Hydrosalpinx and IVF outcome: a prospective, randomized multicentre trial in Scandinavia on salpingectomy prior to IVF*

Hydrosalpinx and IVF outcome: a prospective, randomized multicentre trial in Scandinavia on prior to IVF* A.Strandell, A.Lindhard, U.Waldenstrom, J.Thorburn, P.O.Janson, L.Hamberger 1,4 1 3 1 1 Department of Obstetrics and Gynaecology, Sahlgrenska University Hospital, S-413 45 Göteborg, Sweden, The Fertility Clinic, Copenhagen University Hospital, Rigshospitalet, Blegdamsvej 9, DK-100 Copenhagen, Denmark and IVF Center Falun, Department of Obstetrics and Gynaecology, Falu Hospital, S-791 8 Falun, Sweden 1 3 To whom correspondence should be addressed Many retrospective studies have shown that hydrosalpinx is associated with poor in-vitro fertilization (IVF) outcome. The mechanism of the actual cause is not yet fully under stood. A clinical practice of performing has developed, without any evidence from prospective trials. The aim of the present prospective randomized trial was to test if a prior to IVF was effective in terms of increased pregnancy rates. Patients with hydrosalpinx were randomized to either a laparoscopic or no intervention. A total of 04 patients was available for an intentionto-treat analysis and 19 actually started IVF. Clinical pregnancy rates per included patient were 36.6% in the group and 3.9% in the non-intervention group (not significant, P = 0.067) and the ensuing delivery rates were 8.6% and 16.3% (P = 0.045). The corresponding delivery rates per transfer cycle were 9.5% versus 17.5% (not significant, P = 0.083). A subgroup analysis revealed significant differences in favour of, in implantation rates in patients with bilateral hydrosalpinges (5.6% versus 1.3%, P = 0.038) and in clinical pregnancy rates (45.7% versus.5%, P = 0.09) and delivery rates (40.0% versus 17.5%, P = 0.038) in patients with ultrasound visible hydrosalpinges. The delivery rate was increased 3.5-fold in patients with bilateral hydrosalpinges visible on ultrasound (P = 0.019). Key words: hydrosalpinx/in-vitro fertilization/randomized controlled trial/pregnancy rates/ *The study group for the multicentre trial and responsible investigator at each site were as follows. Sweden: Annika Strandell, Dept Ob/Gyn, Sahlgrenska University Hospital, Göteborg; Urban Waldenstrom, IVF Center Falun, Dept Ob/Gyn, Falun; Niklas Simberg, Dept Ob/Gyn, Uppsala University Hospital, Uppsala; Leif Hagglund, International Fertility Center, Malmo; Gyorgy Csemiczky, Dept Ob/ Gyn, Karolinska Hospital, Stockholm. Denmark: Anette Lindhard, The Fertility Clinic, Rigshospitalet, University Hospital, Copenhagen. Norway: Vidar von Düring, Dept Ob/Gyn, Trondheim; Kari Flo, Dept Ob/Gyn, University of Troms0. Iceland: Gudmundur Arason, Dept Ob/Gyn, National University Hospital, Reykjavik. In-vitro fertilization (IVF) and embryo transfer was originally performed in patients with tubal damage, in order to prove the method's ability. Tubal factor infertility has remained a major cause of IVF treatment, but it has become obvious that patients with severe tubal damage have a poor prognosis (Csemiczky et al., 1996). In many retrospective reports (Strandell et al., 1994; Vandromme et al., 1995; Fleming and Hull, 1996) patients with hydrosalpinges have been identified as a subgroup with significantly lower implantation and pregnancy rates than patients suffering from other types of tubal damage. However, some studies showing a similar tendency do not reach statistical significance (Sharara et al., 1996; Blazar et al., 1997). Two studies have also demonstrated an increased rate ofspontaneous abortion (Andersen etal., 1994; Kassabji etal., 1994) and the only significant finding in another study (Ng et al., 1997) was an increased rate of ectopic pregnancy. Different theories trying to explain the mechanism behind the association between the presence of hydrosalpinges and poor pregnancy outcome have been presented. The main explanations have implied that the leakage of fluid into the uterine cavity renders the endometrium a hostile environment to implantation and could also affect embryo development. Consequently, surgical removal of the diseased tubes would improve pregnancy rates. Without any prospective study testing the effects of surgical intervention and based on only retrospective data, clinical practice changed over time and several IVF programmes have offered their hydrosalpinx patients a prior to IVF treatment. However, a retrospective design does not offer a causal inference and a major intervention like should be investigated in a prospective randomized trial before it is generally applied. The present prospective study was initiated after publishing our first retrospective data (Strandell et al., 1994), demonstrating a 50% reduction in clinical pregnancy rate in the presence of hydrosalpinges compared with other tubal infertility (13 versus 6%). It was postulated that any intervention causing a disruption of fluid leakage would improve IVF results. The present study was designed within Scandinavia as a multicentre randomized controlled trial with as the main intervention. The aim was to test the hypothesis that removal of hydrosalpinx before undergoing IVF treatment would improve pregnancy rates. Materials and methods A power analysis was undertaken to determine an appropriate sample size. The only available published data in 1994, when the prospective study was planned, was our own retrospective study (Strandell et al., 4 Introduction

1994). Assuming a significance level of 0.05 and an expected difference in pregnancy rates of 1 3 % (13 versus 6%) a sample size of 300 patients undergoing one transfer cycle would be sufficient to reach a power of 0.80 (Altman, 1991). We chose a randomization ratio of 3:, in which the intervention was the larger, for the statistical purpose of retrieving as much information as possible on the effects of intervention. All the Nordic IVF centres were invited to participate in the study and nine of them were recruited for the collaborative study. Approval from each country's or each region's ethics committee was obtained. Study criteria Study design Patient recruitment started in 1994. After informed consent was obtained from the patients fulfilling the criteria of the study, randomization was carried out at each centre by using sealed opaque envelopes in blocks of 10-30 depending on the estimated recruit ment ability. The diagnosis of hydrosalpinx was made by the physician who referred the patient to IVF, and the time awaiting treatment varied between 1 month and years. Patients in the intervention group underwent a laparoscopic unior bilateral, depending on whether one or two hydrosalpinges were present. In cases causing technical difficulties intra-operatively, due to the presence of extensive adhesions, a proximal ligation and a distal fenestration were recommended. A period of minimum months between surgery and oocyte aspiration was advised. The rationale for the time span was to give the endometrium a wash-out period after the potential negative influence of hydrosalpingeal fluid of at least one ordinary menstrual cycle before starting down-regulation with gonadotrophin releasing hormone analogue (GnRHa). Patients in the non-intervention group started their IVF treatment according to routine practice at the individual IVF centres. The number of tubes with hydrosalpinx formation per patient was noted. Ultrasound data on the visibility of hydrosalpinx before and during stimulation were recorded. IVF performance The IVF treatment was carried out according to the routines of each centre, although the policies were very similar including a long protocol down-regulation with a GnRHa administered nasally or s.c., stimulation with either human menopausal gonadotrophin (HMG), highly purified or recombinant follicle stimulating hormone (FSH) and transvaginal ultrasound-guided oocyte retrieval. A maximum of two embryos were routinely transferred, but occasionally three embryos were transferred. In order to identify transfers with impaired quality embryos, a two-grade embryo score was used for assessment of embryo quality, where grade 1 included good quality embryos with < 0 % fragmentation of the blastomeres. Doses, administration and duration of all medications were noted, as well as the number of Group allocation If a cycle was disrupted during stimulation or after oocyte retrieval, the first subsequent cycle ending in a transfer was included in the analysis. If a patient did not undergo a transfer on a cumulative basis, her first disrupted cycle reaching to the phase of oocyte aspiration was included. If she did not reach oocyte aspiration, her disrupted stimulation cycle was included. Patients who were randomized but never started a cycle were included and analysed on an intention-totreat basis. Patients who were randomized to surgery, but subsequently were reluctant to undergo the procedure, were included and analysed both according to an intention-to-treat and according to the actual treatment group. If the patient under the same circumstances again changed her mind after the first cycle and underwent, the subsequent transfer cycle in which the randomization group and the actual treatment group were the same, was used for firstcycle analysis. If, at the time of laparoscopy, the diagnosis of hydrosalpinx was found to be incorrect, the allocation to the intervention group was still kept, regardless of the surgical procedure performed. Patients were not excluded if they had an accidental aspiration of hydrosalpingeal fluid at the time of oocyte retrieval (n = 1). Pregnancy definitions Any positive pregnancy test was recorded, but unless the pregnancy was verified by ultrasound as a gestational sac, it was considered to be biochemical and not included. Only clinical pregnancy rates were used for further analysis. Implantation rate was calculated as the number ofgestational sacs visible on ultrasound divided by the number of embryos transferred. An ectopic implantation was considered as one implanted embryo. The outcome of each pregnancy was noted: spontaneous abortion, childbirth (or ongoing pregnancy > 0 weeks) or ectopic pregnancy. Primary endpoint was clinical pregnancy rate calculated per patient, based on her first transfer cycle. Secondary endpoints were implantation and delivery rates. Statistical methods The comparative analyses included only one cycle per patient and were completed both on an intention-to-treat basis and on treatment basis. For comparison between groups Mann-Whitney U-test was used. Fisher's exact test was used for comparison of dichotomous variables between groups. Mantel-Haenszel's c -test was used for test of linear association in ordered contingency tables (Mantel and Haenszel, 1959). Spearman's rank correlation coefficient was used to analyse correlation between data. Two-tailed statistical tests were conducted at the 5% level of significance. Results Randomization and allocation to treatment groups The targeted sample size of 300 was estimated not to be reached within reasonable time, since the incidence of hydrosalpinx in IVF patients appeared to decrease in Scandinavia during the study period, resulting in a decreased recruitment rate. A decision was made to stop recruitment in May 1998, when 04 patients had been randomized. Sufficient power of 80% would then be achieved if the group demon strated a 30% pregnancy rate compared with 13% in the nointervention group. A total of 116 women were randomized 763 In the study, hydrosalpinx was defined as a distally occluded tube which was pathologically dilated or became pathologically dilated when patency was tested by hysterosalpingography (HSG) or laparoscopy. Patient inclusion criteria were presence of uni- or bilateral hydrosalpinges, suitability for IVF treatment, laparoscopic accessibility and age < 3 9 years at the time of randomization. Exclusion criteria were previous IVF treatments and presence of uterine fibroids needing surgical removal. A concomitant male factor that required intracytoplasmic sperm injection (ICSI) was accepted if the centre was running an established ICSI programme with pregnancy rates equivalent to those of their conventional IVF programme. retrieved, fertilized and cleaved oocytes, sperm quality, fertilization method and any cancellation of cycles.

Figure 1. Flow chart of 04 randomized patients. IVF = in-vitro fertilization. to undergo laparoscopic and 88 women were randomized to no intervention. Five women ran domized to did not undergo any laparoscopy and two of them did not undergo any IVF either. These five patients were still allocated to their original randomization group for the intention-to-treat-based analysis but belong to the non-intervention group for the on-treatment-based analysis. Two patients randomized to surgery did not undergo laparoscopy until after their first transfer. Their second cycle was included in the analysis. One woman in the non-intervention group underwent surgery after her first cycle, which did not result in a transfer. Her first cycle was included in the intentionto-treat-based analysis and her second cycle, which resulted in a transfer, was included in the treatment based analysis. Four women in the intervention group and five in the non intervention group did not start any IVF treatment due to various personal reasons. Figure 1 summarizes the patient flow. Table I. Patient characteristics in the two randomization groups (n = 04) Randomization to (n = 116) (n = 88) Age in years 31.8 ± 3.6 (-37) 31.8 ± 3.7 (3-38) Primary infertility (%) 85 (73) 56 (63) Previous delivery (%) 9 (8) 1 (14) Previous ectopic pregnancy (%) 13(11) 1(14) Bilateral hydrosalpinges at 69 (59)* 36 (41)* inclusion (%) Ultrasound visible hydrosalpinges 37/73 (51) 4/73 (57) at inclusion (%) Values are given as mean ± SD (range) or n (%). *P = 0.0. No other statistically significant difference was found between groups. IVF = in-vitro fertilization. Demographic data Patient and cycle characteristics for the two groups are presented in Tables I and II. No statistically significant differ ence in background variables between the groups could be detected, apart from bilateral hydrosalpinges (as opposed to unilateral), which were more frequent in the group who subsequently underwent. Surgical procedures Among the 11 patients undergoing surgery, the different procedures were distributed as follows: 63 bilateral salpingectomies, 40 unilateral salpingectomies (seven addi tional procedures on contralateral tube and seven patients had previously undergone a unilateral ), six bilateral proximal ligations, one salpingostomy, one adhesiolysis, one without surgical intervention due to patent tubes. Although the diagnosis of hydrosalpinx was incorrect in two cases in the laparoscopy group, no patient was excluded because of the assumption that misdiagnoses also occurred in the non-intervention group. In one case, laparoscopy was converted to laparotomy due to technical difficulties and the same patient suffered from a post-operative period of diarrhoea. 764

One patient had a post-operative infection which was success fully treated. No other complications from surgery were reported. Two women became spontaneously pregnant after a unilateral ; one pregnancy reached full term, the other was ectopic and was treated by. Also one woman with a unilateral hydrosalpinx in the non-intervention group conceived before starting IVF and reached full term. The time interval between surgery and transfer varied between 50 and 577 days (mean 178). Reproductive outcome Table II. Cycle characteristics in 187 patients completing oocyte retrieval in the two randomization groups Randomization to (n = 108) (n = 79) Units of gonadotrophin 31 ± 895 (86-5850)186 Duration of ovarian 11.4 ±. (7-0) 11.6 stimulation (days) No. of retrieved oocytes 10.6 ± 5.9 (-34) 10.6 ICSI procedures (%) 14 (13.1) 10 No. of fertilized and 6.8 ± 4.8 (1-7) 7.0 cleaved oocytes No. of transferred embryos.0 ± 0.3 (1-3).0 (per 184 transfer cycles) Transfers with > 0 % 15 (14.0) 5 fragmentation in all embryos (%) ± 781 (500-4875) ±.9 (5-5) ± 6.1 (1-30) (1.6) ± 4.9 (1-3) analyses Two variables potentially predictive of clinical pregnancy were analysed; number of hydrosalpinges per patient and visiblility on ultrasound. The analysis of correlation between pregnancy rate and number of hydrosalpinges per patient starting IVF showed a negative tendency. Clinical pregnancy and delivery rates decreased with the number of hydrosalpinges but without reaching statistical significance. Clinical pregnancy rates decreased from 35.5% to 30.4% and 19.4% in patient groups with none, one or two hydrosalpinges respectively (not signi ficant, P = 0.078) and the corresponding delivery rates decreased from 8.% to 19.6% and 13.9% (not significant, P = 0.060). Within the non-intervention group, clinical pregnancy rate was 1.4% if the hydrosalpinx was visible on ultrasound before ovarian stimulation was initiated, but reached 34.4% if it was non-visible (not significant, P = 0.9). Subgroups of patients with the predictive variables were analysed and results of transfer cycles are presented in Table V Within the subgroup of patients with bilateral hydrosalpinges at inclusion, implantation rate was found to be significantly higher in the group who underwent surgical intervention (P = 0 038) and pregnancy and delivery rates were doubled (not significant, P = 0.057-0.07). Patients with ultrasound-visible hydrosalpinges expressed a similar effect of with an.4-fold increase of the delivery rate ( P = 0.038) whereas the combined subgroup of patients with bilateral ultrasound visible hydrosalpinges increased its delivery rate 3.5-fold ( P = 0.019). Ovarian response Two patients (1.9%) who had undergone bilateral and three patients (3.5%) who had not, among the 19 patients starting IVF, had their ovarian stimulations discontinued due to poor ovarian response. Among the 187 women who reached oocyte aspiration in the actual treatment groups, the number of retrieved oocytes did not correlate with the number of tubes in place, nor with the number of hydrosalpinges. The mean numbers of retrieved oocytes were ± 0.4 (1-3) (6.5) Values are given as mean ± SD (range) or n (%). No statistically significant difference was found between groups. ICSI = intracytoplasmic sperm injection. Table III. Implantation and pregnancy rates in 04 patients' first cycle in the two randomization groups. Intention-to-treat analysis Randomization to Per included patient (n) Clinical pregnancy rate, n (%) Ongoing or delivery rate, n (%) Per started cycle (n) Clinical pregnancy rate, n (%) Ongoing or delivery rate, n (%) Per transfer cycle (n) Implantation rate, n/n (%) Clinical pregnancy rate, n (%) Abortion rate, n/iup (%) No. of ectopic pregnancies Ongoing or delivery rate, n (%) a a 116 40 (34.5) 31 (6.7) 110 38 (34.5) 30 (7.3) 107 46/10 (1.9) 38 (35.5) 6/36 (16.7) 30 (8.0) 88 (5.0) 15 (17.0) 8 1 (5.6) 14 (17.1) 77 8/151 (18.5) 1 (7.3) 5/19 (6.3) 14 (18.) P Relative risk (RR) 9 5 % CI for the RR 0.17 0.13 1.34 1.57 0.9,.1 0.9,.7 0.1 0.1 1.35 1.60 0.9,.1 0.9,.8 0.51 0.6 0.48 1.18 1.30 0.63 0.8, 1.8 0.8,.0 0., 1.8 0.16 1.54 0.9,.7 Includes spontaneous pregnancies. CI = confidence interval for the relative risk; IUP = intrauterine pregnancy. a 765 Implantation and pregnancy rates are presented in Table III (analyses based on intention-to-treat) and Table IV (analyses based on actual treatment given). A statistically significant difference was found between the delivery rates in the ontreatment based analysis, including all patients. Among patients starting IVF, the delivery rates were 8.7% versus 16.7 (P = 0.059), demonstrating a 7% improvement by. Predictive variables and subgroup

10., 10.9 and 10.7 in patients with zero, one or two tubes left in place and 10.5, 10.6 and 10.7 in patients with zero, one or two hydrosalpinges. Discussion The vast majority of retrospective studies on hydrosalpinx and IVF agree upon the negative association of hydrosalpinx with pregnancy results. Meta-analyses have clearly demonstrated a significant reduction in the probability of pregnancy and a significant increase in the probability of spontaneous abortion in the presence of hydrosalpinx (Daya, 1997; Zeyneloglu etal., 1998; Camus etal., 1999). The mechanism that causes the implantation failure is not yet fully understood. The main theories have focused on the embryotoxicity of the fluid or the mechanical leakage of the fluid into the uterine cavity, causing endometrial alterations hostile to embryo implantation and development or simply a mechanical washout of embryos. Although several mouse studies have demonstrated an impaired blastocyst development in the presence of hydrosalpinx fluid (Mukherjee etal., 1996; Beyler etal., 1997; Rawe etal., 1997; Sachdev etal., 1997), studies on human embryos have not shown any embryonic toxicity (Granot etal., 1998; Strandell etal., 1998). Studies on mouse (Murray et al., 1997; Koong et al., 1998) and human embryos (Strandell et al., 1998) reflecting impairment of embryo development detectable only in 100% hydrosalpinx fluid have concluded that there is no potent embryotoxic factor present, but merely a deficiency of essential nutrients causing the impaired development. This theory is further supported by Murray et al. who demonstrated that the impaired blastocyst developmental rate in 100% hydrosalpinx fluid was counteracted by the addition of energy sources (Murray et al., 1998). The theory of the negatively affected endometrium is Table IV. Implantation and pregnancy rates in 04 patients' first cycle in the two actual treatment groups. On-treatment analysis Randomization to P Relative 95% CI for risk (RR) the RR Per included patient (n) 11 9 Clinical pregnancy rate, n (%) a 41 (36.6) (3.9) 0.067 1.53 0.99,.37 Ongoing or delivery rate, n (%) a 3 (8.6) 15 (16.3) 0.045 1.75 1.01, 3.03 Per started cycle (n) 108 84 Clinical pregnancy rate, n (%) 39 (36.1) 1 (5.0) 0.11 1.44 0.9,.59 Ongoing or delivery rate, n (%) 31 (8.7) 14 (16.7) 0.059 1.7 0.98, 3.0 Per transfer cycle (n) 105 80 Implantation rate, n/n (%) 47/06 (.8) 8/149 (18.8) 0.43 1.1 0.80, 1.3 Clinical pregnancy rate, n (%) 39 (37.1) 1 (6.) 0.15 1.41 0.91,.0 Abortion rate, n/iup (%) 6/37 (16.) 5/19 (6.3) 0.48 0.6 0., 1.76 No. of ectopic pregnancies Ongoing or delivery rate, n (%) 31 (9.5) 14 (17.5) 0.083 1.69 0.96,.95 a Includes spontaneous pregnancies. CI = confidence interval for the relative risk; IUP = intrauterine pregnancy. Table V. Implantation and pregnancy rates in first transfer cycle in subgroups of patients with bilateral and/or ultrasound visible hydrosalpinges at inclusion, in the two actual treatment groups. On-treatment analysis Randomization to P Relative risk (RR) 95% CI for the RR Bilateral hydrosalpinges (n) 63 34 Implantation rate, n/n (%) 31/11 (5.6) 8/65 (1.3) 0.038.08 1.0, 4.6 Clinical pregnancy rate, n (%) 5 (39.7) 7 (0.6) 0.07 1.9 0.93, 3.99 Ongoing or delivery rate, n (%) 1 (33.3) 5 (14.7) 0.057.7 0.94, 5.47 Ultrasound visible (n) 35 40 Implantation rate, n/n (%) 1/70 (30.0) 13/78 (16.7) 0.077 1.80 0.98, 3.3 Clinical pregnancy rate, n (%) 16 (45.7) 9 (.5) 0.09.13 1.08, 4. Ongoing or delivery rate, n (%) 14 (40.0) 7 (17.5) 0.038.40 1.09, 5.8 Bilateral and ultrasound visible (n) 0 19 Implantation rate, n/n (%) 16/39 (41.0) 4/37 (10.8) 0.004 3.80 1.40, 10.31 Clinical pregnancy rate, n (%) 1 (60.0) 3 (15.8) 0.008 3.80 1.7, 11.40 Ongoing or delivery rate, n (%) 11 (55.0) 3 (15.8) 0.019 3.48 1.15, 10.59 CI = confidence interval. 766

supported by Cohen et al. where hydrosalpinx was studied in patients in an oocyte donor programme and significantly lower embryo implantation rates were found in hydrosalpinx patients than in normal controls (Cohen etal., 1999). The endometrial receptivity has been studied (Meyer et al., 1996), where integrin expression was examined in endometrial biopsies. Integrins are adhesion molecules that have been identified as markers of uterine receptivity, and women with hydrosalpinges expressed significantly lower amounts of integrins than those without hydrosalpinges. After surgical correction, a return to normal integrin values was demonstrated. Still, there is a lack of knowledge as to what substance or agent or physiological event actually causes the endometrial changes. A strict mechanical explanation has been discussed (Sharara, 1999), suggesting that embryonal apposition to the endometrial surface will be compromised when a fluid interface exists. The only hitherto prospective randomized trial, besides the current study, examining the effect of any surgical interven tion is a pilot study which indicated a positive effect of on implantation rates but without any statistical significance (Dechaud et al., 1998). Due to the small size of the study, the power to detect any true difference between the groups was only 11% and the authors ask for future studies that could form the base for a future meta-analysis. Another alternative to answer the question whether is of clinical benefit, is to conduct a multicentre trial as in the present study. However, a difficulty is that the results of IVF have improved over time and the original sample size calculation based on retrospective data from the early 1990s, gave a smaller sample size than a calculation based on present results. Consequently, the sample size requirements tend to increase over time which results in a progressively smaller likelihood of reaching statistical significance. The present study is the largest of its kind to date. A 7% (relative risk = 1.7) improvement of delivery rates in patients starting In the debate about hydrosalpinx most authors have advocated prophylactic, with one exception (Puttemans and Brosens, 1996) stating that salpingoscopy is mandatory to avoid unnecessary removal of Fallopian tubes, which would leave the patient irreversibly compromised. In the present study, the diagnosis of hydrosalpinx had been made by a previous HSG or diagnostic laparoscopy, at which tubal surgery had already been rejected. Obviously, the judgement of suitability for tubal surgery differs between centres, and, as a referral centre, we found tubes that were more suitable for reconstruction than removal. Another aspect concerning is the risk of ovarian function being disturbed. Retrospective studies using patients with previous do not show any differ ences in ovulation stimulation variables (Kassabji etal., 1994; Vandromme etal., 1995; Shelton etal., 1996; Murray etal., 1998; Ejdrup Bredkjaer et al., 1999) and the data in this prospective study are in agreement. In a case-control study of ovarian function after, no differences in stimulation variables or number of retrieved oocytes could be detected (Verhulst et al., 1994). However, the first study to examine the adnexal sides separately demonstrated a significant reduction in number of developed follicles and number of retrieved oocytes from the ovary ipsilateral to which a unilateral was previously performed, although there were no differences in total numbers when considering both sides (Lass etal., 1998). It was proposed in a retrospective study (Freeman et al., 1998) that hydrosalpinges may have a permanent negative influence on ovarian function, follicular development and oocyte quality since implantation of transferred embryos and normal blastulation of untransferred embryos remained low and in-vitro growth arrest and degeneration remained high despite surgical treatment of hydrosalpinges. The results of this study are partly supportive, since the laparoscopy group to a significantly larger extent had bilateral hydrosalpinges at inclusion and in their results tended to express a higher degree of poor embryo quality than in the non-intervention group in which unilateral hydrosalpinges were more frequent. However, the difference is not of any statistical significance and no signs of impaired follicular development were present in the laparoscopy group to support the observations of Freeman etal. Besides bilaterality, the visibility of hydrosalpinx on ultra sound was analysed, as a potential predictor of fluid accumula tion. Studies using ultrasound to diagnose hydrosalpinx have tended to show larger differences in pregnancy rates between hydrosalpinx patients and controls (Andersen et al., 1994; Katz etal., 1996; Van Voorhis etal., 1998). Anotherretrospec767 Different treatment options with the purpose of disrupting the connection of hydrosalpinx fluid into the uterus have been suggested. Transvaginal aspiration at the time of oocyte retrieval has been described in two retrospective trials (Sowter etal., 1997; Van Voorhis etal., 1998), of which only the latter showed a positive effect. Both studies were of small size, and therefore we believe no definitive conclusions can be drawn. Several retrospective studies have compared the outcome of IVF in patients after with patients carrying hydrosalpinges (Kassabji et al., 1994; Vandromme et al., 1995; Shelton et al., 1996; Meyer et al., 1997) and they all demonstrate a positive effect on pregnancy rates after, although the retrospective design carries an inherent risk of methodological and statistical error. Other laparoscopic procedures, including proximal ligation, neosalpingostomy and, have also been studied retrospectively (Murray et al., 1998). Although the surgical procedures taken together seem to be beneficial, the numbers are too small to draw any conclusions concerning the separate procedures. In a case-control study analysing the effect of bilateral in hydrosalpinx patients as a single procedure, a positive impact on pregnancy rates was shown (Ejdrup Bredkjaer et al., 1999). IVF after clearly indicates a benefit from surgical intervention. Patients with bilateral hydrosalpinges showed a significant increase in implantation rate after and delivery rates were more than doubled (relative risk =.7). The calculated confidence intervals clearly show that the results are in the positive direction only. These results agree well with a previous retrospective study, which demonstrated significantly lower implantation and pregnancy rates in patients with bilateral hydrosalpinges as compared with unilateral (Wainer et al., 1997).

Acknowledgements We are most grateful to the Scandinavian IVF centres participating in the study and the responsible investigator at each site for patient recruitment and data collection. The study was supported by grants from the Göteborg Medical Society, the Hjalmar Svensson Foundation and the society 'Ordensallskapet W:6'. Nils-Gunnar Pehrson is acknowledged for valuable statistical advice. References Aboulghar, M.A., Mansour, R.T. and Serour, G.I. (1998) Controversies in the modern management of hydrosalpinx. Hum. Reprod. Update, 4, 88-890. Altman, G.D. (1991) Practical Statistics for Medical Research. Chapman and Hall, London, pp. 455-460. Andersen, A.N., Yue, Z. and Meng, F.J. et al. (1994) Low implantation rate after in-vitro fertilization in patients with hydrosalpinges diagnosed by ultrasonography. Hum. Reprod., 9, 1935-1938. Andersen, A.N., Lindhard, A., Loft, A. et al. (1996) The infertile patient with hydrosalpinges IVF with or without. Hum. Reprod., 11, 081-084. Beyler, S.A., James, K.P., Fritz, M.A. et al. (1997) Hydrosalpingeal fluid inhibits in-vitro embryonic development in a murine model. Hum. Reprod., 1, 74-78. Blazar, A.S., Hogan, J.W., Seifer, D.B. etal. (1997) The impact of hydrosalpinx on successful pregnancy in tubal factor infertility treated by in vitro fertilization. Fertil. Steril., 67, 517-50. Camus, E., Poncelet, C., Goffinet, F. et al. (1999) Pregnancy rates after IVF in cases of tubal infertility with and without hydrosalpinx: meta-analysis of published comparative studies. Hum. Reprod., 14, 143-149. Cohen, M.A., Lindheim, S.R. and Sauer, M.V. (1999) Hydrosalpinges adversely affects implantation in donor oocute cycles. Hum. Reprod., 14, 1087-1089. Csemiczky, G., Landgren, B.-M., Fried, G. et al. (1996) High tubal damage grade is associated with low pregnancy rate in women undergoing in-vitro fertilization treatment. Hum. Reprod., 11, 438-440. Daya, S. (1997) Hydrosalpinx and fertility potential. Pre-meeting symposium at the Annual Meeting of the Canadian Fertility and Andrology Society, Niagara on the Lake, Canada. 768 Dechaud, H., Daures, J.P., Arnal, F. et al. (1998) Does previous improve implantation and pregnancy rates in patients with severe tubal factor infertility who are undergoing in vitro fertilization? A pilot prospective randomized study. Fertil. Steril., 69, 100-105. dewit, W., Gowrising, C.J., Kuik, D.J. et al. (1998) Only hydrosalpinges visible on ultrasound are associated with reduced implantation and pregnancy rates after in-vitro fertilization. Hum. Reprod., 13, 1696-1701. Ejdrup Bredkjaer, H., Ziebe, S., Hamid, B. et al. (1999) Delivery rates after in-vitro fertilization following bilateral due to hydrosalpinges: a case control study. Hum. Reprod., 14, 101-105. Fleming, C. and Hull, M.G.R. (1996) Impaired implantation after in vitro fertilization treatment associated with hydrosalpinx. Br. J. Obstet. Gynaecol., 103, 68-7. Freeman, M.R., Whitworth, C.M. and Hill, G.A. (1998) Permanent impairment of embryo development by hydrosalpinges. Hum. Reprod., 13, 983-986. Granot, I., Dekel, N., Segal, I. et al. (1998) Is hydrosalpinx fluid cytotoxic? Hum. Reprod., 13, 160-164. Kassabji, M., Sims, J.A., Butler, L. et al. (1994) Reduced pregnancy outcome in patients with unilateral or bilateral hydrosalpinx after in vitro fertilization. Eur. J. Obstet. Gynecol. Reprod. Biol., 56, 19-13. Katz, E. Akman, M.A., Damewood, M.D. et al. (1996) Deleterious effect of the presence of hydrosalpinx on implantation and pregnancy rates with in vitro fertilization. Fertil.Steril., 66, 1-15. Koong, M.K., Jun, J.H., Song, S.J. et al. (1998) A second look at the embryotoxicity of hydrosalpingeal fluid: an in-vitro assessment in a murine model. Hum. Reprod., 13, 85-856. Lass, A., Ellenbogen, A., Croucher, C. et al. (1998) Effect of on ovarian response to superovulation in an in vitro fertilization-embryo transfer program. Fertil. Steril., 70, 1035-1038. Mantel, N. and Haenszel, W. (1959) Statistical aspects of the of the analysis of data from retrospective studies of disease. J. National Cancer Inst.,, 719-748. Meyer, W.R., Castelbaum, A.J., Somkuti, S. et al. (1997) Hydrosalpinges adversely affect markers of endometrial receptivity. Hum. Reprod., 1, 1393-1398. Mukherjee, T., Copperman, A.B., McCaffrey, C. et al. (1996) Hydrosalpinx fluid has embryotoxic effects on murine embryogenesis: a case for prophylactic. Fertil. Steril., 66, 851-853. Murray, C.A., Clarke, H.J., Tulandi, T. et al. (1997) Inhibitory effect ofhuman hydrosalpingeal fluid on mouse preimplantation embryonic development is significantly reduced by the addition of lactate. Hum. Reprod., 1, 504-507. Murray, D.L., Sagoskin, A.W., Widra, E.A. et al. (1998) The adverse effect of hydrosalpinges on in vitro fertilization pregnancy rates and the benefit of surgical correction. Fertil.Steril., 69, 41-45. Ng, E.H., Yeung, W.S. and Ho, P.C. (1997) The presence of hydrosalpinx may not adversely affect the implantation and pregnancy rates in in vitro fertilization treatment. J. Assist. Reprod. Genet., 14, 508-51. Puttemans, P.J. and Brosens, I.A. (1996) Salpingectomy improves in-vitro fertilization outcome in patients with a hydrosalpinx: blind victimization of the fallopian tube? Hum. Reprod., 11, 079-081. Rawe, V.J., Liu, J., Shaffer, S. et al. (1997) Effect of human hydrosalpinx fluid on murine embryo development and implantation. Fertil. Steril., 68, 668-670. Sachdev, R., Kemmann, E., Bohrer, M.K. et al. (1997) Detrimental effect of hydrosalpinx fluid on the development and blastulation of mouse embryos in vitro. Fertil. Steril., 68, 531-533. 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, 53-55. Sharara, F.I. (1999) The role of hydrosalpinx in IVF: simply mechanical? Hum. Reprod., 14, 577-578. Sharara, F.I., Scott Jr, R.T., Marut, E.L. et al. (1996) In-vitro fertilization outcome in women with hydrosalpinx. Hum. Reprod., 11, 56-530. Sowter, M.C., Akande, V.A., Williams, J.A. et al. (1997) Is the outcome of in-vitro fertilization and embryo transfer treatment improved by spontaneous or surgical drainage of a hydrosalpinx? Hum. Reprod., 1, 147-150. Strandell, A., Waldenstrom, U., Nilsson, L. etal. (1994) Hydrosalpinx reduces in-vitro fertilization/embryo transfer rates. Hum. Reprod., 9, 861-863. Strandell, A., Sjögren, A., Bentin-Ley, U. et al. (1998) Hydrosalpinx fluid does not adversely affect the normal development of human embryos and implantation in vitro. Hum. Reprod., 13, 91-95. Wainer, R, Camus, E., Camier, B. et al. (1997) Does hydrosalpinx reduce the pregnancy rate following in vitro fertilization. Fertil. Steril., 68, 10-106. tive study clearly demonstrated that patients with ultrasoundvisible hydrosalpinges carried a far worse prognosis compared with both non-visible hydrosalpinges and other tubal factor infertility (de Wit et al., 1998). The data presented here are in agreement with the results from the studies based on ultrasound. Moreover, an improved outcome after in this patient category has been confirmed. A history of hydrorrhoea combined with a sonographically visible hydrosalpinx may impair the prognosis even more (Andersen et al., 1996). It has to be emphasized that the present study is based only on a first cycle. With that in mind, a general recommendation of preventive for all patients diagnosed with hydrosalpinges cannot be given. It would be valuable to identify a subset of patients who would benefit most from surgery, as suggested (Aboulghar et al., 1998). The present study was able to show that a surgical intervention increased implantation rates in patients with bilateral hydrosalpinges. A clear benefit of in patients with ultrasoundvisible hydrosalpinges was also demonstrated. Based on the results of the present study, it is concluded that a preventive can be recommended for patients with hydrosalpinges enlarged enough to be visible on ultrasound and in particular for those affected bilaterally. The patients in the present study will in the near future have completed all their cycles and the cumulative data will hopefully provide more precise guidelines regarding preventive.

Van Voorhis, B.J., Sparks, A.E., Syrop, C.H. et al. (1998) Ultrasound-guided aspiration of hydrosalpinges is associated with improved pregnancy and implantation rates after in-vitro fertilization cycles. Hum. Reprod., 13, 736-739. Vandromme, J., Chasse, E., Lejeune, B. et al. (1995) Hydrosalpinges in invitro fertilization: an unfavourable prognostic feature. Hum. Reprod., 10, 576-579. Verhulst, G., Vandersteen, N., Van Steirteghem, A.C. et al. (1994) Bilateral does not compromise ovarian stimulation in an in-vitro fertilization/embryo transfer programme. Hum. Reprod., 9, 64-68. Zeyneloglu, H.B., Arici, A. and Olive, D.L. (1998) Adverse effects of hydrosalpinx on pregnancy rates after in vitro fertilization-embryo transfer. Fertil. Steril., 70, 49-499. Received on April 6, 1999; accepted on August 3, 1999 769