INDUCTION OF OVULATION
Several categories of patients who need induction of ovulation exist:
To date, the main agents used for the induction of ovulation are clomiphene citrate, GnRH-analogs and gonadotropins.
Clomiphene citrate is an orally active, non-steroidal agent which is distantly related to diethylstilbestrol. Clomiphene exerts only a very weak biologic estrogenic effect. The structural similarity to estrogen, however, is sufficient to achieve uptake and binding by estrogen receptors including those of the hypothalamus. The clomiphene-receptor binding is defective, but enough to activate the homeostatic negative feed-back relationship between estrogen and gonadotropins. The hypothalamus acts on the false information that the estrogen level in circulation is low, FSH and LH levels increase and this stimulates follicular growth. The effectiveness of the drug seems to be restricted to its ability to induce an appropriate discharge of gonadotropins (8).
In women with anovulatory cycles and an only slightly disturbed hypothalamus-pituitary-ovarian axis, the ovulatory response is present in about 70% of cases, whereas pregnancies occur in only about 30% (8). Several factors can account for this difference, the main one probably being the high level of chromosomally abnormal oocytes following clomiphene induction of ovulation. Doses and regimens can vary, the most popular being those of 100 mg /day for 5 days starting from day 3 to day 7, or from day 5 to day 9 of the cycle. The dose can be increased in a staircase fashion up to 200 mg/day in the case of unsatisfactory or absent ovarian response after at least 2 treatment cycles. Therapy can also be used to manipulate the cycle but administration beyond day 9 may have no effect. Ovulation usually occurs in the interval 5 to 10 days after the last day of medication.
Gonadotropin releasing hormone analogs (GNRH-A)
Gonadotropin releasing hormone analogs (GnRH-a) are synthetic peptides whose structure is similar to the natural GnRH.
The first effect of these molecules that compete with endogenous GnRH molecules for the pituitary receptors is to provoke a massive release of gonadotropins into the circulation before " desensitizing " the hypophysis. The increase of gonadotropins can provoke ovulation, if high enough: this phenomenon is known as the " flare up " effect of GnRH analogs.
These molecules will be discussed below with the regimens used for ovarian stimulation in IVF. The GnRH-a available are presented in Table 1.
The first successful induction of ovulation and pregnancy in the human was described by Gemzell et al. in 1958 (5) utilizing FSH derived from human pituitary glands removed at autopsy. It was also found that the luteinizing properties of human chorionic gonadotropin (hCG) provided a good enough stimulus to ovulation after the follicle had ripened. The real breakthrough came though only when an easy source of gonadotropins became available (9) in the form of human menopausal gonadotropin (hMG) derived from the urine of postmenopausal women. HMG contains equal quantities of FSH and LH. At present, biotechnology and molecular biology tools have also provided us with a highly purified form, and a so-called " recombinant form " of FSH.
HMG is commercialized in 75 IU ampoules im (75 IU of FSH and 75 IU of LH). It must be stressed here that the individual response to induction of ovulation is characterized by extreme variability, as a response can be obtained at very low dosages which accounts for the possibility of such complications as multiple follicular growth with or without hyperstimulation syndrome (2). At the same time, very high doses may not achieve a response (23). For these reasons, treatment must be carefully adjusted to the individual patient which requires accurate monitoring of estradiol levels, and of follicular growth at ultrasound examination.
During the last 15 years, another problem has surfaced in the field of ovulation induction: the need to provoke the recruitment and maturation of numerous follicles for assisted reproductive technologies in order to obtain several follicles at ovum pick-up for in vitro fertilization programs (IVF). The monitoring of these cycles has become of paramount importance and will be discussed by other authors.
In women with functional pituitaries, the use of exogenous gonadotropins to induce multiple follicular development leads to unpredictable responses. This used to make timing of ovum pick-up for IVF programs difficult, and timing of intercourse unreliable for in vivo fertilization (23).
GNRH-A and gonadotropins
GnRH-a provoke reversible blockage of the pituitary thereby suppressing gonadal function. Since the original description of the use of GnRH analogs in combination with human menopausal gonadotropins (hMG) for ovulation induction in assisted reproduction in 1984 (15), the method has achieved widespread use. The introduction of ovarian stimulation for IVF was a major breakthrough which increased the pregnancy rate significantly (25). Induction of ovulation in normally ovulating women, however, presents a higher degree of asynchrony in the development of follicles due to the endogenous gonadotropic activity (3). Moreover, induction of ovulation has a significantly higher success rate in women suffering from hypothalamo-pituitary failure than in women with hypothalamo-pituitary dysfunction (10). It was then proposed to induce a medical reversible hypophysectomy in patients for in vitro fertilization (4). Pretreatment of patients with GnRH analogs prevents premature LH surges (20,22) thus lowering the cancellation rate of IVF cycles significantly; LH surges may in fact occur prematurely—before the follicles have reached maturity—causing luteinization and disruption of normal follicle and oocyte development. Furthermore, a negative correlation has been shown between LH levels in the follicular phase and subsequent oocyte maturity at follicle aspiration (12,16). GnRH administration not only eliminates premature LH surges but also efficiently controls tonic levels of LH which is very important, especially in cases of polycystic ovarian disease (14). Suppression of LH levels by GnRH analogs could result in improved embryo quality (22).
Prospective randomized trials have been performed to compare IVF outcome in patients treated with hMG and patients treated with analogs and hMG (4,15). Pregnancy rates per cycle are significantly higher in the analog group, due to a lower cancellation rate per cycle. Pregnancy rates per embryo transfer, however, do not differ. The number of days of stimulation and the total dose of ampoules of hMG used were significantly higher in the analog group, increasing total IVF costs. Another disadvantage of the use of GnRH analogs is that embryos obtained from these cycles appear to be less resistant to freezing and thawing (24).
Three different protocols have been described for the use of GnRH analogs in combination with hMG for IVF. In the long or blocking protocol (17,13, 26) the analog is administered 7-14 days prior to the addition of hMG. The GnRH analog can be started in the luteal phase of the previous cycle or the follicular phase of the treatment cycle. Once desensitization has been achieved, hMG stimulation is begun. GnRH analog and hMG are discontinued on the day of human chorionic gonadotropin (hCG) administration, 32-36 hours prior to follicle aspiration. Controversy still exists about the shortest interval in which complete desensitization can be achieved: some have reported a shorter interval after a luteal start of analogs, others found estrogen levels to decrease more rapidly after a follicular start (7,14,16). No differences, however, have been shown in subsequent pregnancy rates (14,20). But in patients starting GnRH analog in the follicular phase significantly more cyst formation was found (12,16).
According to the short or flare-up protocol (6) GnRH analog is started on day 1 of the follicular phase, and hMG is added on day 5. The analog and hMG are continued until the day of hCG administration. GnRH analog use according to the short protocol has proved to be efficient in preventing premature LH surges. But studies comparing the long and short protocol report conflicting results about pregnancy rates obtained (12,14). The short protocol seems to have some disadvantages: significantly more cyst formation is seen compared to the long protocol (18). Furthermore, the initial up-regulation caused by the analog coincides with the stimulation phase, resulting in increased LH levels. High LH levels during follicular development seem to have an unfavourable effect on subsequent embryo quality (7,17). Moreover, a negative correlation has been demonstrated between progesterone levels in the follicular phase, due to premature luteinization, and pregnancy rates in IVF patients (20).
In the ultrashort protocol (11), GnRH analog is given on days 3, 4 and 5 of the follicular phase of the treatment cycle only. HMG is started on day 3 and continued to the day of hCG injection. Large clinical studies on the ultrashort protocol are not available yet.
To conclude, GnRH analogs have proved their value as adjuvants in stimulation regimens for assisted reproduction. Although their introduction has not led to increased pregnancy rates, their use is justified by the significant decrease induced in cancellation rates. In future, prospective, randomized trials will reveal whether the long, blocking protocol or the short, flare-up protocol performs better.
Edited by Aldo Campana,