FSH
OVULATION STIMULATION IN ANOVULATORY PATIENTS
Several types of disruptions of ovulation may be encountered, which require different approaches both as far as the therapy and sequence are concerned.
1) Type 1 ovarian dystrophy
This pathology remains enigmatic, even though significant progress has been made in understanding it over the last few years ( 25).
Clomifene citrate
The first line treatment remains antioestrogens (Clomifene citrate or Tamoxifene). As soon as there is a certain level of oestrogen impregnation, their antioestrogenic effect interferes with the hypothalamo-hypophyseal regulation by stimulating the secretion of FSH. On the contrary, if the oestrogenic impregnation is not high enough, they are ineffective as they then behave as would weak oestrogens. This treatment is simple and practically free from risks, but nevertheless raises a certain number of practical questions.
1) Should anti-oestrogens always be used to begin with? Some claim that a single dose of oestradiol is sufficient to predict the effectiveness of Clomifene. Whilst it is established that there is a clear relationship between them, the low cost and simplicity of a treatment by Clomifene are to our mind arguments for it to be tried in all circumstances for one or two cycles.
2) What dose is required? Classically, the initial dose is 100 mg/day (i.e. 2 tablets of Clomid® or Pergotime®) for 5 days (from D3 to D7) . However in practice, only slightly more than half of the patients will respond to such a dose. In the case of failure, and before envisaging more advanced treatments such as FSH, the daily dose and length of the treatment should be tried without hesitation. Few arguments exist for preferring one or the other. Personally, we prefer increasing the dose to begin with then the duration, up to a maximum of 200 mg/day and a duration of 10 days.
3) Should oestrogens be added at the end of the follicular phase? To begin with, it should be underlined that this is specific to France. The main argument is a negative effect of the clomifene on the cervical glair. In practice, this effect is far from common ( 46), however this therapy has above all a huge drawback: administering oestrogens can mimic the pre-ovulation peak of oestradiol, which can trigger the LH peak. If the prescription of oestrogens occurs too soon in the ovocyte maturing process, then the LH peak will occur prematurely. It therefore appears that this prescription can only be defended if the alteration of the glair has been clearly shown and only by identifying it with reference to time by USS (bigger follicle, larger than 17 mm).
4) Should ovulation be started with HCG? The main argument is based on a concommittent anomaly from the positive feed-back from the oestrogens. In practice, this anomaly appears to be quite rare. Furthermore, a premature injection of HCG produces the same effect as previously described. That is to say that again, the injection of HCG may only be envisaged if USS is performed and when the follicle is 20 mm or more in size. Finally, if no studies have been performed on PCO's, Fisch showed in normally ovulating patients that the addition of HCG, even if correctly timed, did not increase the chances of pregnancy ( 21)
5) Should the luteal phase be aided? The main argument is based on the frequent association of Corpus Luteum Insufficiency (CLI) and PCO. The question is complex and has given rise to a number of publications, none of which are more convincing than the others. In general, CLI is the proof of a poor quality follicular phase which the systematic prescription of progesterone tends to conceal.
6) Should the effectiveness of Clomifene be checked and how? We believe this to be essential in most cases to avoid wasting time. Several methods are available: USS and biopsy of the endometrium. However, we consider the basal body temperature curve to be the simplest, associated with a plasmatic dose of Progesterone between the 8th and 10th post-ovulation day.
7) What strategy should be adopted in the case of ovulation under Clomifene but no pregnancy? Of course, everything is dependent on the general context of the infertility and more particularly the association with another cause. If the ovulation problem is isolated, some authors propose using dexamethasone, arguing that the difference between type 1 PCO and type 2 PCO is not always flagrant ( 15). This is not accepted by all, especially since the direct effect of this drug on the metabolism of the androgens is very complex and still not fully understood ( 7).
Hypothalamic anovulations
Pumping of LHRH
The most attractive, logical and cheapest treatment is the pumping of LHRH. The pulsatile IV or sub-cutaneous administration of small doses of LHRH allows a virtually physiological cycle to be re-established. Whilst the IV route is more effective, it requires stricter monitoring to avoid venous or infectious complications. In addition, many prefer using the sub-cutaneous route. The real problem of pumping LHRH is to get the patients to accept it, as they often wrongly consider it to be a method with too many constraints. In terms of results, the pregnancy rates reported are most often between 15 and 25 %, which is to say figures which are close to the natural fertility rate. The second advantage is the low rate of multiple pregnancies, which is very close to that of that of natural reproduction. The largest series was published by Filicorni ( 20). He carried out 600 cycles on 296 patients. Ovulation was obtained in 75 % of cases. In total, 105 pregnancies were obtained, i.e. a rate of 18 % per cycle and 23 % per ovulation. The rate of multiple pregnancies was only 3.7 %. However, 25 % of patients had a miscarriage, if the PCO's are excluded, as they had a higher rate.
Another of the considerable advantages of pumping is its potential to re-establish an ovulation cycle on a long term basis in certain patients, even after the pumping has been stopped. It is thus not uncommon to obtain a pregnancy in the cycle following a treated cycle.
Clomifene citrate
Logically, it should not be administered for this indication. However its low cost, and ease and safety of handling mean that it may be legitimately proposed in particular to anovulative, amenorrhoea-free patients, especially in cases where psychogenetic etiology is clearly predominant. Is it a placebo or pharmacological effect? Nevertheless, 20 to 40 % of patients will ovulate with this drug. ( 4)
Gonadotrophins
In the absence of a response with Clomifene, and in the case of refusal or failure of pumping, this is the preferred treatment. For this indication, classical HMG or purified FSH seem to be equally effective. The question of rec.FSH is to be raised. In patients with a deep depression of their gonadotrope function, it may be feared that the lack of LH is a factor of failure. The results of IVF with stimulation by rec.FSH after hypophyseal blockage are quite reassuring and even encouraging. Nevertheless, the clinical data is still insufficient to be able to affirm this with certainty.
Association of Clomifene and HMG
This does not sem very logical to us for this indication. Its main objective is to perform a certain degree of hyperstimulation, which should not be the aim for this indication.
Hyperprolactinaemiae
With the rare exceptions of hypophyseal macroadenomas, the induction of the ovulation is only based on medical treatments.
All derivatives of ergot of rye, hypoprolactinaemiants have a dopaminergic role in controlling the secretion of prolactine. In addition, they have an anti-tumorous effect (direct or indirect). When used alone, they only re-establish ovulation in a limited proportion of cases (20 to 30 %). In the case of failure (absence of ovulation or absence of pregnancy after 6 apparently normal ovulations), then the addition of other inducers is necessary. We use the following sequence: Clomifene, Clomifene + gonadotrophin, gonadotrophin
Other anovulations
Apart from these well-defined situations and exceptional anovulations of hypophyseal origin, which require the use of HMG, other ovulation problems are encountered which are much more difficult to classify. Amongst them are menopausal type anovulations, which represent a particularly interesting identity.
In fact, 40 % of patients with an early menopause syndrome have a good ovocyte reserve. This "gonadotrophin resistance syndrome" has remained an enigma for a long time, but is finally starting to be broken down. Two main etiologies (which may be non contradictory) have been highlighted recently: the self-immunity hypothesis ( 19) and the genetic hypothesis ( 3). Unfortunately, this progress in the understanding is not accompanied by therapeutic progress. All gonadotrophin or pumping stimulation has proven to be ineffective, without causing the FSH level to fall, either with analogues or steroids. After hypophyseal blockage, HMG stimulation only allows follicular growth in rare cases. Finally, the highest chances of pregnancy lie in ovulation starting again spontaneously when artifical cycles are used.
INDUCING OVULATION IN "NORMALLY OVULATING" PATIENTS
It is surprising to see that therapies which are so commonly used and which have been the subject of so many publications (more than 1100 between 1986 and 1996) are still not unanimously agreed upon.Furthermore, half of the series report ovarian stimulation (OS) and IUI at the same time. It is therefore very difficult to judge the role of the OS tretaments, which some consider to be empirical ( 42).
For didactic reasons, we will break our analysis down by the etiologies of infertility, where these therapies are used.
1) Endometriosis
The arguments in favour of ovulation stimulation after laparoscopic treatment of endometriosis total 3. Harlow ( 24) reported very frequent absence of fertilisation during unstimulated IVF cycles in the case of endometriosis. Pellicer ( 36) showed that the embryos obtained with IVF have fewer blastomers at D3 in case of endometriosis Finally, we have shown that the development of embryos in coculture is not as good in the case of endometriosis, and that it principally concerns a blockage in relation with an ovocyte maturing process ( 38). All this leaves the suspicion that this ovocyte dysmaturity at least partially explains the physiopathology of the infertility due to endometriosis ( 10) and that it may be partially corrected by increasing the level of FSH.
A number of articles have treated OS in this context. However, it is difficult to form a clear idea. The association between the OS and IUI is very frequent. The variable selection of the cases adds to the confusion: certain recommend OS in all cases, whereas others reserve them for stages 1 and 2. Finally, far too many series are retrospective, but nevertheless, we will retain several of those which we consider to be the most valid.
Serta ( 41) compared two groups of 25 patients (stage 1 endometriosis), either treated by IUI alone or by IUI + OS. In total, 132 cycles were carried out, which allowed 16 pregnancies to be achieved. The cumulative pregnancy rates were respectively 13 %, 26 % and 38 % after 1, 2 and 3 months in patients who had IUI. The figures are similar after IUI + OS, i.e. 12 %, 25 % and 34 %. His logical conclusion was that the benefit only comes from IUI.
However, other authors disagree with his findings and conclude that the main benefit comes from OS and not IUI. Thus according to Nulsen ( 34), in a randomised series, the probability of achieving pregnancy is 5.2 times higher (O.R. = 5.2 , CI 1.1 to 22.5) with OS + IUI than with IUI alone. Tummon ( 47) carried out a global comparative study of OS + IUI and abstention, which makes it of limited interest only. The probability of pregnancy is 5.6 times higher (O.R. = 5.6 , CI 1.8 to 17.4) with OS + IUI than with abstention with cumulative pregnancy rates at 4 months of 30 and 10 % respectively. Placed alongside the previous series, this also suggests that it is only the OS which provides the benefit.
However, some studies have shown that OS alone is beneficial, such as that of Fedele ( 18), Kemmann ( 29) and Simpson ( 43). Out of 87 patients with stage 1 and 2 endometriosis and who had just been operated on, Kemmann ( 29) carried out a retrospective analysis on the fertility per cycle, depending on whether the patients were receiving an inductive treatment without IUI or not (Clomifene, or Clomifene + HMG or HMG), or who had had IVF. He was thus able to analyse 605 cycles which led to 37 pregnancies. The conception rate was 2.8 % without treatment, 6.6 % with Clomifene alone, 11.4 % with Clomifene + HMG (p<0.05) and 7.3 % with HMG alone and finally 22 % with IVF(p<0.05). He therefore concluded that an active attitude was beneficial. However, if the IVF is excluded, the global pregnancy rate hardly exceeds 30 %, which is the figure obtained postoperatively and without OS by Marcoux ( 31) in similar cases and in a prospective study. In a retrospective study, Simpson ( 43), compared the pregnancy rates in patients with endometriosis, depending on whether they had not received any postoperative treatment, whether they had been treated with Danazol or if they had had Clomifene stimulation. The relative risk of pregnancy per cycle was significantly higher with Clomifene ( 2.9 CI 1.2 to 7,1) and was not improved by Danazol (1.02 CI 0.5 to 2.3). However, finally the cumulative pregnancy rate was similar in all of the groups.
Out of 49 randomised patients, Fedele ( 18) compared 3 cycles of HMG-HCG with therapeutic abstention for 6 months. The conception rate per cycle was higher during the stimulated cycles, 15 % versus 4.5 %(P<0.05) and the cumulative pregnancy rate at 6 months was higher, 37.4 % as opposed to 24 % (p=ns). He also underlined the high number of multiple pregnancies (33 %) and hyperstimulation in the "stimulated" group.
Lenton ( 30) tried Clomifene for this indication. The fertility rate per cycle was 9.5 % with Clomifene versus 3.3 % without treatment.
It is obvious that there are no glaring conclusions to be drawn. OS increases the conception rate for the treatment cycle. The added use of IUI is very controversial. However, the question may be asked as to whether the sole effect of all of these treatments is simply to accelerate the occurrence of the pregnancies, without significantly changing the frequency. On the other hand, IVF is a much more efficient treatment, with delivery rates per cycle of around 25 % ( 35).
2) Cervical infertility
Paradoxically, there are relatively few publications concerning this etiology. It should be stated that it is not always clearly defined and its reported frequency varies significantly. The main confusion arises from incorrect evaluation at the outset. Cervical infertility is only conceived in the absence of ovulatory problems. From this point, the administration of oestrogesn then a crossed in vitro test allows this diagnosis to be determined more clearly, and finally it is only retained in a few cases. If there is a relationship with the antispermatoid antibody, the choice of IVF is logical. This only leaves the problems of toxic glair or structural problems of the glair, and in these cases, IUI is always indicated. Finally, light OS is also used, which improvces results but increases the risk of multiple pregnancies ( 33).
3) Unexplained infertility
This is a somewhat vague entity, which is influenced by three important parameters: the duration of the infertility, the age and the precise workup beforehand. This has been the subject of an impressive number of articles in the literature, which is difficult to situate when it is considered that the reported spontaneous pregnancy rates vary from 5 % to 87 % (16, 32). Martin thus noted that in untreated couples, there was a cumulative fertilisation rate of 34 % at 6 months, 76 % at 2 years and 87 % at 5 years, which is much lower than a control group which obtained a rate of 74 % at 6 months, but this appears extraordinarily high to us. Others have analysed the probability of spontaneous pregnancies in these cases with regard to age (11) and have shown a drop in this probability of 9 % per year, especially in primary unexplained infertilities. Others did not find any relationship between the duration of the infertility and the chances of success of the treatments ( 17). Finally, Crosignani ( 13) summarised a number of articles on the question of spontaneous fertility in that group and found conception rates per cycle of 0.75 % to 3.97 , with a mean of 1.11 % (out of 393 patients and 18026 cycles).
It can therefore be understood that analysing any of the articles leads to considerable or even inextricable difficulties, especially if this involves a retrospective study.
Finally, one of the rare studies worthy of interest is that directed by Crosignani ( 13, 14) by the ESHRE and which has the advantage of being prospective, randomised and multicentre. This study principally compared IVF, GIFT, intraperitoneal insemination (IPF), IUI (with ovarian stimulation) and isolated ovarian stimulation. This study included 444 patients and 649 cycles. The conception rates (CR) and the delivery rate (DR) per treatment cycle (cycle 1 or 2) were respectively 27.4 % (CR) and 24.5 % (DR) for IUI , 27 % (CR) and 20.8 % (DR) for IPF, 28 % (CR) and 22.8 % (DR) for GIFT, 25.7 % (CR) and 22.9 % (DR) for IVF and 15.2 % (CR) and 13.6 % (DR) and for simple ovarian stimulation (p=0.06 versus the previous figures). Three main conclusions may be drawn: firstly, the pregnancy rate obtained with any technique is infinitely higher than that observed naturally, which in the most favourable hypothesis does not exceed 7.8 % for Collins ( 11), secondly, the results of GIF, IVF, IPF and IUI are very similar, thirdly, ovarian stimulation alone is slightly les efficient. The conclusion of the authors is that the benefit is both due to the number of ovocytes available for fertilisation and the MAP technology.
In a similar study, but which was in a single centre and smaller, Peterson ( 37) found a conclusion which was quite similar, as he found a conception rate per cycle of 1.4 % in the group which had no treatment, 26 % after IVF and 15 % after ovarian stimulation. It was also found that one stimulation cycle provides the same results as 1 year of waiting and that 1 IVF provides the same results as 2 years of waiting or 2 stimulations. He also added to his own cases a meta-analysis of the literature, which turned out to be a little confusing but in which he concluded that one GIFT cycle is more effective than 1 or 2 ovarian stimulation + IUI cycles, that 3 OS + IUI cycles are equivalent to 1 GIFT but more effective than one IVF cycle or one ZIFT cycle, and finally that 4 OS + IUI cycles are more effective than one GIFT, IVF or ZIFT cycle. In complete opposition, in a single centre randomised prospective study, Fedele ( 18) concluded that IUI was ineffective and suggested simply using OS.
Finally, it can be seen that the problem constantly alternates between 2 questions that it would be preferable to separate: should IUI be performed? should ovulation be stimulated? Hughes ( 27), in a very recent publication, made a remarkable meta-analysis of the problem, and highlighted the following essential data: the chances of pregnancy are increased by 2.35 times (CI 1.87 to 2.94) by stimulation with HMG or FSH. They are also increased independently by a factor of 2.82 (CI 2.18 to 3.66) by the use of IUI. The association of the two treatments increases the chances by 6.62. On the other hand, the alteration of the sperm induces a negative factor of 0.48 (CI of 0.37 to 0.60), whereas for endometriosis, the negative factor is 0.45 (Ci of 0.27 to 0.76). In this analysis, Clomifene proved to have little effect, especially when used with IUI, hwoever the author noted that in a previous meta-analysis, he had found that Clomifene had a benficial effect ( 28). Finally, this author concluded that it was of the utmost importantce to carry out further prospective studies with improved definitions of the protocols and more objective randomisations!
4) Male hypofertility
This subject is even more uncertain than the previous one. The definition of male hypofertility is necessarily vague and some class as male hypofertility what others class as unexplained infertility. Here again, confusion reigns between IUI and OS. However, it is clear that for this indication, it is IUI which is the most logical prescription. Should OS be associated? No according to Mélis ( 33) and Fedele ( 18) , and yes according to Tummon ( 47), Nulsen ( 34) and Belaisch-Allart ( 5). The latter recommends limiting IUI to cases where preparation techniques allow 500,000 spermatoids to be collected and not to exceed 4 to 6 IUI before trying IVF. We will not venture further into this subject, which deserves more serious prospective studies rather than rushing headlong and excessively towards ICSI.
|