Alimited degree of vaginal discharge is expected

Size: px

Start display at page:

Download "Alimited degree of vaginal discharge is expected"

Griffin Campbell
5 years ago
Views:

1 ADULT UROLOGY POSTHYSTERECTOMY VAGINAL CUFF FISTULA: DIAGNOSIS AND MANAGEMENT OF AN UNUSUAL CAUSE OF INCONTINENCE DAVID A. GINSBERG, ERIC S. ROVNER, AND SHLOMO RAZ ABSTRACT Objectives. Connection between the vaginal cuff and the peritoneal cavity after hysterectomy is a rare event that can mimic urinary incontinence. The appropriate evaluation and treatment of these patients is discussed. Methods. Five patients underwent excision of the vaginal cuff during a 12-month period. All of these patients had a negative workup for urinary incontinence, except for 1 patient who also had stress incontinence and required a vaginal wall sling at the time of cuff excision. Results. All 5 patients are presently free of excess vaginal drainage or significantly improved, with a mean follow-up of 6 months. One patient developed stress incontinence after cuff excision and later required a vaginal wall sling. There have been no perioperative complications and no evidence of recurrent fistula. Conclusions. Fistula of the vaginal cuff is a diagnosis of exclusion after urinary incontinence has been ruled out. A high index of suspicion is often required to make the diagnosis because these patients often present with symptoms highly suggestive of urinary leakage. Fistula of the vaginal cuff is successfully treated with excision of the vaginal cuff and the fistulous tract (if identified), with minimal morbidity. UROLOGY 52: 61 65, , Elsevier Science Inc. All rights reserved. Alimited degree of vaginal discharge is expected for up to several months after hysterectomy. Leakage that continues after this period, or begins at a later date, is abnormal and deserves a thorough evaluation. Entities that can present as vaginal discharge after hysterectomy include loss of urine secondary to vesicovaginal fistula, ureterovaginal fistula and stress or urge incontinence, lymphatic leak secondary to a lymphatic fistula, 1 and fallopian tube drainage (hydrops tubae profluens). 2 Infectious vaginitis (bacterial, fungal, trichomoniasis) is another common cause of vaginal discharge; however, this is often associated with other symptoms not related to previous hysterectomy and will respond to appropriate therapy. 3 Fistula tract formation between the peritoneal cavity and the vaginal cuff after hysterectomy has been documented 4 ; however, excessive vaginal discharge has not been seen with this process. Most From the Department of Urology, University of California, Los Angeles, School of Medicine, Los Angeles, California Reprint requests: Shlomo Raz, M.D., UCLA Medical Center, 924 Westwood Boulevard, Suite 520, Los Angeles, CA Submitted: November 3, 1997, accepted (with revisions): January 9, 1998 patients with a peritoneovaginal fistula present with complaints of abdominal pain after sexual intercourse; radiographic evidence of pneumoperitoneum has consistently been seen in these patients as well. We describe 5 patients with fistula of the vaginal cuff who presented with excessive vaginal discharge. All initially complained of urinary incontinence and several patients had undergone previous anti-incontinence surgery without resolution of their symptoms. Excision of the vaginal cuff resulted in resolution of their leakage, and we theorize that these patients had an abnormal connection between the vaginal cuff and the peritoneal cavity. MATERIAL AND METHODS Five patients underwent excision of the vaginal cuff during the past 12 months (Table I). Their ages ranged from 43 to 59 years, with an average age of 51 years. The patients were all postmenopausal and were receiving hormone replacement therapy. All the patients complained of urinary incontinence at initial presentation. Continuous daytime urinary leakage that began after hysterectomy was the primary complaint in 2 patients; continuous urinary leakage that began 2.5 and 15 years 1998, ELSEVIER SCIENCE INC /98/$19.00 ALL RIGHTS RESERVED PII S (98)

2 Pt. No. Age (yr) TABLE I. Patient characteristics Hysterectomy (Date) SI Surgery (Date) Symptoms Procedure (Date) F/U (mo) Outcome 1 48 TAH/BSO (1/94) Burch (1/94) VWS (3/95) CI since TAH VCE/E (10/95) 6 Dry Occasional UI 2 59 TAH (1981) None CI 1 mo VCE/E (4/96) 12 Dry 3 52 VH (4/93) BNS (1972, 1979, SI/UI 20 yr VCE/E (5/96) 5 VWS 9/963dry 1992) 4 53 TAH/BSO (4/93) MMK (4/93) Collagen (1/96, 3/96) CI 9 mo VCE (9/96) 3 Dry 5 43 TAH/BSO (1/96) MMK (1/96) SI 2yr CI since TAH VCE/VWS (10/96) 5 Positional incontinence KEY: Pt. No. patient number; SI stress incontinence; F/U follow-up duration; TAH total abdominal hysterectomy; BSO bilateral salpingo-oophorectomy; CI continuous incontinence (unrelated to stress or urge); VCE vaginal cuff excision; E enterocele repair; VWS vaginal wall sling; UI urge incontinence; VH vaginal hysterectomy; BNS bladder neck suspension; MMK Marshall-Marchetti-Krantz operation. after hysterectomy was the primary complaint in 2 other patients. One of these patients also presented with an increase in her urinary incontinence with positional change. A worsening of symptoms of urge and stress incontinence, which were intermittently present for more than 20 years, was noted in 1 patient who had undergone three previous bladder neck suspensions. All 5 patients had undergone previous hysterectomy (one vaginal and four abdominal with three bilateral salpingo-oophorectomies), with 4 patients undergoing concomitant antiincontinence surgery at the time of the hysterectomy. Three of these patients subsequently underwent five additional antiincontinence procedures (two collagen injections, two bladder neck suspensions, and one vaginal wall sling) before vaginal cuff excision. The interval from hysterectomy to excision of the vaginal cuff ranged from 9 months to 15 years. Physical examination revealed induration of the vaginal cuff in 2 of the patients; the actual fistula site at the vaginal cuff was identified in 3 patients when a more aggressive examination could be accomplished at the time of cuff excision. Three patients were noted to have a concomitant enterocele and subsequently underwent repair at the time of vaginal cuff excision. There was no evidence of vaginitis or infection of the vaginal cuff in any of the patients. The Pyridium pad test was performed in all 5 patients. Patients were given Pyridium 100 mg three times a day for 3 days and were instructed to continuously use a pad and a tampon. Four of the patients had a wet tampon that was not stained orange with Pyridium-dyed urine and a dry pad. One patient had a positive pad test (and a diagnosis of stress urinary incontinence) in addition to a large amount of moisture on a tampon that was not stained orange. This patient also received a vaginal wall sling at the time of vaginal cuff excision. Urinary incontinence, as well as a vesicovaginal fistula, was ruled out with formal urodynamics testing, voiding cystourethrography (VCUG), cystoscopy, and a Marshall test in each patient. Two patients also underwent magnetic resonance imaging (MRI) to rule out hydrops tubae profluens. An attempt to radiographically identify a peritoneovaginal fistula with contrast placed in the vagina was unsuccessful in 1 patient. Pneumoperitoneum was not identified in any of the 5 patients. Exposure for vaginal cuff excision is obtained with the use of silk labial retraction sutures and a Scott retractor. The vaginal cuff, with any identifiable fistulous tract, is then grasped with several long Allis clamps and the cuff is sharply excised with the use of Metzenbaum scissors. If needed, standard enterocele repair is performed as previously described with excision of the peritoneal sac and repair with placement of modified McCall culdoplasty sutures. 5 The vaginal cuff is then closed with a running, locking 2-0 polyglycolic acid suture and antibiotic-impregnated vaginal packing is placed. The vaginal packing is removed several hours postoperatively, and patients are discharged home approximately 6 hours after surgery. RESULTS With a mean follow-up of 6.2 months (range 3 to 12), all 5 patients are presently dry or substantially improved without evidence of recurrent fistula of the vaginal cuff or vaginal discharge (Table I). One patient complains of occasional urinary urgency and urge incontinence in the morning; her preoperative complaints of continuous incontinence remain completely resolved after cuff excision. Another patient was found to have stress urinary incontinence 4 months after cuff excision and she subsequently underwent a vaginal wall sling with resolution of her complaints. The last patient, with positional incontinence, is scheduled for collagen injection in the future; her complaint of continuous incontinence has resolved since cuff excision. A fistulous tract was identified in 3 of the patients at the time of cuff excision. The fistulous tract appeared to extend from the vaginal cuff toward the peritoneal cavity. Pathologic results were available from one procedure, with the specimen revealing a dilated space lined with squamous metaplasia. Routine specimens were not sent on the other 4 patients and were, therefore, not available for review. COMMENT A small to moderate amount of vaginal discharge is commonly seen for up to several months after vaginal surgery. However, prolonged drainage is clearly abnormal and deserves appropriate evaluation and treatment. Fistula of the vaginal cuff after hysterectomy is a rare cause of prolonged vaginal 62 UROLOGY 52 (1), 1998

3 discharge that results in constant, gravitational, vaginal wetness in the absence of other pelvic abnormality. This entity can be confused with urine, lymph, or fallopian tube drainage. The most common diagnosis associated with posthysterectomy urinary leakage, vesicovaginal fistula, is still a relatively uncommon event, with an incidence of 0.1% to 0.2% after hysterectomy. 6 Ureterovaginal fistulas are a less commonly seen cause of posthysterectomy urinary leakage. 7 Stress urinary incontinence and urge incontinence (especially in patients who undergo concomitant bladder neck suspension at the time of hysterectomy) need to be considered. Vaginal fistulas involving lymphatic 1 or fallopian tube 2 fluid have been reported in published studies and are extremely rare causes of vaginal discharge after hysterectomy. Infectious vaginitis may present with vaginal discharge and is often accompanied by additional signs and symptoms, such as vaginal malodor, vulvar and perivaginal pruritis, and physical findings (eg, vulvar and perivaginal erythema and colpitis macularis). 3 There is no association between previous hysterectomy and infectious vaginitis. The diagnosis of vaginal cuff fistula is challenging. If a fistula of the vaginal cuff is to be diagnosed, a high index of suspicion needs to be maintained in patients who present with complaints suggestive of urinary incontinence. The difficulties associated with the diagnosis of excessive vaginal discharge as a vaginal cuff fistula are certainly exemplified by the patients in this series, 3 of whom saw us for persistent incontinence after a previous failed procedure. The diagnosis of vaginal cuff fistula is one of exclusion. Physical examination may reveal an inflammed cuff; however, there is nothing pathognomonic on pelvic examination in these patients that indicated a fistula. A Pyridium pad test should initially be performed to rule out urinary incontinence. If positive, urodynamics testing, VCUG, and cystoscopy may be used to identify the etiology and severity of the problem. If these tests are negative in the presence of a positive Pyridium pad test, a ureterovaginal fistula should be suspected and an upper tract study should be performed. Other rare causes of vaginal discharge after hysterectomy that should be considered include colovaginal fistula secondary to diverticular disease, 8 hydrops tubae profluens, 2 and lymphatic vaginal fistula. 1 Colovaginal fistula secondary to diverticular disease can be diagnosed with computed tomography or with a fistulogram/vaginogram and is often treated with a period of antibiotic therapy followed by surgical resection of the involved colon. Fallopian tube drainage into the vaginal cuff (hydrops tubae profluens) may be identified by physical examination, vaginogram, or MRI and is managed with excision of the involved fallopian tube and vaginal cuff closure. A report of lymphatic vaginal fistula after radical hysterectomy was documented by the presence of chyle in the vaginal fluid and successfully treated with bed rest and a high protein diet. Peritoneovaginal fistula after hysterectomy has been described in published reports; however, in a review of published reports by Massad et al., 4 the presence or absence of vaginal discharge was not mentioned. Patients have typically presented with abdominal pain after sexual intercourse and were radiographically found to have pneumoperitoneum. Radiographic identification of a fistula tract between the vagina and the peritoneum with the use of contrast placed in the vagina has been described in several of these patients. Patients have been successfully treated surgically, with fistula closure, and conservatively, with avoidance of sexual intercourse and antibiotics. The question remains as to the origin of the leakage from the vaginal cuff in these 5 patients. Urinary incontinence was identified in 1 patient and she underwent a vaginal wall sling procedure at the time of vaginal cuff excision. The other 4 patients had no evidence of urinary incontinence after extensive preoperative evaluation, including videourodynamics, cystoscopy, Marshall test, and Pyridium pad test. Yet, despite this evaluation, 2 patients had complaints of urinary incontinence after vaginal cuff excision. It is possible that this incontinence was present before vaginal cuff excision and not identified; however, the patient with complaints of positional incontinence clearly has had resolution of her complaints of continuous incontinence since cuff excision. Another potential site of leakage from the vaginal cuff would be a fistulous tract from a fallopian tube remnant. However, 3 of the patients in this study had bilateral salpingo-oophorectomies at the time of hysterectomy and fallopian tube abnormality was not identified with pelvic MRI in the remaining 2 patients. In addition, there was no evidence of lymphatic vaginal or colovaginal fistula in any of these patients. A possible source of leakage is peritoneal fluid. The peritoneal reflection is adjacent to the vaginal cuff, and the peritoneal cavity produces a sufficient amount of fluid to account for the patients complaints and is one site in the pelvis that had not been ruled out as a potential fistula source. The presence of a concomitant enterocele could certainly be a contributing factor to formation of a peritoneovaginal fistula. The peritoneal reflection is firmly adhered to the vaginal cuff after hysterectomy; with enterocele formation, this could conceivably lead to increased pressure at the vaginal cuff and fistula formation. The peritoneal reflec- UROLOGY 52 (1),

tion is commonly juxtaposed to the vaginal cuff after hysterectomy and enterocele formation is not a rare occurrence; what predisposed these particular patients to fistula formation is not clear.

4 tion is commonly juxtaposed to the vaginal cuff after hysterectomy and enterocele formation is not a rare occurrence; what predisposed these particular patients to fistula formation is not clear. If peritoneovaginal fistula is indeed the source of vaginal discharge, these patients, presenting with complaints suggestive of urinary incontinence, behaved differently than what has previously been documented. There was no relationship to the onset of symptoms and sexual relations and pneumoperitoneum was not radiographically identified in any of the patients in this series. The pathophysiology of peritoneovaginal fistula in previous reports is thought to be related to acute disruption of the vaginal cuff during intercourse. The pathogenesis resulting in vaginal discharge secondary to a fistula of the vaginal cuff is not as well defined in all of the patients in this series. Two of the patients complaints clearly began after hysterectomy. However, 3 of the patients difficulties with vaginal discharge began a variable length of time after hysterectomy, with one vaginal cuff fistula diagnosed 15 years after previous total abdominal hysterectomy without any additional pelvic surgery in the intervening years. It is hard to imagine that a fistula of the vaginal cuff would present this many years after hysterectomy; yet, after simple excision of the vaginal cuff and enterocele repair, this patient has clearly had a resolution of her symptoms and no longer requires pads for protection. The pathogenesis of this fistula is probably a chronic process and could be related to factors such as poor cuff closure, a persistent tract after hysterectomy, concomitant enterocele putting excessive pressure on the previous suture line at the level of the vaginal cuff, chronic infection, and postmenopausal changes to the tissue that may act independently, or together, resulting in a fistulous tract and excessive vaginal discharge. diagnosis and treatment. AUA Update Series 9: , Massad LS, Santoriello K, and Berchuck A: Pneumoperitoneum due to peritoneovaginal fistula following hysterectomy. Int J Gynecol Obstet 29: , Raz S, Nitti VW, and Bregg KJ: Transvaginal repair of enterocele. J Urol 149: , Harris WJ: Early complications of abdominal and vaginal hysterectomy. Obstet Gynecol Survey 50: , Gerber GS, and Schoenberg GS: Female urinary tract fistulas. J Urol 149: , Cross SB, and Copas PR: Colovaginal fistula secondary to diverticular disease. J Reprod Med 38: , EDITORIAL COMMENT The authors are to be congratulated for an excellent review of a rather rare and always perplexing condition. I have referred to this in past presentations as posthysterectomy pseudoincontinence. However, although having encountered several patients with vaginal cuff fistulas that drain serous fluid and mimic urinary incontinence, I have not yet been able to identify a true connection between the vaginal cuff and the peritoneal cavity, nor have I been able to prove that the serous fluid that leaked out of these patients was peritoneal fluid. What we have found, using exactly the techniques described by the authors, along with retrograde injection of the vaginal fistula, is that a distensible cavity is present that is lined with heaped up, mixed secretory and squamous vaginal epithelium that produces fluid and leaks variably on a daily basis. This leakage is often accentuated by increased intra-abdominal pressure, thus mimicking stress incontinence. Some patients have reported an increase in leakage after a long tub bath, suggesting the possibility that the cavity might have filled during the bath and leaked out slowly once they were up and walking. Figure 1 shows the findings in just such a patient. A small fistula was found at the most superior aspect of the right side CONCLUSIONS Vaginal discharge secondary to vaginal cuff fistula is an uncommon and difficult diagnosis to make that can masquerade as urinary incontinence. It is an entity that must be considered during the evaluation of posthysterectomy incontinence and further illustrates the importance of appropriate documentation of stress incontinence before definitive surgical repair. REFERENCES 1. Lau TK, and Wong WSF: Lymphatic vaginal fistula after Wertheim-Taussig hysterectomy: a case report. Gynecol Oncol 52: , Leach GE, Yip C-M, Donovan BJ, et al: Tubovaginal leakage: an unusual cause of incontinence. J Urol 137: , Krieger JN: Vaginitis syndromes: a practical approach to FIGURE 1. Cone-tipped catheter filling an epithelialized cavity at the apex of the vagina. Patient was posthysterectomy. Presenting complaint was incontinence. Fluid was not urine and there was no connection to the peritoneum. 64 UROLOGY 52 (1), 1998

5 of the patient s vaginal cuff. This was injected with contrast material using a standard large ureteral bulb catheter. The large cavity shown was identified. Fluid in the cavity was neither urine nor lymph. The ostium was resected sufficiently to allow inspection, biopsy, and fulguration of the rather exuberant epithelial lining. Postoperatively the patient had no further leakage. The pathologic findings as described for the above patient would be compatible with both the symptoms and abnormalities of the patients in this report. Also, the authors too were unable to demonstrate an actual fistula into the peritoneal cavity. It might explain the patient whose symptoms did not appear for 15 years and the absence of pneumoperitoneum in this series. The treatment described by the authors is curative and perhaps more definitive than our method of simple exteriorization and fulguration. Until a definite peritoneal connection can be demonstrated, however, I prefer the term posthysterectomy pseudoincontinence to describe the leakage and vaginal cuff fistula to describe the pathologic finding. REPLY BY THE AUTHORS The experience described in the Editorial Comment appears to be similar to what we have found. We have not been able to identify the fistula cavity radiographically as shown by Dr. Ball, which may reflect the lack of an obvious fistulous tract at the vaginal cuff when examined in the office setting. Subsequent examination at the time of surgery did reveal the fistula site in 3 of the 5 patients. We prefer resection of the vaginal cuff for these patients, to ensure adequate removal of the entire fistula tract and any epithelialized cavities with which the fistula communicates. However, simple exteriorization and fulguration of the tract, as described by Dr. Ball, may be sufficient, especially if preoperative studies are able to adequately delineate the full extent of the fistula tract and cavity. Thomas P. Ball, Jr., M.D. Department of Urology University of Texas Health Science Center San Antonio, Texas UROLOGY 52 (1),

Case Based Urology Learning Program

Case Based Urology Learning Program Resident s Corner: UROLOGY Case Number 19 CBULP 2011 044 Case Based Urology Learning Program Editor: Associate Editors: Manager: Case Contributors: Steven C. Campbell,