International Journal of Gynecology and Obstetrics 123 (2013) 135–138

Contents lists available at ScienceDirect

International Journal of Gynecology and Obstetrics

journal homepage: www.elsevier.com/locate/ijgo

CLINICAL ARTICLE

Early surgical repair of iatrogenic ureterovaginal fistula secondary to

gynecologic surgery
Shicheng Yu, Haiyang Wu, Liwei Xu, Gonghui Li ⁎, Zhigen Zhang
Department of Urology, Sir Run Run Shaw Hospital, School of Medicine, Zhejiang University, Hangzhou, China

a r t i c l e i n f o a b s t r a c t

Article history: Objective: To describe the early surgical repair of iatrogenic ureterovaginal fistula (UVF) secondary to gynecologic
Received 24 January 2013 surgery, and evaluate the efficacy and feasibility of early surgical intervention. Methods: Data were retrospective-
Received in revised form 15 May 2013 ly reviewed for 52 patients with iatrogenic UVF who underwent early surgical repair between January 1996 and
Accepted 23 July 2013 January 2011 at the Sir Run Run Shaw Hospital, Hangzhou, China. Preoperative patient characteristics and post-
operative endpoints, including type of index procedure, repair technique, operative time, perioperative compli-
Keywords:
cations, and postoperative follow-up data were assessed. Results: All early surgical repairs were uneventful. The
Early surgical repair
Gynecologic surgical procedure
mean time from diagnosis to surgery was 15.7 days (range, 14–21 days). The mean operative time was 73.4 min
Urinary fistula (range, 51–110 min) for ureteroneocystostomy, and 104 min (range, 91.5–153 min) for Boari flap procedure.
Vaginal fistula The mean estimated blood loss was 57.8 mL (range, 35–80 mL). No major intraoperative complications occurred,
although 5 patients had postoperative surgical site infections. Forty-three patients were followed for a mean of
18.4 months (range, 10.5–24.0 months). Ureteral stricture with asymptomatic hydronephrosis occurred in 2
(4.7%) patients. Ipsilateral renal function was preserved in all patients. Conclusion: Early surgical repair of UVF
secondary to gynecologic surgery was safe and effective, and preserved ipsilateral renal function.
© 2013 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Ureterovaginal fistula (UVF) is a rare but physically, psychologically,
and socially devastating condition. The etiology of UVF varies geograph-
ically. In the United States and other high-resource countries, UVF com-
monly occurs as an iatrogenic complication of gynecologic procedures
[1,2], whereas in low-resource countries where hygienic conditions
are poor, it may be attributable to mismanaged labor [3,4]. Several
cases related to gynecologic surgery have been reported during the
past 20 years in China, where UVF used to be mainly of obstetric origin
[5], and the incidence seems to be rising with the increased application
of endoscopic surgery in gynecologic procedures.
Most patients affected with UVF present with urinary leakage per
vaginam. They often show great mental distress owing to the odor
that accompanies leakage. Diagnosis of UVF is straightforward. It con-
sists of a negative methylene blue dye test during cystoscopy, and con-
trast extravasation when retrograde or intravenous pyelography is
performed. However, the preferred and optimal management of iatro-
genic UVF is debatable [6–8].
⁎ Corresponding author at: Department of Urology, Sir Run Run Shaw Hospital, Zhejiang
University, No.3 East Qing Chun Road, Hangzhou 310016, China. Tel./fax: +86
57186002150.
E-mail address: ligonghui1228@gmail.com (G. Li).
The aim of the present study was to summarize the early surgical re-
pair of iatrogenic UVF secondary to gynecologic surgeries occurring at a
single hospital in Hangzhou, China, over the past 15 years.
2. Materials and methods
In a retrospective analysis, data were reviewed from 63 patients
who were treated for iatrogenic UVF at the Sir Run Run Shaw Hospital,
Hangzhou, China, between January 1, 1996, and January 1, 2011. The re-
search protocol of the study was approved by the ethics committee of
the Sir Run Run Shaw Hospital. All patients were given information on
the surgical procedure prior to surgery, and informed consent was
obtained from each patient.
The procedure was performed under general anesthesia and the
patient was placed in the supine position with the ipsilateral flank
supported by a cushion. A transabdominal, extraperitoneal approach
with Gibson incision was adopted. The expanded ureter was initially
identified and carefully dissected at or above the crossing of the com-
mon iliac vessels. The lesion was approached distally after encircling
the uninvolved ureter with a small vessel loop. The ureter was sharply
transected obliquely above the lesion and the distal ureter stump was
ligated with size 2-0 polyglactin sutures (Vicryl; Ethicon, Bridgewater,
NJ, USA). Ureteroneocystostomy or flap replacement was performed
depending on the length of the ureter.
During ureteroneocystostomy, the ureter underwent anastomosis
to the lateral wall of the bladder using interrupted size 4-0 polyglactin

0020-7292/$ – see front matter © 2013 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.ijgo.2013.05.008
136 S. Yu et al. / International Journal of Gynecology and Obstetrics 123 (2013) 135–138

sutures (Vicryl; Ethicon) by the Lich-Grégoir extravesical non-refluxing Table 2

technique. If the ureter was not long enough to perform a tension-free Indications for index operation and number of ureterovaginal fistula cases in each
category.
anastomosis, the patient underwent a ureteral reimplantation via the
Boari flap technique. A 6 F double pigtail stent was indwelled in all Diagnosis No. of cases
patients and removed 6 weeks after surgery. The bladder was drained Laparoscopic Abdominal Vaginal
by a Foley catheter for 14 days. approach approach approach
Data were retrospectively reviewed with respect to the etiology of Uterine leiomyoma 7 (3 LAVH) 3 3
UVF, the surgical technique performed, the overall cure rate, and periop- Uterine adenomyosis 1 1 4
erative complications of early surgical repair. Data are reported descrip- Endometrial carcinoma 16 2 –
tively as mean (range), number, or number (percentage). Data were Cervical carcinoma – 6 –
Postpartum hemorrhage – 3 –
tabulated and analyzed using Prism version 5.0 (GraphPad, La Jolla, Ruptured uterus – 1 –
CA, USA). Metrorrhagia – 2 –
Hemorrhage during cesarean – 1 –
3. Results Hemorrhage after LEEP – 2 –

Abbreviations: LAVH, laparoscopy-assisted vaginal hysterectomy; LEEP, loop electrosurgical

Among the 63 cases of iatrogenic UVFs identified during the study
period, 52 occurred after gynecologic surgery and underwent early
surgical repair. Of these 52 women, 9 were lost from the sample pop-
ulation due to lack of follow-up, and data from the remaining 43
were retrospectively reviewed.
Of the 52 patients, 30 were referred from local community hospitals
in Zhejiang Province, and the others were previous gynecologic in-
patients at Sir Run Run Shaw Hospital. The mean age of the patients
was 41.3 years (range, 22–63 years). All patients had unilateral UVF
(right side, 23; left side, 29). The vaginal leakage began between 6
and 15 days (mean, 10.2 days) after the index gynecologic surgery.
Twenty-eight patients reported varying degrees of flank discomfort.
Among the 30 patients referred from local hospitals, none had under-
gone a previous attempt at repair.
The main causes of fistula formation were hysterectomy for ma-
lignancy (24/52), either in the transabdominal or laparoscopic ap-
proach, in addition to hysterectomy under emergent conditions such
as a ruptured uterus or uncontrolled massive bleeding (9/52) (Tables 1
and 2). UVF seemed more likely to occur with laparoscopic surgeries
(24/52), particularly those carried out between 2000 and 2003
(Table 1 and Fig. 1). This time period coincided with initial use of
the laparoscopic technique for gynecologic surgeries in Zhejiang
Province, China. All UVF diagnoses were confirmed by cystoscopy
with a concurrent methylene blue dye test and intravenous or retro-
grade pyelography.
All 52 patients received surgical repair at an early stage. The time
from leakage onset to referral for surgical repair ranged from 14 to
21 days (mean, 15.7 days). Surgical procedural data are summarized
in Table 3. All patients received intraoperative antibiotic prophylaxis
with third-generation cephalosporins, and treatment was continued
postoperatively for 72 hours. Vaginal douching with povidone-iodine
was performed the night before surgery. Surgical procedures were car-
ried out successfully for all 52 women, with a mean operative time of
73.4 minutes (range, 51–110 minutes) for ureteroneocystostomy and
excision procedure.
104 minutes (range, 91.5-153 minutes) for the Boari flap procedure.
The mean estimated blood loss was 57.8 mL (range 35–80 mL), and
no transfusion was required for any patient. No major intraoperative
complications occurred. Five patients had minor postoperative com-
plications (i.e. surgical site infections). The mean hospital stay was
9.4 days (range, 7–15 days). All surgeries were successful at the first
attempt (52/52) and no patients had vaginal leakage after surgery.
Forty-three patients (43/52) had regular follow-up visits in the
outpatient clinic at the hospital every 3 months for an average peri-
od of 18.4 months (range 10.5–24.0 months). Ultrasonography of
the urinary tract was performed and serum creatinine levels were esti-
mated at each visit. Follow-up intravenous pyelography was performed
6 months after surgery. All 43 patients reported continence during
follow-up, and none reported vesicoureteral reflux. Only 2 patients
had ureteral stricture with asymptomatic, mild hydronephrosis that
did not require intervention. All patients had preserved ipsilateral
renal function with symmetric excretion contrast revealed by intrave-
nous pyelography (Table 3).
As mentioned above, 9 patients were lost to follow-up. They were all
referred from local hospitals in remote regions of Zhejiang Province be-
tween 1996 and 1999, and one of the reasons leading to lack of regular
follow-up might have been inconvenient access to the study hospital.
Case information and results from the treatment of these 9 patients
are shown in Table 4. Although the surgery was successful for all 9 pa-
tients, only 2 returned for follow-up 3 months after surgery. Ultrasonog-
raphy was performed for these 2 patients with uneventful outcomes.

Table 1
Index operations that caused ureterovaginal fistulas.

Type of index operation No. of cases

(n = 52)

Abdominal approach 21
Abdominal hysterectomy 4
Abdominal extrafascial hysterectomy + 2
salpingo-oophorectomy + lymphadenectomy
Radical hysterectomy 6
Emergent hysterectomy 9
Laparoscopic approach 24
Laparoscopic hysterectomy 5
Laparoscopic extrafascial hysterectomy + 16
salpingo-oophorectomy + lymphadenectomy
Laparoscopic assist vaginal hysterectomy 3
Vaginal approach 7
Fig. 1. Annual number of ureterovaginal fistula repairs during the period of review
Vaginal hysterectomy 7
(1996–2011).
 S. Yu et al. / International Journal of Gynecology and Obstetrics 123 (2013) 135–138 137

Table 3 surgeries ranges from 0.4% to 2.5%, and 30% of these cases coincide with
Clinical features of ureterovaginal fistula and long-term outcomes at follow-up.a hysterectomy [12]. Because the morbidity associated with UVF may be
Clinical feature or outcome of ureterovaginal fistula Value b serious, resulting in potential loss of ipsilateral renal function and dete-
rioration of the patient’s quality of life, management continues to chal-
Clinical feature (n = 52)
Timing of surgical repair, d 15.7 (14–21) lenge the ingenuity of any surgeon.
Technique for repair Endoscopic techniques for management of iatrogenic UVF via retro-
Ureteroneocystostomy 43 grade placement of ureteral stents have been reported with a success
Boari flap 9
rate varying from 15% to 84% [6,13]. Many surgeons advocate the use
Initial operative outcomes (n = 52)
Operative time, min of percutaneous nephrostomy as an option for the management of
Ureteroneocystostomy 73.4 (51–110) UVF. However, Schmeller et al. [14] reported a fistula persistence rate
Boari flap 104 (91.5–153) of 55% and a ureteral stricture rate of an 18% in their series of patients,
Estimated blood loss, mL 57.8 (35–80) for whom percutaneous nephrostomy was the primary modality of
Intraoperative complications
treatment. When associated morbidities, such as ureteral stricture for-
Hemorrhage need transfusion 0 (0)
Peripheral organs injuries 0 (0) mation and limited indications, are taken into consideration, the effec-
Postoperative complications tiveness of these techniques for management of UVF is questionable.
Surgical site infection 5 (9.6) Although surgical repair is generally accepted as an effective strategy
Urinary leakage 0 (0)
for management of UVF, there are controversies over the timing, ap-
Hematuria 0 (0)
Ileus 0 (0) proach, and techniques of repair [15,16]. It has long been thought that
Results of initial repair immediate surgery for women with UVF is not preferable, and most sur-
Cure at the first attempt 52 (100) geons advocate waiting for at least 3 months before surgical repair.
Need a second attempt 0 (0) However, many patients run the risk of losing ipsilateral renal function
Long-term outcomes at follow-up (n = 43)
owing to ureteral stricture, and/or experiencing distress because of
Time of follow-up, months 18.4 (10.5–24.0)
Results of fistula repair persistent vaginal leakage during the waiting period. In addition, de-
Vaginal leakage 0 (0) layed repair can cause dense adhesions or scarring of perivesical tissues
Incontinence 0 (0) that hamper the procedure. Dissection and lysis of the scarred tissues
Renal function
might result in secondary fibrosis, ischemia, and anastomotic break-
Serum creatinine (μmol/L) 75.2 (52–104)
Intravenous urography
down, which may contribute to a failure of UVF repair. Optimal timing
Vesicoureteral reflux 0 (0) of surgical repair depends on various factors including the causative
Ureteral stricture 2 (4.7) procedure, the type of ureteral injury, the condition of the surrounding
Hydronephrosis 2 (4.7) pelvic tissue, and the patient’s overall health. Because most UVFs
Symmetric excretion of ipsilateral kidney 43 (100)
related to gynecologic surgery are due to direct injury to the ureter by
a
Clinical features were evaluated among 52 women with ureterovaginal fistula (n = electrocoagulation, improper suturing, or injudiciously placed hemo-
52) and long-term outcomes were evaluated among 43 women.
b
static clamps, the fistulas tend to be smaller than those of obstetric ori-
Values are given as mean (range), number, or number (percentage).
gin and have a mild surrounding inflammatory reaction [17]. Moreover,
most cases of UVF can be managed by ureteroneocystostomy, which
does not require too much dissection surrounding the fistula area. It is
4. Discussion therefore feasible to implement early surgical repair of gynecologic
surgery-related UVF.
Ureterovaginal fistula is a serious sequel to unrecognized ureteral in- In the present series, we usually performed surgical repair 2–3 weeks
jury and is usually iatrogenic. The close proximity of the ureter to the fe- after fistula formation when the surrounding inflammatory reaction
male reproductive organs renders it particularly vulnerable to injury had just subsided, but fibrosis and tissue scarring had not yet occurred.
during gynecologic surgeries. This is especially true for hysterectomy Knowing that abdominal adhesions are often associated with UVF re-
for malignancy or emergent hysterectomy for massive hemorrhage, pair as a result of previous gynecologic surgery, we chose the extra-
where much more parametrium needs to be dissected or ligated peritoneal approach with a Gibson incision. The dilated ureter could
[9,10]. The present data and other studies [11] indicate that the ureter be easily identified at the bifurcation of the common iliac artery and
is most at risk during laparoscopic gynecologic procedures because it sharply dissected with the retention of abundant adventitia. Although
cannot be palpated in the cardinal ligament as it would be during adhesions were encountered, they were easily divided with a combina-
open operations. The incidence of iatrogenic UVF related to gynecologic tion of sharp and blunt dissection, and an identifiable cleavage plane
could be made during the dissection.
Lee et al. [18] reported 9 cases of UVF after gynecologic surgery with
delayed management, of which 4 underwent surgical repair via ure-
Table 4
Initial outcomes of patients lost to follow-up (n = 9).
teroureterostomy. However, the success rate of ureteroureterostomy
in this delayed group was only 50%. The failure of repair might be attrib-
Outcome No. of women uted to excessive tissue dissection and difficult anastomosis under ten-
Initial operative outcomes (n = 9) sion. In the present study, the ureter was transected just above the
Repair by ureteroneocystostomy 9 lesion and no dissection was required in the surrounding fistula area,
Intraoperative complications
making the procedure relatively less difficult. Kumar et al. [19] analyzed
Hemorrhage needing transfusion 0
Peripheral organs injuries 0 data from 84 patients with UVF of gynecologic origin, and all patients
Postoperative complications except 1 received conventional delayed surgical repair. Although there
Surgical site infection 0 was no surgical failure, postoperative complications such as wound in-
Urinary leakage 0
fections and hematuria were still encountered. Rafique et al. [20] report-
Hematuria 0
Ileus 0
ed comparable results of both delayed and early repair in their study of
Successful repair at discharge 9 17 women. However, 2 patients had major complications: 1 had a lower
Follow-up at the first 3 months (n = 2) respiratory tract infection, and 1 had prolonged ileus. We propose that
Vaginal leakage/incontinence 0 early surgical repair using the extraperitoneal approach has the associ-
Hydronephrosis in ultrasonography 0
ated benefit of decreased blood loss, avoidance of bowel manipulation,
138 S. Yu et al. / International Journal of Gynecology and Obstetrics 123 (2013) 135–138
and reduced postoperative complications, all of which are comparable
to findings in previous studies.
The present study shows that early surgical repair of UVF secondary
to gynecologic surgery via the extraperitoneal approach is effective and
safe. It also preserves ipsilateral renal function. To date, most reports in-
clude a small number of cases with limited experience and definitive
conclusions are hard to make. Although the present study includes a rel-
atively large patient cohort, it was not randomized or controlled. How-
ever, that type of study would be difficult to carry out for a disease that
is so rare. The present conclusions might not be statistically valid, but
will represent treatment options based on a relatively large experience.
Long-term results from future multi-center, case-controlled studies will
help in the comparison of early surgical repair of UVF and the current
standard of care, which encompasses a waiting period of 3 months.
Conflict of interest
The authors have no conflicts of interest.
References
[1] 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(1):82–6.
[2] Hilton P. Urogenital fistula in the UK: a personal case series managed over 25 years.
BJU Int 2012;110(1):102–10.
[3] Onsrud M, Sjøveian S, Mukwege D. Cesarean delivery-related fistulae in the Demo-
cratic Republic of Congo. Int J Gynecol Obstet 2011;114(1):10–4.
[4] Hawkins L, Spitzer RF, Christoffersen-Deb A, Leah J, Mabeya H. Characteristics and
surgical success of patients presenting for repair of obstetric fistula in western
Kenya. Int J Gynecol Obstet 2013;120(2):178–82.
[5] Rao D, Yu H, Zhu H, Duan P. The diagnosis and treatment of iatrogenic ureteral and
bladder injury caused by traditional gynaecology and obstetrics operation. Arch
Gynecol Obstet 2012;285(3):763–5.
[6] Selzman AA, Spirnak JP, Kursh ED. The changing management of ureterovaginal
fistulas. J Urol 1995;153(3 Pt 1):626–8.
[7] Elabd S, Ghoniem G, Elsharaby M, Emran M, Elgamasy A, Felfela T, et al. Use of en-
doscopy in the management of postoperative ureterovaginal fistula. Int Urogynecol
J Pelvic Floor Dysfunct 1997;8(4):185–90.
[8] Waaldijk K. The immediate management of fresh obstetric fistulas. Am J Obstet
Gynecol 2004;191(3):795–9.
[9] Likic IS, Kadija S, Ladjevic NG, Stefanovic A, Jeremic K, Petkovic S, et al. Analysis of uro-
logic complications after radical hysterectomy. Am J Obstet Gynecol 2008;199(6):
644.e1–3.
[10] Hwang JH, Lim MC, Joung JY, Seo SS, Kang S, Seo HK, et al. Urologic complications of
laparoscopic radical hysterectomy and lymphadenectomy. Int Urogynecol J 2012;23(11):
1605–11.
[11] Tamussino KF, Lang PF, Breinl E. Ureteral complications with operative gynecologic
laparoscopy. Am J Obstet Gynecol 1998;178(5):967–70.
[12] Neuman M, Eidelman A, Langer R, Golan A, Bukovsky I, Caspi E. Iatrogenic injuries to the
ureter during gynecologic and obstetric operations. Surg Gynecol Obstet 1991;173(4):
268–72.
[13] Labasky RF, Leach GE. Prevention and management of urovaginal fistulas. Clin Obstet
Gynecol 1990;33(2):382–91.
[14] Schmeller NT, Göttinger H, Schüller J, Marx FJ. Percutaneous nephrostomy as primary
therapy of ureterovaginal fistula. Urologe A 1983;22(2):108–12.
[15] Shelbaia AM, Hashish NM. Limited experience in early management of genitourinary
tract fistulas. Urology 2007;69(3):572–4.
[16] Demirci U, Fall M, Göthe S, Stranne J, Peeker R. Urovaginal fistula formation after
gynaecological and obstetric surgical procedures: clinical experiences in a Scandina-
vian series. Scand J Urol 2013;47(2):140–4.
[17] Karram MM. Lower Urinary Tract Fistulas. In: Walters MD, Karram MM, editors.
Urogynecology and Reconstructive Pelvic Surgery. 3rd ed. St Louis, MO: Mosby
Elsevier; 2007. p. 450–9.
[18] Lee JS, Choe JH, Lee HS, Seo JT. Urologic complications following obstetric and gyne-
cologic surgery. Korean J Urol 2012;53(11):795–9.
[19] Kumar A, Goyal NK, Das SK, Trivedi S, Dwivedi US, Singh PB. Our experience with
genitourinary fistulae. Urol Int 2009;82(4):404–10.
[20] Rafique M, Arif MH. Management of iatrogenic ureteric injuries associated with
gynecological surgery. Int Urol Nephrol 2002;34(1):31–5.