Introduction

Genitourinary fistulae are one of the most devastating complications in the urogynecology setting. The commonest type of these fistulae is the vesicovaginal fistula (VVF). In developed countries, iatrogenic injury during hysterectomy or pelvic surgery is the primary cause of fistulae [1]. However, the etiology differs in developing countries, being secondary to prolonged, obstructed, complicated labor [2].

Vesicouterine fistulae (VUF) are not as common as VVF. In 1908, the first case was reported by Knipe [3], and later in 1957, Youssef [4] reported on the classic symptoms of VUF. These symptoms included amenorrhea and cyclic hematuria coinciding with the time of the menstrual cycle, or menouria; known afterward as Youssef syndrome. The etiology is iatrogenic, secondary to bladder injury during low uterine segment cesarean delivery [5]. The prevalence of these fistulae is estimated to be 1–4% of all genitourinary fistulae [6].

Ureterovaginal fistulae (UVF) occur as a result of injury to the distal ureter during pelvic surgery. It occurs in about 0.5–2.5% of major gynecological surgical procedures. There are several risk factors for the development of UVF, including endometriosis, pelvic inflammatory disease, radiotherapy, and pelvic malignancy [7]. Concomitant VVF and UVF occur in about 12% of cases. The most common presenting symptomatology is continuous urinary incontinence, occurring 1–4 weeks postoperatively [8]. This may be preceded by several days of flank pain and low-grade fever, presumably as a result of urinoma and/or obstructed kidney [9]. However, these patients continue to have a normal voiding pattern, as bladder filling is maintained from the contralateral kidney.

Most cases of VVF can be repaired via the transvaginal approach [10]. However, in cranially located supratrigonal VVF, VUF, and UVF, the classic repair is through a laparotomy with its associated significant morbidities in previously operated patients [11]. Recently, with the development of minimally invasive surgery, laparoscopy and robotics-assisted laparoscopy have become an attractive tool in genitourinary fistulae repair. They offer low morbidity while preserving the same success rates of the open surgical approaches [12]. More recently, laparoendoscopic single-site surgery (LESS) has been reported to be a valid alternative to the standard conventional laparoscopy for treating such difficult urological procedures [12, 13]. Difficulties in the learning curve are the main obstacles in the practice of these minimally invasive approaches [14].

We present our 10-year single-center experience of laparoscopic repair of different types of female genitourinary fistulae. We performed both conventional and LESS repair of genitourinary fistulae.

Materials and methods

This is a retrospective study that included our records over the last 10 years in the faculty of medicine, Alexandria University. The study was conducted after approval of the Alexandria University Review Board for medical research. Informed consent was taken from all patients including the option for conversion to laparotomy or conventional laparoscopy. All fistulae not accessible through the vaginal approach were selected (high supra-trigonal VVF, all VUF and UVF). Overall, 46 patients with different genitourinary fistulae were reported. All patients had pelvic examination and were investigated by retrograde cystography, cystoscopy, intravenous urography, and retrograde pyelography was done only if there was any suspicion of ureteric injury.

In all cases, the patients were given general anesthesia and placed in the lithotomy position and a nasogastric tube was inserted. In patients with VVF, cystoscopy was carried out at the start with bilateral ureteral stenting using 5-French ureteral catheters and another 5-French ureteric catheter with a different color was inserted into the fistula. A sterile Foley catheter (22-French) was then fixed in the urinary bladder with slight traction in VVF cases, and a betadine-soaked vaginal gauze was inserted to prevent gas leak. After pneumoperitoneum, the Trendelenburg position was applied to the patients to allow for movement of the intestinal loops away from the dissection site.

In the conventional group, pneumoperitoneum was established via a closed technique and three to four ports were used. These included a camera port infra-umbilically, a 10-mm port in the left side midway between the umbilicus and left anterior superior iliac spine, and a third 5-mm port at the same point on the right side. An additional 5-mm port was inserted in the iliac fossa as needed. In the LESS group, about an umbilical incision measuring 2–2.5 cm was made for either the TriPort (10 patients) or the QuadriPort (3 patients). Both straight and pre-bent instruments were used. At the time of suturing, an additional 5-mm port was inserted midway between the right anterior superior iliac spine and the umbilicus, which was used for suturing owing to a lack of a pre-bent needle holder. All suturing were performed using 3/0 Vicryl sutures mounted on a 22-mm needle.

Repair of VVF was similar in both the conventional and the LESS group. We started by incising the peritoneum between the bladder and vagina to develop a plane through which extravesical dissection was completed. The site of the fistulous communication was then sharply incised by the scissors. The posterior bladder wall was adequately mobilized and separated from the anterior vaginal wall. The fistula edges on both sides were trimmed and the bladder was then closed longitudinally in two layers in a water-tight fashion using 3/0 Vicryl, where the first layer was continuous and the second layer was interrupted. The integrity of the suturing was tested by filling the bladder with 250 cc of saline and making sure that no urine leaked. Then the vagina was closed transversally using 2/0 Vicryl sutures in a single layer in continuous fashion. A well-vascularized interposing omental flap was mobilized and interposed between the bladder and the vagina and fixed to the vagina with two interrupted 3/0 Vicryl sutures. A tube drain (Nelaton catheter) was fixed at the end of the operation. Postoperatively, patients were given anticholinergic agents to prevent bladder spasms, and the urethral catheter was left in place for 3 weeks. After 3 weeks, a retrograde cystogram was performed to ensure complete closure of the fistula.

Repair of VUF was started by developing the plane between the uterus and urinary bladder by sharp and blunt dissection. After identification of the fistula, a limited cystotomy was performed and the fistulous tract was completely excised. No electrocautery was utilized to avoid tissue ischemia. After complete mobilization of the posterior bladder wall from the uterus, the bladder wall was trimmed and closed in a two-layer watertight fashion using 3/0 Vicryl sutures. Then, the uterine opening was closed by single figure of eight of 3/0 Vicryl sutures. A well-vascularized omental flap was then mobilized and fixed between the two organs with two 3/0 Vicryl sutures. Finally, a tube drain was inserted into the pelvis in all patients.

Repair of UVF was performed by conventional laparoscopy only, using three to four ports. Extra-vesical dissection was performed by incising the posterior peritoneum at the level of iliac vessels and the ureter was identified. Dissection was continued until the ureterovesical junction or the fistula site. We made a extravesical submucosal tunnel to re-implant the ureter in the bladder in all cases. The ureter was divided, spatulated, and anastomosed to the new hiatus in the bladder using 4–0 Vicryl sutures after a JJ stent was inserted in retrograde fashion. A tube drain was inserted and a urethral catheter left for 7 days. The JJ stent was removed after 6 weeks. CT urography was carried out in all cases and compared with the preoperative films (Fig. 1).

Fig. 1
figure 1

a Preoperative CT urography in a patient with a left ureterovaginal fistula with significant hydroureteronephrosis. b Postoperative CT urography in the same patient after 3 months showing patency of the left ureter and marked reduction of the preoperative hydronephrosis

Full size image

On the first postoperative day, analgesics were given according to the physician-directed schedule to all patients. After that, analgesics are given on demand according to the patients’ requirements.

Results

Overall, 46 patients with laparoscopic repair of genitourinary fistulae were reported. Laparoscopic VVF repair was done in 25 patients, 18 by conventional laparoscopy and 7 by LESS repair. VUF was reported in 14 patients; 8 of them were repaired using conventional laparoscopy whereas the remaining 6 cases were repaired by LESS. All 7 of the reported cases of UVF were repaired via conventional laparoscopy. All patients developed their fistula following either obstetric or gynecological surgeries, except for one patient whose VVF was a radiation fistula.

In all patients with VVF and VUF, extravesical repair was carried out where the fistulous tract was excised and the bladder, vagina, and uterus were closed in separate layers with interposing tissue in between. Follow-up cystography 6 weeks after the operation showed complete healing of the fistula in all patients. In patients with UVF, extravesical ureteric re-implantation was performed and follow-up CT urography showed successful reimplantation and marked reduction of hydronephrosis in all patients (Fig. 1b).

The overall mean operative time was 176 ± 25 min. The mean blood loss was 105 ± 25 cc. There were no intraoperative complications in any of the patients. Minor postoperative complications were observed in the form of ileus, mild wound infection, and fever. All were managed conservatively. None of the patients was converted into open surgery. However, in all patients who had LESS repair of their fistula, except for one with VVF, an extra 5-mm port was added. The overall mean postoperative hospital stay was 3.2 ± 1.2 days. Three patients of conventional laparoscopic VVF repair stayed for 5 days postoperatively owing to persistent fever and prolonged urine leakage. After a mean follow-up of the patients of 6.3 ± 3.1 years, all patients had had successful repair, except for one patient with a complex VVF (two large fistulae measuring 2 and 2.5 cm in the LESS group).

Comparison of the two VVF and VUF groups (conventional and LESS repair) is shown in Tables 1 and 2. Analysis of data showed that LESS repair of VVF and VUF has comparable results to conventional laparoscopic repair as regards operative time, blood loss, and success rate. However, LESS has the advantage over conventional repair by offering the patients a shorter hospital stay and lower analgesic requirements (p < 0.003).

Table 1 Comparison of the two groups with vesicovaginal fistulae (VVF; total number: 25)
Full size table
Table 2 Comparison of the two vesicouterine fistulae (VUF) groups (total number: 14)
Full size table

Discussion

Laparoscopy has been demonstrated to be equivalent or superior to open surgery in most urological and gynecological diseases. Almost all the ablative and reconstructive urological procedures can now be performed via laparoscopy or robotics-assisted laparoscopy, with the advantage of better cosmesis, minimal analgesic requirements, rapid convalescence, and shorter hospital stay [15, 16].

Laparoscopic repair of VVF was first reported by Nezhat et al. [17], who reported a single case of VVF repaired laparoscopically using the O’Connor technique and omental flap interposition. This was followed by many published clinical studies, including case reports and other case series [18, 19]. Our study included the largest reported number of laparoscopic repairs of VVF until now (25 cases) with the longest period of follow-up.

For VUF, the first laparoscopic repair was reported in 2001 by Hemal et al. as a case report [20]. Several case series then followed; however, all of these reports included a small number of patients [14, 21]. To our knowledge, the first time in the literature a series of LESS repair of VUF was reported was by our group in 2013 [13]. In this recent study, we continue to expand the number of laparoscopic repairs of VUF using both conventional laparoscopy (8 cases) and LESS (6 cases) with a longer period of follow-up.

Robotics-assisted laparoscopic repair of vesicovaginal and VUF has been recently reported [22, 23]. In 2015, Miklos et al. published a systematic review of 44 studies in which laparoscopic or robotics-assisted laparoscopic repair of vesicovaginal fistula were performed. They concluded that robotics-assisted surgery had comparable results with conventional laparoscopy, with the advantages of improved visualization of the operative field and fine tissue manipulation and dissection [24]. However, robotics surgery is not cost-effective, especially in developing and low-income countries.

The LESS procedure was introduced to provide more cosmetic minimally invasive laparoscopy and has now been adopted by urologists and gynecologists worldwide since its introduction in 2007 [25, 26]. It has been reported to be used in the management of different urological reconstructive and ablative procedures. However, the first reported LESS repairs of a female genitourinary fistula were by our group in 2011 (vesicovaginal fistula) and 2013 (vesicouterine fistula [12, 13].

Our technique of laparoscopic and LESS repair of VVF and VUF followed the basic principles of fistula repair [8, 27]. We found that, in hands experienced in laparoscopy, all the steps of fistula repair that are usually performed in open surgery could be performed in both conventional laparoscopy and LESS. Laparoscopic repair of UVF is most technically challenging, as it also entails laparoscopic re-implantation of the ureter. In the literature, few case reports were found with either a laparoscopic Boari flap or extravesical repair [28, 29]. Again, our study included a larger number of patients, but all were repaired by extravesical reimplantation. We feel that extravesical re-implantation is more amenable to being performed via a laparoscopic approach rather than other types of re-implantation.

The reported operative time for laparoscopic repair of VVF and VUF in the literature ranged from 140 to 220 min. This was comparable with our results, where the overall mean operative time was 176 ± 25 min. The mean blood loss in our series was also comparable to that reported in the literature (105 ± 25 cc and 135 ± 30 cc respectively). The same applies to the mean hospital stay [5, 14, 17].

The limitations of this study include its retrospective nature, the high cost of LESS, especially in developing countries, and the need for a high level of training and skills to perform LESS repair. We tried to reduce the cost by using the reusable Triport or Quadriport. Our center is one of the referral centers and has been experienced in LESS since 2009. This allowed us to perform LESS repair of different genitourinary fistulae with a high success rate. To the best of our knowledge, this is the largest reported series that included a variety of different types of female genitourinary fistulae repaired by either conventional laparoscopy or LESS. Moreover, it carried the longest period of follow-up of more than 6 years.

Conclusion

Laparoscopic repair of VVF, VUF, and UVF is a technically feasible and safe procedure with a high success rate and low morbidity. LESS repair of VVF and VUF is a valid alternative with a comparable success rate with conventional laparoscopic repair, but with shorter hospital stay and fewer analgesic requirements. However, laparoscopic repair of female genitourinary fistulae is a technically challenging procedure that requires good laparoscopic skills.