorite approach among
clinicians except for extreme situations. Rizk and Nawar (2004) highlighted
that the cancelation of the IVF cycle creates a frustrating situation for both
the physician and the patient. In ovulation induction cycles when GnRH
antagonists or agonists are not used, Delvigne and Rozenberg (2000b) advise
caution since a spontaneous LH peak may occur, resulting in a pregnancy that
may be associated with OHSS complications.

At What Point Should hCG Be Withheld?
The serum estradiol levels above which hCG should be withheld vary widely
among different centers (Rizk et al., 1991a, b). Schenker and Weinstein (1978)
withheld hCG when serum estradiol levels exceeded 800 pg/ml. Blankstein et al.
(1987) suggested 1700 pg/ml, and Haning et al. (1983) accepted 4000 pg/ml as
the upper limit.

Withholding hCG and Continuation of GnRH Agonist
or Antagonist
Forman et al. (1990) suggested that hCG should be withheld if the serum
estradiol level exceeded 2000 pg/ml in association with a total of more than 15
follicles, each more than 12 mm in mean diameter. GnRH-a was continued and
hMG commenced at a lower dosage after a further period of desensitization.
An interesting case report, where the patient was at high risk of severe OHSS,
was treated by a high dose of GnRH antagonist in addition to discontinuation
of gonadotrophins and withholding hCG (de Jong et al., 1998).
I and others believe that, when guidelines for withholding hCG are
suggested, more than one parameter should be considered, namely the presence
of polycystic ovaries on ultrasound, occurrence of OHSS in previous cycles,
serum estradiol level of ¸3500 pg/ml, the slope of the estradiol rise and the
presence of 20 or more follicles (Rizk and Aboulghar, 1991, 1999, 2005;
Aboulghar and Mansour, 2003).

Coasting or Delaying hCG Administration
Coasting introduced a new philosophy in ovarian stimulation. It has
demonstrated that you can switch off the exogenous gonadotrophin supply
without an immediate detrimental effect on follicular development and
granulosa cell function. Withholding hMG and delaying hCG has been
attempted for two decades. Serum estradiol levels at the time of ovulation
triggering has been thought of as a predictor of the risk of developing OHSS.
It has therefore been proposed to postpone hCG administration to allow the
serum estradiol levels to drop below a certain threshold. This has been termed
˜˜coasting™™ or ˜˜controlled drift period.™™
Coasting has been employed in ovulation induction since the late 1980s
and early 1990s (Rizk, in press; Rabinovici et al., 1987; Urman et al., 1992).
Shortly afterwards, coasting was used to prevent severe OHSS in IVF cycles (Sher
et al., 1993). Currently, coasting is the most popular method among physicians
to prevent OHSS (Delvigne and Rozenberg, 2001). More than 15 studies
(Tables VII.9“VII.11) have been published and several reviews have critically
evaluated the effect of coasting on OHSS (Delvigne and Rozenberg, 2002a, b;
Aboulghar and Mansour, 2003; Rizk and Aboulghar, 2005). The differences
between the American approach and the European approach for prevention of
OHSS has generated scienti¬c discussion (Rizk, in press). It is generally
perceived that the American approach is more aggressive, using a higher dose of
gonadotrophins, retrieving a large number of oocytes, and replacing a larger
number of embryos, resulting in a signi¬cantly higher number of multiple
pregnancies that has attracted signi¬cant media attention with serious criticism
(Adamson et al., 2005). That approach is changing but is still very tardy com-
pared with Europe, and is at least a decade behind Europe. However, coasting
will continue to play a major role in both the United States and Europe.

Coasting is currently the most popular method for prevention of OHSS because
of many advantages (Delvigne et al., 2001; Delvigne and Rozenberg, 2002a, b;
Grace et al., 2005). First, this method has the great attraction of the cycle not
being cancelled. Second, fresh embryos are still being transferred in contrast to
cryopreservation. Finally, no additional gonadotrophins or medications are
used (Delvigne, 2004).

It is well-established that high estradiol levels are associated with OHSS;
however, it is very unlikely that the high estradiol levels are the cause of OHSS.
Therefore, reduction of the estradiol levels themselves is not the main goal
of coasting (Dhont et al., 1998; Aboulghar and Mansour, 2003). Coasting may
diminish the functional granulosa cell cohort, resulting in the gradual decline
in the circulating estradiol levels but, more importantly, reduction of the
chemical mediators that augment capillary permeability as vascular endothelial
Table VII.9 Population characteristics of the study, design and selection criteria used for coasting
Reproduced with permission from Delvigne & Rozenberg (2002b). Hum Rep Update 8:291“6

Study (n) Mean age (years) + SD (range) Design/control group(s)
Selection criteria E2 (pg/ml), additional criteria

Sher et al. (1995) (51) 37.3 (28“42) descriptive/NA*
E2 43000, follicle number 429
and 30% follicles ¸ 15 mm
Benadiva et al. (1997) (22) 34.5 + 3.6 retrospective/cryopreserved patients
E2 ¸3000
Tortoriello et al. (1998) (44) 32.6 + 0.7 retrospective/subgroup of coasted
E2 43000, follicle number ¸ 5
¸ 16 mm and two follicles ¸19 mm vs. two control groups
Dhont et al. (1998) (120) NA retrospective/historical cohort
E2 ¸ 2500, follicle number ¸ 20
Lee et al. (1998) (20) NA retrospective/IVF patients
E2 ¸ 2777 and many immature follicles
of which less than three follicles 418 mm
Fluker et al. (1999) (63) 32.2 + NA descriptive/NA
E2 rose rapidly and generally 43000
Waldenstrom et al. (1999) (65) 31.5 (23“39) descriptive/NA
˜very high™ E2 and 425 ˜large™ follicles
of which the three largest ¸ 17 mm
Egbase et al. (1999) (15) 33.5 + 2.8 prospective randomized early
E2 4 6000 and 415 follicles/ovary and
two or more 4 18 mm follicular aspiration
Dechaud et al. (2000) (14) NA descriptive/NA
E2 ¸5000 and/or 420 follicles of which
three or more follicles ¸ 18 mm without
abdominal pain
Ohata et al. (2000) (5) 32 (25“37) ¸ 30% follicles ¸ 16 mm and severe descriptive/NA
OHSS in a previous cycle
Aboulghar et al. (2000) (24) 29.9 + 4.6 retrospective/historical group
E2 43000 and 420 follicles with a
dominant follicle ¸ 16 mm
Al-Shawaf et al. (2001) (50) 32.5 + 4.5 (23“41) prospective/observational normal cycle
420 follicles and 25% ¸ 15 mm, E2 43596

* NA ¼ not available
Table VII.10 IVF data for the coasted cycle. The E2 data are in pg/ml. Means (ranges) or + SD are given
Reproduced with permission from Delvigne Rozenberg (2002). Hum Rep Update 8:291“6

Fertilization Pregnancy
Study Coasting duration Oocytes rate (%) rate (%)**
E2 day of coasting E2 day of hCG retrieved ÁE2

Sher et al. (1995) NA* 2163 (560“2920) 5487 6.1 (3“11) 21 69 41a
Benadiva et al. (1997) 3803 + 731 2206 + 731 1597 1.9 + 0.9 15 + 6.5 62.2 63.6b
Tortoriello et al. (1998) 4015 + 112 2407 + 130 2475 2.6 + 0.3 15.8 + 1.2 59.8 44.45a
Dhont et al. (1998) 3834 + 872 2348 + 472.2 1486 1.94 + 0.8 19.7 + 0.6 NA 37.5b
Lee et al. (1998) NA NA NA 2.8 + 1.3 NA 63 40c
Fluker et al. (1999) NA 2832 + 129 2245 5.3 + 0.2 10.8 + 0.5 71 36.5b
Waldenstrom et al. (1999) 46483 (3541“7764) 1569 (472“2507) 4576 4.3 (3“6) 10 (3“21) 61 42b
4471 (2821“7353)
Egbase et al. (1999) 10 055 + 965 1410 + 246 NA 4.9 + 1.6 9.6 + 3.2 58.4 + 2.1 33.3
Dechaud et al. (2000) 5761 3596 NA 1.6 (1“3) 15 36.7 30a
Ohata et al. (2000) NA 1242.6 (425“1800) NA 4 (3“6) 9.2 (6“15) NA 20b
Aboulghar et al. (2000) 7150 + 1050 4640 + 1100 NA 2.92 + 0.92 16 + 3.5 59 35a
Al-Shawaf et al. (2001) NA NA NA 3.4 + 1.6 11.0 + 5.5 (0“22) 55.1 40b

* NA ¼ not available
Pregnancy rate/retrieval
Pregnancy rate/cycle
Pregnancy rate/embryo transfer

Table VII.11 Clinical description of registered OHSS cases
Reproduced with permission from Delvigne & Rozenberg (2002). Hum Rep Update 8:291“6

Study (no. coasted patients) No. patients with ascites Hemoconcentration Comment/OHSS classi¬cation

Sher et al. (1995) (51) 12 0
Benadiva et al. (1997) (22) NA* NA 1 moderate/classi¬cation not de¬ned
Tortoriello et al. (1998) (44) 6 (5 clinically and 3
1 at ultrasound)
Dhont et al. (1998) (120) NA 1 5.8% moderate (involving ascites) and severe OHSS
Lee et al. (1998) (20) 4 (2 paracentesis) NA 4 severe OHSS (distress with ovarian enlargement and ascites)
Fluker et al. (1999) (63) 1 1 Cumulated results of two groups (classical and modi¬ed coasting n ¼ 93);
9/93 had nausea and vomiting, 2 had ascites
Waldenstrom et al. (1999) (65) <300 ml: 6/61 300“800 ml: 2/61 1 paracentesis
3/61 4800 ml: 2/61
Egbase et al. (1999) (15) 3 NA 3 additional cases of moderate OHSS when using the classi¬cation of Schenker
Dechaud et al. (2000) (14) NA 0/10 refers only to severe forms of OHSS
Ohata et al. (2000) (5) 5 0 ascites at ultrasound
Aboulghar et al. (2000) (24) 4 0 ascites at ultrasound (moderate according to the classi¬cation of Golan)
Al-Shawaf et al. (2001) (50) 1**/50 NA 2 moderate according to the classi¬cation of Navot
Total 46/283 7/378
% 16.3 2.8

* NA ¼ not available
** This patient was excluded because of a protocol violation

growth factor (VEGF) (Rizk et al., 1997). In an interesting study from Spain,
Garcia-Velasco et al. (2004) suggested that coasting acts through down-
regulation of VEGF gene expression and protein secretion. The fact that
medium and small follicles are more likely to undergo atretic changes is of
crucial relevance in both steroid and vasoactive mediator secretion. They also
observed that a signi¬cantly higher percentage of granulosa lutein cells become
apoptotic after coasting. This difference is even greater for immature follicles.
During coasting, estradiol levels initially increased because dominant follicles
may continue their growth despite the lack of hormonal stimulus, whereas
intermediate follicles may undergo atresia.
The characteristics of granulosa cells in the follicles of women undergoing
coasting in controlled ovarian stimulation for IVF has recently been charac-
terized (Tozer et al., 2004a). The effect of withholding gonadotrophins
during controlled ovarian stimulation in women at risk of developing OHSS
was recently evaluated (Tozer et al., 2004b). Individual follicles of variable sizes
were assessed in relation to the granulosa cell number, oocyte retrieved,
fertilization and embryo quality. The authors acknowledged that the
ideal control group of women would be those identi¬ed to be at risk of
developing OHSS but not coasted. However, this was not ethically possible
since coasting has been successfully used to prevent severe OHSS in their unit
for several years (Al-Shawaf et al., 2001). The control group was selected from
optimally responsive women that excluded all poor and hyper-responders.
Twenty-two women who had been coasted and the control group of optimally
responding women were studied in detail. At the time of oocyte retrieval,
the follicular ¬‚uid from four to six individual follicles of different sizes was
collected for VEGF analysis. The authors observed wide variations of follicular
¬‚uid levels of VEGF in follicles of the same size, both in different patients
and in the same patient, which re¬‚ects the unique and individual composition
of each follicular environment. Despite these wide variations, VEGF levels in
the follicular ¬‚uid of follicles in the coasted group were constantly lower than
the VEGF follicular ¬‚uid levels in the control group.
VEGF concentration in follicular ¬‚uid may depend on the quality and
number of granulosa cells (Van Blerkom et al., 1997). As discussed above, a
negative correlation was observed between follicular ¬‚uid VEGF and granulosa
cell number, which was independent of follicle size (Tozer et al., 2004b).
Greater granulosa cell numbers have been associated with more competent
follicles (McNatty et al., 1979) and lower follicular ¬‚uid VEGF levels with more
oocytes (Friedman et al., 1998) and better embryo quality (Barroso et al., 1999).
Tozer et al. (2004b) suggested that this correlation, which was more signi¬cant
in the coasted group, may be due to the differential effect of gonadotrophin
withdrawal on individual follicles in favor of those follicles with greater number
of granulosa cells and/or that are more competent. The study did not con¬rm
or refute VEGF as the cornerstone of OHSS pathophysiology, but established
that VEGF follicular ¬‚uid concentrations in highly responsive women who
have undergone coasting are signi¬cantly lower than in the control group of
women studied.

There is no question about the popularity of coasting in Europe and the
United States. Delvigne and Rozenberg (2001) assessed whether physicians
would modify their preventive attitude in relation to clinical factors and to
the estradiol response. They constructed case scenarios with three levels of risk
factors for OHSS. At random, three out of the 12 arti¬cially constructed case
scenarios were sent to 573 physicians who are members of the European Society
for Human Reproduction and Embryology (ESHRE). Among the selected
preventive measures, coasting was by far the most popular choice (60%),
followed by the use of intravenous albumin; or hydroxyethyl starch solution
(36%) and cryopreservation of all embryos (33%).

Several authors have reported successful reduction of severe OHSS by delaying
hCG or coasting. Rabinovici et al. (1987) from Israel were the ¬rst to
report their experience with rescue of 12 gonadotrophin-induced cycles that
were liable to develop hyperstimulation. Treatment with hMG was stopped
in 12 patients who either had overt biochemical overstimulation or were at
an increased risk of hyperstimulation. The duration of the pause in treatment
ranged from 2 to 10 days. In nine patients, including the six who were
overstimulated, the plasma estradiol levels declined despite the continuing
growth of most follicles. None of these patients conceived following hCG.
The pregnancies occurred in three patients whose estradiol levels continued
to rise until the day of hCG. They therefore concluded that, although rescue of
the overstimulated cycles is sometimes possible, resulting conceptions seem
to be associated with a continuing rise of estradiol during the period of
treatment pause.
Urman et al. (1992) from Canada studied 40 cycles in 32 patients with
PCOS. The authors used a controlled drift period to avoid cancellation. The
clinical pregnancy rate per cycle was 25% (10 out of 40). OHSS occurred in
2.5% (1 out of 40). The authors did not share the same conclusion about the
relation between pregnancy and the rise of estradiol suggested by Rabinovici
and colleagues.
A more recent Canadian study was published by Fluker et al. (1999). The
patients considered for this study were 51 women undergoing superovulation
who had estradiol levels of 43000 pg/ml. In 4 of the 51 women, excessive
follicular diameter was observed with the presence of 8 to 10 follicles ¸ 18 mm
(n ¼ 2) or more than 30 follicles 410 mm (n ¼ 2). For religious reasons, none
of the four women would consider converting to IVF. The cycles were canceled
and hCG administration was withheld. Among the remaining 47 women who
received the hCG, serum estradiol levels continued to rise for at least one day
after the onset of the coasting period, then gradually reached a plateau on the
second day and began to fall precipitously on the third day (Figure VII.10).
Human gonadotrophin was administered on the evening of the third to the
¬fth day (mean 3.4 days). The fall from peak E2 levels of 2824 pg/ml to ¬nal
levels of 1246 pg/ml on the day of hCG administration represents a mean

Fig. VII.10: Serum estradiol and LH concentrations before and during the coasting period
in superovulation cycles
Reproduced with permission from Fluker et al. (1999). Fertil Steril 71:294“301

reduction in serum estradiol concentrations of 56%. Mean LH levels rose near
the onset of the coasting period and decreased spontaneously, while follicular
growth continued. Moderate OHSS occurred in three (6%) of the 47 women
to whom hCG was administered. A small amount of ascites was noted sono-
graphically and ovaries were enlarged to 6“10 cm. This was not accompanied by
signi¬cant abnormalities in renal function or hemotologic parameters.
Spontaneous resolution occurred with bed rest at home. Eleven pregnancies
occurred among the 47 women (23.4%), including eight singletons, one twin,
one triplet, and one ectopic pregnancy.

An extensive literature has been published on the role of coasting to prevent
hyperstimulation in GnRH agonist IVF cycles (Rizk, in press). Rizk and Smitz
(1992) observed that the GnRH agonist is associated with a higher incidence
of OHSS. In most of the published studies, the long agonist protocol was used,
with the exception of a study by Dhont (1998) when the short protocol was
used (Tables VII.9“VII.11).
The credit for the ¬rst observation study on coasting in IVF goes to the
west coast of the United States in San Francisco. Sher et al. (1993) suggested
that prolonged coasting in GnRH-a/hMG/FSH cycles could prevent the
life-endangering complications of OHSS. They withheld gonadotrophins in
17 patients whose serum estradiol exceeded 6000 pg/ml, and continued daily
GnRH-a until estradiol levels had fallen below 3000 pg/ml. HCG (10 000 IU)
was administered to trigger ovulation. The estradiol levels continued to
rise rapidly in the 48 h following the initiation of the coasting period, then
plateaued and began to fall 96“168 h after the gonadotrophins were stopped.
The coasting period lasted between four and nine days and the day of

<< . .

( 30)

. . >>

Copyright Design by: Sunlight webdesign