Recanalization of obstructed fallopian tube by selective salpingography and transvaginal bougie dilatation: outcome and cost analysis*

FERTILITY AND STERILITY Vol. 66, No.2, August 1996 Copyright 1996 American Society for Reproductive Medicine Printed on acid-free paper in U. S. A. Recanalization of obstructed fallopian tube by selective salpingography and transvaginal bougie dilatation: outcome and cost analysis* Erich K. Lang, M.D.t Heber H. Dunaway, Jr., M.D. Louisiana State University Medical Center and Center for Reproductive Gynecology, Lakeside Hospital, New Orleans, Louisiana Objective: To investigate effectiveness and cost of trans cervical salpingography and recanalization in the management of infertility caused by tubal occlusion. Design: Retrospective analysis of patients investigated with trans cervical selective salpingography and, in some instances, treated by transcervical recanalization. Setting: Four hundred patients with hysterosalpingography diagnosis of obstructed fallopian tubes (clinical environment) are investigated. A repeat hysterosalpingogram after administration of a prostaglandin antagonist demonstrated patency ofthe tubes in 82 patients and selective trans cervical salpingography demonstrated patency in an additional 131 patients. Intervention: Transcervical tubal recanalization. Of the remaining 187 patients, recanalization by transcervical technique was successful in 145 patients. The underlying etiology for tubal obstruction was salpingitis isthmica nodosa in 62, salpingitis and perisalpingitis in 71, endometriosis in 8, failed surgical anastomosis in 43, and undeterminate cause in 3 patients. Pregnancy was attained in 24 patients, there were 10 minor and 1 major complication. Outcome Measure: Attained and maintained patency oftubes, pregnancy, attendant complications. Conclusion: A pregnancy rate of 12.8% was attained after transcervical recanalization of obstructed tubes. An attendant increased rate of pregnancy in patients proven patent after selective salpingography, valuable detailed information about proximal and distal tubes after recanalization of the obstructed proximal tube segment, the low rate of complications, and low cost are factors recommending the use of this technique. Moreover, tubal surgery or IVF treatment are not influenced adversely by prior transcervical tubal recanalization and remain an option for patients who failed to attain pregnancy. Fertil Steril 1996;66:210-5 Key Words: Infertility, fallopian tubes-recanalization, selective salpingography Increasing emphasis on cost containment calls for a reappraisal of our concepts guiding the selection of assisted reproductive technology (ART). DeCherney (1) has raised the question of a capitated system, basic workup, and preliminary therapy at a fixed fee to get the patient pregnant. Patient pressure for Received January 8, 1996; revised and accepted March 21, 1996. * Presented at the 51st Annual Meeting ofthe Anlerican Society for Reproductive Medicine, Seattle, Washington, October 7 to 12, 1995. t Reprint requests and present address: Erich K. Lang, M.D., Department of Radiology, University of Medicine and Dentistry of New Jersey, University Hospital, 150 Bergen Street, Newark, New Jersey 07103-2405 (FAX: 201-982-7429). maximal success, however, introduces a dilemma with demands for high-cost, high-yield procedures deployed early (2). Moreover, there is continued disagreement as to the diagnostic sensitivity, specificity, and accuracy inherent to our tests and required for decisions choosing between different therapeutic pathways. A recent statistical analysis proclaims the hysterosalpingogram (HSG) a reliable test for tubal obstruction, though not for assessment of peritubal disease (3). However, this paper does not take into consideration the known large number offalse-positive obstructions of the proximal tube attendant to spasm or debris, which if properly identified eliminate need for further workup or intervention. Moreover, assessment of the distal tubes, unless proven normal on HSG, largely is relegated to laparoscopy. 210 Lang and Dunaway Recanalization of fallopian tubes Fertility and Sterility

Laparoscopy and chromopertubation are not only the gold standard for diagnostic evaluation, but often are used as a primary diagnostic modality. With the advent of transcervical selective tubal catheterization, a more detailed assessment of the fallopian tubes as well as recanalization of obstructed segments of the proximal fallopian tubes, short of surgical intervention, became feasible. Categorization of disease of the proximal tubes and, after overcoming obstruction in the proximal tubes, assessment of the distal tubes, is possible (4-6). Moreover, in many instances, the lumen of an obstructed proximal tube can be recanalized by this technique (6,7). The aim of this study is to determine the value of transcervical selective salpingogram for evaluation of the proximal and distal tubes and transcervical recanalization for re-establishing patency of obstructed proximal tubes in patients with underlying inflammatory disease or with strictures occurring after prior surgical intervention or attendant upon reversal surgery (7). MATERIALS AND METHODS Four hundred patients with the diagnosis of obstruction of one or both tubes on HSG were entered into this study. The protocol and the informed consent developed specifically for this study were approved by the institutional review board. These patients were either referred by outside physicians with a diagnosis of nonfilling of fallopian tubes on HSG or this diagnosis was established at our clinic. All patients had a complete endocrine workup, excluding endocrine causes for their infertility. Likewise, the male partners had been examined and eliminated as probable cause for the infertility problem. The patients ranged in age from 21 to 46 years, with a median age of 33.2 years. A problem of infertility had been established in all patients over a period of :2: 18 months. Two hundred thirty-four of the patients had experienced prior pregnancies and delivered a live baby. Another 22 patients give a history of a prior blighted ovum or tubal pregnancy. Superovulation therapy had been carried out in 92 patients. Hysterosalpingogram was performed under sedation on patients seen at our clinic. Fentanyl (25 mg) or 2 mg Versed (Elkins Sinn, Cherry Hill, NJ) were given IV. After proper prepping and draping, a balloon catheter (Bard, Billerica, MA) was introduced through the cervical canal into the uterine cavity. The balloon was inflated and seated at the rostral end of the cervical canal. Subsequent to that, aqueous contrast medium, nonionic or ionic (Conray 60, Omnipaque 350; Mallinckrodt, St. Louis, MO) was infused until the endometrial cavity was delineated Vol. 66, No.2, August 1996 adequately. With initial filling of the tubes, injection into the-uterine cavity was continued until spillage into the free peritoneal cavity was documented. The procedure was carried out under fluoroscopic control. Films were obtained of the filled endometrial cavity in anteroposterior, oblique projection as well as after deflation of the balloon to study the endocervical canal. Oblique projections were obtained to demonstrate the distal end of the tubes and the pattern of spillage in the periovarian space. Patients who were referred with the diagnosis of nonfilling of one or both tubes or who exhibited such findings on our preliminary studies were rescheduled for a second HSG to confirm the findings. These patients were premedicated with 325 mg aspirin for 2 days (a prostaglandin antagonist) and with 100 mg doxycycline (Warner Chilcott, Morris Plains, NJ) twice daily for 2 days. The latter was instituted to permit us to follow the HSG if nonfilling of the tubes persevered with selective salpingography and transcervical recanalization of the tubes at the same sitting. The patients also were premedicated with 25 mg Fentanyl and 2 mg Versed IV. After proper prepping and draping, a tenaculum was attached to the cervix at 12 o'clock and the HSG was performed in the customary fashion. If, once again, there was failure of filling of either or both tubes, we proceeded to perform a selective salpingogram. For this purpose, a 9 French straight Teflon tube was introduced through the endocervical canal. A 3.5 to 5.5 French curved catheter (varied manufacturers) (with angulation of its tip variable from 90 to 120 ) then was advanced through the endocervical Teflon tube and the right and left cornu, respectively, were engaged. Under fluoroscopic control, the catheter tip usually marked with an opaque ring marker was advanced into the cornu and engaged into the ostium of the fallopian tube. Injections of 1 ml of aqueous contrast medium were carried out with the catheter tip engaged. These were recorded in the projections suggested above for the HSG. If once again the tube or tubes fail to fill, a 0.014- inch platinum tip guidewire with a highly flexible 7- cm tip (Target Therapeutics, Santa Monica, CA) was introduced through the catheter and cannulation of the tube was attempted under fluoroscopic guidance. If possible, the guidewire was advanced through the interstitial and isthmic segment into the ampullary segment. On rare occasions a 0.018-inch guidewire with central channel was used to monitor the advance of the guidewire while performing multiple injections through the central channel. Once the guidewire had been advanced into the isthmic or ampullary segments, a 1.2 French catheter (Target Therapeutics) was advanced over the guidewire through the perceived stenosis and, ifpos- Lang and Dunaway Recanalization of fallopian tubes 211

Table 1 Fallopian Tube Obstruction No. of patients N onfilling of tubes on initial HSG Filling of tubes after prostaglandin antagonist premedication Tube patent on selective salpingogram Patent after trans cervical recanalization Refractory to trans cervical recanalization One tube Both tubes 400 2 398 82 21 61 131 93 38 145 88 57 42 Figure 1 An injection through 1.2 French catheter (arrowheads) advanced to the ampullary segment of tube shows a normal distal tube and free spillage into the peritoneal cavity. sible, into the ampullary segment of the tubes. At this point, the guidewire was removed and injection of approximately 0.75 to 2 ml of aqueous contrast medium was carried out. This served to document the distal tubes and, in particular, sl:lowed any abnormalities of the distal tubes, such as hydrosalpinx or pyosalpinx, loss of rugal pattern, or abnormal spillage with adhesions in the periovarian space. This injection again was recorded in anteroposterior and oblique projections (Fig. 1). If a stricture or stenosis was demonstrated, the guidewire was reinserted and progressive dilatation with bougie catheters from 1.2 to 2.4 French and, in rare instances, 3.5 French was carried out, (the latter size is reserved for dilatation of strictures occurring at uterotubal anastomosis) (Fig. 2). In all patients in whom catheterization of a tube was carried out or passage of a guidewire attempted, antibi- otic therapy with 100 mg doxycycline twice per day was continued for another 3 days. In addition, if there was evidence of a hydrosalpinx or pyosalpinx, an attempt was made to aspirate a sample through the 1.2 French catheter and to obtain culture and sensitivity studies. In 20 patients, spasm was thought to be responsible for continued nonfilling of the tubes and glucagon or terbutalin were given IV during the HSG. Follow-up of all patients was in our clinics or by physicians who had referred the patients to us. In the latter case, we continued follow-up by phone with either the physician or the patient to ascertain pregnancy and birth of a well baby if this occurred. Patients who did not become pregnant within 6 months were scheduled for re-examination by HSG. This was performed either at our clinics or by the referring physicians, who made their results available to us. Some patients with a documented restenosis were subjected to a repeat transcervical recanalization. RESULTS Figure 2 A 3.5 catheter has been advanced into the isthmic segment across stricture at a utero-tubal reimplantation. Note lack of spillage due to adhesions around the fimbriated end. Obstruction of both tubes was confirmed in 398 patients. Obstruction of only one tube was confirmed in two patients, both of whom had a prior unilateral salpingectomy. However, repeat HSG after premedication with a prostaglandin antagonist found one (n = 21) or both (n = 61) tubes to be pa't ent. Initial nonfilling was attributed to spasm in these 82 patients (Table 1). The remaining 318 patients then were subjected to selective salpingography for further categorization of the obstruction. Once again, one tube (n = 93) or both tubes (n = 38) proved patent on selective salpingograms (Table 1). Residual debris or spasm was incriminated as cause for the obstruction in the majority of these 131 patients. However, the pres- Lang and Dunaway Recanalization of fallopian tubes Fertility and Sterility 212

Table 2 Transcervical Recanalization of Obstructed Fallopian Tubes Failed Failed Salpingitis microsurgery microsurgery isthmica Salpingitis- inflammatory reversal Undetermined nodosa perisalpingitis Endometriosis etiology surgery etiology No. of patients 62 72 Technically successful 51 62 Pregnancies 8 8 Patent on follow-up 21 27 HSG Complications 2 4 Median follow-up 39 (2 to 112) 35 (2 to 109) months* Lost to follow-up 5 4 Disease of distal tubes 18 16 Tubal pregnancy 1 8 25 18 3 6 19 6 1 3 4 1 0 2 14 5 0 0 0 7 0 28 (4 to 76) 31 (1 to 72) 23 (1 to 52) 16 (4 to 28) 1 1 3 2 1 8 4 0 0 *Values in parentheses are ranges. ence of inflammatory manifestations involving the isthmic segment in 16 of these patients suggested a combination of debris and inflammatory disease to be responsible for the initial nonfilling. Moreover, in 28 patients, there was evidence of disease of the distal tubes, loss of rugal pattern, mild hydrosalpinx or pyosalpinx, and periovarian adhesions, which may have contributed to the initial nonfilling of the tubes. In 187 patients, there was an organic occlusion of the tubes; in 9 it afflicted one tube and in 178 it afflicted both (Table 1). The etiologies responsible for organic tube obstruction were failed microsurgical procedures in 43 patients, salpingitis isthmica nodosa in 62 patients, endosalpingitis and perisalpingitis in 71 patients, and endometriosis in 8 patients, whereas in 3 patients a precise etiology cannot be established (Table 2). Transcervical recanalization of the obstructed tubes was attempted in all of these patients. Transcervical recanalization was technically successful in 145 of 187 patients or 202 tubes. Fortytwo patients proved refractory to attempts at transcervical recanalization (Table 1). Intrauterine pregnancies were attained in 24 patients. Complications occurred in 13 patients (perforations without sequelae in 9, gram-negative septicemia in 1, temperature elevation> 38 C in 3). Gram-negative septicemia developed in a patient who, after successful transcervical recanalization, showed evidence of bilateral pyosalpinx. Septic shock and a life-threatening condition developed within 2 hours after the procedure and was treated vigorously with N fluids (1,000 ml Ringer's solution), N corticosteroids, decadron 12 mm N, and the antibiotics clindamycin and gentamicin. The patient recovered rapidly, however, vigorous antibiotic therapy was continued for 10 days. Amoxillin clavulanate were added to the antibiotic management. Vol. 66, No.2, August 1996 Aspirates were obtained from 24 tubes suggesting a hydrosalpinx or pyosalpinx. In 14 of these the culture was sterile. Trachomatis was cultured in 6, enterococci was cultured in 2, enterobacter, klebsiella, proteus, Escherichia coli, and streptococcus viridans were cultured in 1 each; multiple organisms were cultured in 2 of 10 patients. Of 120 patients examined 2:6 months after initial transcervical recanalization who had not attained pregnancy, 69 demonstrated patency of one or both tubes on repeat HSG (48 one tube, 21 both tubes). In 19 patients who showed reocclusion of the tubes, repeat transcervical recanalization was carried out, which was technically successful in 11. None of these patients, however, subsequently became pregnant. Abnormal distal tubes or periovarian adhesions were demonstrated in 44 of 145 patients with successfully recanalized lumen of the proximal tubes. Pregnancy after IVF and/or ET occurred in an additional 17 patients. Sixteen patients were lost to follow-up. When analyzing the results of trans cervical recanalization for various subgroups it is noteworthy that the technical success rate is highest in patients with occlusion attributable to salpingitis isthmica nodosa or salpingitis and perisalpingitis and endometriosis (Table 2, Fig. 3). The procedure was successful in 51 of 62 patients with salpingitis isthmica nodosa and resulted in eight intrauterine pregnancies (Table 2). However, only 21 of 43 patients who did not attain pregnancy and were re-examined by HSG demonstrated continued patency of the tubes 2:6 months. Only one tubal pregnancy occurred in this subgroup. The procedure was technically successful in 62 of 71 patients with salpingitis and perisalpingitis as cause for obstruction. Eight ofthese patients became pregnant. Continued patency was documented in 27 of 54 patients who had not become pregnant on follow-up HSGs 2:6 months. Lang and Dunaway Recanalization of fallopian tubes 213

Effectiveness and cost are the two major factors that influence the choice of treatment modalities ad- vocated for patients with occlusive disease of the fallopian tubes. The success rate for tubal surgery is variously quoted from 16% to 69% (8, 9). A large study in 1993 suggests that single cycle IVF may result in approximately 16% live births (4,835 of 30,132 cycles initiated) (2). However, three cycles IVF may well result in a live birth rate of approximately 40%. In our patients with proven occlusion of the tubes, a pregnancy rate of 12.8% (24/187) resulted after transcervical recanalization (4,5,7, 10). Cost data for tubal surgery varied widely. A N orwegian study calculates the cost per live birth as $17,000 for tubal surgery compared to $12,000 after IVF treatment (11). Cooper (12) calculated the cost of a successful pregnancy after tubal surgery as $31,842 in 1983, which would equate to a cost of $72,763 per pregnancy in 1993 dollars. The cost per live birth attendant upon IVF treatment is heavily dependent upon utilization rate and cost sharing (13, 14). Therefore, a substantial reduction in cost for this technique might be anticipated (12). Thus, the old computed cost for a single birth from IVF in excess of $40,000 may be erroneous (15). Moreover, concurrent tubal plasty and assisted reproduction further may decrease cost and increase the rate of live birth (16). The cost per live birth in our patients treated with trans cervical recanalization of the occluded fallopian tubes is approximately $6,400. Moreover, the cost is relatively independent from the rate of utilization of this interventional procedure. Moreover, the ability of selective salpingography to eliminate approximately 41 % of false-positive diagnoses of tubal occlusion safeguards against erroneously beginning costly IVF treatment cycles on such misdiagnosed patients. Admittedly, the majority of these false positives can be eliminated by other techniques, such as chromopertubation. In addition, selective salpingography identifies at an early time coexistent disease of the distal tubes or parovarian adhesions, which may mandate IVF treatment or perhaps laparoscopic fimbriolysis or lysis of adhesions (17). Transcervical tubal recanalization does not complicate subsequent surgery or IVF treatment nor reduce effectiveness of these techniques. On the contrary, the detailed inform'a tion rendered by selective salpingography may facilitate appropriate choice of IVF treatment, microsurgery, or a combination thereof and facilitate planning of contemplated complex interventions. (18). Our experience indicates a disparity in the to be anticipated success rate attainable for various underlying etiologies. Both technical success rate and rate of attained pregnancies were highest for patients with obstruction secondary to endometriosis and treated by recanalization (75% and 37%, respec- Lang and Dunaway Recanalization of fallopian tubes Fertility and Sterility Figure 3 Transcervical bougie dilatation (2 French) recanalizes an obstructed right tube. Note residual strictures (arrowheads) of the isthmic segment attesting to prior salpingitis isthmic nodosa. Six of eight patients with occlusion caused by endometriosis were recanalized successfully by transcervical recanalization. Three of these patients became pregnant. Follow-up HSGs showed the tubes to remain patent in two of three patients who had not attained pregnancy. Abnormal distal tubes were demonstrated in one of the patients (Table 2). Transcervical recanalization was technically successful in 19 of25 patients in whom a prior microsurgical anastomosis correcting occlusion by underlying inflammatory disease had failed (Fig. 2). Four of these patients became pregnant after our intervention. Fourteen of 15 patients who had not become pregnant, however, demonstrated patency of the recanalized tubes on follow-up HSG. Eight patients demonstrated significant abnormalities of the distal tubes (Table 2). Transcervical recanalization was technically successful in 6 of 18 patients with reocclusion after surgical reversal of a prior tuballigation (Table 2). The procedure was successful only in patients who demonstrated a stenotic occlusion and failed in eight patients who showed a fistula at the microsurgical reanastomotic site or a combination of fistula and stenosis. Significant disease of the distal tubes was present in four of six patients in whom we successfully dilated the stenosis at the anastomotic site. Pregnancy occurred in only one patient in this group. However, follow-up HSG showed one or both tubes to be patent in all five patients in whom we had not attained pregnancy. One of these patients, subsequently, successfully conceived after IVF. Our follow-up ranged from 1 to 112 months, with a mean follow-up of 34 months. DISCUSSION 214

tively). However, in the management of occlusions secondary to inflammatory disease, results were gratifying also (83% and 12%, respectively). The results were particularly salutary in the group of patients with reocclusion of a surgical anastomosis attempting correction of inflammatory occlusion of the fallopian tubes. In this subgroup, 76% were technically successful, attaining a pregnancy rate of 16%. Conversely, this technique failed universally in patients who developed fistulae after reversal surgery attempting correction of prior tubal ligation. Even in the subgroup that developed stenosis, the technique was only marginally effective. However, the diagnostic information and road mapping is invaluable for contemplated microsurgical corrective procedures or to make the choice of IVF and ET. An attendant benefit difficult to gauge is the relatively large number of pregnancies occurring within 1 to 6 months after selective salpingography in patients who presented with tubal occlusion on HSG. In our experience, approximately 30% of these patients became pregnant during the ensuing five cycles. It may be assumed that clearing of the tubes of debris by retrograde flush facilitated subsequent passage of an ovum and, thereby, lead to pregnancy (18). A low incidence of tubal pregnancy, 1 in 145 patients in whom transcervical recanalization succeeded is noted. The mix of pathology in this patient material, massive distal disease making passage of an ovum unlikely, and relatively minimal disease of the proximal tubes may explain at least in part the low incidence of this complication. The valuable diagnostic information derived from selective salpingograms on the status of both proximal and distal tubes, the adjuvant effect toward attaining pregnancy after this procedure, the relatively high pregnancy rate resulting after correction of occlusive disease of the fallopian tubes by transcervical tubal recanalization, the relatively low cost, and the safety and extremely low rate of complications recommend the use of trans cervical selective salpingography and tubal recanalization as the initial measure in the management of infertility attributable to tubal disease.. REFERENCES 1. DeCherney AH. Infertility: we are not taking new patients. Fertil Sterile 1995;64:470-3. 2. Collins JA, Bustillo M, Visscher RD, Lawrence LD. An estimate of the cost of in vitro fertilization services in the United States in 1995. Fertil SteriI1995;64:538-45. 3. Swart P, Mol BWJ, van der Veen F, van Beurden M, Redekop WK, Bossuyt PMM. The accuracy of hysterosalpingography in the diagnosis of tubal pathology: a meta-analysis. Fertil SteriI1995;64:486-91. 4. Lang EK Organic vs functional obstruction of the fallopian tubes: differentiation with prostaglandin antagonist and B2 agonist mediated hysterosalpingography and selective ostial salpingography. Am J RoentgenoI1991;157:77-80. 5. Lang EK, Doody MC, Dunaway HE. Ostial salpingography and transcervical percutaneous recanalization of the proximal tubes, Lippincott's reviews. Radiology 1992; 1:547-61. 6. Confino E, Tur-Kaspa I, DeCherney AH, Corfman R, Coulam C, Robinson E, et al. Transcervical balloon tuboplasty, a multi-centered study. JAMA 1990;64:2079-82. 7. Lang EK, Dunaway HH. Transcervical recanalization of strictures of the post-operative fallopian tube. Radiology 1994; 191:507-12. 8. Fayez YA. Comparison between tubouterine implantation and tubouterine anastomosis for repair of cornual occlusion. Microsurgery 1987;8:78-82. 9. McComb P. Microsurgical tubocornual anastomosis for occlusive cornuial disease: reproducible results without the need for tubouterine implantation Fertil Steril 1986;46:571-7. 10. Risquez F, Confino E. Transcervical tubal recanalization of past, present and future. Fertil Steril 1993;60:211-26. 11. Holst N, Maltau IN, Forsdahl F, Hansen LJ. Handling of tubal infertility after introduction of in vitro fertilization: changes and consequences. Fertil Steril 1991;55:140-3. 12. Cooper GW. An analysis of the cost of infertility treatment. Am J Public Health 1986;76:1018-9. 13. Shushan A, Eisenberg VH, Schenker JG. Subfertility in the era of assisted reproduction: changes and consequences. Fertil Steril 1995;64:459-69. 14. Burner ST, Waldo DR, McKusic DR. National health expenditures projections through 2030. Health Care Financ Rev 1992; 14:1-29. 15. Neumann PJ, Weinstein MC, Gharib SD. The cost of a successful delivery with in-vitro fertilization. N Engl J Med 1994;331:239-43. 16. Novy MJ. Concurrent tuboplasty and assisted reproduction. Fertil Steril 1994;62:242-5. 17. Bateman BG, Nunley WC Jr, Kitchin JD III. Surgical management of distal tubal obstruction-are we making progress. Fertil Steril 1987;48:523-42. 18. Sulak PJ, Letterie GS, Coddington CC, Hayslip CC, Woodward JE, Klein TA. Histology of proximal tubal occlusion. Fertil SteriI1987;48:437-40. Vol. 66, No.2, August 1996 Lang and Dunaway Recanalization of fallopian tubes 215