Effect of hydrosalpinx fluid on secretion of trophoblastic matrix metalloproteinases

FERTILITY AND STERILITY VOL. 77, NO. 3, MARCH 2002 Copyright 2002 American Society for Reproductive Medicine Published by Elsevier Science Inc. Printed on acid-free paper in U.S.A. Effect of hydrosalpinx fluid on secretion of trophoblastic matrix metalloproteinases Nicole Jastrow, M.D., Didier Chardonnens, M.D., Melvina Araman, Ph.D., Arielle Meisser, Ph.D., Aldo Campana, M.D., and Paul Bischof, Ph.D. Infertility and Gynaecological Endocrinology Clinic, WHO Collaborating Centre, University Hospital of Geneva, Geneva, Switzerland Objective: To determine if hydrosalpinx fluid affects trophoblastic metalloproteinases (MMPs) secretion. Design: Measurement of the effect of hydrosalpinx and peritoneal fluids (as controls) added to the medium on the MMPs secreted by cytotrophoblastic cells. Setting: Academic research center. Patient(s): Five samples of hydrosalpinx fluid were obtained at the time of ovocyte retrieval. Three samples of peritoneal fluids were collected at laparoscopic sterilization. Main Outcome Measure(s): The concentration and activity of MMP-2 and MMP-9, the concentration of the tissue inhibitor of metalloproteinases (TIMP-1), and the total gelatinolytic activity of the cytotrophoblastic cells were measured in the culture medium. Result(s): Hydrosalpinx significantly stimulated MMP-2, MMP-9, and TIMP-1. The net result was a significant stimulation of the total gelatinolytic activity. Peritoneal fluids increased MMP-2, MMP-9, and TIMP-1 concentrations, but the total gelatinolytic activity was not modified. Conclusion(s): In contrast to peritoneal fluids, hydrosalpinx stimulates the total gelatinolytic activity of cytotrophoblastic cells. This might indicate that the effect of hydrosalpinx on implantation rates may not be due to an inhibition of the capacity of an embryo to invade the endometrium. However, the stimulatory effect of hydrosalpinx on the net gelatinolytic activity could partly explain the increased incidence of ectopic pregnancies that have been described in the presence of hydrosalpinx. (Fertil Steril 2002;77:588 94. 2002 by American Society for Reproductive Medicine.) Key Words: Hydrosalpinx, MMP, gelatinases, infertility, trophoblast Received February 15, 2001; revised and accepted September 5, 2001. Supported by grant 32-49257.96 from the Swiss National Fund for Scientific Research to PB. Reprint requests: Paul Bischof, Ph.D., Laboratoire d Hormonologie, Maternité, 1211 Geneva 14, Switzerland (FAX: 41 22 382 43 10; E-mail: paul. bischof@hcuge.ch). 0015-0282/02/$22.00 PII S0015-0282(01)03011-4 Numerous studies have investigated in vitro fertilization (IVF) outcome in patients with hydrosalpinges, and all have observed reduced pregnancy rates and/or implantation rates in these patients (1 5). Similar observations were reported following the transfer of cryopreserved embryos (6). Various hypotheses have been proposed to explain the lower implantation and pregnancy rates in the presence of hydrosalpinx. The leakage of the fluid into the uterine cavity may either render the endometrium hostile to embryo implantation or interfere with embryo development, or it may simply prevent interaction between the embryo and the endometrium through a washout mechanism. The present study explored whether hydrosalpinx fluid prevented implantation of the embryo by inhibiting the secretion of the enzymes responsible for the invasive behavior of the blastocyst. Human embryos are not available for research, so the invasive behavior of the embryo has been mimicked with cytotrophoblastic cells isolated from legal abortions in early pregnancy (6 to 12 weeks of gestation); these cells have been shown to invade an extracellular matrix (7). Cytotrophoblastic cells secrete many proteolytic enzymes, among them the metalloproteinases (MMPs), the only enzymes capable of digesting the components of extracellular matrix (8, 9). MMPs are divided into three subfamilies according to their substrate specificity: collagenases, gelatinases, and stromelysines. Gelatinases are represented by two enzymes, MMP-2 and MMP-9, which digest collagen IV (the major constituent of basement membranes). They are thus considered as the rate-limiting factors of cytotrophoblastic cell invasion (7). Cytotrophoblastic cells secrete gelatinases very 588

early in embryo development, as human blastocysts (10) or even triploid eight-cell human embryos (11) produce MMPs. MMP activity is regulated by many factors, such as the tissue inhibitors of MMPs (TIMPs). So far, four TIMPs (TIMP-1, TIMP-2, TIMP-3, and TIMP-4) have been described. They are secreted locally in the extracellular space where they control the activity of MMPs (12). Thus, they allow a precise regulation of invasion. The aim of the present study was to determine whether hydrosalpinx fluid affects trophoblastic gelatinases and/or TIMP-1 secretion, as it has been shown to be the principal inhibitor of MMP-9 (13). To the best of our knowledge, this is the first investigation of the effect of hydrosalpinx on the enzymes responsible for invasion of the embryo. MATERIALS AND METHODS Patients Five hydrosalpinx fluids were collected from patients who underwent a long protocol for in vitro fertilization. The mean age of the women was 35.2 years (range: 30 to 40 years). One woman had tubal factor infertility; other causes of infertility were endometriosis (n 1) or male factor (n 1). The protocol included a pituitary downregulation with at least 10 days of GnRH analogs followed by an ovarian stimulation with either hmg or recombinant FSH until the size of one of the follicles was 20 mm. At that point, hcg (5,000 10,000 IU) was given, 34 hours before pick-up of the oocytes. Hydrosalpinx fluids were obtained at the time of oocytes retrieval by transvaginal echoguided aspiration. Three samples of peritoneal fluids were collected from three healthy women at the time of a laparoscopic sterilization. This was performed through a suprapubic puncture. The mean age of the women was 38.7 years (range: 38 to 40 years). One of them was using oral contraceptives until sterilization was performed. The hydrosalpinx fluids and peritoneal fluids were immediately centrifuged and stored at 20 C until used. The ethics committee at our institution approved this study, and written informed consent was obtained from the patients. Preparation of Cytotrophoblastic Cells and Culture Conditions Cytotrophoblastic cells were isolated, purified, and cultured as previously described (8). Briefly, trophoblastic villi obtained from induced abortions (6 to 12 weeks) were digested by trypsin. Cytotrophoblastic cells were separated from blood cells and syncytia on a discontinuous Percoll gradient, and the contaminating leukocytes were removed by immunopurification with an antibody to CD45 coupled to magnetic particles. These cytotrophoblastic cells were counted in a Neubauer cell in presence of Trypan Blue and diluted to 10 6 cells/ml. Cells (6 10 5 /wells) were cultured overnight in Dulbecco s modified Eagle s medium (DMEM; Life Technologies, Basel, Switzerland) containing 2 mmol/l of L-glutamine, 4.2 mmol/l of magnesium sulphate, 2.5 mmol/l of HEPES, 1% gentamycin, 1% Amphotericin B, 100 g/ml of streptomycin (Grünenthal, Stolberg, Germany), and 100 U/mL of penicillin (Hoescht-Pharma, Zürich, Switzerland) in the presence of 10% fetal calf serum (FCS; Life Technologies). The next morning (day 0), the cells were incubated in the presence or absence of increasing concentrations of hydrosalpinx fluid or peritoneal fluid (0, 10%, 20%, 50%, 80% v/v in culture medium without FCS). Incubation was performed under a 5% CO 2 and 95% air atmosphere in a humidified incubator at 37 C. The culture was stopped on day 2. The supernatants were divided into aliquots and stored at 20 C until assayed. The cells were lysed with 200 L Triton X-100 (25% in water) and stored at 20 C for total cell protein measurements. Gelatinolytic Assays Zymography, which dissociates the gelatinases from their inhibitors, measures the activity of prommp-2 and prommp-9 as well as the activity of MMP-2 and MMP-9. This assay was performed as previously described (14). Zymograms were scanned in an Apple Onescanner and the surface of the digestion bands was measured by the NIH Image 1.60 program on the Power Macintosh 7100/66 computer. All zymograms were evaluated using the same preset standards and results expressed in arbitrary units (AU). MMP-9 zymographic activity (MMP-9z) represents the addition of the surface of the digestion band of prommp-9 and of activated MMP-9. Similarly, the zymographic activity of MMP-2 (MMP-2z) represents the addition of prommp-2 and activated MMP-2. Quantitative estimation of total gelatinolytic activity was performed by measuring the degradation of heat-denatured ( 3 H)-collagen type IV using a method already reported by us (15). The standard curve was built with the use of collagenase from Clostridium histolyticum (Sigma, Buchs, Switzerland) and ranged from 0.8 to 50 ng/ml (0.26 16.5 IU/mL). Protein Assays The MMP-2 and MMP-9 immunoreactivities (MMP-2i and MMP-9i) were measured in culture supernatants and diluted hydrosalpinx or peritoneal fluids (without cytotrophoblastic cells) using our own enzyme immunoassays as described elsewhere (16). Tissue inhibitor of metalloproteinases (TIMP-1) was measured by a commercially available enzyme immunoassay (Biotrak, Amersham, Dübendorf, Switzerland). Statistical Analysis The concentrations of MMP-9i, MMP-2i, MMP-9z, MMP-2z, TIMP-1, and the total gelatinolytic activity in the hydrosalpinx and peritoneal fluids were measured at the different dilutions used (10% to 80% v/v in the culture FERTILITY & STERILITY 589

TABLE 1 Comparison between hydrosalpinx and peritoneal fluids. Hydrosalpinx fluid Peritoneal fluid Statistics (Mann-Whitney nonparametric) Total gelatinolytic activity (ng/ml) 4.0 0.8 (20.0) 1.7 0.4 (23.5) P.0293 MMP-9i a (ng/ml) 17.2 3.5 (20.3) 19.7 4.0 (20.3) P.533 MMP-2i b (ng/ml) 378.8 158.8 (41.9) 604.3 133.2 (22.0) P.0617 MMP-9z c (AU) 10101.8 4004.3 (39.6) 10238.3 2407.5 (23.5) P.228 MMP-2z d (AU) 1059.5 315.7 (29.8) 4244.0 1546.3 (36.4) P.0565 TIMP-1 e (ng/ml) 273.5 109.2 (39.9) 562.1 161.7 (28.8) P.533 Note: Values are mean SEM; values between brackets are coefficients of variation in percent. a MMP-9i, immunoreactive MMP-9. b MMP-2i, immunoreactive MMP-2. c MMP-9z, zymographic activity of MMP-9. d MMP-2z, zymographic activity of MMP-2. e TIMP-1, tissue inhibitor of metalloproteinases. medium) before adding cytotrophoblastic cells. To evaluate the effects of hydrosalpinx and peritoneal fluids on the different trophoblastic parameters, the values of each parameter found in the culture medium in the absence of cytotrophoblastic cells but in the presence of hydrosalpinx or peritoneal fluids were subtracted from the values found in the culture media in the presence of cytotrophoblastic cells and hydrosalpinx or peritoneal fluid. These results were expressed as a percentage of the respective controls (cytotrophoblastic cells in absence of hydrosalpinx or peritoneal fluid). Because the same dilutions (10% to 80%) were used with hydrosalpinx and peritoneal fluids, the data were pooled to compare the overall effect of hydrosalpinx and peritoneal fluids on the different parameters measured in cytotrophoblastic cell supernatants. All experiments were run in duplicates. Statistical analyses were performed by analysis of variance (ANOVA), Mann-Whitney, and simple regression analysis using the StatView 4.5 program (Abascus) on a Power Macintosh 7100/66 computer. RESULTS Characterization of Hydrosalpinx and Peritoneal Fluids The concentrations of MMP-9i, MMP-2i, MMP-9z, MMP-2z, TIMP-1, and the total gelatinolytic activity were compared for hydrosalpinx and peritoneal fluids (Table 1). Hydrosalpinx produced a statistically significant increase in the total gelatinolytic activity, greater than found with peritoneal fluids (P.0293). The immunoreactivity of MMP-9 and MMP-2 and the zymographic activity of MMP-9 and MMP-2 showed no statistically significant differences in hydrosalpinx compared to peritoneal fluids. The TIMP-1 concentration was similar in hydrosalpinx and peritoneal fluids. Effects of Hydrosalpinx Fluids Hydrosalpinx produced a statistically significant increase in the total gelatinolytic activity of cytotrophoblastic cells in a dose-dependent manner (R 0.601; P.0001). This increase was significant when compared to a concentration of none to 80% of hydrosalpinx (P.0003; see Fig. 1A). Hydrosalpinx also significantly increased MMP-9i and MMP-2i in a dose-dependent manner (R 0.353; P.013, R 0.422; P.002, respectively). The increase of MMP-9i was statistically significant for each concentration compared to no hydrosalpinx (P.016.003; see Fig. 1B), whereas the stimulation of MMP-2i was only significant for the concentration of 80% of hydrosalpinx when compared to no hydrosalpinx (P.0042; see Fig. 1C). MMP-9z was significantly increased in a dose-dependent manner (R 0.374; P.0001) by hydrosalpinx, and this increase was statistically significant when compared with a concentration of 50% and 80% to zero hydrosalpinx (P.0001 and 0.0001, respectively; see Fig. 1D). MMP-2z was not stimulated in a dose-dependent manner by hydrosalpinx; however, 20% hydrosalpinx stimulated MMP-2z significantly (P.0023; see Fig. 1E) as compared to no hydrosalpinx. Hydrosalpinx significantly stimulated TIMP-1 in a dose-dependent manner (R 0.262; P.05), with a statistically significant increase for a hydrosalpinx concentration of 80% when compared to no hydrosalpinx (P.0083; see Fig. 1F). Effects of Peritoneal Fluids Peritoneal fluids did not modify the total gelatinolytic activity of cytotrophoblastic cells (Fig. 2A). MMP-9i was 590 Jastrow et al. Effect of hydrosalpinx on gelatinases Vol. 77, No. 3, March 2002

FIGURE 1 Effects of hydrosalpinx fluid. (A), On trophoblastic gelatinolytic activity. (B), On immunoreactive MMP-9. (C), On immunoreactive MMP-2. (D), On the zymographic activity of MMP-9. (E), On the zymographic activity of MMP-2. (F), On the immunoreactivity of TIMP-1. Results are expressed as percent of the respective controls (cytotrophoblastic cells without hydrosalpinx fluids), mean and SEM were calculated from five different hydrosalpinx fluids tested in duplicates, and P values were obtained by ANOVA. significantly increased in a dose-dependent manner (R 0.550; P.002), and this increase was statistically significant for a concentration of 50% and 80% of peritoneal fluids when compared to no increase (P.0056 and.0024, respectively; see Fig. 2B). MMP-2i was significantly increased in a dose-dependent manner (R 0.650; P.0001), with a statistically significantly increase for a concentration of 80% of peritoneal fluids (P.0007; see Fig. 2C) when compared to no peritoneal fluids. MMP-9z was significantly increased in a dose-dependent manner (R 0.388; P.034). When compared to zero peritoneal fluids, a concentration of 80% of peritoneal fluids stimulated MMP-9z in a statistically significant manner (P.0268; see Fig. 2D). MMP-2z was not modified by peritoneal fluids (see Fig. 2E). None of the concentrations of peritoneal fluids significantly modified the concentration of TIMP-1. However, the overall effect was a statistically significant dose-dependent stimulation (R 0.995; P.0004; see Fig. 2F). Comparison of the Effects of Peritoneal Fluids and Hydrosalpinx on Cytotrophoblastic Cells The comparison between the effects of hydrosalpinx and peritoneal fluids on the trophoblastic parameters is shown in Figure 3. The total gelatinolytic activity of cytotrophoblastic cells was significantly higher in presence of hydrosalpinx than peritoneal fluids (P.0011; see Fig. 3A). MMP-9i was also significantly higher when hydrosalpinx was added to the culture medium (P.0204; see Fig. 3B). In contrast, peritoneal fluids increased significantly more MMP-2i than hydrosalpinx (P.0033; see Fig. 3C). The stimulatory effect on MMP-9z was similar for hydrosalpinx and peritoneal fluids, whereas MMP-2z was significantly higher in presence of FERTILITY & STERILITY 591

FIGURE 2 Effects of peritoneal fluid. (A), On trophoblastic gelatinolytic activity. (B), On immunoreactive MMP-9. (C), On immunoreactive MMP-2. (D), On the zymographic activity of MMP-9. (E), On the zymographic activity of MMP-2. (F), On the immunoreactivity of TIMP-1. Results are expressed as percent of the respective controls (cytotrophoblastic cells without peritoneal fluids), mean and SEM were calculated from three different peritoneal fluids tested in duplicates, and P values were obtained by ANOVA. hydrosalpinx (P.0035; see Fig. 3E). In contrast, TIMP-1 was significantly higher when peritoneal fluid was added to the cells instead of hydrosalpinx (P.0490; Fig. 3F). DISCUSSION Hydrosalpinx has deleterious effects on IVF outcome. Different hypotheses trying to explain the association between the presence of hydrosalpinx and reduced implantation and pregnancy rates have been presented. Hydrosalpinx, the volume of which increases during ovarian stimulation (17, 18), may leak from the fallopian tubes into the uterus and inhibit embryonal apposition to the endometrium (1, 19, 20). Hydrosalpinx may also impair endometrial physiology. Indeed, the potential presence of cytokines, prostaglandins, and other inflammatory components in the hydrosalpinx fluid may result in reduced endometrial receptivity. Meyer et al. (21) showed a decreased expression of v 3 integrin in endometrial biopsies of women with hydrosalpinx. Hydrosalpinx could furthermore impair embryonic development. Several investigators have suggested that hydrosalpinx contains embryotoxic factors that are deleterious to the normal development of murine embryos (22 25). In contrast, hydrosalpinx does not seem to affect human embryo development (26, 27) but rather exerts a negative effect on the oocyte quality of women in IVF programs (28). One study showed that hydrosalpinx even improved human trophoblast viability in vitro and enhanced the production of trophouteronectin and -hcg by these cells (29). The present study investigated another potential explanation for the deleterious effects of hydrosalpinx on pregnancy rates. We hypothesized that hydrosalpinx fluid could inhibit the invasive potential of embryos by inhibiting the synthesis and/or activity of cytotrophoblastic MMPs. Peritoneal fluid was used as the control because it is in 592 Jastrow et al. Effect of hydrosalpinx on gelatinases Vol. 77, No. 3, March 2002

FIGURE 3 Comparison of the effects of hydrosalpinx and peritoneal fluids. (A), On trophoblastic gelatinolytic activity. (B), On immunoreactive MMP-9. (C), On immunoreactive MMP-2. (D), On the zymographic activity of MMP-9. (E), On the zymographic activity of MMP-2. (F), On the immunoreactivity of TIMP-1. Results are expressed as percent of the respective controls (cytotrophoblastic cells without hydrosalpinx or peritoneal fluids), mean and SEM were calculated from five different hydrosalpinx and three different peritoneal fluids tested in duplicates at four different concentrations, and P values were obtained by unpaired Student s t-test. contact with the tubes and does not seem to impair implantation. The parameters measured here varied considerably between the different hydrosalpinx or peritoneal fluid samples. The reason for this variability is unknown so far, but may reflect a wide heterogeneity depending on multiple factors such as the extent of inflammatory response and/or different susceptibility to hormonal factors. However, when comparing these two fluids, we found the gelatinolytic activity to be the only statistically significant difference. This could be due to the presence of inflammatory cytokines in hydrosalpinx (16). The effects of hydrosalpinx and peritoneal fluids on the secretion of cytotrophoblastic MMPs and TIMP-1 are significantly different. The data show that hydrosalpinx fluid stimulates the secretion of MMP-2, MMP-9, TIMP-1, and the total gelatinolytic activity of the cells. This could reflect a higher stimulation of MMPs as compared to their inhibitors, as the net result is a significant stimulation of the total gelatinolytic activity. In contrast, peritoneal fluid increases MMP-2, MMP-9, and TIMP-1 concentrations, but the total gelatinolytic activity remains unmodified. This could reflect a similar stimulation of the MMPs and their inhibitors. Thus, hydrosalpinx fluid could probably stimulate cytotrophoblastic cell invasion more than peritoneal fluids. This would indicate that the effect of hydrosalpinx on implantation rates is not due to an inhibition of the capacity of an embryo to invade. Our results support the concept that hydrosalpinx would rather favor the implantation process. This would fit the observation of Sawin et al. (29) that hydrosalpinx improved human trophoblast viability in vitro and enhanced the production of trophouteronectin and -hcg by these cells. Moreover, it is known that tubal disease or hydrosalpinx increases the risk of ectopic pregnancy after IVF (6, 30 34). In our study, we observed that hydrosalpinx fluid, in contrast to peritoneal fluids, stimulates the gelatinolytic activity of cytotrophoblastic cells. The relative increase of the invasive FERTILITY & STERILITY 593

potential of the embryo in the presence of an altered tubal physiology could partly explain the increased risk of having ectopic pregnancies in the presence of hydrosalpinx. In contrast, in the uterine cavity, implantation is not increased but rather decreased by hydrosalpinx. This might be explained by other factors, such as a washout mechanism or an alteration of endometrial physiology. The lack of correlation between MMP-2i and MMP-2z, when comparing the effects of hydrosalpinx and peritoneal fluids, is a surprising observation for which we have no explanation at the present time. This needs further investigation. In conclusion, our study found that an inhibitory effect of hydrosalpinx on the capacity of an embryo to invade the endometrium could be excluded. Further investigations are clearly necessary to understand the mechanism of action responsible for the lower implantation and pregnancy rates in hydrosalpinx. The stimulatory effect that hydrosalpinx fluid exerts on the invasive potential of the embryo could partly explain the increased rate of ectopic pregnancies seen with hydrosalpinx. This also needs further investigation. References 1. Andersen AN, Yue Z, Meng FJ, Petersen K. Low implantation rate after in-vitro fertilization in patients with hydrosalpinges diagnosed by ultrasonography. Hum Reprod 1994;9:1935 8. 2. Strandell A, Waldenstrom, Nilsson L, Hamberger L. Hydrosalpinx reduces in vitro fertilization/embryo transfer pregnancy rates. Hum Reprod 1994;9:861 3. 3. Katz E, Akman MA, Damewood MD, Garcia JE. Deleterious effect of the presence of hydrosalpinx on implantation and pregnancy rates with in vitro fertilization. Fertil Steril 1996;66:122 5. 4. Vandromme J, Chasse E, Lejeune B, Van Rysselberge M, Delvigne A, Leroy F. Hydrosalpinges in in-vitro fertilization: an unfavorable prognostic feature. Hum Reprod 1995;10:576 9. 5. Fleming H, Hull MGR. Impaired implantation after in vitro fertilization treatment associated with hydrosalpinx. Br J Obstet Gynecol 1996;103: 268 72. 6. Akman MA, Garcia JE, Damewood MD, Watts LD, Katz E. Hydrosalpinx affects the implantation of previously cryopreserved embryos. Hum Reprod 1996;11:1013 4. 7. Librach CL, Werb Z, Fitzgerald ML, Chiu K, Corwin NM, Esteves RA, et al. 92 kda type IV collagenase mediates invasion of human cytotrophoblasts. J Cell Biol 1991;113:437 49. 8. Bischof P, Friedli E, Martelli M, Campana A. Expression of extracellular matrix degrading metalloproteases by cultured human cytotrophoblast cells: effect of cell adhesion and immunopurification,. Am J Obstet Gynecol 1991;65:1791 801. 9. Fisher SJ, Leitch MS, Kantor MS, Basbaum CB, Kramer RH. Degradation of extracellular matrix by the trophoblastic cells of first trimester placentas. J Cell Biochem 1985;27:31 41. 10. Puistola U, Ronnberg L, Martikainen H, Turpeenniemi-Hujanen T. The human embryo produces basement membrane collagen (type IV collagen)-degrading protease activity. Hum Reprod 1989;4:309 11. 11. Bischof P, Martelli M, Campana A, Itoh Y, Ogata Y, Nagase H. Importance of metalloproteinases (MMPs) in human trophoblast invasion. Early Pregnancy 1995;1:263 9. 12. Bischof P, Campana A. A model for implantation of the human blastocysts and early placentation. Hum Reprod 1996;2:262 70. 13. Nagase H. Activation mechanism of matrix metalloproteinases. Biol Chem 1997;378:151 60. 14. Martelli M, Campana A, Bischof P. Secretion of matrix metalloproteinases by human endometrial cells in vitro. J Reprod Fertil 1993;98: 67 76. 15. Bischof P, Haenggeli L, Campana A. Gelatinase and oncofetal fibronectine secretion is dependent upon integrin expression on human cytotrophoblasts. Hum Reprod 1995;10:734 72. 16. Meisser A, Chardonnens D, Campana A, Bischof P. Effects of tumour necrosis factor alpha, interleukin-1 alpha, macrophage colony stimulating factor and transforming growth factor beta on trophoblastic matrix metalloproteinases. Mol Hum Reprod 1999;5:252 60. 17. Sharara FI, McClamrock HD. Endometrial fluid collection in women with hydrosalpinx after human chorionic gonadotrophin administration: a report of two cases and implications for management. Hum Reprod 1997;12:2816 9. 18. Schiller VL, Tsuchiyama K. Development of hydrosalpinx during ovulation induction. J Ultrasound Med 1995;14:799 803. 19. Mansour RT, Aboulghar MA, Serour GI, Riad R. Fluid accumulation of the uterine cavity before embryo transfer: a possible hindrance for implantation. J In Vitro Fert Embryo Transf 1991;8:1157 9. 20. Sawin SW, Loret de Mola JR, Monzon-Bordonaba F, Wang CL, Feinberg RF. Effect of hydrosalpinx on IVF outcome mechanism? Fertil Steril 1998;70:787 8. 21. Meyer WR, Castelbaum AJ, Somkuti S, Sagoskin AW, Doyle M, Harris JE, et al. Hydrosalpinges adversely affect markers of endometrial receptivity. Hum Reprod 1997;12:1393 8. 22. Mukherjee T, Copperman AB, McCaffrey C, Cook CA, Bustillo M, Obasaju MF. Hydrosalpinx fluid has embryotoxic effects on murine embryogenesis: a case for prophylactic salpingectomy. Fertil Steril 1996;66:851 3. 23. Rawe VJ, Liu J, Shaffer S, Compton MG, Garcia JE, Katz E. Effect of human hydrosalpinx fluid on murine embryo development and implantation. Fertil Steril 1997;68:668 70. 24. Beyler SA, James KP, Fritz MA, Meyer WR. Hydrosalpingeal fluid inhibits in vitro embryonic development in a murine model. Hum Reprod 1997;12:2724 8. 25. Sachdev R, Kemmann E, Bohrer MK. Detrimental effect of hydrosalpinx fluid on the development and blastulation of mouse embryos in vitro. Fertil Steril 1997;68:531 3. 26. Granot I, Dekel N, Segal I, Fieldust S, Shoham Z, Barash A. Is hydrosalpinx fluid cytotoxic? Hum Reprod 1998;13:1620 4. 27. Strandell A, Sjogren A, Bentin-Ley U, Thorburn J, Hamberger L, Brannstrom M. Hydrosalpinx fluid does not adversely affect the normal development of human embryos and implantation in vitro. Hum Reprod 1998;13:2921 5. 28. Freeman MR, Whitworth CM, George AH. Permanent impairment of embryo development by hydrosalpinges. Hum Reprod 1998;13:983 6. 29. Sawin SW, Loret de Mola JR, Monzon-Bordonaba F, Wang CL, Feinberg RF. Hydrosalpinx fluid enhances human trophoblast viability and function in vitro: implications for embryonic implantation in assisted reproduction. Fertil Steril 1997;68:65 71. 30. Dubuisson JB, Aubriot FX, Mathieu L, Foulot H, Mandelbrot L, de Joliere JB. Risk factors for ectopic pregnancy in 556 pregnancies after in vitro fertilization: implications for preventative management. Fertil Steril 1991;56:686 90. 31. Herman A, Ron-El R, Golan A, Weinraub Z, Bukovsky I, Caspi E. The role of tubal pathology and other parameters in ectopic pregnancies occurring in in vitro fertilization and embryo transfer. Fertil Steril 1990;54:864 8. 32. Martinez F, Trounson A. An analysis of risk factors associated with ectopic pregnancy in a human in vitro fertilization program. Fertil Steril 1986;45:79 87. 33. Zouves C, Erenus M, Gomel V. Tubal ectopic pregnancy after in vitro fertilization and embryo transfer: a role for proximal occlusion or salpingectomy after failed distal tubal surgery? Fertil Steril 1991;56: 691 5. 34. Barmat LI, Rauch E, Spandorfer S, Kowalik A, Sills ES, Schattman G, et al. The effect of hydrosalpinges on IVF-ET outcome. J Assist Reprod Genet 1999;16:350 4. 594 Jastrow et al. Effect of hydrosalpinx on gelatinases Vol. 77, No. 3, March 2002