Hydrosalpinges adversely affect markers of endometrial receptivity

Transcription

1 Human Reproduction vol.12 no.7 pp , 1997 OUTSTANDING CONTRIBUTION Hydrosalpinges adversely affect markers of endometrial receptivity W.R.Meyer 1, A.J.Castelbaum 2, S.Somkuti 3, fertilization (IVF) embryo transfer (Yaron et al., 1994). A.W.Sagoskin 4, M.Doyle 5, J.E.Harris 1 and Despite the fact that IVF embryo transfer was initially B.A.Lessey 1,6 developed in a population of women with tubal factor 1 infertility, recent evidence strongly suggests that the presence Department of Obstetrics and Gynecology, University of North Carolina, Chapel Hill, NC, 2 Northern Fertility and Reproductive of a communicating hydrosalpinx may have a deleterious Associates, Meadowbrook, PA, 3 Abington Memorial Hospital, effect on pregnancy rates in IVF and cryopreserved thawed Abington, 4 The Shady Grove Fertility Center, Rockville, MD, and embryo transfer cycles (Kassabji et al., 1994; Koong et al., Norwalk, CT, USA 1994; Strandell et al., 1994; Vandromme et al., 1995; Akman 6 To whom correspondence should be addressed et al., 1996; Fleming and Hull, 1996; Katz et al., 1996; Blazer et al., 1997). These studies, complemented by reports While in-vitro fertilization (IVF) was initially developed in demonstrating increased implantation rates and decreased women with tubal factor infertility, recent clinical studies have suggested that the presence of hydrosalpinges lowers miscarriage rates after surgical extirpation, drainage, or implantation and pregnancy rates. We postulated that these proximal ligation of hydrosalpinges (Vandromme et al., 1995; hydrosalpinges cause impaired endometrial receptivity. A Shelton et al., 1996) suggest that retrograde spillage of total of 103 women with hydrosalpinges were prospectively hydrosalpingeal tubal fluid may affect either embryo growth evaluated, and compared with 55 infertile and 44 fertile (Mukherjee et al., 1996) and/or endometrial receptivity controls. All women had endometrial biopsies during the (Lessey et al., 1994b). The specific mechanisms of these window of implantation, analysed by conventional histointerference with implantation due to the intrauterine putative effects are unknown but may include physical logical criteria, and also stained for three integrin markers of endometrial receptivity (α1β1, α4β1 and αvβ3). Women accumulation of refluxed hydrosalpinx fluid (Mansour et al., with hydrosalpinges (cases) expressed significantly less of 1991), chronic endometritis (Sharara et al., 1996), and toxic the αvβ3 integrin compared with controls. There was no effects of hydrosalpinx fluid on embryo quality (Mukherjee difference in expression of α1β1 orα4β1 among groups. A et al., 1996). In addition, tubal epithelium may secrete significantly greater number of cases had out of phase cytokines, leukotrienes or prostaglandins into the sequestered histology and missing αvβ3 (type I defects) and absent fluid that could directly alter endometrial function (Toth integrin expression despite normal histological maturation et al., 1992). (type II) defects, compared with controls. Of 20 women Endometrial receptivity, previously defined only in histowith impaired endometrial receptivity who were also bioptemporal logical terms, has more recently been characterized by the sied after hydrosalpinx surgery, 70% demonstrated appearance of specific endometrial integrins during increased αvβ3 expression. Seventy-seven percent of type the putative window of implantation (Lessey et al., 1992, I and 57% of type II defects were corrected postoperatively. 1994a; Tabibzadeh, 1992). The epithelial endometrial integrin Using markers of endometrial receptivity, this study demon- αvβ3 appears coincident with the opening of this window strates that inflammatory hydrosalpinges have an adverse (cycle days 19 20). Decreased αvβ3 expression has been effect on endometrial receptivity, which in some cases may reported in connection with luteal phase deficiency (LPD; be overcome by surgical treatment of the hydrosalpinx. Lessey et al., 1992), endometriosis (Lessey et al., 1994c), Key words: endometrial receptivity/hydrosalpinges/implanta- and unexplained infertility (Lessey et al., 1995). Based on tion/integrins/markers our preliminary data (Lessey et al., 1994b), we suspected that a similar endometrial dysfunction might also occur secondary to the presence of hydrosalpinges. To investigate a role for αvβ3 in this disorder, we evaluated endometrial Introduction histology and integrin expression during the mid-luteal phase The endometrium plays a critical role in the establishment in women with persistent hydrosalpingeal disease. A subset of normal pregnancy. Recent identification of markers of of these women also underwent endometrial assessment after endometrial receptivity have allowed a better understanding surgical treatment of their hydrosalpinges. To our knowledge, of the functional human endometrium. Nowhere is this issue this is the first investigation of the deleterious effects of of greater importance than in the setting of in-vitro hydrosalpinges on markers of endometrial receptivity. European Society for Human Reproduction and Embryology 1393

2 W.R.Meyer et al. Materials and methods Table I. Demographic and HSCORE characteristics of case and control Study design subjects undergoing endometrial biopsy during the mid-luteal phase A prospective, case-controlled study was performed comparing the Characteristic Hydrosalpinx Infertile control Fertile control expression of cycle-specific integrins in women with hydrosalpinges (n 103) (n 55) (n 44) (cases) to fertile and infertile controls. In accordance with each of the respective Institutional Review Boards, endometrium was obtained Age (years; mean SD) by endometrial biopsy from 103 women with hydrosalpinges and Mean c compared with 55 samples from women with a partner with male Minimum Maximum factor infertility (infertile controls) and 44 women with proven fertility HSCORE (mean SD) (fertile controls), during the period from July 1991 to October α Inclusion criteria required that the date of the preceding menstrual α period was certain and that all patients had regular menstrual β a,b cycles. Each of the cases had documented unilateral or bilateral a P versus infertile control. hydrosalpingeal disease, detected by hysterosalpingography and/or b P versus fertile controls. laparoscopy. Salpingoscopy was not used in the infertility evaluation. c P 0.01 versus cases and infertile controls. All biopsies were obtained during the mid-luteal phase between cycle days 21 and 24 based on a urinary luteinizing hormone (LH) surge HSCORE 0.7 as the cut-off of a positive test for the evaluation (LH 7 to 10). Male factor infertility was defined as a total motile of epithelial αvβ3 immunostaining, as this cut-off was clearly sperm count of /ml or total motility 30%. Fertile controls distinguishable from negative staining and gave the maximal sensitivwere women with regular cyclic menses and at least one successful ity and specificity. Categorical assignment of putative endometrial pregnancy, without a known history of infertility. receptivity defects was based on previous definitions (Lessey et al., Twenty women with persistent unilateral or bilateral hydrosalpinges 1995): the absence of the αvβ3 integrin in samples obtained during and defects in endometrial receptivity (based on immunohistochemical cycle days was defined as a type I defect in the presence of a staining showing the lack of the endometrial αvβ3 integrin) underwent histological delay of 3 days, or a type II defect in the presence of an endometrial biopsy before and after surgical extirpation, distal neosalpingostomy, transvaginal needle drainage or proximal tubal ligation of the hydrosalpinges. in-phase endometrium. Statistical analysis Comparisons between groups for continuous variables (HSCORE or age) was performed using ANOVA with Scheffé s correction. A χ 2 analysis was used to compare the incidence of putative defects in endometrial receptivity in cases and controls. Tissue evaluation Endometrial histological dating was performed according to the criteria of Noyes et al. (1950). Histological dating was based on the most advanced glandular dating criteria. Gland/stromal dyssynchrony of 3 days was interpreted as out of phase. Immunoperoxidase Results staining was performed for the α1, α4, and β3 integrin subunits, using the monoclonal antibodies TS2/7, B5G10, and SSA6, respectively, as The demographics of the total population, composed of 202 previously described (Lessey et al., 1994a). The monoclonal antiinges (cases) were of similar age ( years) to those women are shown in Table I. Infertile women with hydrosalp- bodies TS2/7 and B-5G10 directed against α1 and α4 integrin subunits were generously provided by Dr Martin Hemler (Dana Farber Cancer women with identifiable male factor infertility ( Institute, Boston, MA, USA). Serial cryosections 5 µm thick were years). Fertile controls tended to be of younger age ( placed on poly-l-lysine-coated slides, fixed in acetone at 20 C years; P 0.01). for 10 min, and stained using Vectastaine Elite ABC kits (Vector Immunohistochemical staining for the three epithelial cycle- Laboratories, Burlingame, CA, USA). Diaminobenzidine (DAB; dependent integrin subunits was performed on each biopsy. Sigma Chemical Co., St Louis, MO, USA) was used as the chromagen. We have previously demonstrated that the co-expression of Primary antibodies (TS2/7, B5G10, or SSA6) were placed on cryosecthese integrin subunits frames the window of implantation. As tions following blocking with 1% bovine serum albumin (Sigma Chemicals, St Louis, MO, USA) in phosphate buffered saline (PBS) shown in Figure 1, there was significantly less overall β3 ph , and allowed to bind at room temperature for 1 h. A expression in the endometrial epithelium of women with rinse with PBS was followed by secondary antibody consisting of hydrosalpinges compared to infertile (P 0.012) and fertile biotinylated goat anti-mouse antibody (Vector Laboratories, Burlingame, (P 0.014) controls; similar expression of α1 orα4 subunits CA, USA) for 30 min before adding diaminobenzidine for was noted between groups (Table I). A total of 103 women 3 min to complete the reaction. Samples were subsequently washed with hydrosalpinges (cases) was compared to infertile and in PBS and mounted. The resultant staining was evaluated on a Nikon fertile controls and assigned to one of the following categories: microscope, by a single blinded observer. The histological score normal, type I and type II defects. In type I defects, the lack (HSCORE) was calculated using the following equation: HSCORE of the β3 subunit was felt to be due to out of phase endometrium ΣPi(i 1), where I intensity of staining with a value of 1, 2 or 3 with delayed expression of the αvβ3 integrin. Type II defects (weak, moderate or strong respectively), Pi is the percentage of represented an occult endometrial receptivity defect because stained epithelial cells for each intensity, varying from 0 to 100% and 1 is a correction for optical density. This yields a range of results the lack of epithelial αvβ3 occurred in histologically normal from 0 for no staining to 4 for maximal staining. The use of HSCORE endometrium. has been used clinically with low inter- and intra-observer variation As shown in Figure 2, of the 103 endometrial biopsies (Budwit-Navotny et al., 1986). Based on previous Receiver Operator obtained in women with hydrosalpinges, 33 (32%) demon- Characteristic (ROC) analysis (Lessey et al., 1994c), we used an strated delayed histological development and impaired β3 1394

3 Hydrosalpinges and markers of uterine receptivity Figure 3. Scattergram of the HSCORE for αvβ3 immunostaining before and after surgical treatment of hydrosalpinges in 20 women with suspected impairment in endometrial receptivity. Endometrial biopsies, all obtained during the window of implantation in LH- timed cycles, were compared before and after attempted surgical correction of hydrosalpinges. All women had suspected defects in uterine receptivity based on an overall HSCORE for αvβ3 immunostaining of 0.7. In 15 of 20 cases, there was improvement in expression of the αvβ3 integrin in the subsequent endometrial biopsy. Figure 1. Semi-quantitative HSCORE for αvβ3 integrin expression in women with hydrosalpinges and infertile and fertile controls. Immunohistochemical assessment (of integrin expression) was performed on luteinizing hormone (LH)-timed endometrial biopsies from 103 women with documented hydrosalpinx (cases) and compared with 55 women with the diagnosis of male factor infertility (infertile controls) or 44 women with proven fertility (fertile controls). The cases exhibited a significantly lower mean HSCORE for the αvβ3 integrin compared to either infertile (P 0.012) or fertile (P 0.014) controls. The other cycle specific integrins, α1β1 and α4β1, were not different between groups (Table I). Table II. Outcome after surgical treatment for hydrosalpinges by surgical procedure Procedure Corrected (%) Salpingectomy 12/13 (92.3) Salpingostomy 1/3 (33.3)* Tubal interruption 2/3 (66.7) Transvaginal aspiration 0/1 (0) *Both patients with failure to correct in this group had documented reaccumulation of hydrosalpinx secondary to re-occlusion. Figure 2. The relative incidence of type I and II endometrial receptivity defects in women with hydrosalpinges compared with infertile and fertile controls. In 103 endometrial biopsies from women with hydrosalpinges (cases), only 38.8% had normal expression of the αvβ3 integrin compared with 65.4 and 68.1% for infertile (n 55) and fertile controls (n 44) respectively. There were significantly higher incidences of type I and II defects in the cases than were observed in the controls (P 0.01). integrin expression (type I) and 30 cases (29.1%) had aberrant αvβ3 expression despite in phase endometrial histology (type II). In comparison, infertile controls had fewer type I (11 cases; 20%) and type II defects (eight cases; 18.1%). Compared with cases, fertile controls also exhibited fewer defects in uterine receptivity (eight type I; 14.5% and six type II; 13.3%). Statistical evaluation using χ 2 demonstrated a significantly (Lessey et al., 1994c). higher incidence of both defects in the cases compared to the controls (χ ; P 0.01). In 63 women with hydrosalpinges who were found to have impaired expression of the αvβ3, 20 underwent surgical correction followed by repeat assessment of integrin immunostaining in an LH-timed endometrial biopsy, most within 3 months of surgery in a normal cycle. As shown in Figure 3, 15 of 20 biopsies demonstrated an increase in overall αvβ3 expression after surgery. Overall, the number of endometrial defects was reduced from 20/20 (13 type I and seven type II defects) to 5/20 (two type I and three type II defects) in the post-surgical group. As shown in Table II, most of the women received salpingectomy and over 90% successfully corrected. The other procedures used including neosalpingostomy, ligation or aspiration were less successful in restoring normal endometrial receptivity. Of the five women that did not correct, three re-accumulated hydrosalpinx fluid after neosalpingostomy or aspiration, one demonstrated histological delay in the second biopsy, and one patient had concomitant endometriosis, another identifiable cause of occult in phase type II defects 1395

4 W.R.Meyer et al. et al., 1997). Thus, direct exposure to hydrosalpinx fluid, as shown in Figure 4, might account for the maturational delay or otherwise occult defects of endometrial development demonstrated in this study. Previous studies have validated the use of integrins as markers of endometrial receptivity (Castelbaum et al., 1996; Lessey et al., 1992, 1994c). In an infertile population, we have previously described two distinct types of endometrial dysfunction based on alterations in integrin expression. The first, defined as a type I defect, is characterized by a retardation in histological development consistent with luteal phase deficiency (LPD). On this basis, these samples fail to express the β3 integrin subunit (Lessey et al., 1992). Perhaps more intriguing is a second and otherwise occult endometrial defect defined as a type II defect. This phenomenon was first noted in the endometrium of women with minimal and mild endometriosis (Lessey et al., 1994c) and later in women with unexplained infertility (Lessey et al., 1995). In this scenario, though histologically in phase and otherwise normal in appearance, these women have a decreased expression of the αvβ3 integrin and also appear to have decreased cycle fecundity in both IVF (Castelbaum et al., 1994) and natural cycles (Lessey et al., 1996). In the present study of women with hydrosalpinges, both defects are found at significantly increased rates when compared with fertile and infertile controls. Not unexpectedly, there are also a number of women in each control group that exhibit both type I and II defects. This may reflect a percentage of spontaneous cycles, in normal women, that demonstrate sporadic sub-optimal endometrial development and/or quality (Batista et al., 1996). The appearance of αvβ3 is most variable around the early window of implantation. In cases in which normal, otherwise fertile women were found to be negative for this marker, a later biopsy might display normal expression. In women with hydrosalpinges, repeat biopsies, when performed on β3 negative patients, have revealed consistently negative results, even when obtained later in the cycle. For this reason, we now recommend performing the endometrial biopsy closer to the putative closure of the window of implantation (LH 9 to 10), to avoid the potential for false negative results. These data also suggest that not all women with hydrosalpinges exhibit abnormal endometrium. In the present study, over one-third of the women with hydrosalpinx were found to be normal and may, therefore, not exhibit an endometrial defect or have associated decreased fecundity in an assisted reproduction cycle. The need to avoid the practice of indiscrim- inate salpingectomy in all women with hydrosalpinges considering IVF makes the ability to identify women at risk for sub-optimal IVF success increasingly important (Blazer et al., 1997). Opponents to prophylactic salpingectomy of hydrosalp- inges have suggested the use of preoperative salpingoscopy in the IVF candidate (Puttermans and Brosens, 1996). Since this procedure was not used in this investigation, a direct study of the correlation between the pathological type of hydrosalpinx and the presence or absence of a defect in endometrial receptivity was not possible. As demonstrated by most prior investigations, surgical correction or extirpation of a hydrosalpinx appear to be Figure 4. Transvaginal uterine ultrasonography in a patient with communicating hydrosalpinges, obtained at the time of implantation (cycle day 22). Note the presence of fluid in the endometrial cavity, presumed to come from the fluid-filled Fallopian tubes, demonstrating that the endometrium may become exposed to this fluid at the time that embryos normally implant. Discussion The data presented in this study represent the first published report describing an effect of hydrosalpinges on markers of endometrial receptivity. Our results are consistent with accumulating data that support a detrimental effect of hydrosalpinx fluid on implantation in human reproduction (Andersen et al., 1994; Kassabji et al., 1994; Vandromme et al., 1995; Blazer et al., 1997). As demonstrated in Figure 4 and first suggested by Mansour et al. (1991), an endometrial cavity bathed in hydrosalpingeal fluid at the time of implantation may adversely effect outcome in the setting of IVF. Subsequently, Andersen et al. (1994), Strandell et al. (1994) and Kassabji et al. (1994) noted a lower pregnancy rate in IVF cycles in women with unilateral or bilateral hydrosalpinx. Vandromme and co-workers (1995) noted an overall difference in pregnancy rate of 28% by comparing IVF cycles in women with hydrosalpinges to cycles performed after surgical correction. Similar improvements were noted by others (Katz et al., 1996; Shelton et al., 1996). Pregnancy after both fresh and frozen embryo transfer was been reported to be adversely affected in patients with a unilateral or bilateral hydrosalpinx (Akman et al., 1996). Ironically, this group of women was that for which the process of IVF embryo transfer was initially proposed as an attempt to circumvent altered reproductive function. The mechanism by which hydrosalpinx fluid may affect implantation rates remains unsettled. A direct effect of Chlamydia trachomatis, or the indirect effect of protein expression as a result of infection with this pathogen, has been postulated adversely to affect endometrial receptivity (Sharara et al., 1996). This hypothesis fails to explain our observed correction of endometrial αvβ3 integrin expression after neosalpingostomy. Bioactive substances such as tumour necrosis factor (TNF)-α and other potential growth factors known to be associated with salpingitis (Toth et al., 1992) have been shown to alter both endometrial stromal and epithelial cell integrin expression (Grosskinsky et al., 1996; Castelbaum 1396

5 Hydrosalpinges and markers of uterine receptivity associated with improvement in pregnancy and implantation References rates (Kassabji et al., 1994; Vandromme et al., 1995; Lessey Akman, M.A., Garcia, J.E., Damewood, M.D. et al. (1996) Hydrosalpinx affects the implantation of previously cryopreserved embryos. Hum. Reprod., et al., 1996; Shelton et al., 1996). These claims, however, have 11, not been without controversy (Bergh and Hamberger, 1996; Andersen, A.N., Yue, Z., Meng, F.J. and Petersen, K. (1994) Low implantation Muasher, 1996). In the present study, we observed a dramatic rate after in-vitro fertilization in patients with hydrosalpinges diagnosed by ultrasonography. Hum. Reprod., 9, improvement in endometrial receptivity assessed by the pres- Batista, M.C., Cartledge, T.P., Zellmer, A.W. et al. (1996). A prospective ence of the αvβ3 integrin and histological dating after surgical controlled study of luteal and endometrial abnormalities in an infertile treatment of hydrosalpinges. In 20 cases in which the αvβ3 population. Fertil. Steril., 65, integrin was missing from the window of implantation, 15 had Bergh, C. and Hamberger, L. (1996) Salpingostomy improves pregnancy rate in patients with hydrosalpinx. Hum. Reprod., 11, substantial improvement in the expression of this integrin in Blazer, A.S., Hogan, J.W., Seifer, D.B. et al. (1997) The impact of hydrosalpinx a subsequent biopsy after tubal surgery, possibly accounting on successful pregnancy in tubal factor infertility treated by in vitro for the noted restoration of a normal pregnancy rate during fertilization. Fertil. Steril., 67, IVF embryo transfer. The absence of uniform correction of Boutten, A., Dehoux, M., Edelman, P. et al. (1992). IL6 and acute phase plasma proteins in peritoneal fluid of women with endometriosis. Clin. all putative defects in these 20 women may explain why some Chim. Acta, 210, investigators have failed to observe improvement in pregnancy Budwit-Navotny, D.A., McCarty, Sr K.S., Cox, E.B. et al. (1986) rates after surgical treatment of hydrosalpinx (Sharara et al., Immunohistochemical analyses of estrogen receptor in endometrial adenocarcinoma using a monoclonal antibody. Cancer Res., 46, ). Clearly, treatment failures were most associated with Castelbaum, A.J., Riben, M., Howarth, J. et al. (1994) Minimal endometriosis re-accumulation of the hydrosalpinx fluid which occurred more impairs endometrial αvβ3 integrin expression and cycle fecundity compared frequently with neosalpingostomy. to tubal factor patients in an IVF program. Am. Fertil. Soc. Annual Meeting, 7, S89. The most striking effect of surgery was a reduction in the Castelbaum, A.J., Somkuti, S.G., Ying, L. et al.. (1996) Characterization of presence of either gland/stromal dyssynchrony or otherwise integrin expression in a well-differentiated endometrial cancer cell line out of phase endometrium. It is possible that this delay (Ishikawa). J. Clin. Endocrinol. Metab., 82, in endometrial histology may reflect sub-acute endometritis Fakih, H., Baggett, B., Holtz, G. et al. (1987) Interleukin-1: a possible role in the infertility associated with endometriosis. Fertil. Steril., 47, secondary to the contact of endometrium with tubal fluid. Like Fleming, C. and Hull, M. (1996) Impaired implantation after in vitro endometriosis, inflammatory cytokines present in the peritoneal fertilisation treatment associated with hydrosalpinx reply. Br. J. Obstet. fluid (Fakih et al., 1987; Boutten et al., 1992; Rier et al., 1994; Gynaecol., 103, 851. Ryan et al., 1995) may alter normal endometrial function and Grosskinsky, C.M., Yowell, C.W., Sun, J.H. et al. (1996) Modulation of integrin expression in endometrial stromal cells in vitro. J. Clin. Endocrinol. account for the aberrant integrin expression in otherwise in Metab., 81, phase endometrium (Lessey et al., 1994c). Kassabji, M., Sims, J.A., Butler, L. and Muasher, S.J. (1994) Reduced In conclusion, this study presents direct immunohistochemiin vitro fertilization. Eur. J. Obstet. Gynecol. Reprod. Biol., 56, pregnancy outcome in patients with unilateral or bilateral hydrosalpinx after cal evidence to support the contention that women with Katz, E., Akman, M.A., Damewood, M.D. and García, J.E. (1996) Deleterious hydrosalpingeal tubal disease demonstrate decreased implanta- effect of the presence of hydrosalpinx on implantation and pregnancy rates tion rates in the setting of IVF embryo transfer. Evaluation of with in vitro fertilization. Fertil. Steril., 66, epithelial integrins in the uterus offers unique insights into Koong, A.C., Chen, E.Y. and Giaccia, A.J. (1994) Hypoxia causes the activation of nuclear factor kb through the phosphorylation of IkBa on endometrial receptivity. The current study demonstrates abnor- tyrosine residues. Cancer Res., 54, mal endometrial phenotypes in women with hydrosalpinges Lessey, B.A., Damjanovich, L., Coutifaris, C. et al. (1992) Integrin adhesion compared to both fertile and infertile controls. Assessment of molecules in the human endometrium. Correlation with the normal and abnormal menstrual cycle. J. Clin. Invest., 90, endometrial histology and integrin expression may identify a Lessey, B.A., Castelbaum, A.J., Buck, C.A. et al. (1994a) Further subset of women with hydrosalpinges who would benefit from characterization of endometrial integrins during the menstrual cycle and in salpingectomy prior to IVF embryo transfer. pregnancy. Fertil. Steril., 62, Lessey, B.A., Castelbaum, A.J., Riben, M. et al. (1994b) Effect of hydrosalpinges on markers of endometrial receptivity and success in IVF. Am. Fertil. Soc. Annual Meeting, Abstr. O-091, S45. Acknowledgements Lessey, B.A., Castelbaum, A.J., Sawin, S.J. et al. (1994c) Aberrant integrin expression in the endometrium of women with endometriosis. J. Clin. The authors would like to thank Dr Jinghai Sun for his excellent Endocrinol. Metab., 79, technical assistance and Wendy Greene from the Department of Lessey, B.A., Castelbaum, A.J., Sawin, S.J. and Sun, J. (1995) Integrins as Biostatistics for her assistance with statistical analysis. We would markers of uterine receptivity in women with primary unexplained infertility. also like to thank the many investigators that provided patient samples, Fertil. Steril., 63, including Drs Mark Damario (Cornell Medical Center, NY, NY, Lessey, B.A., Castelbaum, A.J., Somkuti, S.G. et al. (1996) Improvement in USA), Martin F. Freedman (Northern Fertility and Reproductive pregnancy rates with GnRH agonist in women with infertility, minimal or Associates, Meadowbrook, PA, USA), Kathryn Honea and Virginia mild endometriosis and aberrant αvβ3 expression. Am. Soc. Reprod. Med. Houserman (Birmingham, AL, USA), Glen Hofmann (Cincinatti, OH, Annual Meeting. USA), Dub Howard (Carrollton, TX, USA), Brad Hurst (University Mansour, R.T., Aboulghar, M.A., Serour, G.I. and Riad, R. (1991) Fluid accumulation of the uterine cavity before embryo transfer: a possible of Colorado Health Sciences Center, Denver, CO, USA), Daniel hindrance for implantation. J. In Vitro Fert. Embryo Transfer, 8, Kenigsberg (Life IVF Associates, Port Jefferson, NJ, USA), Richard Muasher, S.J. (1996) Salpingectomy improves pregnancy rate in patients with Scott, Jr (St Barnabas Medical Center, Livingston, NJ, USA) and hydrosalpinx. Hum. Reprod. 11, Robert Tamis, Arizona Institute of Reproductive Medicine, Phoenix, Mukherjee, T., Copperman, A.B., McCaffrey, C. et al. (1996) Hydrosalpinx AZ, USA). This research was supported in part by the National fluid has embryotoxic effects on murine embryogenesis: a case for Institutes of Health grants HD-29449, HD-30476, and HD prophylactic salpingectomy. Fertil. Steril., 66, (B.A.L.), Bethesda, MD and the ASRM-TAP Pharmaceutical Fellow- Noyes, R.W., Hertig, A.I. and Rock, J. (1950) Dating the endometrial biopsy. ship (S.S.). Fertil. Steril., 1,

6 W.R.Meyer et al. Puttermans, P.J. and Brosens, I. (1996) Preventive salpingectomy of hydrosalpinx prior to IVF salpingectomy improves in-vitro fertilization outcome in patients with a hydrosalpinx: blind victimization of the Fallopian tube? Hum. Reprod., 11, Rier, S.E., Parsons, A.K. and Becker, J.L. (1994) Altered interleukin-6 production by peritoneal leukocytes from patients with endometriosis. Fertil. Steril., 61, Ryan, I.P., Tseng, J.F., Schriock, E.D. et al. (1995) Interleukin-8 concentrations are elevated in peritoneal fluid of women with endometriosis. Fertil. Steril, 63, Sharara, F.I., Scott, R.T. Jr, Marut, E.L. and Queenan, J.T. Jr (1996) In-vitro fertilization outcome in women with hydrosalpinx. Hum. Reprod., 11, Shelton, K.E., Butler, L., Toner, J.P. et al. (1996) Salpingectomy improves the pregnancy rate in in-vitro fertilization patients with hydrosalpinx. Hum. Reprod., 11, Strandell, A., Waldenström, U., Nilsson, L. and Hamberger, L. (1994) Hydrosalpinx reduces in-vitro fertilization/embryo transfer pregnancy rates. Hum. Reprod., 9, Tabibzadeh, S. (1992) Patterns of expression of integrin molecules in human endometrium throughout the menstrual cycle. Hum. Reprod., 7, Toth, M., Jeremias, J., Ledger, W.J. and Witkin, S.S. (1992) In vivo tumor necrosis factor production in women with salpingitis. Surg. Gynecol. Obstet., 174, Vandromme, J., Chasse, E., Lejeune, B. et al. (1995) Hydrosalpinges in invitro fertilization: an unfavourable prognostic feature. Hum. Reprod., 10, Yaron, Y., Botchan, A., Amit, A. et al. (1994) Endometrial receptivity in the light of modern assisted reproductive technologies. Fertil. Steril., 62, Received on January 15, 1997; accepted on April 22,