| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Gynecology (S.R., G.B., A.B., J.-M.F., A.P.), Centre Hospitalier Universitaire, and Laboratory of Tumor and Development Biology (C.M., S.B., S.L., A.B., J.-M.F., A.P.), University of Liege, Tour de Pathologie (B23), Center of Experimental Cancer Research, ’GIGA-R’, B-4000 Liège, Belgium; Department of Obstetrics, Gynecology, Reproductive Medicine and Public Health (N.C.-B.), Hospital Tenon, Assistance Publique-Hôpitaux de Paris, Paris 75020, France; Equipe d’ Accueil 1533 Université Pierre et Marie Curie Université Paris 06 (N.C.-B., P.B.), 75005 Paris, France; and Endocrinology Unit (P.B.), Hospital Saint-Antoine, Assistance Publique-Hôpitaux de Paris, 75012 Paris, France
Address all correspondence and requests for reprints to: Jean-Michel Foidart, Département de Gynécologie et Obstétrique, Boulevard 12e de Ligne, 1, B-4000 Liège, Belgium. E-mail: jmfoidart{at}ulg.ac.be.
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Objective: Our objective was to quantify the effects of VA-2914 on endometrial vascularization, fibrillar matrix, and vascular endothelial growth factor (VEGF)-A expression in endometrial biopsies from 41 women before and after 12 wk daily treatment with a placebo, or 2.5, 5, or 10 mg VA-2914.
Methods: Collagen fibrillar network was stained by silver impregnation. Vessel area, density, and structure were quantified with a computer-assisted image analysis system after double immunostaining using an anti-von Willebrand factor (endothelial cells) and an anti-
smooth muscle actin (vascular smooth muscle cells) marker antibody. VEGF-A mRNAs were quantified by RT-PCR and localized by immunohistochemistry.
Results: The endometrial vessels, collagen network, and mRNA levels of VEGF-A were identical during the luteal phase at baseline and in VA-2914 treated women. VEGF-A distribution was unchanged.
Conclusions: VA-2914 does not alter the endometrial matrix and cells, and does not modify the endometrial vessel morphology as compared with baseline biopsies.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Progestin administration causes endometrial vessel remodeling characterized by an increased number of dilated and "naked" vessels devoid of pericytes that are associated with endometrial bleeding (2).
Progesterone receptor ligands include not only pure agonists or antagonists but also compounds with mixed activity. They all belong to the class of selective progesterone receptor modulators (PRMs). Many PRMs display direct antiproliferative effects in the endometrium, justifying their possible use in the treatment of endometriosis and myoma-associated bleeding. They also suppress late follicular development, block the LH surge, and retard endometrial maturation, which renders them potential estrogen-free contraceptive drugs (3).
VA-2914 is a steroid derived from 19-norprogesterone with potent in vitro and in vivo progesterone antagonist activity (4, 5). Acute and chronic toxicology studies in animals indicate a satisfactory safety profile, with considerably less antiglucocorticoid activity in cell culture (4) and in the rat (6) than that of mifepristone. It inhibits ovulation in rats in a dose-dependent manner upon single-dose oral administration and exhibits, by a decrease in implantation sites, an antifertility activity during continuous low-dose administration (7). VA-2914 is also effective as postcoital contraception in animal models (6, 7) and women (8).
A series of phase I clinical studies has been performed to determine the effects of VA-2914 on ovulation and endometrial maturation in women (9). We recently evaluated the effects of VA-2914, administrated continuously for 3 months, on ovulation and endometrial maturation in 46 normal women. Anovulation was observed in 80% of women in the 5 and 10 mg groups. Plasma estradiol (E2) levels remained in the physiological follicular phase range throughout treatment. No endometrial hyperplasia was noted. Amenorrhea occurred in 82 and 90% of women in the 5 and 10 mg groups (10).
In this study we evaluate the effects of low-dose VA-2914 on the degradation of the endometrial matrix, as well as on endometrial vessel characteristics and the expression of vascular endothelial growth factor (VEGF)-A, an angiogenic cytokine known to play a key role on blood vessel growth and maturation (11).
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The study was approved by the Ethics Committees at the University of Liege and Saint-Antoine Hospital of Paris. Written informed consent was obtained from each subject. There were 46 women, aged 18–35 yr, with documented regular ovulatory menstrual cycles (25–32 d), within 20% of ideal body weight, not currently using an intrauterine device, and not having used reproductive hormonal medications or glucocorticoids for at least 12 wk before the study, randomly assigned to either a placebo or one of the three treatment groups (2.5, 5, or 10 mg/d VA-2914; HRA Pharma, Paris, France). The therapy was started on the first or second day of menstrual bleeding after a baseline observational cycle and taken daily for 12 wk. Women agreed to prevent pregnancy by abstinence or use of barrier contraception, or they were protected by tubal ligation or partner vasectomy. An endometrial biopsy was performed on d 6–8 after the LH surge determined by a urinary test during baseline cycle and treatment month 3. Endometrial biopsy was performed on treatment d 77 in the absence of detectable urinary LH surge during treatment month 3.
Hormonal analysis
E2 was measured by a direct RIA using Diasorin reagents (Anthony, France). The cross-reactivity of estrone and estriol was only 0.6%. Coefficients of variation at the level of 50 pmol/liter were 3.6 and 10.4%, respectively, and at the level of 260 pmol/liter, 2.9 and 6.8%, respectively.
Progesterone was measured by RIA using CIS Bio International reagents (Gif sur Yvette, France). The only significant cross-reacting steroids were: deoxycorticosterone 6.2%, 20-dihydroprogesterone 2.2%, and 6β-dihydroprogesterone 2.1%. Coefficients of variation at the level of 10 ng/ml were 3.5 and 4.5%, respectively, and at the level of 18.8 ng/ml, 4.4 and 4%, respectively (10).
Endometrial tissues
Human endometrial biopsies were obtained from 41 women who completed the study with a Cornier Pipelle suction curette (C.C.D. International, Paris, France). A fragment was immediately snap frozen and stored in liquid nitrogen until analysis. The other part of the biopsy was fixed in 4% formaldehyde solution in PBS and embedded in paraffin. Histological sections were stained with hematoxylin and eosin. Specimens were evaluated in a blinded fashion by two independent pathologists who classified them as proliferative or early (d 17–19), mid (d 20–22), or late (d 24–26) secretory according to classical criteria (12). The immunohistology and quantitative analysis were performed on an entire tissue section (5 µm thick) with a mean size of 25 mm2 and a mean number of gland sections of 100.
von Willebrand factor (vWF)/-smooth muscle actin (SMA) double immunostaining
Serial sections were mounted on silanized slides and used for immunohistochemistry. Endometrial vessels were identified by double immunostaining of endothelial cells and smooth muscle cells. The primary antibodies (Abs) used were a rabbit polyclonal antihuman vWF (A082; Dako Denmark A/S, Glostrup, Denmark) and a mouse monoclonal anti-SMA (clone 1A4, A2547; Sigma-Aldrich Corp., St. Louis, MO). Slides were washed in Tris-HCl (pH 7.6) between all steps unless otherwise stated. Endogenous peroxidase was blocked with 3% H2O2 for 20 min at room temperature. Slides were incubated with normal sheep serum (Hormonology Laboratory, Marloie, Belgium) for 30 min at room temperature, directly followed by vWF Ab (1:500 in normal sheep serum/Tris 10%) overnight at 4 C. A swine antirabbit Ig conjugated to peroxidase (P0217; Dako Denmark A/S) was used as a secondary Ab. 3'3-diaminobenzidine HC1 (DAB+) (Liquid DAB+ substrate chromogen system, K3468; Dako Denmark A/S) was applied for 15 min at room temperature in the dark as a chromogen, and sections were rinsed in H2O. Slides were incubated with normal goat serum (Hormonology Laboratory) before incubation with the second primary Ab, SMA (1:400 in normal goat serum/Tris 10%), for 90 min at 37 C. A goat antimouse Ab conjugated to biotin (E 0433; Dako Denmark A/S) diluted 1:400 in Tris buffer for 30 min at room temperature was used as a secondary Ab, followed by an incubation for 30 min at room temperature with streptavidin-alkaline phosphatase (D 0396; Dako Denmark A/S) diluted 1:500. Finally, Fast Red chromogen system (K4016; Dako Denmark A/S) was applied for 10 min at room temperature in the dark. Sections were rinsed in H2O and mounted in Aqua Polymount (Polysciences, Inc., Warrington, PA). Negative control was performed by replacing each primary Ab with normal serum.
VEGF-A immunohistochemistry
VEGF-A protein was localized in sections of paraffin-embedded endometrial biopsies. Sections were stained with VEGF-A Ab by the following steps. After dewaxing and rehydrating, tissue sections were subjected to antigen retrieval in pressure cook at 1.4 bar and 126 C for 11 min using citrate buffer, and were then allowed to cool for 20 min. Endogenous peroxidases were blocked in 3% H2O2 for 20 min at room temperature, and nonspecific binding of the primary Ab was blocked with 10% BSA. Incubation with rabbit polyclonal antihuman VEGF-A Ab (A-20; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) diluted 1:50 was conducted 60 min at room temperature. Sections were washed in PBS (5x 5 min) before incubation with the secondary Ab (goat antirabbit EnVision/HRP, K4003; Dako Denmark A/S) for 30 min at room temperature. Finally, staining was revealed with the 3'3-diaminobenzidine HCl system, and sections were counterstained with hematoxylin.
Location and intensity of immunostaining were measured using a semiquantitative scoring system. This scoring system is a standard method used in previous studies (13, 14, 15, 16), and a high correlation has been demonstrated between objectively measured immunoreactivity (image analysis) and semiquantitative scoring of immunostaining patterns (16). Immunostaining intensity and distribution of epitopes in all tissue sections were assessed on a four-point scale: 0 = no staining, 1 = mild staining, 2 = moderate staining, and 3 = intense staining, and expressed as a mean score and range.
VEGF-A mRNA expression
Frozen tissue was pulverized using a Dismembrator (B. Braun Biotech International, GmBH, Melsungen, Germany) and suspended in lysis buffer [5 M guanidine thiocyanate, 25 mM sodium citrate, 17 mM N-lauroylsarcosine, and 0.1 M 2-mercaptoethanol (pH 7)]. Total RNA was purified by the QIAGEN method (QIAGEN Benelux B.V., Venlo, The Netherlands). The concentration and purity of RNA were determined by spectrophotometry. RT-PCR amplification was performed using the GeneAmp Thermostable rTth Reverse Transcriptase RNA PCR Kit (PerkinElmer, Branchburg, NJ), and the following specific pair of primers: VEGF-A, 5'-TCTACCTCCACCATGCCAAGT-3' and 5'-GCTGCGCTGATAGACATCCA-3' (104 bp product); and 28S, 5'-GTTCACCCACTAATAGGGAACGTGA-3' and 5'-GATTCTGACTTAGAGGCGTTCAGT-3' (212 bp product). Reverse transcription was performed from 10 ng RNA total at 70 C for 15 min, followed by 2 min at 95 C to denature DNA. PCR amplification was run as follows: 15 sec at 94 C, 30 sec at 60 C, and 15 sec at 72 C during 31 cycles for VEGF-A; and 15 sec at 94 C, 20 sec at 68 C, and 10 sec at 72 C during 15 cycles for 28S and followed by a final 2-min extension step at 72 C. RT-PCR products were resolved on 10% polyacrylamide gels and analyzed using a Fluor-S Multimager (Bio-Rad Laboratories, Inc., Hercules, CA) after staining with Gelstar dye (FMC BioProducts, Rockland, ME). Products were quantified by normalization with respect to 28S rRNA. The levels of expression were compared at baseline and in a combined group of women exposed to various doses of VA-2914.
Staining of the argyrophilic fibrillar network
Silver impregnation of histological sections was performed as previously described (17).
Image analysis and measurement
Slides were observed with an Olympus microscope (Omnilabo, Aartselaar, Belgium), and the entire section (mean area 25 mm2) was analyzed at magnification of x400. Vessels were identified by positive vWF staining, and the presence of pericytes was documented according to the SMA staining as previously described in detail (2). Briefly, a coating of the endothelial cells with SMA positive cells (pericytes) was considered as terminal venules or arterioles, whereas "naked" vessels were considered to be capillaries. A score (zero or one) was attributed to blood vessels according to the absence or presence of SMA staining. Vessels were counted in the entire biopsy.
The contour of each vessel type was drawn manually at 400 x magnification using Photoshop software (Adobe Systems, Inc., San Jose, CA) by two different colors: blue for SMA negative vessels (score 0), and red for SMA stained vessels (score 1), Tissue boundaries were delineated in green and (Fig. 1A
). The color image was then decomposed into its three components: red, green, and blue (18). Figure 1, B and C, shows binary images of vessels scored zero and one, respectively.
|
Statistical analysis
The parameters were expressed in terms of the mean ± SE. Analyses for statistical significance were evaluated by the Kruskal-Wallis test with Dunn’s correction for multiple comparisons. Then, for comparison between two groups, the Mann-Whitney U test and paired Wilcoxon were used. Statistical significance was set at P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Patient’s characteristics and endometrial histology are described in Table 1 and have been reported in detail elsewhere (10). Briefly, all women in the placebo group and most patients (nine of 11) exposed to VA-2914 2.5 mg/d experienced normal menstrual cycles, whereas amenorrhea was observed in the majority of women receiving 5 or 10 mg/d (Table 1). During the baseline cycle, 32 biopsies showed typical features of the midsecretory phase, and one showed a polyp. No tissue could be obtained from eight women. After 12 wk treatment with VA-2914, 17 biopsies were classified as a secretory pattern, and one as proliferative. No tissue was obtained from 11 patients. Patients with polyps were excluded for further analysis.
|
) and after 12 wk continuous treatment with VA-2914 (2.5–10 mg/d) (Fig. 2, B–D) were similar.
|
. All women in the placebo-treated group exhibited an endometrial secretory pattern with E2 and progesterone plasma levels typical of a midluteal phase. Among the six women exposed to VA-2914 2.5 mg/d, three had progesterone plasma levels above 6 ng/ml, indicative of an ovulation, and their endometrium also showed the expected secretory pattern. Three other women in that group had progesterone plasma levels below the range of an ovulatory cycle (0, 0, and 2 ng/ml, respectively). In two of them, an endometrial pattern of a secretory phase was documented. Among the 13 women exposed to VA-2914 5 or 10 mg/d, only three had progesterone plasma levels above 6 ng/ml (24, 16, and 13 ng/ml, respectively). In the 10 other women, progesterone plasma levels were below the detection limit. Therefore, the histological demonstration of a secretory pattern in these women should be ascribed to a VA-2914 injury. In all women, the E2 plasma levels were in the range of a spontaneous midfollicular phase. Altogether, these data suggest that exposure to 5–10 mg/d VA-2914 blocks ovulation but does not induce the histological changes classically elicited by antigonadotropic progestins.
|
Staining of argyrophilic fibrillar network documented the integrity of the collagen-rich fibrillar network in all biopsies at baseline and after treatment with placebo or all doses of VA-2914 (Fig. 2
, E and F). This is in sharp contrast with the stromal breakdown and extensive endometrial matrix proteolysis in women with progestin-only contraceptives (1, 19). This indicates that VA-2914 does not elicit the extensive endometrial stromal breakdown observed after progestin exposure.
Microvessel density and area
Altogether, more than 12,000 vessel sections were counted and characterized. Capillaries were identified by the isolated presence of vWF positive endothelial cells (Fig. 2G), whereas a continuous layer of SMA positive cells coating the endothelial cells characterized small arterioles and venules (Fig. 2H).
The relative vascular area (ratio of vessel section surface to total tissue area) was determined in endometrial biopsies of 32 women at baseline during the midsecretory phase and after treatment with VA2914 (Fig. 3, A and B). At baseline, about 80% of the vascular areas were occupied by vessels coated with SMA positive cells. After 12 wk treatment with various doses of VA-2914, this proportion remained unchanged (Fig. 3B).
|
). Vessels coated with smooth muscle cells were larger than "naked" capillaries in all groups. Their respective sizes were identical at baseline and during treatment with VA-2914. No difference in vessels counts, density, and size was observed with a Mann-Whitney U test. The comparison of values from each individual woman at baseline and after treatment by a paired Wilcoxon test failed also to show a significant difference. VEGF-A immunolocalization
During the baseline cycle, VEGF-A was strongly localized in the apical part of the cytoplasm of the surface epithelium [score 2.5 (range 1–3)] and of glandular epithelial cells [score 2.9 (range 2–3)]. Mild expression was also found in the stroma [score 0.8 (range 0–1)] and endothelial cells [score 0.8 (range 0–1)] (Table 3 and Fig. 2I). VEGF-A localization remained unchanged in all cellular compartments after 12 wk treatment with placebo or VA-2914 (Table 3 and Fig. 2J).
|
The mRNA expression of VEGF-A was analyzed by RT-PCR in endometrial tissue from 11 women at baseline and from eight women treated with various doses of VA-2914. The VEGF-A mRNA levels remained unchanged (data not shown).
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
). Atrophy has been described at 10 mg dosage (Table 1). In this study, 82 and 90%, respectively, of women became amenorrheic during the 12 wk daily treatment with 5 or 10 mg (see also Ref. 10). Although a comparison between endometria after treatment with oral VA-2914 and after treatment with an oral progestin such as oral levonorgestrel (LNG) is an appropriate design for a subsequent clinical study, this phase IIA protocol aimed at evaluating in a limited set of women the impact of VA-2914 upon the endometrial histology, ovulation, and endocrine ovarian activity in women with documented spontaneous ovulatory cycle. This first clinical study indicates that a 12-wk exposure to VA-2914 5 or 10 mg/d causes anovulation, amenorrhea, and a persistent E2 secretion with an endometrial pattern that closely resembles that of a spontaneous ovulatory cycle during the midluteal phase. This pattern is strikingly distinct from that observed with other PRMs that cause endometrial hyperplastic changes that are worrisome for chronic administration (22). Argyrophilic staining of the collagen fibrillar network showed no evidence of endometrial matrix degradation during the midsecretory phase of spontaneous ovulatory cycle (control) or in women exposed to VA-2914. Focal stromal breakdown, i.e. menstrual-like tissue collapse and fragmentation, and lysis of the collagen-rich argyrophilic fibrillar network are classically observed in endometrial biopsies performed at menstruation or at the start of a bleeding episode in women receiving sc or intrauterine LNG as progestin (1). Such focal and stromal breakdown and collagen fiber lysis are evidenced in bleeding endometria, and associated with higher activities of collagenase-1 and gelatinases A and B together with lower tissue inhibitor of metalloproteinase-1 than in nonbleeding endometria (1, 19). The absence of matrix remodeling and collagen lysis in normal secretory endometria and in women receiving VA-2914 suggests that the endometrial changes elicited by menstruation and/or prolonged progestin exposure are absent in our patients.
We have shown previously that extensive endometrial remodeling was induced by exogenous LNG administration in the form of an intrauterine system (2). Endometrial samples obtained in eight women exposed for 1 month to the LNG-releasing intrauterine system displayed an 11.5-fold increase in small naked vessel number and a 6-fold increase in pericyte-coated vessels. In women using the long-term LNG-releasing intrauterine system, a 4-fold increase in the number and size of fragile capillaries was observed (2). We hypothesized that such an increased number of fragile "naked" capillaries could easily be injured and responsible for bleeding. Therefore, we compared in this study the impact of VA-2914 or of endogenous progesterone during the midsecretory phase of spontaneous ovulatory cycle (baseline cycle) on the pattern of endometrial vessels. Endometrial VA-2914 exposed women exhibited identical blood vessel number, area, density, and maturation in comparison to the pattern of normal spontaneous secretory endometrium under the influence of endogenous E2 plus progesterone. This pattern of endometrial vessels was observed independently of the endogenous levels of E2 and progesterone (Table 2). In the same patients, we documented previously the persistence of physiological E2 levels (10). The similarity of pattern in the presence of endogenous progesterone or of VA-2914 is somewhat surprising because VA-2914 is not a progestin but a selective modulator of the progesterone receptor. This pattern may contribute to explain the high level of amenorrhea observed in our patients. Additional clinical study comparing the impact of orally given synthetic progestin in the presence or absence of VA-2914 will be necessary to delimit more precisely the impact of this new drug on the endometrial physiology.
The presence of VEGF-A mRNA and protein have been demonstrated in the human endometrium throughout the menstrual cycle (23). Most studies, including ours, have shown predominant localization of VEGF-A in the apical part of surface and glandular epithelial cells compared with the stroma (24). In this study, VA-2914 did not modify the VEGF-A distribution or its relative staining intensity. Thus, this PRM acts in a way distinct from that of mifepristone, which increases microvessel density and decreases stromal VEGF in endometria of women exposed for 120 d to mifepristone (13).
Altogether, this preliminary phase II study performed on a limited number (46) of women confirms that the PRM VA-2914 maintains matrix and vessel stability, does not impair the expression of the angiogenic cytokine, VEGF-A, and allows the maintenance of a stable vascular network that closely resembles that observed during the secretory phase of spontaneous ovulatory cycle despite anovulation and amenorrhea. This drug could, in this way, be associated with an improved clinical tolerance profile because it is devoid of the endometrial matrix and vascular remodeling associated with menstruation or with the bleeding conveyed by withdrawal of progesterone.
|
Footnotes |
|---|
First Published Online August 26, 2008
Abbreviations: Ab, Antibody; DABt, 3'3-diaminobenzidine HC1; E2, estradiol; LNG, levonorgestrel; PRM, progesterone receptor modulator; SMA, smooth muscle actin; VEGF, vascular endothelial growth factor; vWF, von Willebrand factor.
Received April 2, 2008.
Accepted August 18, 2008.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
, is present in the vascular endothelium of the human and nonhuman primate endometrium. J Clin Endocrinol Metab 86:1370–1378
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
