Scandinavian Journal of Urology and Nephrology, 2012; Early Online, 1–5
ORIGINAL ARTICLE
Urovaginal fistula formation after gynaecological and obstetric surgical procedures: Clinical experiences in a Scandinavian series
Scand J Urol Nephrol Downloaded from informahealthcare.com by Goteborgs University on 12/09/12 For personal use only.
UMIT DEMIRCI, MAGNUS FALL, SOFIA GÖTHE, JOHAN STRANNE & RALPH PEEKER Department of Urology, Sahlgrenska Academy at The University of Gothenburg, Sweden
Abstract Objective. The aim of this retrospective study was to review what kinds of surgical procedures are most frequently complicated by urovaginal fistulae, to find out how they were diagnosed and managed, and to study the outcome after surgical reconstruction. Material and methods. Nineteen women who underwent fistula repair at Sahlgrenska University Hospital between 2003 and 2009 were retrospectively studied by reviewing the medical records. Results. For 17 of the 19 patients hysterectomy was the causative procedure. Fourteen patients developed vesicovaginal and five developed ureterovaginal fistula. Urethrocystoscopy was sufficient for the diagnosis in nearly 50% of the patients and when combined with methylene blue instillation 90% of all fistulae were found. Several patients sought medical advice due to vaginal leakage following gynaecological surgery without the doctor suspecting a fistula, and for these patients the diagnosis was delayed. Eighteen patients were operated on with an abdominal approach and one with a vaginal approach, in all cases a minimum of 3 months after primary surgery. The reconstruction technique included the interposition of vascularized tissue. None of the patients reported leakage or relapse at follow-up after fistula repair. Conclusions. Hysterectomy was the most common cause behind the formation of urovaginal fistulae. Misinterpretation of symptoms after gynaecological surgery was common even in cases where the symptoms were indicative of a urovaginal fistula. Delayed fistula repair after a minimum of 3 months, via the abdominal route and with the interposition of vascularized tissue, yielded an excellent final outcome.
Key Words: Fistula repair, gynaecological surgery, urovaginal fistula, vesicovaginal fistula
Introduction Urovaginal fistula is a devastating condition in which an abnormal channel is created between the vagina and the urinary tract, most frequently to the bladder but sometimes to the urethra or to the ureter. It results in continuous leakage of urine from the vagina, which for most women is a major personal and social tragedy. For countries in the industrialized world, gynaecological surgery is the most common cause of the development of urovaginal fistulae. Hysterectomies are the major cause of this complication. Nevertheless, the incidence of urovaginal fistula formation is lower in the Western world than in a number of developing countries where there are much higher figures for both prevalence and incidence. Here, the
aetiology is also different, fistulae mainly being due to complicated obstructive labour [1,2]. The aim of this retrospective study was to review what kinds of surgical procedures were most frequently complicated by urovaginal fistulae, to find out how they were diagnosed and managed, and to study the outcome after surgical reconstruction. Material and methods The patient material was selected by reviewing medical records according to diagnostic code. The time interval chosen was from 1 September 2003 to 31 August 2009, a timespan of 6 years. Patients with surgery as the single cause of fistula formation were included in the study; women who had undergone previous irradiation were excluded. In
Correspondence: R. Peeker, Department of Urology, Sahlgrenska University Hospital, SE-413 45 Göteborg, Sweden. Tel: +46 313429013. E-mail:
[email protected]
(Received 17 May 2012; accepted 9 July 2012) ISSN 0036-5599 print/ISSN 1651-2065 online Ó 2012 Informa Healthcare DOI: 10.3109/00365599.2012.711772
2
U. Demirci et al.
total, 19 women with a median age of 49 (range 22–59) years remained for final evaluation. These patients’ records were again reviewed in detail, particularly in order to answer the following questions: . .
Scand J Urol Nephrol Downloaded from informahealthcare.com by Goteborgs University on 12/09/12 For personal use only.
.
.
What kinds of surgical procedures caused fistula formation? How long time had elapsed before the diagnosis was established; was the diagnosis unduly delayed? Which diagnostic methods were used? For how long did the patient wait for reconstructive surgery, which technique was used and were there any complications after reconstruction? What was the outcome of reconstruction?
The study protocol was approved by the ethics committee of Sahlgrenska University Hospital. Results Causative surgical procedures and diagnostic delay The surgery that most frequently led to the development of urovaginal fistulae was hysterectomy. In total, 17 patients (90%) underwent total hysterectomy; in two of them the procedure started out as an acute caesarean section but was later converted to a hysterectomy because of uncontrollable bleeding. Of the remaining 15 hysterectomies (79%), 11 were performed on malignant and four on benign indication. Of the remaining two patients, one was operated on for urethral diverticulum and the other for ovarian cancer metastases with radical excision of the distal ureter and a small portion of the bladder, and reimplantation of the ureter. One of the vaginal hysterectomies was converted to laparoscopy owing to ureteral damage, and one laparoscopy to open abdominal surgery for the same reason. A third one with the vaginal approach was also converted to open surgery after the detection of a bladder rift. A total of 14 patients developed vesicovaginal and five developed ureterovaginal fistula. All of them sought emergency medical advice because of leakage from the vagina but the time between onset of the leakage and diagnosis varied considerably. In one woman the course of events was exceptional. She had leakage from the vagina 1 week after surgery, which subsided after a few days. After about 6 months she experienced urinary leakage through the vagina and a fistula was diagnosed. Several patients sought medical advice due to vaginal leakage following gynaecological surgery (especially hysterectomy) without the doctor suspecting a fistula. In one case the patient went to the healthcare centre because of urinary leakage with onset 1 week after acute caesarean section and
hysterectomy. A positive urinary dipstick was found and the patient’s symptoms were interpreted as a urinary tract infection (UTI) that could be treated with antibiotics, without contemplation of possible fistula development. After 2 months, the patient consulted the gynaecological emergency clinic for recurrent UTIs and involuntary leakage, whereupon suspicion of a vesicovaginal fistula was raised for the first time. The diagnosis was finally established 1 month later, i.e. 3 months and 17 days after the primary operation. Another patient reported urinary leakage from the vagina. The consulting doctor suspected fistula, but he could not visualize a fistula at examination with ultrasound and provocation. The patient’s problems were judged to be due to mixed incontinence, and a diagnosis of ureterovaginal fistula was established at a later stage, 3 months after primary surgery, by the “three-swab test” and the instillation of methylene blue. In this patient, kidney ultrasonography revealed a hydronephrosis on the left side with reduction of the kidney parenchyma and, on a chrome clearance and MAG-3 study, the function on the affected kidney was estimated to 10 ml/min/1.73 m2. Diagnostic methods A combination of vaginal speculum examination and urethrocystoscopy was sufficient for the diagnosis in nearly 50% of the patients in the study. When combined with methylene blue instillation and the three-swab test, 90% of all fistulae in the preset series were found. In one patient, cystography primarily
B
V
Figure 1. A cystography revealing a vesicovaginal fistula. The contrast medium passes from the bladder (B) to the vagina (V), clearly demonstrating the fistula (arrow).
Urovaginal fistula after gynaecological surgery revealed the vesicovaginal fistula (Figure 1). One of the patients with ureterovaginal fistula reported total vaginal incontinence 4 days postoperatively. A three-swab test with methylene blue was performed. The diagnosis was thus established, supplemented by a colpography that visualized a fistula from the vagina to the right ureter (Figures 1 and 2).
Scand J Urol Nephrol Downloaded from informahealthcare.com by Goteborgs University on 12/09/12 For personal use only.
Time to reconstruction The time to reconstruction very much depended on the time of diagnosis. Late symptoms, or sometimes early but misinterpreted symptoms, rendered a late diagnosis, and this in turn led to a delayed reconstruction. Reconstruction was intentionally performed at least 3 months postoperatively, thereby allowing time for the postoperative tissue reaction and oedema to subside. One woman, who was reconstructed quite late, after 16 months, was initially reluctant to undergo repeated surgery owing to a previous traumatic sectio with large blood loss and subsequent hysterectomy. Reconstruction method All patients but one were operated on by the same surgeon (RP). Eighteen patients were operated on with an abdominal approach and one with a vaginal approach. The reconstruction technique for all
V
3
operations included the interposition of vascularized tissue, except in one case (Table I). Outcome after reconstructive surgery None of the reconstructed patients reported, or had signs of, leakage or relapse at the time of the followup visit 2–3 months after surgical reconstruction. Postoperative complications were few and mild; a few cases of UTI responding promptly to standard antibiotics. Over the years, there has been no report of fistula recurrence from any of the patients in this series. Discussion The close anatomical relation between the ureter/ bladder and the female reproductive system makes the area particularly susceptible to the development of fistulae after pelvic surgery. In the non-industrialized world, prolonged and obstructive labour accounts for the majority of fistulae [3]. In Western countries, gynaecological surgery is the most common aetiology, and in the present series hysterectomies in particular seem causative. It is difficult to say exactly why hysterectomy is the most common reason, but it is likely that the close connection between the urinary tract and the distal part of the uterus plays a major role. Moreover, in hysterectomy on benign indication the ureters and the posterior aspect of the bladder are, as a rule, not completely dissected and exposed. Speculatively, this may be a contributory reason behind inadvertent injury to these structures, with subsequent fistula formation. It is reasonable to believe, however, that there are certain, preoperatively identifiable, circumstances that increase the risk for fistula development. Studies addressing this complex problem are scant. The reason for fistula development is probably in all cases due to damage, detected or undetected, to the bladder or ureter. Aggravating factors and complications during surgery are plausible risk factors for urovaginal fistula development; anatomical alterations, previous inflammatory conditions or radiation treatment and large tumours are among the conditions that may Table I. Vascularized tissue used in the 19 fistula repairs. Tissue Omentum Caecum
Figure 2. Colpography on a hysterectomized patient. The investigation was, in principle, performed in a similar manner to hysterosalpingography. Upon instillation of the contrast medium into the vagina (V) a ureterovaginal fistula was revealed with dye outlining the right ureter (arrow).
No. of patients 16 1
Martius flap
1
Ureteroneocystostomy only
1
a
a
In this patient complete excision and tissue interposition was refrained from owing to unfavourable intraoperative circumstances.
Scand J Urol Nephrol Downloaded from informahealthcare.com by Goteborgs University on 12/09/12 For personal use only.
4
U. Demirci et al.
predispose to fistulae. Bleeding that is difficult to control, in this series highlighted by the sectio cases that were converted to hysterectomies, may cause difficult conditions for the operating surgeon and thereby increase the risk of accidental damage to the urinary tract. Anatomical conditions, such as adipositas and narrow space in the lesser pelvis, can prejudice the dissection and thus increase the risk of damage. Iatrogenic ureteral injury during pelvic surgical procedures is a well-known complication and important cause of significant morbidity. In a recent report by Hove et al., registered ureteral injuries were reviewed to identify factors of importance to prevent such injuries [4]. They critically evaluated claims concerning ureteral injuries reported to the Danish Patient Insurance Association. In 44 out of 136 submitted claims, the injury was found to be caused by failure to expose the ureters. Most of the ureteral injuries were discovered too late postoperatively and a considerable number of the patients in the report by Hove et al. suffered from chronic renal failure on the affected side. This is at variance with the present series, in which only one kidney had compromised function at the time of diagnosis. Most probably, and contrary to accidental ureteral ligation, for example, the development of a ureterovaginal fistula in general relieves a potential obstruction, thereby decreasing the risk of renal deterioration. In this report it is worth noting a diagnostic delay for some patients despite that they had sought emergency medical advice after gynaecological surgery, presenting symptoms indicative of a urovaginal fistula. There is, the authors believe, room for improvement in this context. It must be firmly stressed that one should always have a very attentive attitude towards patients who recently have undergone surgery in the pelvic region. It is suggested that every inaugural leakage of urine or prolonged vaginal discharge noted after such surgery should be regarded as urovaginal fistula formation until unquestionably ruled out with the use of relevant diagnostic measures. Diagnostically, the addition of methylene blue instillation and the three-swab test, which is a relatively simple and tolerable procedure, revealed more fistulae than just the standard methods. In the typical case the cystoscopy provides the diagnosis. The delay in many of the patients in the present series might have been due to the limited availability of cystoscopy. Several of the patients in this series were treated with catheter drainage of the bladder immediately upon recognition of the fistula, to make the fistula close spontaneously. This has been previously suggested to have a good outcome [5], but in the present series these efforts were in vain. Other authors have reported
a successful outcome with endoscopic injection of fibrin glue for the treatment of urovaginal fistulae [6], but this technique was not found to be suitable for any of the subjects in this series. Almost all of the patients in this study were reconstructed with an open abdominal approach, although some of them probably could have been operated on with a vaginal approach. Moreover, interposition of vascularized tissue was used in all cases, except for the patient in whom the fistula was just bypassed. These facts most certainly mirror the present lack of confidence with the vaginal approach, in contrast to a very favourable experience with abdominal operation and omentum interposition, results that tally with those reported by others using a similar surgical approach [7]. Nevertheless, excellent outcome has also been reported after reconstruction via the vaginal route [8] and these authors also demonstrated that the interposition of a peritoneal flap yielded reconstruction patency comparable with Martius flap interposition. The necessity of using interposition plasty at fistula repair has been debated, but the authors believe that this measure is of vital importance, an opinion shared by others [9]. The jury is, however, still out on the issue concerning the ideal material. The authors use the omentum majus whenever possible, but other techniques have been advocated [10,11], including the use of collagen graft [12]. The good final outcome for all patients is probably due to a number of factors. One important factor may be the timing of reconstruction at least 3 months postoperatively. Such a rigid and non-negotiable attitude has, however, been challenged, with other authors reporting favourable results also after immediate reconstruction [13,14]. At present, it is not possible to take a firm scientific standpoint on this issue, principally because of the lack of solid evidence concerning which strategy is the most advantageous. There is still a need to “play it safe”; after such a major complication as fistula formation the following surgical remedy has to be successful. Elkins reported on a large series of patients subjected to fistula repair in which the second and the third attempts at repair succeeded in only 50% and 33%, respectively [15]. This report supports the authors’ contention that success at the first attempt is of paramount importance. Furthermore, even though the final outcome in the present series was very good, the authors believe that it is time to embark on less invasive techniques for the reconstruction of urovaginal fistulae. In recent years, there have been many reports in the literature on successful outcomes after laparoscopic and robotic assisted fistula repair, although most series were small [16–21].
Urovaginal fistula after gynaecological surgery In conclusion, hysterectomy is by far the most common cause behind the formation of urovaginal fistulae. Misinterpretation of symptoms after gynaecological surgery is common even in cases where the symptoms are indicative of a urovaginal fistula. Delayed fistula repair after a minimum of 3 months, via the abdominal route and with the interposition of vascularized tissue, yields an excellent final outcome.
Scand J Urol Nephrol Downloaded from informahealthcare.com by Goteborgs University on 12/09/12 For personal use only.
Acknowledgements This study was supported by Anna-Lisa and Bror Björnssons Research Foundation, Märtha and Gustaf Ågrens Research Foundation and Sahlgrenska University Hospital project number ALF 7582. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper. References [1] Goodwin WE, Scardino PT. Vesicovaginal and ureterovaginal fistulas: a summary of 25 years of experience. J Urol 1980; 123:370–4. [2] Langkilde NC, Pless TK, Lundbeck F, Nerstrom B. Surgical repair of vesicovaginal fistulae – a ten-year retrospective study. Scand J Urol Nephrol 1999;33:100–3. [3] Kelly J. Ethiopia: an epidemiological study of vesico-vaginal fistula in Addis Ababa. World Health Stat Q 1995;48:15–17. [4] Hove LD, Bock J, Christoffersen JK, Andreasson B. Analysis of 136 ureteral injuries in gynecological and obstetrical surgery from completed insurance claims. Acta Obstet Gynecol Scand 2010;89:82–6. [5] Davits RJ, Miranda SI. Conservative treatment of vesicovaginal fistulas by bladder drainage alone. Br J Urol 1991;68:155–6. [6] Sharma SK, Perry KT, Turk TM. Endoscopic injection of fibrin glue for the treatment of urinary-tract pathology. J Endourol 2005;19:419–23. [7] Mondet F, Chartier-Kastler EJ, Conort P, Bitker MO, Chatelain C, Richard F. Anatomic and functional results of transperitoneal–transvesical vesicovaginal fistula repair. Urology 2001;58:882–6.
5
[8] Eilber KS, Kavaler E, Rodriguez LV, Rosenblum N, Raz S. Ten-year experience with transvaginal vesicovaginal fistula repair using tissue interposition. J Urol 2003;169:1033–6. [9] Evans DH, Madjar S, Politano VA, Bejany DE, Lynne CM, Gousse AE. Interposition flaps in transabdominal vesicovaginal fistula repairs: are they really necessary? Urology 2001; 57:670–4. [10] Alagol B, Gozen AS, Kaya E, Inci O. The use of human dura mater as an interposition graft in the treatment of vesicovaginal fistula. Int Urol Nephrol 2004;36:35–40. [11] Reynolds WS, Gottlieb LJ, Lucioni A, Rapp DE, Song DH, Bales GT. Vesicovaginal fistula repair with rectus abdominus myofascial interposition flap. Urology 2008;71:1119–23. [12] Robles JE, Saiz A, Rioja J, Brugarolas X, Berian JM. Collagen graft interposition in vesicovaginal fistula treatment. Urol Int 2009;82:116–18. [13] Lee JH, Choi JS, Lee KW, Han JS, Choi PC, Hoh JK. Immediate laparoscopic nontransvesical repair without omental interposition for vesicovaginal fistula developing after total abdominal hysterectomy. JSLS 2010; 14:187–91. [14] Waaldijk K. The immediate management of fresh obstetric fistulas. Am J Obstet Gynecol 2004;191:795–9. [15] Elkins TE. Surgery for the obstetric vesicovaginal fistula: a review of 100 operations in 82 patients. Am J Obstet Gynecol 1994;170:1108–18; discussion 18–20. [16] Das Mahapatra P, Bhattacharyya P. Laparoscopic intraperitoneal repair of high-up urinary bladder fistula: a review of 12 cases. Int Urogynecol J Pelvic Floor Dysfunct 2007; 18:635–9. [17] Dorairajan LN, Hemal AK. Lower urinary tract fistula: the minimally invasive approach. Curr Opin Urol 2009;19: 556–62. [18] Gupta NP, Mishra S, Hemal AK, Mishra A, Seth A, Dogra PN. Comparative analysis of outcome between open and robotic surgical repair of recurrent supra-trigonal vesico-vaginal fistula. J Endourol 2010;24:1779–82. [19] Hemal AK, Kolla SB, Wadhwa P. Robotic reconstruction for recurrent supratrigonal vesicovaginal fistulas. J Urol 2008; 180:981–5. [20] Laungani R, Patil N, Krane LS, Hemal AK, Raja S, Bhandari M, et al. Robotic-assisted ureterovaginal fistula repair: report of efficacy and feasibility. J Laparoendosc Adv Surg Tech A 2008;18:731–4. [21] Otsuka RA, Amaro JL, Tanaka MT, Epacagnan E, Mendes JB Jr, Kawano PR, et al. Laparoscopic repair of vesicovaginal fistula. J Endourol 2008;22:525–7.
