FERTILITY AND STERILITY ~' Vcd. 67, No. 3, March 1997 Copyright ~ 1997 American Society for Reproductive Medicine Printed oil ~lcid.frt, e paper in U. S. A. Pregnancy outcome after laparoscopic fimbrioplasty in nonocclusive distal tubal disease* Walid A. Saleh, M.D.t Alexander M. Dlugi, M.D.S Division of Reproductive Endocrinology and InfertiliO,, Department of Gynecology and Obstetrics, Heno' Ford Hospital, Detroit, Michigan Objective: To determine pregnancy rates (PR) after fimbrioplasty and salpingostomy in nonocclusive distal tubal disease. To evaluate the relative impact of various factors using contemporary st~itistical analysis. Design: Prospective cohort. Setting: Tertiary institutional infertility clinic. Patient(s): Infertility patients. Intervention(s): Fimblioplasty and salpingostomy. Main Outcome Measure(s): Cumulative PR, monthly fecundity rates, monthly probability of pregnancy, crude PR, and cure rates. Result(s): Thirty-five percent of patients conceived with a cure rate of 72.2%, monthly probability of pregnancy of 3.9%, and monthly fecundity rate of 3.9%. Cumulative PRs were 22%, 35%, and 58% at 6, 12, and 24 months, respectively. Pairwise comparisons (unilateral, bilateral, or either) failed to detect any statistical difference between the salpingostomy and fimbrioplasty groups. Salpingostomy patients initially may have a higher tendency to become pregnant but appear to lose that advantage after the first few months. When patients with tubo-ovarian adhesions are excluded from the analysis, patients who underwent a bilateral salpingostomy as their sole procedure had better outcome compared with those who only underwent bilateral fimbrioplasty. There was no significant association between pregnancy outcome and the presence of endometriosis, other infertility factors, or tubo-ovarian adhesions. The staging of adnexal adhesions and endometriosis did not predict pregnancy outcome. Conclusion(s): Laparoscopic fimbrioplasty and salpingostomy are clinically efficacious for the treatment of nonocclusive distal tubal disease. After accounting for statistical interactions of various factors among them, no particular association with pregnancy outcome could be identified. This illustrates the need for a revision of the classification of patients with distal tubal disease. Fertil Steril v 1997;67:474-80 Key Words: Laparoscopy, fallopian tube diseases, infertility, adhesions, fimbrioplasty, salpingostomy Tuboperitoneal factors are implicated in up to 40% of infertility couples. In the presence of complete distal tubal occlusion, where pregnancy outcome is sig- Received No,]ember 16, 1995; revised and accepted October 16, 1996. * Presented at the 51st Annual Meeting of the American Society For Reproductive Medicine, Seattle, Washington, October 7 to 12, 1995. t Present address: The Women's Health Center, Lisbon, Ohio. $ Present address and reprint requests: Alexander M. Dlugi, M.D., Center for Reproductive Endocrinology, Atlantic Health System, 95 Mt. Kemble Avenue, Morristown, New Jersey 07962 (FAX: 201-971-4601). nificantly reduced, IhrF offers a greater chance of success (1). Laparoscopic microsurgery continues to be the mainstay of therapy for the treatment of nonocclusive tubal infertility. Unfortunately, most series have included patients with distal obstructions in their analysis. Consequently, it has been difficult to identify the contribution of isolated surgical findings in tubal infertility. Several factors restrict the analysis of treatment success and comparison between studies: lack of consensus on scoring pelvic disease, variety of nomenclature of surgical techniques, study designs, length of follow-up, clustering of additional infertility factors, subjective outcome 474 Saleh and Dlugi Pregnancy outcome following fimbrioplasty Fertility and Sterility '~
variables, and lack of actuarial statistical methods. We prospectively followed a group of infertility patients with distal tubal disease managed laparoscopically with the KTP/532 laser (Laserscope, Santa Clara, CA). We tested the following hypotheses: [1] bilateral disease has a worse prognosis compared to unilateral disease, [2] patients undergoing a fimbrioplasty have a better pregnancy outcome compared with patients undergoing a salpingostomy, [3] tubo-ovarian adhesions have a negative impact on pregnancy outcome, [4] interfimbrial adhesions have a better prognosis compared with phimosis, [5] adhesions secondary to endometriosis have less negative impact compared with adhesions from other causes, and [6] other infertility factors have a negative effect on pregnancy outcome. We carefully identified, stratified, and statistically analyzed the various factors with contemporary statistical methods, paying special attention to the above problems and report the results herein. MATERIALS AND METHODS Between October 1988 and July 1992, 769 consecutive patients underwent operative endoscopic surgery by the senior author (A.M.D.). Of those, 291 infertility patients underwent a surgical procedure for nonocclusive distal tubal disease. The duration of infertility was -> 12 months and the follow-up up to 41.2 months. In the absence of pregnancy, the follow-up was ->6 months. Twenty-nine patients with nonocclusive disease underwent a second operative laparoscopy and were censored at the time of their second surgery. Before surgery, all patients underwent a thorough infertility evaluation. Patients with additional factors contributing to their infertility were treated for them according to the diagnosis. A male factor was defined by a count < 20 x 106/mL, motility < 50% or <50% normal morphology. A cervical factor was defined by at least two appropriately timed postcoital tests demonstrating less than five progressively motile sperm per high power field. Patients with a male or cervical factor underwent an IUI with or without empiric ovarian stimulation. Patients with ovulation dysfunction underwent ovulation induction with either clomiphene citrate or hmg. Patients with uterine synechiae or filling defects by hysterosalpingography underwent concurrent hysteroscopic intervention. A KTP/532 laser was used in all cases to restore gross pelvic anatomy to normal. The laser power ranged from 7 to 10 W with continuous pulses. A 600-pm core flexible quartz fiber delivery system was used to vaporize peritoneal and ovarian surface implants of endometriosis while a 400-pm fiber was used for fimbrioplasty, cuff salpingostomy, and lysis of adhesions. Endometriosis was staged according to the revised American Fertility Society (AFS) classification system (2). Care was taken to pass a very fine manipulating instrument gently through the fimbriated ends in a brushing manner to identify microadhesions between fimbrial folds (interfimbrial adhesions) (3). Traditionally, the term fimbrioplasty is used to define broadly any procedure that involves reconstruction of the fimbria. It does not differentiate between situations that require only deagglutination and dilatation of the fimbria and those that require serosal incision to relieve distal constriction of a nonoccluded tube. Because neosalpingostomy represents the surgical creation of a new ostium when the fimbrial end is totally occluded, we have used the term salpingostomy to describe this procedure in a nonoccluded tube and reserved the term fimbrioplasty for procedures that involve deagglutination and dilatation of the fimbrial end. All patients with interfimbrial adhesions and/or mild phimosis (defined as narrowing of the distal tube) of the tube had a fimbrioplasty. Phimosis was considered as moderate to severe when the tube was constricted partially and required a cuff salpingostomy and mild when dilatation was deemed sufficient. In the statistical analysis, salpingostomy was considered the predominant treatment for tubes that had both a salpingostomy and a fimbrioplasty performed on them. All paratubal and paraovarian adhesions were divided with the KTP laser (salpingovariolysis). Proximal tubal patency was confirmed with transcervical injection of methylene blue through a uterine-manipulating instrument (HUMI; Unimar, Canoga Park, CA). We excluded fi'om the study patients with proximal and midtubal occlusion. Patients with hydrosalpinges also were excluded and are the subject of a separate publication (1). Patients with a history of unilateral salpingectomy or those with an unsuccessfully operated side were counted as bilateral disease. At the time of surgery, care was taken to record the side, place, extent, and nature of disease and whether the tube was patent before and after intervention. The AFS classification of adnexal adhesions was used and a grade was assigned to each pelvic side accordingly (4). Adhesions were classified as filmy or dense, and as fimbrial, tubo-ovarian, or other pelvic adhesions. At the termination of the procedure, copious irrigation of the pelvis was carried out and the deperitonealized areas of the pelvic sidewalls were covered with Interceed (TC7) adhesion barriers (Johnson and Johnson Medical, Inc., New Brunswick, NJ). After laparoscopy, patients were discharged routinely on the day of surgery. For patients who conceived, the time of interest Vol. 67, No. 3, March 1997 Saleh and Dlugi Pregnancy outcome following fimbrioplasty 475
was from the date of surgery to the last menstrual period before conception. For patients who did not become pregnant, the time interval of interest was from the dateof surgery to the last follow-up. For analysis purposes, the patients who did not become pregnant were censored at last follow-up. The log-rank survival test was used to evaluate the association between disease grade, study factors, and pregnancy. A stepwise Cox proportional hazards regression analysis evaluated the association between pregnancy outcome and the various study factors of interest in a single multivariate analysis. Cumulative pregnancy rates (PRs) were calculated from the life-table using the Kaplan-Meier approach (5). The two-parameter exponential model of Guzick and Rock (6) yielded the cure rate and the monthly probability of pregnancy. Monthly fecundity rates were calculated using the ratio of total pregnancies to total person-months of exposure from Cramer et al. (7). For the cnade PR comparisons, the ~(2 test was used under assumption of a binomial distribution. The two-sample t-test with logarithmic transformation was used for the comparison of time to achieve pregnancy. All statistical tests were two tailed. Patient records were electronically stored on a Macintosh computer (Apple Computer, Inc., Cupertino, CA) using the M.E.D. Patient medical record system (JAM Software PTY Limited, Leichardt, New South Wales, Australia). This allowed for ready access to and analysis of the clinical parameters under investigation. All statistical analyses were performed using S.A.S. release 6.07 from Statistical Analysis System Institute, Inc. (Cary, NC). RESULTS Effect of Age and Duration of Infertility The mean _ SD age of the patients was 31.8 _ 4.2 years and the mean age of patients who conceived was not different from those who did not. Patients who conceived were infertile for a shorter time compared with those who did not (3.0 _+ 1.8 versus 4.2 _ 3.0 years; P < 0.001). Pregnancy outcome was similar for patients with primary compared with patients with secondary infertility. The waiting time to surgery after the initial visit was similar in the two groups and averaged 0.7 _+ 0.3 months. Patients who achieved pregnancy had a shorter time of followup compared with those who did not conceive (7.7 _ 7.7 versus 9.8 _+ 7.5 months; P = 0.004). Overall PRs Overall, 102 patients conceived (35.1%). There was one ectopic pregnancy (1.0%) and 13 spontaneous abortions (12.7%). For the overall study group, 100" 80" o - 60- y ~ 4o- E 0 20-6 12 18 24, i = 30 36 42 Months Figure 1 Cumulative PR for all the patients with nonocclusive distal tubal disease. the cumulative PRs were 22%, 35%, 58%, and 78% at 6, 12, 24, and 36 months, respectively (Fig. 1). Fifty percent of patients who eventually conceived did so within 5.1 months. Overall, the monthly fecundity rate, monthly probability of pregnancy, cure rate, and time to conception were 3.9%, 3.9%, 72.2% and 7.7 _ 0.8 months, respectively (Table 1). Salpingostomy Versus Fimbrioplasty The log-rank test was used to perform various pairwise combinations involving the fimbrioplasty and salpingostomy categories. Cox proportional hazards regression analysis then was used to account for patient age, total duration of infertility, presence of other factors, endometriosis, and tubo-ovarian adhesions. Statistical significance was not detected at the 0.05 level for any of the comparisons (unilateral versus bilateral salpingostomy, unilateral versus bilateral fimbrioplasty, unilateral salpingostomy versus unilateral fimbrioplasty, and bilateral salpingostomy versus bilateral fimbrioplasty). Similarly, when both surgical procedures were counted as one in the statistical analysis, pregnancy outcome for patients with unilateral tubal disease was not different from those with bilateral disease. The only possible trend suggested by the data is an initial higher PR for the salpingostomy patients, although they appear to lose this advantage after the first few months. However, when patients with tubo-ovarian adhesions were excluded from the analysis, patients who underwent a bilateral salpingostomy as their sole procedure had a better outcome compared with those who only underwent a bilateral fimbrioplasty (Fig. 2, Table 1). A statistically significant difference was detected from both the log-rank (P = 0.01) and the Cox regression test (P = 0.03). For comparison with other studies, we also ana- 476 Saleh and Dlugi Pregnancy outcome following fimbrioplasty Fertility and Sterility ~
Table 1 Pregnancy Rates for all the Patients Combined and for Patients Stratified by Sole Surgical Procedure Performed Bilaterally* Crude Monthly probability Cure Mean months to Category PRt of pregnancy+ rate + pregnancy Overall (n = 291) 35.1 (102) 3.9 (3.3 to 4.5) 72.2 (45.6 to 98.8) 7.7 _+_ 0.8 Fimbrioplasty in = 88) 39.8 (35) 4.1 (0.6 to 7.6) 73.3 (39.1 to 100.0) 9.7 +_ 1.7 Salpingostomy in = 9) 55.6 (5) 16.1 (0.0 to 36.3) 91.0 (17.4 to 100.0) 2.4 _+ 0.9 * Patients with tubo-ovarian adhesions are excluded from the stratified analysis. t Values in parentheses are the number of pregnancies. $ Numbers in parentheses are 95% confidence intervals. Values are means +_ SEM. lyzed the 181 patients who underwent a bilateral fimbrioplasty or salpingostomy with or without salpingovariolysis as a subgroup classified under "timbrioplasty" by the International Federation of Fertility and Sterility Classification of tubal procedures (3). The crude PR, monthly fecundity rate, monthly probability of pregnancy, cure rate, and mean time to conception were 35.9%, 3.8%, 4.2%, 73.0%, and 8.3 months, respectively. Cumulative PRs for this particular grouping were 21%, 34%, 59%, and 73% at 6, 12, 24, and 36 months, respectively, and similar to the overall study group. Impact of Tubo-ovarian and Pelvic Adhesions Surgical findings initially were recorded as tubal, ovarian, and tubal-ovarian adhesions and described as filmy or dense. These were pooled together into a tubo-ovarian adhesions category after the Cox regression analysis failed to find a statistical difference among these separate types of adnexal adhesions. Similarly, for all four variables, pregnancy parameters were similar for the none and unilateral subgroups. Therefore, these were grouped together 60 o 20 in the rest of the analysis. The results indicate that no significant individual main effects were detected for interfimbrial adhesions, phimosis, and tubo-ovarian adhesions. We evaluated the AFS classification of adnexal adhesions (minimal, mild, moderate, and severe) (4). Patients with bilateral adhesions were analyzed first by stage of the least severe side and then by stage of most severe side. Neither of these staging methods yielded a relationship between AFS staging and pregnancy outcome. Patients with one totally unaffected side by laparoscopy were not significantly different from patients with bilateral disease of any kind (Table 2). Impact of Additional Infertility Factors Four different combinations involving endometriosis and other infertility factor categories are shown in Table 2. The log-rank test was used to perform the various pairwise comparisons of interest. Patients with endometriosis but no other infertility factor tended to have better outcome compared with the other three categories (P = 0.05). However, when Cox proportional hazards regression is used to account for patient age, duration of infertility, presence of salpingostomy, fimbrioplasty, and tubo-ovarian adhesions, that apparent advantage largely disappears. Similarly, there was no association between the revised AFS staging of endometriosis and pregnancy outcome. Bilateral tubo-ovarian adhesions were present in 38 of 198 patients (19.2%) with endometriosis whereas 34 of 93 patients (36.6%) without endometriosis had similar adhesions (P = 0.001). DISCUSSION 0 0 6 1'2 Monlhs 1'8 Figure 2 Cumulative PR when bilateral fimbrioplasty CA), and bilateral salpingostomy (O) were the sole surgical procedures performed. Patients with tubo-ovarian adhesions are excluded. Patients who underwent a bilateral salpingostomy as their sole procedure had better outcome compared with those who only underwent bilateral fimbrioplasty (P < 0.03). Many authors have attempted to evaluate the value of various surgical procedures for the correction of tubal infertility (1, 8-22). However, conclusions about the contribution of specific surgical findings or procedures and comparisons with other multivariate analyses (8-10) cannot be made from most of these series mainly because of the grouping Vol. 67, No. 3, March 1997 Saleh and Dlugi Pregnancy outcome following fimbrioplasty 477
Table 2 Pregnancy Rates for Patients Stratified for Presence of Endometriosis and Other Infertility Factors Crude Monthly probability Cure Category PR* of pregnancy~ rater Mean months to pregnancy$ Endometliosis only {n = 99} 42.4 142) Other factors only (n = 62) 33.9 (21) Both In = 99) 34.3 134) Neither In = 31) 16.1 (5) 12.2 (10.2 to 14.1) 9.2 (5.0 to 13.8) 5.1 (4.9 to 15.1) 4.0 (1.5 to 6.6) 64.7 (62.0 to 67.4} 5.8 + 1.0 56.5 {47.9 to 65.1} 9.9 _+ 1.9 65.8 (39.8 to 91.8} 8.0 _+ 1.2 58.5 {36.5 to 80.5} 12.3 ~ 6.0 * Values in parentheses are the number of pregnancies. Numbers in parentheses are 95~ confidence intervals. Values are means _+ SEM. of patients with different surgical findings, the inclusion of patients with tubal obstructions, and the limited statistical analysis provided. In this study, 295 reconstructive tubal surgeries performed with the KTP/532 laser on nonocclusive distal tubal disease were analyzed. We conducted this study in a prospective manner to limit the recall bias faced by retrospective studies when attempting to grade the severity of pelvic disease from previously dictated operative notes. All surgeries were performed by the same surgeon (A.M.D.). Patients with bilateral hydrosalpinges or proximal obstructions were not included in order to focus upon the impact of fimbrial and adnexal pathology and the various procedures applied to their correction. The trend introduced by the Ad Hoc Committee of the International Federation of Fertility and Sterility has been to group together linear serosal incisions of a partially occluded fimbrial opening (that we referred to as salpingostomy), deglutinations of interfimbrial microadhesions, and dilatation of a phimosed tube as part of a single fimbrioplasty category (3). E~:en in its broad definition of the term, fimbrioplasty rarely has been reported as a separate procedure. Most series also failed to exclude patients with salpingovariolysis in their fimbrioplasty data analysis. Therefore, data on isolated surgical procedures and isolated surgical findings in the absence of other pathology are lacking. To avoid further confounding influences, we were particularly careful in differentiating between the various types of fimbrioplasties and salpingovariolyses. The use of the life-table analysis facilitates comparative analysis between studies. The best estimates of PRs are the cure rate and monthly probability of pregnancy obtained from the two-parameter model (6). Results from this model are not widely available in the literature. The monthly probability of pregnancy represents the monthly probability of conceiving among those who eventually would get pregnant. The cure rate represents the proportion of patients who would conceive eventually given an infinite follow-up. The monthly fecundity rate uses a single-parameter model and is easy to calculate. Its limitation for infertility populations results from its assumption of a 100% success rate given an infinite follow-up. It represents the probability a patient has to conceive within a single month and can be obtained by dividing the number of pregnancies by the sum of the person-months of exposure over all intervals (7). The 0.05 significance level has been used for each of the statistical tests used, although P values between 0.1 and 0.05 should be considered borderline because of the multiple testing performed across the study group. The crude PR of 35.1% for bilateral fimbrioplasties performed with the KTP/532 laser were within the range reported for conventional laparotomy (11), microsurgery by laparotomy (9, 12, 13), CO2 laser by laparotomy (14), laparoscopic sharp dissection (15), laparoscopic electrocautery (16, 17), and CO2 laser laparoscopy (18, 19). Few studies address the analysis from the standpoint of fimbrial pathology alone. Our rates were comparable to the rates reported for interfimbrial adhesions alone (20), phimosis (8, 9, 12), and severe phimosis with partial obstruction (12). Cumulative PR obtained from the life-table analysis, when available for fimbrioplasty, were comparable to ours (9, 10, 12-14, 17). To our knowledge, cure rates and monthly probability of pregnancy are not available for comparison. A stepwise Cox regression analysis was used to evaluate the interactions and impact of the various study factors. This is a statistically powerful way to evaluate whether any of the factors either individually or in combination with each other were a statistically significant predictor of pregnancy. The P values were not computed from any individual parameter estimate, but rather from the overall structure of the cumulative pregnancy curve of each group being compared. Based on the results of this analysis, none of the factors studied (interfimbrial adhesions, phimosis, fimbrioplasty, salpingostomy, tuboovarian adhesions, endometriosis, or other infertility factors) represented a significant independent prognostic parameter when the analysis was conducted on the overall study population. In patients requiring a salpingovariolysis, preg- 478 Saleh and Dlugi Pregnancy outcome following fimbrioplasty Fertility and Sterility '~
nancy outcome did not correlate with laparoscopic assessment of quantity and quality of adnexal adhesive disease when the AFS scoring system was used. There was no statistical difference between patients who had a completely unaffected side and those with bilateral disease of any type. A multicenter controlled study already demonstrated the value of salpingovariolysis in the treatment of nonocclusive tubal infertility (21). Furthermore, in our study, exclusion of patients with completely inoperable bilateral adnexa ("frozen pelvis") may have created a selection bias. However, several authors have stressed the lack of correlation between adhesion scoring and pregnancy outcome (13, 19, 22). Several scoring systems previously have been reported, but none have gained universal acceptance. None have taken into account tubal occlusion or actual tubal damage as criteria. A valid scoring system might not be available until one can incorporate the internal damage of the tube, the degree of constriction, and the loss of normal internal tubal architecture. Recent studies have revealed an advantage of salpingoscopic assessment of tubal mucosal damage over a conventional adhesion scoring system when predicting pregnancy outcome (9, 23). A recent prospective clinical trial suggests that tubes with minimal pathology at laparoscopy have more significant intraluminal disease at salpingoscopy whereas laparoscopic findings correlate well in the presence of more severe disease (23). When patients with tubo-ovarian adhesions are excluded, those who underwent a salpingostomy as their sole procedure had a better outcome compared with those who only underwent a bilateral fimbrioplasty. This suggests that, in the presence of pelvic adhesions or extrinsic distal tubal disease, concomitant tubal endothelium damage is severe enough to mask the individual impact of surgical procedures aimed at correcting serosal pathology. The fact that, despite the lack of homogeneity between study populations, most authors have published similar results for endoscopic treatment of distal tubal disease also supports this fact. In patients without tubo-ovarian adhesions, the cumulative pregnancy curve (Fig. 2) suggests that a subgroup of patients with partial obstruction (salpingostomy group) may achieve maximal therapeutic effect at the 6 month postoperative period, whereas patients with only mild to moderate fimbrial pathology (fimbrioplasty group) may achieve pregnancy independently. Even though the numbers are limited, this may suggest that this subgroup benefits from surgical correction of the partial obstruction, whereas patients with less obstruction (fimbrioplasty group) may benefit from treatmentindependent pregnancies. Whether this conclusion is valid or not remains to be determined. In previous reports, cumulative PRs after tubal reconstructive surgery continued to increase beyond the 1-year point (1, 24). On the other hand, a multicenter, controlled study by Collins et al. (25) reported that 68% of pregnancies among women with "incomplete tubal occlusion" occurred independently of treatment. In their study, pregnancies occurring > 12 months after surgery were considered as treatment independent. In any case, based on our findings, we must question the recommendation of the International Federation of Fertility and Sterility to pool patients with partially occluded tubes in the same category as patients having a fimbrioplasty (3), especially when no tubo-ovarian adhesions can be identified. This study illustrates the complexity of interpreting statistical analysis of tubal infertility data. Comparison between rates obtained after stratification must be viewed with caution, as the numbers become smaller and the statistical power weaker. The methodology used to obtain cure rate estimates also can result in misleading P values. Because of the referral nature of our practice, our patients tend to have extensive pathology and multiple infertility factors. An analysis of larger cohorts of"pure" categories are needed to clarify the individual impact of and interactions among these various factors. To conclude, we stress the importance of multivariate and life-table analysis in reaching a realistic understanding of prognosis in tubal surgery. REFERENCES 1. Dlugi AM, Reddy S, Saleh WA, Mersol-Barg MS, Jacobsen G. Pregnancy rates after operative endoscopic treatment of total (neosalpingostomy) or near total {salpingostomy) distal tubal occlusion. Fertil Steril I994;62:913-20. 2. The American Fertility Society. Revised American Fertility Society classification of endometriosis: 1985. Fertil Steril 1985;43:351-2. 3. Murphy AA. Reconstructive surgery of the oviduct. In: Rock JA, Murphy AA, Jones HW, editors. Female reconstructive surgery. Baltimore: Williams & Wilkins, 1992:146-69. 4. The American Fertility Society. The American Fertility Society classifications of adnexal adhesions, distal tubal occlusion, tubal occlusion secondary to tubal ligation, tubal pregnancies, Mullerian anomalies and intrauterine adhesions. Fertil Steril 1988; 49.'-944-55. 5. Cutlet- S J, Ederer F. Maximum utilization of the life table method in analyzing survival. J Chronic Dis 1958;8:699-702. 6. Guzick DS, Rock JA. Estimation of a model of cumulative pregnancy following infertility therapy. Am J Obstet Gynecol 1981; 140:573-8. 7. Cramer DW, Walker AM, SchiffI. Statistical methods in evaluating the outcome of infertility therapy. Fertil Steril 1979;32:80-6. 8. Monrozies X, Zervoudis S, Elefterion A, Berberi A, Parinaud J, Reme JM. Etude multiparametrique des sterilites tubaires mecaniques. A propos de soixante-dix cas de plastie distales soit la microchirurgie et ses limites. Ann Chir 1987;41:203-8. 9. Henry-Suchet J, Veluyre M, Pia P. Etude statistique des fac- Vol. 67, No. 3, March 1997 Saleh and Dlugi Pregnancy outcome following fimbrioplasty 479
teurs influencant le pronostic des plasties tubaires. Importance de l'etat de la muqueuse ampullaire et de l'infection chlamydienne. J Gynecol Obstet Biol Reprod (Paris) 1989; 18: 571-80. 10. Eyraud B, Erny R, Vergnet F. Chirurgie tubaire distale par clioscopie. J Gynecol Obstet Bio] Reprod (Paris) 1993;22:9-14. 11. Wallach EE, Manara LR, Eisenberg E. Experience with 143 cases of tubal surgery. Fertil Steril 1983;39:609-17. 12. Donnez J, Casanas-Roux F. Prognostic factors offimbrial microsurgery. Fertil Steril 1986;46:200-4. 13. Jacobs LA, Thie J, Patton PE, Williams TJ. Primary microsurgery for postinflammatory tubal infertility. Fer~il Steril 1988;50:855-9. 14. Kelly RW, Diamond MP. Intra-abdominal use of the carbon dioxide laser for microsurgery. Obstet Gynecol Clin North Am 1991; 18:537-44. 15. Mettler L, Irani S, Kapamadzija A, Serum K. Pelviscopic tubal surgery: the acceptable vogue. Hum Reprod 1990;5: 971-4. 16. Fayez JA. An assessment of the role of operative laparoscopy in tuboplasty. Fertil Steril 1983;39:476-9. 17. Audibert F, Hedon B, Axnal F, Humeau C, Boulot P, Bachelard B, et al. Therapeutic strategies in tubal infertility with distal pathology. Hum Reprod 1991;6:1439-42. 18. Aub~iot FX, Dubuisson JB, Henrion R, Bouquet de Joliniere J. Infertilite tubo-peritoneale et clioscopie operatoire (endometriose exclue). Chirurgie 1990; 116:481-4. 19. Donnez J, Nisolle M. CO~ laser laparoscopic surgery: adhesiolysis, salpingostomy, laser uterine nerve ablation and tubal pregl~ancy. Baillieres Clin Obstet Gynaecol 1989;3:525-43. 20. Nordenskjold F, Ahlgl'en M. Intevfimbrial adhesions: detection and treatment of an easily overlooked cause ofinfertility. J Reprod Med 1984;29:595-6. 21. Tulandi T, Collins JA, Burrows E, Jarrell JF, McInnes RA, Wrixon W, et al. Treatment-dependent and treatment-independent pregnancy among women with periadnexal adhesions. Am J Obstet Gynecol 1990; 162:354-7. 22. Bruhat MA, Nage G, Manhes H, Soualhat C, Ropert JF, Pouly JL. Laparoscopy procedures to promote fertility. Ovariolysis and salpingolysis. Result of 93 selected cases. Acta Era" Fertil 1983; 14:113-5. 23. Surrey ES, Surrey MW. Correlation between salpingoscopic and laparoscopic staging in the assessment of the distal fallopian tube. Fertil Steril 1996;65:267-71. 24. Dlugi AM, Saleh WA, Jacobsen G. KTP/532 laser laparoscopy in the treatment ofendometriosis-associated infertility. Fertil Steril 1992; 57:1186-93. 25. Collins JA, Wrixon W, Janes LB, Wilson EH. Treatmentindependent pregnancy among infertile couples. N Engl J Med 1983;309:1201-6. 480 Saleh and Dlugi Pregnancy outcome following fimbrioplasty Fertility and Sterility ~