Following initial dissection, the next step involves fistula localization and excision (Fig. 13.2). Two techniques have been described in the literature. The traditional transperitoneal approach to VVF repair was popularized by O’Conor and involves a transvesical approach. This technique requires bivalving the bladder to localize the fistula [21]. The advantage of the transvesical O’Conor dissection is the ability to identify the fistula tract and completely dissect it free. Liberal cystostomy can also help identify the ureteric orifices. The transvesical approach was performed via laparotomy until the first laparoscopic case was published in 1994 [22].

However, improved visualization and angled camera offered by the robotic approach has opened up debate as to whether a large opening of the bladder is necessary at all. Long posterior vesical incisions may compromise bladder function and capacity and increase operative time and blood loss. Laparoscopic and robotic VVF repair with extravesical approaches have been performed with good success. Pooled data from laparoscopic and robotic repairs indicate nearly equal representation in the literature of both the transvesical and extravesical approaches, and that both techniques have similar success rates [23]. Three cases of robotic extravesical repair have been reported with recurrence-free results in each [8, 24, 25].

Numerous methods have been described to identify the fistula location in order to perform VVF repair without bivalving the bladder. Transvaginal manipulation of the previously placed fistula catheter can facilitate extravesical fistula localization. Transillumination, via cystoscopy or vaginoscopy, has also been reported. In one series, the authors described the use of concomitant cystoscopy to aid with developing the vesicovaginal plane and localizing the fistula tract [26]. Other authors have described the use of vaginal transillumination via vaginoscopy to facilitate dissection and localization of the fistula [27]. Focusing the light of the cystoscope or vaginoscope on the fistula while switching off the robotic camera light allows for improved intra-abdominal visualization and localization.

The robotic approach to fistula repair provides exceptional magnification, making identification of the fistula much easier than a vaginal approach. Since more fistulas occur at the time of hysterectomy and involve the vaginal cuff, we take the same approach to fistula repair as we do a colpopexy and separate the bladder from the anterior vaginal wall. The area that is most adherent is usually the fistula tract. Once it is identified and opened and fluid is seen, it is the urethral catheter. Then the critical component of the dissection is further separating the bladder and vagina distal to the fistula tract, so that the bladder and vagina are no longer adherent.

No randomized or comparative trials exist to compare the results of transvesical and extravesical approach to laparoscopic or robotic VVF repair. It is likely that adherence to the proper techniques of fistula surgery is likely more important than the approach. Regardless of technique, it is important to have clear and wide exposure of the fistula tract to allow for closure of the fistula edges (and, based on surgeon preference, excision of the fistula tract ), while preserving both ureteric orifices.

Once the fistula tract is identified and the posterior wall of the bladder is dissected off the anterior vaginal wall, the fistula tract can excised using monopolar scissors. The excised portion is sent for pathological examination. Bladder mobilization allows for a tension-free closure. Blunt and wide dissection should be limited to avoid injury to the trigone and ureteral orifices. Bleeding in the region of the fistula is best controlled with bipolar cautery in order to minimize excessive tissue necrosis from monopolar energy.

The bladder should be closed in vertical fashion to minimize the contact between the planned horizontally closed vaginal suture line. The bladder is closed in two layers using absorbable suture (2-0 Vicryl). Following bladder closure, the integrity of the repair should be tested. Data pooled from 44 studies of laparoscopic and robotic VVF showed that success rates were 6% higher in cases where a bladder fill test was performed [23]. The bladder should be filled with 180–200 mL of saline through the urethral catheter to assess for the absence of extravasation. Any defects should be closed with interrupted, absorbable suture.

Once the bladder is closed, reapproximation of the vagina begins (Fig. 13.3). A horizontal closure with absorbable suture (2-0 Vicryl) in a running, locking fashion is usually used (re-creating a new vaginal cuff). However, if the bladder repair was performed horizontally, the vaginal closure should be vertical such that the suture lines do not overlap. The integrity of the vaginal closure can be tested by removing the vaginal sponge or sizer and assessing for the preservation of pneumoperitoneum.

Following vaginal closure in a non-radiated patient, mobilization and placement of tissue interposition is usually not necessary (Fig. 13.4). The need for flap interposition for VVF repair has been evaluated during time of open transabdominal VVF repair. Several retrospective series have shown success rates ranging from 63 to 97.5% for fistulas less than 10 mm repaired without interposition flaps [28, 29]. Laparoscopic repair of VVF without tissue interposition has been reported with good success, though numbers are limited. Two separate series of five and two VVFs, each less than 10 mm, repaired laparoscopically without interposition showed 100% success at one-year follow-up [13, 30]. As noted by Miklos et al., flap coverage is not a substitute for careful dissection and closure. However, omental or peritoneal interposition in the vesicovaginal space takes limited time and morbidity to perform and provides an additional layer to prevent recurrence.

The advantage of a transperitoneal approach is the ability to use a vascularized pedicle for flap coverage between the bladder and vagina. If one is planning to use omentum, it is wise to suture the omentum to the abdominal wall (to be used for the later interposition) before placing the patient in steep Trendelenburg. Often the omentum retracts entirely and is difficult to find robotically. One disadvantage of the robotic approach is the relative lack of flexibility when preparation of the omentum would be necessary. As such, several authors have described the use of regional flaps and other techniques of interposition during robotic VVF repair. A commonly reported technique is a peritoneal flap interposition, which avoids the need for omental preparation or colon mobilization [10, 11]. Other authors have described success with interposition using the epiploic appendices of the sigmoid colon [8, 10]. Fibrin glue was used as interpositioning material in the first reported case of robotic VVF repair [31]. A cadaveric amniotic graft has also been used for robotic-assisted VVF following radiation-induced damage [32].

An omental interposition can also be done robotically by taking the patient out of Trendelenberg position, but this requires undocking and repeat docking. If omentum is unavailable or cannot be adequately mobilized, the epiploic appendices of the sigmoid colon or a peritoneal flap from a nearby location (usually the peritoneum overlying the bladder) can be used as tissue for interposition. The interpositioned tissue should be fixed using absorbable suture on the distal and lateral vaginal walls. The fixation sutures should be adequately distanced from the suture line of the vaginal closure. Importantly, the graft needs to be located such that it prevents the suture lines from re-establishing the fistula tract. When interposition tissue is not available, as in the case of a fistula with surrounding inflammation or large fistula, we have used acellular cadaveric dermis as interposition tissue with good success.

Following satisfactory fistula closure, a drain is left in the rectovaginal pouch. At conclusion, the robotic and assistant trocars are removed under optical guidance. The fascia of the 12-mm trocars is closed with monofilament non-absorbable, number one running suture .

Bladder drainage following repair is important for successful outcomes . Most authors have foregone the use of suprapubic drainage, especially in cases where the bladder is not bivalved. If hemostasis is a concern, a suprapubic catheter should be placed. In the absence of injury or extensive peri-ureteral dissection, the ureteral catheters can usually be removed at the conclusion of the procedure. The surgical drain can be removed once output has minimized. The reported duration of urethral catheter drainage following robotic VVF repair ranges from 10 to 21 days. Most authors perform imaging with either retrograde cystogram or voiding cystourethrogram (VCUG) at the time of catheter removal.

Ten reports of robotic VVF repair in 31 patients have been reported. Follow-up ranged from three to 24 months with no recurrences reported (Table 13.1). Reported cases of robotic VVF have shown a mean operative time ranging from 141 to 305 min. The length of stay ranged from one to seven days with the majority of patients having a post-operative stay less than 3 days. The largest systematic review of laparoscopic and robotic VVF repair found a 2% rate of conversion to laparotomy. The rates of urinary tract infection, wound infection, and enterotomy were all less than 1% [23].