1 Human Reproduction Unit, Department of Obstetrics and Gynaecology and 2 Department of Pathology, School of Medicine, Universidade Estadual de Campinas (UNICAMP), Campinas, Brazil
3 To whom correspondence should be addressed at: Caixa Postal 6181, 13084-971, Campinas, SP, Brazil. Email: cpetta{at}attglobal.net
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: endometrial bleeding/intrauterine levonorgestrel/leukocytes/MMP-3/microvessels
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The mechanism leading to episodes of bleeding remains to be elucidated (D'Arcangues et al., 1992; Findlay, 1996
; Hickey et al., 1999
; Fraser and Hickey, 2000
). Changes in the endometrial steroid receptor profile, with down-regulation of the receptor for estrogen and both isoforms of the progesterone receptor (Critchley et al., 1998b
), alterations in endometrial histology, endometrial vascular morphology, function and haemostasis, and endometrial repair mechanisms are all possible mechanisms to explain bleeding (Fraser et al., 1996
).
The role of matrix metalloproteinases (MMP) in endometrial breakdown that triggers menstruation has been described previously (Salamonsen and Woolley, 1996, 1999
). This study aimed to assess the histological appearance, including the number of leukocytes, the microvascular density and calibre, and the quantification of MMP-3 in the endometrium of long-term users of LNG-IUS who maintain some degree of endometrial bleeding or among those who were amenorrhoeic.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The sample size was established in 28 women in each group. This figure was estimated based on a 40% difference between users of LNG-IUS with and without bleeding and with alpha of 0.05 and beta of 0.20. A total of 58 users of LNG-IUS (Mirena®; Leiras/Schering Oy, Finland) for contraception for >3 years were enrolled.
Endometrial specimens were carefully obtained by biopsy (Pipelle de Cornier, Prodimed, Neuilly-en-Thelle, France) at the outpatient clinic without anaesthesia to try to obtain material from the whole wall of the uterine cavity. In the cases in which the material was scanty it was useful for the histological quantitative analysis but not for other analyses. The other findings were obtained from 22 out of 29 samples from the group with bleeding and 18 out of 29 samples from the group with amenorrhoea, because the tissue was atrophic and the material was scant (Mazur and Kurman, 1995) that did not allow for an adequate analysis of those parameters. Usually, these specimens are often reported as insufficient or inadequate for analysis (Wang et al., 1995
; Marbaix et al., 2000
; Vincent et al., 2000
). However, scant tissue may be representative of the lining of the uterine cavity as a whole (Mazur and Kurman, 1995
).
The women included in this study answered a questionnaire about their menstrual pattern after the insertion of the LNG-IUS. Women were considered belonging to the group of no bleeding if they had not presented any uterine bleeding in the 3 months prior to the biopsy. The women assigned to the group with bleeding were those who had presented some uterine bleeding in the 3 months prior to joining the study but were not bleeding at the time the endometrial biopsy was taken.
The biopsies were fixed in 10% buffered neutral formalin, dehydrated, embedded in paraffin and cut at 4 mm, deparaffinized, rehydrated, stained with haematoxylin and eosin (HE) according to standard protocols and submitted to immunohistochemical analysis. The qualitative histological analysis was performed with the use of a Nikon E-200 microscope. The histological and immunohistochemical quantitative analyses were performed on the captured image, with the use of a Nikon E-200 microscope, a Nikon Coolpix-95 videocamera coupled to a computer with a special plate and software for image analysis (Imagelab®, Brazil).
Endometrial histology
The morphometric assessment was similar to that used by Wang et al. (1995): five randomly selected high power fields (x400 objective) were assessed for the qualitative and quantitative analyses. Each sample was classified into the histological pattern, and they were considered suitable for the qualitative and quantitative analyses if the section contained both glands and stroma, and if at least five random fields could be counted at a x400 magnification.
The qualitative analysis included the following histological patterns: (i) proliferative: features consistent with the proliferative phase of the normal cycle; (ii) secretory: consistent with the secretory phase of the normal cycle; (iii) progestogenic: evidence of exogenous progestogenic effects, small glands with cuboidal or low columnar epithelia and decidualized stroma; (iv) atrophic: endometria with very little dense stroma, reduced glands with small cuboidal epithelial cells (Hickey et al., 1999) or scant or absent material (Mazur and Kuman, 1995
). Endometritis (defined by the presence of plasma cells or eosinophils), stromal necrosis, mucosal collapse, superficial erosion with epithelial re-epithelization, glandular pseudostratification and edema, were the other qualitative characteristics analysed in this study.
The number of stromal mitoses, leukocytes, eosinophils, glands, glandular mitoses and measurements of the glandular area and perimeter, major and minor glandular diameter, and height of the glandular epithelium were the parameters morphometrically assessed in five high power fields (x400). The leukocytes were morphologically identified by the size of the cell and the characteristic nuclear pattern. When presenting cytoplasmic granules, they were classified as eosinophil leukocytes and when presenting eccentric nuclei and chromatin with typical distribution they were identified as plasma cells.
Endometrial vessels
The endometrial microvasculature was visualized using immunohistochemical reactions by the avidinbiotinperoxidase method, using antigen recovery with antigen exposure in a pressure steam cooker (15 min, citrate buffer 0.01 mmol/l, pH 6.0). The primary antibody utilized was the mouse monoclonal antibody against human CD34 antigen (Neomarkers, Labvision, USA), at a dilution of 1:50, which is expressed on the endothelial cell membrane (Fina et al., 1990; Hickey et al., 1999
). The secondary antibody was the Envision (Dako) reaction amplifier. Positive and negative reactions were performed by using human placenta as control.
Vessel counting
The greater endometrial vascular density field in each case was chosen in order to standardize the vessel counting. At a magnification;x100, all brown-coloured structures were considered positive, even if a lumen could not be identified (Hickey et al., 1999). The measurements were calculated by the average number of vessels/mm2. The vascular calibre, based on the CD34 immunoexpression, was evaluated using a magnificationx100 objective by selecting the major vessels in each case, and the diameter and area measurements were calculated with the image software.
MMP-3 expression quantification
MMP-3 immunohistochemistry was performed with the use of a monoclonal mouse anti-human MMP-3 antibody (Neomarkers; Labvision, USA), Ab-2 (Clone SL-1 IID 4), at a dilution of 1:50. Stromal cells with fine granular brown-coloured cytoplasm were considered MMP-3 positive. Negative and positive reactions were carried out with human placentas used as reaction control. Perl's histochemical reaction was carried out after the immunostain reaction in order to distinguish the positive immunostaining brown colour from the haemosiderin of the endometrial mucosa. In each case, the most intensely stained regions were selected and 10 fields at magnificationx400 were counted. The numbers of positive and negative cells were counted. All the stroma cells were counted until reaching 1000 cells in each case.
Statistical analysis
The quantitative variables and the qualitative evaluation of histological findings were analysed using the 2-test. The variables for age, duration of contraceptive use, and body mass index (BMI) (kg/m2) were analysed using the MannWhitney U-test. As for the quantitative histological evaluation, the MannWhitney U-test was used for the abnormal distribution and Fisher's t-test for the normal distribution. The MannWhitney U-test was also utilized for the evaluation of density, microvascular calibre and quantification of MMP-3. The significance was established at P<0.05.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Stromal edema was present in the majority of the cases in both groups (P=0.238). Endometritis (morphological identification of plasma cells or eosinophils) was observed in 54.5 and 61.1% in the groups with and without bleeding respectively (P=0.923). Stromal necrosis (P=1), superficial erosion with focal re-epithelialization (P=0.197), mucosal breakdown (P=0.464), and glandular pseudostratification (P=0.187) were absent in the majority of cases in both groups.
Other alterations were observed at low frequency. In the group with bleeding, there was one case of polyp, with intense fibrinous exudate, presence of giant cell and focal proliferative endometrium; two cases with intense neutrophilic exudate; and four cases (18.2%) with many macrophages with haemosiderin. In the group with amenorrhoea, we observed one case with giant cell inflammatory reaction, hyalin material in an enlarged interstitial space, fusiform cells and hyalin decidua; two cases with microcalcifications, and six cases (33.3%) with macrophages with haemosiderin. Stromal mitosis (P=0.191), number of eosinophils (P=0.489) and plasma cells (P=0.911) were similar in both studied groups. Figure 1a shows that the total number of leukocytes was significantly higher in the group with bleeding (P=0.014).
|
Vascular aspects of endometrial biopsies
Vessel counting was performed by two observers, and the variation between them was minimal. The final counting of vascular density was taken as the mean of these values. Endometrial vascular density was (mean ± SD) 25±18.1 and 33±16.3 vessels/mm2 in the groups with and without bleeding respectively (P=0.817) (Figure 1d). Major vascular diameter was (mean ± SD) 2614±1826 µm in the group with bleeding, and 1781±1354 µm in the group in amenorrhoea (P=0.211) (Figure 1e).
MMP-3 immunohistochemical data
Figure 1f shows that in the group with bleeding, the rate of cells expressing MMP-3 by 1000 stromal counted cells (mean ± SD) was 162±119.4, thus being significantly higher than in the group in amenorrhoea (52±52.6) (Figure 2a and b) (P<0.005).
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
It is necessary to take into account that the group of women presenting uterine bleeding had several patterns, and the samples of the endometrium in this group were not collected during the bleeding episode. Therefore, the findings cannot be interpreted as an explanation for the irregular bleeding experienced by some LNG-IUS users, Unpredictable breakthrough bleeding is the most frequent cause of discontinuation of the method of contraception, but our findings could provide an insight into the mechanism of action for bleeding during the use of LNG-IUS. Additionally, in vivo, MMP-3 is secreted and diluted in the extracellular space, where weak immunostaining was reported (Jeziorska et al., 1996).
The finding of no correlation between the bleeding patterns and the histological appearance of the endometrium was described previously (Nilsson et al., 1978; Silverberg et al., 1986
; Hickey et al., 1999
; Pengdi et al., 1999
; Skinner et al., 1999
). In both groups, the majority of samples showed a progestin-modified appearance (Critchley et al., 1998a
; Phillips et al., 2003
) and atrophic endometrium was the second most common finding. These specimens are often reported as insufficient or inadequate for analysis (Wang et al., 1995
; Marbaix et al., 2000
; Vincent et al., 2000
), however, it may be representative of the uterine cavity (Mazur and Kurman, 1995
). These results were similar to previous publications showing that approximately half of the samples obtained from users of Norplant® implants and depot medroxyprogesterone acetate (D-MPA) were not adequate for analysis (Hadisaputra et al., 1996
; Vincent et al., 1999
).
No difference was found between the two groups regarding vascular density. A marked decrease in the number of vessels was observed among women with amenorrhoea (32 vessels/mm2), similar to that observed in users of high doses of oral progesterone (Song and Fraser, 1995) or in the endometrium exposed to progesterone-releasing IUD (Progestasert®) (Shaw et al., 1981
).
Others described an increase in vascular density in users of Norplant implants and LNG-IUS (Hickey et al., 1999; McGavigan et al., 2003
). Our results with the use of CD34 are discordant with other studies that found a great increase in vascular density in the users of Norplant immediately after the insertion (Hickey et al., 1999
) or in users for periods of 2 and 4 months (McGavigan et al., 2003
). The different findings were probably related to the short length of LNG-IUS use in these other series. In agreement with our results, one other study found that LNG was not able to stimulate angiogenesis in mice (Hague et al., 2002
).
Leukocytes were significantly higher in the group with bleeding. This is in agreement with previous opinions that menstruation arises as a result of an inflammatory process (Finn, 1986; Salamonsen and Lathbury, 2000
). Withdrawal or down-regulation of the receptors for progesterone (as occurred with the use of LNG-IUS) may induce the entry or activation of leukocytes into the endometrium, via production of chemokines by endometrial stromal or epithelial cells, which may up-regulate the production of cytokines by means of any of these cell types (Salamonsen and Woolley, 1996
, 1999
).
An increase in infiltrated leukocytes was also observed after 1 month of the insertion of LNG-IUS (Jones and Critchley, 2000) and more frequently in the endometria of women who had bled recently and less frequently in the atrophic endometria of users of lower doses of progestogen orally for contraception (Ludwig, 1982).
The number of leukocytes has been described as varying according to the time of LNG-IUS use (Critchley et al., 1998b) and having no association with the menstrual pattern of the user. It is necessary to recall that our women used the LNG-IUS for >3 years and, on the other hand, in the group of women with bleeding only seven samples out of 29 presented scant or atrophic endometria, which may be one of the explanations for the differences between both groups (Vincent and Salamonsen , 2000
).
Leukocyte number has been demonstrated as important in the normal function of the endometrium, including its role in the regulation of MMP and in endometrial bleeding (Salamonsen and Lathbury, 2000; Salamonsen et al., 2000
). However, few studies have correlated its number with the presence of endometrial breakdown bleeding, a fact observed in our study that showed that the women who maintained some level of bleeding presented a higher number of leukocytes, even with the biopsies not being taken during an episode of bleeding. Our number of cases is bigger than the one observed and can explain our findings. However, one limitation of our study was that leukocytes were evaluated only by morphology, without immunohistochemical staining as performed in one previous study (Vincent et al., 2000
). Our results and those of previous studies (Vincent and Salamonsen, 2000
; Salamonsen et al., 2000
; Salamonsen and Lathbury, 2000
) show the complexity of the relationship between leukocytes and the response to the exogenous steroids even in the case of intrauterine administration.
MMP-3 was significantly higher in women with bleeding. MMP-3 is a protease involved in endometrial breakdown at menstruation, playing a central role in establishing a cascade of MMP activation (Salamonsen and Lathbury, 2000). Once activated, MMP-3 is able to digest various extracellular matrix proteins, including type IV collagen, laminin and fibronectin of the basement membranes (Nagase, 1998
). Proteases released from migratory cells or activated mast cells may activate the latent MMP. Some active MMP can also activate other MMP, thus setting up a cascade of MMP activation. After secretion, MMP are also controlled through the inhibition of enzyme activity by endogenous tissue inhibitors of metalloproteinases (TIMP). The TIMP-1, -2, -3 are located in the endometrium and their role would appear to be maintenance of tissue integrity (Zhang and Salamonsen, 1997
; Skinner et al., 1999
). The MMP degradation of endometrial extracellular matrix and basement membranes results in the fragility of blood vessels and the loss of endometrial tissue (Salamonsen and Woolley, 1996
, 1999
).
The zymography and western blotting studies in users of Norplant implants showed the release of pro-MMP-3, and its levels of activation increased when bleeding endometria were compared with non-bleeding samples. Immunolocalization of MMP-3 occurs in 36% of the bleeding endometria, but virtually no detection was made in endometrial samples during non-bleeding intervals (Marbaix et al., 2000). In our study, the MMP-3 was found in most women who maintained some degree of endometrial bleeding pattern (except in two cases) although the samples were taken during the interval between bleedings. It could be speculated that our results were misinterpreted due to the presence of haemosiderin. However, Perl's reaction was used to avoid biases in the analysis. Our results are in agreement with the observation of intense MMP-9 immunoreactivity in the stromal cells of all endometrial samples in users of LNG-IUS (Skinner et al., 1999
).
The MMP-3 expression was different from that observed with other progestin-only contraceptives. The LNG-IUS delivers higher local concentrations of LNG to the uterine cavity than Norplant implants and D-MPA. Another difference in the present study was that the MMP-3 was found in focal points in all endometrial stromal cells, and not restricted to areas showing stromal breakdown and tissue shedding. The difference could be explained by the previous finding which showed that the high local concentration of progestin could down-regulate the progesterone receptor (Critchley et al., 1998b) and also because progesterone withdrawal was described as the optimal stimulus for inducing all the MMP in the endometrium, including MMP-3 (Salamonsen et al., 1997
).
In conclusion, this study showed that the endometria of long-term users of LNG-IUS who presented some degree of bleeding through its use presented a higher number of leukocytes and MMP-3, compared to women with amenorrhoea, thus suggesting a pivotal role of both in the triggering of endometrial bleeding. The molecular mechanisms leading to the deregulation of leukocytes and MMP, as well as to their activation, need to be further investigated.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Critchley HO, Wang H, Jones RL, Kelly RW, Drudy TA, Gebbie AE, Buckley CH, McNeilly AS and Glasier AF (1998a) Morphological and functional features of endometrial decidualization following long-term intrauterine levonorgestrel delivery. Hum Reprod 13, 12181224.[Abstract]
Critchley HO, Wang H, Kelly RW, Gebbie AE and Glasier AF (1998b) Progestin receptor isoforms and prostaglandin dehydrogenase in the endometrium of women using a levonorgestrel-releasing intrauterine system. Hum Reprod 13, 12101217.[Abstract]
D'Arcangues C, Odlind V and Fraser IS (1992) Dysfunctional uterine bleeding induced by exogenous hormones. In Alexander NJ and D'Arcangues C (eds) Steroid Hormones and Uterine Bleeding. AAAS Press, Washington, pp. 81105.
Fina L, Molgaard HV, Robertson D, Bradley NJ, Monaghan P, Delia D, Sutherland DR, Baker MA and Greaves MF (1990) Expression of the CD34 gene in vascular endothelial cells. Blood 75, 24172426.[Abstract]
Findlay JK (1996) Future directions for research on endometrial bleeding. Hum Reprod 11 (Suppl 2), 179183.[ISI][Medline]
Finn CA (1986) Implantation, menstruation and inflammation. Biol Rev 61, 313328.[ISI][Medline]
Fraser IS and Hickey M (2000) Endometrial vascular changes and bleeding disturbances with long-acting progestins. Steroids 65, 665670.[CrossRef][Medline]
Fraser IS, Hickey M and Song J-Y (1996) A comparison of mechanisms underlying disturbances of bleeding caused by spontaneous dysfunctional uterine bleeding on hormonal contraception. Hum Reprod 11 (Suppl 2), 165179.[ISI][Medline]
Hadisaputra W, Affandi B, Wituaksono J and Rogers PA (1996) Endometrial biopsy collection from women receiving Norplant. Hum Reprod 11 (Suppl 2), 3134.[Medline]
Hague S, Mackenzie IZ, Bicknell R and Rees MC (2002) In-vivo angiogenesis and progestogens. Hum Reprod 17, 786793.
Hickey M, Simbar M, Young L, Markham R, Russel P and Fraser IS (1999) A longitudinal study of changes in endometrial microvascular density in Norplant® implants users. Contraception 59, 123129.[CrossRef][ISI][Medline]
Hidalgo MM, Bahamondes L, Perrotti M, Diaz J, Dantas-Monteiro C and Petta C (2002) Bleeding patterns and clinical performance of the levonorgestrel-releasing intrauterine system (Mirena) up to two years. Contraception 65, 325328.[CrossRef][ISI][Medline]
Jeziorska M, Nagase H, Salamonsen LA and Woolley DE (1996) Immunolocalization of the matrix metalloproteinases gelatinase B and stromelysin 1 in human endometrium throughout the menstrual cycle. J Reprod Fertil 107, 4351.[Abstract]
Jones RL and Critchley HO (2000) Morphological and functional changes in human endometrium following intrauterine levonogestrel delivery. Hum Reprod 15 (Suppl 3) 162172.[Medline]
Ludwig H (1982) The morphologic response of the human endometrium to long-term treatment with progestational agents. Am J Obstet Gynecol 142, 796808.[ISI][Medline]
Marbaix E, Vekemans M, Galant C, Rigot V, Lemoine P, Dubois D, Picquet C, Henriet P, Twagirayezu P, Sufi S et al. (2000) Circulating sex hormones and endometrial stromelysin-1 (matrix metalloproteinase-3) at the start of bleeding episodes in levonorgestrel-implant users. Hum Reprod 15 (Suppl 3), 120134.[Medline]
Mazur MT and Kurman RJ (1995) Effects of hormones. In Mazur MT and Kurman RJ (eds) Diagnosis of Endometrial Biopsies and Curettings. A Practical Approach. Springer Verlag, New York, pp. 109130.
McGavigan CJ, Dockery P, Metaxa-Mariatou V, Campbell D, Stewart CJ, Cameron IT and Campbell S (2003) Hormonally mediated disturbance of angiogenesis in the human endometrium after exposure to intrauterine levonorgestrel. Hum Reprod 18, 7784.
Milson I, Andersson K and Andersch B (1991) A comparison of fluriprofen, tranexamic acid and a levonorgestrel-releasing intra-uterine contraceptive device in the treatment of idiopathic menorrhagia. Am J Obstet Gynecol 169, 879883.
Nagase H (1998) Stromelysins 1 and 2. In Parks WC and Mecham RP (eds) Matrix Metalloproteinases. Academic Press, San Diego, pp. 4384.
Nilsson CG, Luukkainen T and Arko H (1978) Endometrial morphology of women using a D-norgestrel releasing intrauterine device. Fertil Steril 29, 397401.[ISI][Medline]
Pengdi Z, Xiaoqun L, Hongzhi L, Zhao G, Jie C, Ruhua X, Shizhu L, Shangchun W and Jiedong W (1999) The effect of a levonorgestrel-releasing intrauterine device on human endometrial oestrogen and progesterone receptors after one year of use. Hum Reprod 14, 970975.
Phillips V, Graham CT, Manek S and McCluggage WG (2003) The effects of the levonorgestrel intrauterine system (Mirena coil) on endometrial morphology. J Clin Pathol 56, 305307.
Salamonsen LA and Lathbury L (2000) Endometrial leukocytes and menstruation. Hum Reprod Update 6, 1627.
Salamonsen LA and Woolley DE (1996) Matrix metalloproteinases and their tissue inhibitions in endometrial remodeling and menstruation. Reprod Med Rev 5, 185203.
Salamonsen LA and Woolley DE (1999) Menstruation: induction by matrix metalloproteinases and inflammatory cells. J Reprod Immun 44, 127.[CrossRef]
Salamonsen LA, Butt AR, Hammond FR, Garcia S and Zhang J (1997) Production of endometrial matrix metalloproteinases, but not their tissue inhibitions, is modulated by progesterone withdrawal in an in vitro model for menstruation. J Clin Endocrinol Metab 82, 14091415.
Salamonsen LA, Zhang J, Hampton A and Lathbury L (2000) Regulation of matrix metalloproteinases in human endometrium. Hum Reprod 15 (Suppl 3), 11121119.
Shaw ST, Jr, Macaulay LK, Aznar R, Gonzalez-Angulo A and Roy S (1981) Effects of a progesterone-releasing intrauterine contraceptive device on endometrial blood vessels: a morphometric study. Am J Obstet Gynecol 114, 821827.
Silverberg SG, Haukkamaa M, Arko H, Nilsson CG and Luukkainen T (1986) Endometrial morphology during long-term use of levonorgestrel-releasing intrauterine devices. Int J Gynecol Pathol 5, 235241.[ISI][Medline]
Skinner JL, Ruley SC, Gebbie AE, Glasier AF and Critchley HO (1999) Regulation of matrix metalloproteinase-9 in endometrium during the menstrual cycle and following administration of intrauterine levonorgestrel. Hum Reprod 14, 793799.
Song JY and Fraser IS (1995) Effects of progestogens on human endometrium. Obstet Gynecol Surv 50, 385394.[CrossRef][Medline]
Vincent AJ, Malakooti N, Zhang J, Rogers PA, Affandi B and Salamonsen LA (1999) Endometrial breakdown in women using Norplant is associated with migratory cell expressing of matrix metalloproteinase-9 (gelatinose B). Hum Reprod 14, 807815.
Vincent AJ, Zhang J, Ostor A, Rogers PA, Affandi B, Kovacs G and Salamonsen LA (2000) Matrix metalloproteinase-1 and -3 and mast cells are present in the endometrium of women using progestin-only contraceptives. Hum Reprod 15, 123130.
Wang IYS, Russel P and Fraser IS (1995) Endometrial morphometry in users of intrauterine contraceptive devices and women with ovulatory dysfunctional uterine bleeding: a comparison with normal endometrium. Contraception 51, 234243.
Zhang J and Salamonsen LA (1997) Tissue inhibitor of metalloproteinases (TIMP)-1, -2 and -3 in human endometrium during the menstrual cycle. Mol Hum Reprod 3, 735741.[Abstract]
Submitted on September 17, 2003; accepted on May 7, 2004.