Suppression of LH during ovarian stimulation: analysing threshold values and effects on ovarian response and the outcome of assisted reproduction in down-regulated women stimulated with recombinant FSH

Juan Balasch1,3, Ester Vidal1, Joana Peñarrubia1, Roser Casamitjana2, Francisco Carmona1, Montserrat Creus1, Francisco Fábregues1 and Juan A. Vanrell1

1 Institut Clínic of Gynecology, Obstetrics and Neonatology and 2 Hormonal Laboratory, Faculty of Medicine, University of Barcelona, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona, Spain


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: It has been recently suggested that gonadotrophin-releasing hormone agonist down-regulation in some normogonadotrophic women may result in profound suppression of LH concentrations, impairing adequate oestradiol synthesis and IVF and pregnancy outcome. The aims of this study, where receiver-operating characteristic (ROC) analysis was used, were: (i) to assess the usefulness of serum LH measurement on stimulation day 7 (S7) as a predictor of ovarian response, IVF outcome, implantation, and the outcome of pregnancy in patients treated with recombinant FSH under pituitary suppression; and (ii) to define the best threshold value, if any, to discriminate between women with `low' or `normal' LH concentrations. METHODS: A total of 144 infertile women undergoing IVF/intracytoplasmic sperm injection (ICSI) treatment were included. Seventy-two consecutive patients having a positive pregnancy test (including 58 ongoing pregnancies and 14 early pregnancy losses) were initially selected. As a control non-pregnant group, the next non-conception IVF/ICSI cycle after each conceptual cycle in our assisted reproduction programme was used. RESULTS: The median and range of LH values in non-conception cycles, conception cycles, ongoing pregnancies, and early pregnancy losses, clearly overlapped. ROC analysis showed that serum LH concentration on S7 was unable to discriminate between conception and non-conception cycles (AUCROC = 0.52; 95% CI: 0.44 to 0.61) or ongoing pregnancy versus early pregnancy loss groups (AUCROC = 0.59; 95% CI: 0.46 to 0.70). To assess further the potential impact of suppressed concentrations of circulating LH during ovarian stimulation on the outcome of IVF/ICSI treatment, the three threshold values of mid-follicular serum LH proposed in the literature (<1, <=0.7, <0.5 IU/l) to discriminate between women with `low' or `normal' LH were applied to our study population. No significant differences were found with respect to ovarian response, IVF/ICSI outcome, implantation, and the outcome of pregnancy between `low' and `normal' S7 LH women as defined by those threshold values. CONCLUSIONS: Our results do not support the need for additional exogenous LH supplementation in down-regulated women receiving a recombinant FSH-only preparation.

Key words: assisted reproduction/early pregnancy loss /FSH/LH/ovarian stimulation


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Assisted reproduction has evolved rapidly over the past two decades with significant progress in our understanding of the endocrinology of the assisted reproductive process. At present, it is well established that successful IVF and embryo transfer requires both stimulation of the ovary and suppression of the pituitary. Thus, exogenous gonadotrophins and gonadotrophin-releasing hormone (GnRH) analogues are the key hormones required to maximize IVF success (Barbieri and Hornstein, 1999Go).

Human menopausal gonadotrophin (HMG) containing different FSH/LH ratios has been used effectively for ovulation induction in IVF patients, and for years this was the only preparation available for clinical use. Because of advances in the urinary purification techniques that have become available, HMG was purified further so that, in the early 1980s, purified FSH (with <1% LH contamination) was made possible, whereas in the 1990s highly purified FSH (with <0.1% LH contamination) became a therapeutic option for ovarian stimulation. More recently, biotechnology has made available for the first time a recombinant human FSH preparation (Howles, 1996Go; Olijve et al., 1996Go) which is completely devoid of LH activity.

As previously stressed (Hull et al., 1994Go; Daya, 1998Go; FIVNAT, 1999Go; Fleming et al., 2000Go), it is clear that changes in stimulation regimens with widespread use of urinary or recombinant FSH preparations, without LH supplementation, has not led to a drop in overall programme success rates, and it has been suggested that there is no need to explore the requirement for LH supplementation further (Hull et al., 1994Go; Daya, 1998Go). However, when considering the relative roles of FSH and LH preparations in folliculogenesis, it has been stressed that the potential benefit of adding LH to FSH-only regimens might depend on the extent to which the endogenous plasma LH is suppressed by concomitant GnRH analogue therapy (Hillier, 1994Go).

Whereas the relative importance of FSH and LH in inducing follicular growth and maturation in the human is still being investigated, considerable debate exists as to whether the LH activity contained in HMG preparations could affect the outcome of assisted reproduction treatment in GnRH agonist down-regulated women (Hull et al., 1994Go; Westergaard et al., 1996Go; Loumaye et al., 1997Go; Daya, 1998Go; Söderström-Anttila, 1998Go; Filicori, 1999Go; Lévy et al., 2000Go). Thus, the concentrations of residual endogenous LH remaining during full-dose GnRH agonist pituitary suppression are certainly sufficient to achieve adequate follicular maturation during ovarian stimulation with purified human urinary FSH or recombinant FSH (Lévy et al., 2000Go). However, it has been suggested that GnRH agonist down-regulation in some normogonadotrophic women may result in profound suppression of LH concentrations, impairing adequate oestradiol synthesis (Westergaard et al., 1996Go; Fleming et al., 1998Go, 2000Go; Janssens et al., 2000Go), fertilization rates (Westergaard et al., 1996Go) or the final clinical treatment outcome by increasing the risk of early pregnancy loss (Westergaard et al., 2000Go).

The amount of LH activity actually necessary for a normal follicle and oocyte development is not known, but is likely to be very low, since <1% of follicular LH receptors need to be occupied in order to allow normal steroidogenesis (Chappel and Howles, 1991Go; Lévy et al., 2000Go). Threshold values of serum LH of <1, <=0.07 and <0.5 IU/l in the mid-follicular phase (days 7–8) of gonadotrophin ovarian stimulation cycles, in down-regulated women have been proposed to diagnose the group of `profoundly' suppressed LH patients (Chappel and Howles, 1991Go; Fleming et al., 2000Go; Westergaard et al., 2000Go).

Following the recently published papers on this subject (Fleming et al., 2000Go; Westergaard et al., 2000Go), we have examined in our own clinic the impact of hormonal parameters on the outcome of assisted reproduction with a retrospective analysis of prospectively collected material. The specific aims of this study, where receiver-operating characteristic (ROC) analysis was used, were: (i) to assess further the usefulness of mid-follicular serum LH concentration as a predictor of ovarian response, IVF outcome, implantation, and the outcome of pregnancy in pituitary-suppressed patients undergoing assisted reproduction; and (ii) to define the best threshold value, if any, to discriminate between women with `low' and `normal' LH concentrations.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patients
A total of 144 women was included in the present investigation. For this study we have collated the data from a cohort of women having a positive pregnancy test (clinical or biochemical pregnancy) after IVF/ICSI where we have frozen blood samples still stored for measurement of mid-follicular hormone concentrations as reported below. As it is the policy of our laboratory to store blood samples from cycles carried out during the previous 6–7 months, as of July 2000, we were able to gather data on 72 consecutively pregnant patients who, at the time of preparing this analysis, would have reached >20 weeks gestation to be defined as ongoing pregnancies (see below). As a control non-pregnant group, the next non-conception IVF/ICSI cycle after each conceptual cycle in our assisted reproduction programme was used. The overall mean age of the women was 34.0 ± 0.4 years (mean ± SE) (range 23–42). All of them were regularly menstruating premenopausal women and none of them had occult ovarian failure on the basis of their cycle day 3 FSH concentrations of <12 IU/l [standard International Reference Preparation (IRP) 78/549]. No woman had polycystic ovarian disease according to ultrasound examination of the ovaries or basal gonadotrophin measurement. Patient indications for IVF/ICSI included the following main diagnoses: male factor infertility (46% of patients), tubal infertility (23%), endometriosis (17%), and unexplained infertility (14%). For the specific purpose of this investigation, blood samples drawn on stimulation day 7 (S7) were assayed for serum LH (which was measured at the completion of the study) in addition to the routine serum oestradiol measurement for monitoring ovarian response. The S7 was used because monitoring of ovarian response (ultrasonography and oestradiol measurement) is routinely performed on this day in our assisted reproduction programme.

Stimulation regimen
Ovarian stimulation was carried out with FSH under pituitary suppression with GnRH agonist. In all women, pituitary desensitization was achieved by s.c. administration of leuprolide acetate (Procrin; Abbott Laboratories, Madrid, Spain). This treatment was started in the midluteal phase of the previous cycle and given 1 mg daily, then reduced to 0.5 mg after ovarian arrest was confirmed. Gonadotrophin stimulation of the ovaries was started when serum oestradiol concentrations declined to <50 pg/ml and a vaginal ultrasonic scan showed an absence of follicles >10 mm diameter. On days 1 and 2 of ovarian stimulation, 6 and 4 ampoules/day of recombinant human FSH (75 IU per ampoule) (Gonal-F, Serono SA, Madrid, Spain) respectively, were administered subcutaneously. On days 3 and 4 of ovarian stimulation, two ampoules per day of FSH were administered to each patient. From day 5 onward, FSH was administered on an individual basis according to the ovarian response as assessed by sequential transvaginal ultrasonography and serum oestradiol measurements. Human chorionic gonadotrophin (HCG; 5000 IU) (Profasi; Serono SA) was administered i.m. when a consistent rise in serum oestradiol concentration was associated with the presence of two or more follicles >18 mm in diameter with >=4 follicles measuring >=14 mm. Oocyte aspiration was performed with vaginal ultrasonography 35–36 h after HCG administration. The maturational status of the oocytes and the embryo grading were recorded according to published criteria (Veeck, 1986Go); embryos of Veeck grades 1 or 2 were considered high quality. Two days later up to three embryos per patient were replaced. Two additional doses of 2500 IU of HCG on the days of embryo transfer and 3 days later were given to supplement the luteal phase in all patients.

Pregnancy outcome
The following categorization of pregnancy outcome was made: (i) biochemical pregnancy: four women showed serum ß-HCG concentrations >50 IU/l 13–14 days after embryo transfer and 7 days later but no intrauterine or extrauterine pregnancy could be demonstrated by ultrasonography and menses ensued; (ii) spontaneous abortion: ten women had increasing concentrations of ß-HCG on days 13–14 and 20–21 after embryo transfer, and pregnancy was confirmed by ultrasonic demonstration of an intrauterine gestational sac but the gestation subsequently spontaneously terminated; (iii) ongoing pregnancy: this category comprised 58 patients achieving >20 weeks gestation.

For the evaluation of the results, the following groups of IVF/ICSI outcome were considered: (i) no conception (n = 72); (ii) conception (n = 72) which included ongoing pregnancy (n = 58) and biochemical (n = 4) and spontaneous abortions (n = 10); (iii) ongoing pregnancy (n = 58); (iv) early pregnancy loss (n = 14) which included biochemical (n = 4) and first trimester spontaneous abortions (n = 10).

Hormone analyses and ultrasonography
For this study, serum concentration of LH was measured retrospectively on S7 in all treatment cycles studied. This was done using frozen blood samples which were examined in one run. Hormones were measured using commercially available kits as reported previously (Balasch et al., 1998Go; Fábregues et al., 2000Go; Peñarrubia et al., 2000Go). Oestradiol concentrations in serum were estimated by a competitive immunoenzymatic assay (Immuno 1, Technicon; Bayer, Tarrytown, NY, USA). The sensitivity was 10 pg/ml and the interassay coefficient of variation (CV) was 5%.

FSH and LH serum concentrations were measured by an immunoenzymatic assay with two monoclonal antibodies (Immuno 1, Technicon; Bayer) and data are expressed in terms of IRP 78/549 and 68/40 respectively. The sensitivity of the assays was 0.1 IU/l for FSH and 0.3 IU/l for LH, and interassay CV were 2.7 and 3.1% respectively. The intra-assay CV of the LH assay with varying concentrations of LH were found to be (mean concentration, intra-assay CV): (i) 1.13 IU/l, 1.76%; (ii) 0.66 IU/l, 2.12%; (iii) 0.53 IU/l, 3.10%; (iv) 0.32 IU/l, 3.85%. There was not cross-reactivity of LH with free alpha subunit. Since it has been stressed (Fleming, 1999Go) that a statistically normal distribution of concentrations of serum LH should be observed within the `measurable' range (>0.3 IU/l in our study) in order to validate an LH assay, we tested this point in our study population. Using the one-sample Kolmogorov–Smirnov test which is a test of the goodness of fit between a distribution of observed frequencies and a theoretical frequency distribution, we found that LH values were normally distributed in the whole study population, the non-conception group, and pregnant patients (data not shown). This supports the validity of the present investigation. Total ß-HCG was measured by a solid-phase, two-site chemiluminescent enzyme immunometric assay standardized against the Third International Standard 75/537 (Immulite, Diagnostic Products Co., Los Angeles, CA, USA), with a detection limit of 2 IU/l. The interassay CV was 5.8%.

Ultrasonic scans were performed using a 5 mHz vaginal transducer attached to an Aloka sector scanner (Model SSD-620; Aloka Co. Ltd, Tokyo, Japan).

Statistics and probability testing
For statistical analysis Student's t-test, the Mann–Whitney U-test, Fisher's exact test and {chi}2-test were used as appropriate. Results are expressed as mean ± SE, and as medians and interquartile (25 and 75%) ranges. P < 0.05 was considered significant. The discrimination attained between two study groups (conception versus non-conception assisted reproduction treatment cycles, and ongoing pregnancy versus early pregnancy loss groups) was evaluated with ROC analysis (Hanley and McNeil, 1982Go; Zweig and Campbell, 1993Go). ROC curves are plots of all the sensitivity and specificity pairs which are possible for all levels of a particular parameter. They are constructed by plotting the false positive rate or (100 – specificity) on the x-axis. The y-axis shows the true positive rate or sensitivity. The best threshold value discriminating between two conditions is the value located at the greatest distance from the diagonal.

Calculation of the area under the ROC curve (AUCROC) provides the quantitative measure of accuracy, i.e. the ability of a particular parameter (e.g. LH concentrations on S7) to discriminate between two conditions (e.g. ongoing pregnancy versus early pregnancy loss). An ROC curve representing a parameter with no discrimination at all is a 45° diagonal line from the left lower corner (0% true positive rate and 0% false positive rate) to the upper right corner (100% true positive rate and 100% false positive rate) with an area under the curve of 0.5. A parameter with no overlap between the two conditions will discriminate perfectly and has an ROC curve passing along the y-axis to the upper left corner (100% true positive rate and 0% false positive rate) to end again in the upper right corner with an area under the curve of 1.0.

Data were analysed by Statistics Package for Social Sciences (SPSS, Chicago, IL, USA) statistical software. The Medcalc software (Mariakerke, Belgium) was used to construct ROC curves.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Table IGo shows patient characteristics, gonadotrophin treatment, LH on S7, ovarian response, ovum retrieval, and outcome of IVF/ICSI in conception and non-conception groups. No significant differences were found for any of the variables investigated including LH concentrations on S7. Remarkably, the median and range of LH values in these two groups of patients clearly overlapped (Figure 1Go). The same was true when ongoing pregnancies were compared with early pregnancy losses (Table IGo and Figure 1Go).


View this table:
[in this window]
[in a new window]
 
Table I. Patient characteristics, gonadotrophin treatment, LH on stimulation day 7 (S7), ovarian response, ovum retrieval, and IVF/ICSI outcome in conception versus non-conception groups and ongoing pregnancies versus early pregnancy losses
 


View larger version (16K):
[in this window]
[in a new window]
 
Figure 1. Comparison of serum LH concentration on stimulation day 7 in conception versus non-conception groups (A) and according to the outcome of pregnancy (B). Data are presented as box-plots. Each box represents the middle 50% of the data (25–75% or interquartile range). The central horizontal line represents the median. Vertical lines represent the 10–90% range of data.

 
To analyse the diagnostic accuracy of S7 LH concentration to discriminate between conception versus non-conception cycles, and ongoing pregnancies versus early pregnancy losses, the AUCROC curves determined with ROC analysis are presented (Figure 2Go). Clearly, serum LH concentration on S7 was unable to discriminate between conception and non-conception cycles (AUCROC = 0.52; 95% CI: 0.44 to 0.61) or ongoing pregnancy versus early pregnancy loss groups (AUCROC = 0.59; 95% CI: 0.46 to 0.70).



View larger version (16K):
[in this window]
[in a new window]
 
Figure 2. Receiver operating characteristic curves of LH concentration on stimulation day 7 for discriminating conception versus non-conception cycles (solid line) and ongoing pregnancies versus early pregnancy losses (dashed line).

 
To assess further the potential impact of suppressed concentrations of circulating LH during ovarian stimulation on the outcome of IVF/ICSI treatment, the three threshold values of mid-follicular serum LH proposed in the literature (Chappel and Howles, 1991Go; Fleming et al., 2000Go; Westergaard et al., 2000Go) to discriminate between women with `low' and `normal' LH were applied to our study population (Table IIGo). The proportion of patients having `low' LH on S7 was similar for the whole patient cohort, the conception group, or non-conception patients irrespective of the threshold value considered. Interestingly enough, no significant differences were found with respect to early pregnancy loss rates between `low' and `normal' S7 LH (20 versus 19%; 18 versus 20%; 33 versus 18%; see Table IIGo) women as defined by those threshold values despite that, as expected, mean LH serum concentrations in `low' LH groups were significantly different as compared with `normal' LH groups. When early pregnancy loss rates obtained in the three `low' LH groups as defined by the different threshold values were compared, no significant differences were found (Table IIGo). The similar lack of capacity of LH concentration to discriminate between ongoing pregnancy and early pregnancy loss groups is further stressed by the analysis of oestradiol serum concentrations on day S7 in conception cycles. Oestradiol concentrations on day S7 were higher, albeit not statistically different, in women having `normal' LH as compared with patients with `low' LH, again irrespective of the LH threshold value considered. This was also true when the whole patient cohort (conception and non-conception groups) was analysed (data not shown). Oestradiol concentrations on the day of HCG injection were similar in `low' and `normal' LH groups as were also patient characteristics, gonadotrophin treatment, ovarian response, ovum retrieval and outcome of IVF/ICSI (Table IIIGo).


View this table:
[in this window]
[in a new window]
 
Table II. Patients with `low' or `normal' LH on stimulation day 7 (S7) according to three different threshold values
 

View this table:
[in this window]
[in a new window]
 
Table III. Patient characteristics, gonadotrophin treatment, ovarian response, ovum retrieval, and IVF/ICSI outcome in groups dictated by the LH concentration on stimulation day 7 according to three different threshold values
 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Experimental and clinical experience indicates that LH is not required for follicular growth, but exogenously administered LH probably plays a primary role in complete maturation of the follicle and oocyte competence in patients with long-standing hypogonadotrophic hypogonadism (Balasch et al., 1995Go; Fox et al., 1997Go; Cortvrindt et al., 1998Go). It is not clear, however, whether the resting concentrations of LH after pituitary suppression with GnRH agonists are sufficient to fulfil these requirements in patients displaying different concentrations of LH activity, and whether these low endogenous LH concentrations may in some cases amplify any possible differences in outcome during treatment with HMG and FSH preparations. Thus, in experimental studies in macaques using recombinant FSH and LH, it was attempted to determine if the FSH-only preparation in inducing multiple follicular development will be compromised by the lack of LH activity (Zelinski-Wooten et al., 1995Go; Weston et al., 1996Go). Although normal follicular growth was obtained with recombinant FSH alone after pituitary suppression with a GnRH antagonist, the combination of results was confusing with respect to the IVF and embryo outcome. Similarly, a number of studies comparing FSH and HMG for ovulation induction in patients undergoing assisted reproduction have led to contradictory results (Daya, 1998Go; Söderström-Anttila, 1998Go; Filicori, 1999Go).

Two very recent studies exploring the effects of suppression of LH during ovarian stimulation (days 7–8) in normogonadotrophic women undergoing assisted reproduction (Fleming et al., 2000Go; Westergaard et al., 2000Go) have added further to the controversy. Westergaard et al. (2000) reported that as many as 49% of women stimulated with recombinant FSH under pituitary suppression had low concentrations of LH (<0.5 IU/l) in the mid-follicular phase. In comparison with the normal LH group, these women had serum oestradiol concentrations significantly lower on S8. A total of 64 positive pregnancy tests were recorded and the proportion of them in the two groups (30 versus 34% per started cycle in the low and normal LH groups respectively) was similar. However a 5-fold higher risk of early pregnancy loss (biochemical pregnancies and clinical spontaneous abortions) was seen in the low LH group (45 versus 9%). It has been reported that 26% of women treated with highly purified or recombinant FSH and GnRH agonist had suppressed LH concentration (<=0.7 IU/l) on S7 (Fleming et al. 2000Go). Patients with suppressed LH had much lower oestradiol concentrations at this point and at HCG administration in cycles treated with either urinary or recombinant FSH. However, the gross ovarian response, as became evident by FSH dose demands, duration of stimulation, and also oocyte and embryo yields and embryo cryopreservation, were influenced only in cycles treated with urinary FSH.

The present study adds further to the debate. First, the proportion of `profoundly' suppressed women in our study population (31, 15 and 7% of patients had mid-follicular LH serum concentration <1, <=0.7 and <0.5 respectively) was lower than previously reported (Fleming et al., 2000Go; Westergaard et al., 2000Go). Second, no significant effect of LH suppression on oestradiol biosynthesis by the follicle was observed. Oestradiol serum concentrations on S7 were not significantly different when `low' and `normal' LH groups delineated by the three different LH thresholds were compared. Although, as expected, oestradiol concentration at S7 was somewhat lower in the `low' LH groups, this did not imply a negative influence of low LH activity on the ovary. The following facts support this contention: (i) peak oestradiol concentrations reached after ovarian stimulation in IVF cycles depends both on the number of growing follicles and the presence of sufficient LH action capable of stimulating androgen substrate production from theca cells to be transformed into oestrogen by FSH-stimulated aromatase activity in granulosa cells (Hillier, 1994Go); (ii) oestradiol serum concentrations at the end of ovarian stimulation (HCG injection day) were similar in `low' and `normal' LH groups; (iii) final follicular development and oocyte yield were also similar; (iv) duration of stimulation and total dose of recombinant FSH were similar irrespective of the S7 LH threshold considered.

The above discrepancies between previous studies and ours with respect to ovarian steroidogenesis could be explained on the basis of the GnRH agonist used. A number of studies in the literature have shown that different GnRH agonists have yielded satisfactory results in IVF patients. For hydrophobic GnRH agonists, however, there is a correlation between the hydrophobicity of substituents in position 6 and potency; in general, the potency increases with increasing hydrophobicity (Conn and Crowley, 1991Go). Thus, they may have different effects on different human cells. Both in-vitro (Miró et al., 1992Go) and in-vivo (Balasch et al., 1992Go) studies have shown that the inhibitory effect on ovarian steroidogenesis and follicular development is more evident with the more potent buserelin used by Fleming et al. (2000) and Westergaard et al. (2000) than with leuprolide acetate used in our study. Fleming et al. (2000) reported that with the more potent recombinant FSH treatment, suppression of LH failed to show the impact upon gross ovarian responses that were evident with highly purified urinary FSH. In fact, the same recombinant FSH preparation was used in Fleming's and our studies.

In concordance with previous reports (Fleming et al., 2000Go; Westergaard et al., 2000Go) no differences were observed in our study with respect to ovarian response and IVF/ICSI outcome in `low' and `normal' LH groups irrespective of the LH threshold considered according to those values proposed in the literature. In agreement with the study by Fleming et al. (2000) but in contrast with data from Westergaard et al. (2000), we found that serum LH measurements in the mid-follicular phase did not predict implantation and the outcome of pregnancy. It is notable that in those previous studies (Fleming et al., 2000Go; Westergaard et al., 2000Go) the threshold values for LH (<=0.7 and <0.5 respectively) were established according to the detection limit of LH assays used. We have, on the other hand, used a statistical method using ROC curve analysis to test the usefulness of mid-follicular LH concentration to discriminate between conception versus non-conception cycles and ongoing prenancies versus early pregnancy losses. In ROC curve analysis, many efficiencies of all decision levels can be calculated, resulting in an overall quantification of accuracy which is not affected by the prevalence of a condition (Hanley and McNeil, 1982Go; Zweig and Campbell, 1993Go). Thus, our analysis shows clearly that mid-follicular LH concentrations are not predictive of implantation or early pregnancy loss after IVF/ICSI. In fact, a marked overlapping was found for mid-follicular LH concentrations in the present study when non-conception, conception, ongoing pregnacy, and early pregnancy loss groups were compared.

In conclusion, the present study has shown that serum LH measurements, in the mid-follicular phase during ovarian stimulation with recombinant FSH under pituitary suppression with leuprolide in normogonadotropic women undergoing assisted reproduction, cannot predict ovarian response, IVF/ICSI outcome, implantation, and the outcome of pregnancy. Therefore, our results do not support the need for additional exogenous LH supplementation in down-regulated women receiving a recombinant FSH-only preparation. Finally, irrespective of the role of the agonist, we would stress that researchers who are going to carry out analyses on LH concentrations and the outcome of assisted reproduction should adopt rigorous statistical methods for the evaluation of their results.


    Notes
 
3 To whom correspondence should be addressed at: Institut Clínic of Gynecology, Obstetrics and Neonatology, Hospital Clínic, C/Casanova 143, 08036-Barcelona, Spain. E-mail: jbalasch{at}medicina.ub.es Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Balasch, J., Jové, I.C., Moreno, V. et al. (1992) The comparison of two gonadotropin-releasing hormone agonists in an in vitro fertilization program. Fertil. Steril., 58, 991–994.[ISI][Medline]

Balasch, J., Miró, F., Burzaco, I. et al. (1995) The role of luteinizing hormone in human follicle development and oocyte fertility: evidence from in-vitro fertilization in a woman with long-standing hypogonadotrophic hypogonadism and using recombinant human follicle stimulating hormone. Hum. Reprod., 10, 1678–1683.[Abstract]

Balasch, J., Fábregues, F., Peñarrubia, J. et al. (1998) Follicular development and hormonal levels following highly purified or recombinant follicle-stimulating hormone administration in ovulatory women and WHO group II anovulatory infertile patients. J. Assist. Reprod. Genet., 15, 552–559.[ISI][Medline]

Barbieri, R.L. and Hornstein, M.D. (1999) Assisted reproduction–in vitro fertilization success is improved by ovarian stimulation with exogenous gonadotropins and pituitary suppression with gonadotropin-releasing hormone analogues. Endocr. Rev., 20, 249–252.[Abstract/Free Full Text]

Chappel, S.C. and Howles, C. (1991) Reevaluation of the roles of luteinizing hormone and follicle-stimulating hormone in the ovulatory process. Hum. Reprod., 6, 1206–1212.[Abstract]

Conn, P.M. and Crowley, W.F. (1991) Gonadotropin-releasing hormone and its analogues. N. Engl. J. Med., 324, 93–103.[ISI][Medline]

Cortvrindt, R., Hu, Y. and Smitz, J. (1998) Recombinant luteinizing hormone as a survival and differentiation factor increases oocyte maturation in recombinant follicle stimulating hormone-supplemented mouse preantral follicle culture. Hum. Reprod., 13, 1292–1302.[Abstract]

Daya, S. (1998) hMG versus FSH: is there any difference? In Filicori, M. and Flamigni, C. (eds), Ovulation Induction: Update `98. Parthenon, New York, pp. 183–192.

Fábregues, F., Balasch, J., Manau, D. et al. (2000) Circulating levels of nitric oxide in successful and unsuccessful implantation after in vitro fertilization and embryo transfer. Relationship to estradiol and progesterone. Acta Obstet. Gynecol. Scand., 79, 564–569.[ISI][Medline]

Filicori, M. (1999) The role of luteinizing hormone in folliculogenesis and ovulation induction. Fertil. Steril., 71, 405–414.[ISI][Medline]

FIVNAT (1999) Dossier FIVNAT-99, Bilan de l'année 98.

Fleming, R. (1999) Follicular phase serum LH levels and assisted reproduction outcome. In Filicori, M. (ed.), The Role of Luteinizing Hormone in Folliculogenesis and Ovulation Induction. Monduzzi Editori, Bologna, Italy, pp. 189–199.

Fleming, R., lloyd, F., Herbert, M. et al. (1998) Effects of profound suppression of luteinizing hormone during ovarian stimulation on follicular activity, oocyte and embryo function in cycles stimulated with purified follicle stimulating hormone. Hum. Reprod., 13, 1788–1792.[Abstract]

Fleming, R., Rehka, P., Deshpande, N. et al. (2000) Suppression of LH during ovarian stimulation: effects differ in cycles stimulated with purified urinary FSH and recombinant FSH. Hum. Reprod., 15, 1440–1445.[Abstract/Free Full Text]

Fox, R., Ekeroma, A. and Wardle, P. (1997) Ovarian response to purified FSH in infertile women with long-standing hypogonadotropic hypogonadism. Aust. NZ J. Obstet. Gynaecol., 37, 92–94.[ISI][Medline]

Hanley, J.A. and McNeil, B.J. (1982) The meaning and use of the area under receiving operating characteristic (ROC) curve. Radiology, 143, 29–36.[Abstract]

Hillier, S.G. (1994) Current concepts of the roles of follicle stimulating hormone and luteinizing hormone in folliculogenesis. Hum. Reprod., 9, 188–191.[Abstract]

Howles, C.M. (1996) Genetic engineering of human FSH (Gonal-F). Hum. Reprod. Update, 2, 172–191.[Free Full Text]

Hull, M.G.R., Armatage, R.J. and McDermott, A. (1994) Use of follicle stimulating hormone alone (urofollitropin) to stimulate the ovaries for assisted conception after pituitary desensitization. Fertil. Steril., 62, 997–1003.[ISI][Medline]

Janssens, R.M.J., Lambalk, C.B., Wermeiden, J.P.W. et al. (2000) Dose-finding study of triptorelin acetate for prevention of a premature LH surge in IVF: a prospective, randomized, double-blind, placebo-controlled study. Hum. Reprod., 15, 2333–2340.[Abstract/Free Full Text]

Lévy, D.P., Navarro, J.M., Schattman, G.L. et al. (2000) Exogenous LH: let's design the future. Hum. Reprod., 15, 2258–2265.[Abstract/Free Full Text]

Loumaye, E., Engrand, P., Howles, C.M. et al. (1997) Assessment of the role of serum luteinizing hormone and estradiol response to follicle-stimulating hormone on in vitro fertilization outcome. Fertil. Steril., 67, 889–899.[ISI][Medline]

Miró, F., Sampaio, M.C., Tarín, J.J. et al. (1992) Steroidogenesis in vitro of human granulosa-luteal cells pretreated in vivo with two gonadotropin releasing hormone analogs employing different protocols. Gynecol. Endocrinol., 6, 77–84.[ISI][Medline]

Olijve, W., de Boer, W., Mulders, J.W.M. et al. (1996) Molecular biology and biochemistry of human recombinant follicle stimulating hormone (Puregon). Mol. Hum. Reprod., 2, 371–382.[Abstract]

Peñarrubia, J., Balasch, J., Fábregues, F. et al. (2000) Day 5 inhibin B serum concentrations as predictors of assisted reproductive technology outcome in cycles stimulated with gonadotrophin-releasing hormone agonist-gonadotrophin treatment. Hum. Reprod., 15, 1499–1504.[Abstract/Free Full Text]

Söderström-Anttila, V. (1998) Clinical outcome of ovulation induction: highly purified FSH versus hMG. In Filicori, M. and Flamigni, C. (eds), Ovulation Induction: Update `98. Parthenon, New York, pp. 193–200.

Veeck, L.L. (1986) Atlas of the Human Oocyte and Early Conceptus. Williams & Wilkins, Baltimore.

Westergaard, L.G., Erb, K., Laursen, S. et al. (1996) The effect of human menopausal gonadotrophin and highly purified, urine-derived follicle stimulating hormone on the outcome of in-vitro fertilization in down-regulated normogonadotrophic women. Hum. Reprod., 11, 1209–1213.[Abstract]

Westergaard, L.G., Laursen, S.B. and Andersen, C.Y. (2000) Increased risk of early pregnancy loss by profound suppression of luteinizing hormone during ovarian stimulation in normogonadotrophic women undergoing assisted reproduction. Hum. Reprod., 15, 1003–1008.[Abstract/Free Full Text]

Weston, A.M., Zelinski-Wooten, M.B., Hutchinson, J.S. et al. (1996) Developmental potential of embryos produced by in-vitro fertilization from gonadotrophin-releasing hormone antagonist-treated macaques stimulated with recombinant human follicle stimulating hormone alone or in combination with luteinizing hormone. Hum. Reprod., 11, 608–613.[Abstract]

Zelinski-Wooten, M.B., Hutchinson, J.S., Hess, D.L. et al. (1995) Follicle stimulating hormone alone supports follicle growth and oocyte development in gonadotrophin-releasing hormone antagonist-treated monkeys. Hum. Reprod., 10, 1658–1666.[Abstract]

Zweig, M.H. and Campbell, G. (1993) Receiver-operating characteristic (ROC) plots: a fundamental evaluation tool in clinical medicine. Clin. Chem., 39, 561–577.[Abstract/Free Full Text]

Submitted on February 19, 2001; accepted on April 26, 2001.