Laparoscopic repair of urogenital fistulae: A single centre experience.

CONTEXT: Sparse literature exists on laparoscopic repair of urogenital fistulae (UGF). AIMS: The purpose of the following study is to report our experience of laparoscopic UGF repair with emphasis on important steps for a successful laparoscopic repair. SETTINGS AND
DESIGN: Data of patients who underwent laparoscopic repair of UGF from 2003 to 2012 was retrospectively reviewed.
MATERIALS AND METHODS: Data was reviewed as to the aetiology, prior failed attempts, size, number and location of fistula, mean operative time, blood loss, post-operative storage/voiding symptoms and episodes of urinary tract infections (UTI).
RESULTS: Laparoscopic repair of 22 supratrigonal vesicovaginal fistulae (VVF) (five recurrent) and 31 ureterovaginal fistulae (UVF) was performed. VVF followed transabdominal hysterectomy (14), lower segment caesarean section (LSCS) (7) and oophrectomy (1). UVF followed laparoscopy assisted vaginal hysterectomy (18), transvaginal hysterectomy (2) and transabdominal hysterectomy (10) and LSCS (1). Mean VVF size was 14 mm. Mean operative time and blood loss for VVF and UVF were 140 min, 75 ml and 130 min, 60 ml respectively. In 20 VVF repairs tissue was interposed between non-overlapping suture lines. Vesico-psoas hitch was done in 29 patients of urterovaginal fistulae. All patients were continent following surgery. There were no urinary complaints in VVF patients and no UTI in UVF patients over a median follow-up of 3.2 years and 2.8 years respectively.
CONCLUSION: Laparoscopic repair of UGF gives easy, quick access to the pelvic cavity. Interposition of tissue during VVF repair and vesico-psoas hitch during UVF repair form important steps to ensure successful repair.

INTRODUCTION

The urinary tract is at risk of injury during pelvic operations and complicated labour. Such injuries may lead to urogenital fistula (UGF) whereby the urine bypasses the continence mechanisms and involuntarily leaks through genital tract. It affects physical, mental, social and sexual life of the patient. Most common UGF are vesicovaginal fistulae (VVF) or ureterovaginal fistulae (UVF). In developing countries with low level of obstetrical care, complicated labour is a common cause of VVF. In developed world, iatrogenic injury to urinary tract is the most common cause of VVF.[1] Ureteral injury occurs in 0.3-2% patients after gynaecological surgery.[23] Hysterectomy per se accounts for the vast majority of bladder (2.9%) or ureteric (1.8%) injuries[4] and subsequent UGF.[1] During laparoscopic surgery, inadvertent injuries to urinary tract most commonly occurs during electro-coagulation, hence 70% of these injuries are diagnosed in the post-operative period.[5]

There are multiple approaches to manage UGF ranging from conservative management to open surgical repair. The choice of operation for VVF is predominantly a matter of surgeon's preference. For UVF, the choice is essentially between endourological, open, laparoscopic or robot assisted repair. Laparoscopy has evolved to be an efficient surgical modality as it has the advantages of less post-operative pain, blood loss, shorter convalescence and minimal scar. We report our experience of laparoscopic repair of UGF with emphasis on the salient steps that increase the probability of a successful reconstruction.

MATERIALS AND METHODS

Data of all the patients who underwent laparoscopic repair of UGF from July 2003 to September 2012 was collected and retrospectively reviewed. Aetiology, prior failed attempts at repair, size, number and location of fistula, mean operative time, mean blood losses were noted. Regular follow-up of these patients was done and urinary symptoms and urinary tract infections (UTI) were noted if any. The demographic, clinical and operative data of patients is tabulated below [Table 1].

Table 1

Demographic, clinical and operative data of patients

A total of 53 patients underwent laparoscopic repair of UGF during this period. Five patients with VVF had undergone previous failed attempts of repair. Two of these five patients were operated transvaginally, one transabdominally; one was operated once by transvaginally and then transabdominally. In one patient endoscopic fulguration of the VVF was attempted. In all UVF patient initially endoscopic procedures were attempted, which were unsuccessful.

Workup of patients included a thorough history, physical examination including a pelvic exam and radiological investigation such as micturating cystourethrogram, intravenous urography, cystoscopy and retrograde pyelography in selected cases prior to surgical repair. These examinations confirmed and characterised the fistula between the vagina and bladder/ureter. In all UVF patients, percutaneous nephrostomy was placed to preserve renal function. It also aided in obtaining nephrostomogram to accurately assess the distance between the point of ureteral injury and the bladder.

All VVF were located in supratriagonal portion of the urinary bladder. All patients of UVF had injury in the lower third of the ureter and a normal bladder capacity. Laparoscopic O'Connor's repair and ureteroneocystotomy with/without vesico-psoas hitch was done in all VVF and UVF cases respectively.

The technique of VVF repair and UVF repair was same as published previously.[67]

Briefly, all the patients were intubated and a Foley's catheter was inserted in the urethra.

In patients of VVF, a wet pack was placed in the vagina to prevent leakage of saline during bladder filling and of gas after opening urinary bladder. It also caused the anterior vaginal wall to bulge into the bladder making the visualization easier and provided a stable, uniform and firm plane to dissect off the posterior bladder wall at and beyond the fistula using sharp dissection. A 10-mm port for the telescope was placed at the upper edge of the umbilicus after creating pneumoperitoneum using a veress needle. Two additional 5-mm ports were placed in each iliac fossa and a 4th 5-mm port for retraction when necessary. The patient was then put in the Trendelenberg position with a head-down tilt of about 20° to allow the bowel to fall away from the pelvis. Adhesiolysis was done and the peritoneum dissected off the dome of the bladder. The bladder was filled with 100-150 ml of saline until it was seen to bulge. A limited midline cystotomy was made in posterior bladder using laparoscopic shears and hook. The opening was extended down to fistula, as described by O'Connor. The fistula was then circumscribed and the vagina separated from the bladder for 1-1.5 cm [Figure 1a]. Bladder flaps were created to allow tension-free suturing of healthy tissue. The edges of the fistula were not excised, but closed horizontally with interrupted sutures of 2-0 polyglactin [Figure 1b]. An additional horizontal mattress suture of 2-0 polyglactin was taken and the edges left long, keeping the needle attached. The omentum was brought down and a stitch taken through it with the needle of the mattress suture, after which the ends were tied, thus fixing the omentum over the repaired fistula [Figure 1c]. When omentum was short a lateral peritoneal flap was used for interposition. The cystotomy was then closed with a continuous 2-0 polyglactin suture in a single layer [Figure 1d]. The bladder was checked for water tightness by instilling 100 ml of saline through the Foley catheter and any large leaks were closed with additional sutures. Suprapubic cystostomy was not employed in any case. A drain was placed in the pelvis through one of the 5-mm ports and the telescope and ports were removed. The 10-mm port sites were closed with the 1-0 polyglactin suture.

Figure 1

(a) Bladder flaps have been dissected off the fistula. (b) Horizontal closure of fistula. (c) Interposing the omentum between suture lines. (d) Vertical closure of cystotomy

In patients of UVF, position of patient, method and position of ports placement was same as above. On the ipsilateral side the posterior peritoneum over the iliac vessels was incised and the ureter was dissected lower down, up to the stricture segment with care taken to preserve the periureteral tissue. The distal ureter was clipped and the normal healthy proximal part was spatulated at the 6 O'clock position. Bladder was filled with sterile saline and adequately mobilized. The bladder wall was hitched to ipsilateral psoas tendon with interrupted 3-0 polypropylene sutures. Non-refluxing Lich-Gregoir onlay technique was used for ureteroneocystostomy. The detrusor was incised using electrocautery and a tunnel was created. Subsequently cystotomy was made using scissors. The ureteroneocystostomy was done using interrupted 4-0 polyglactin sutures without a double-J (DJ) stent [Figure 2]. Drain was placed in all patients.

Figure 2

(a) Left vesico-psoas hitch with detrussorotomy. (b) Cystotomy with jet of urine seen. (c and d) Apical stitch of ureteroneocystotomy and subsequent suturing

RESULTS

The operation was successfully completed in all cases.

A total of 22 patients underwent laparoscopic VVF repair. There were no major intraoperative complications. A total of 20 patients accepted oral fluids on the evening of the operation. One patient of VVF repair developed abdominal distension on the evening of surgery and another patient of VVF repair had recurrent vomiting following fluid intake. These patients were kept nil orally until next morning and subsequently oral intake was resumed gradually starting with clear liquids. All patients were ambulatory on the morning after surgery. No patients developed symptomatic bladder spasms.

A total of 31 patients underwent laparoscopic ureteroneocystostomy. Vesico-psoas hitch was done in 29 cases to obtain a tensionless ureteroneocystostomy. In one patient with bilateral ureteral injury bladder was hitched on the right side only [Figure 3]. Tensionless ureteroneocystostomy could be done in this case without hitching the bladder and without excessively mobilizing the ureter on the left side. In another patient, the ureteral injury was in juxtavesical portion of the ureter and bladder was having a large capacity. Vesico-psoas hitch was not deemed necessary in this case for performing an anti-refluxing ureteroneocystotomy.

Figure 3

(a) Intravenous urography (IVU) showing bilateral ureteral injury with extravasation of contrast from lower third of ureter. (b) Retrograde ureteropyelogram showing right lower ureteral narrowing and complete cut off in lower third of left ureter. (c) Completed repair showing bilateral ureteroneocystotomy with right side vesico-psoas hitch. (d) Post-operative IVU showing prompt drainage of contrast into urinary bladder without any extravasation

The abdominal drain was removed in all cases after 48 h when there was minimal or no output. In patients operated for UVF, the urethral catheter was removed on the day 7 post-operatively after a nephrostomogram using the existing nephrostomy tube under antibiotic coverage. In patients operated for VVF, cystogram was done on 14th post-operative day and subsequently per urethral catheter was removed.

Intraoperative and post-operative details of these patients are tabulated below [Table 1].

VVF patients were periodically followed-up with history particularly storage/voiding symptoms, urinary incontinence, examination and urine analysis. Follow-up of UVF patients was performed with voiding cystography at 3 months, followed by intravenous pyelogram (IVP). Ultrasonography kidney, ureter and bladder region and urine examination was done subsequently at 3 monthly intervals for 2 years. IVP was repeated at 1 year.

DISCUSSION

This case series describes laparoscopic repair of VVF and UVF with good post-operative outcome in the long-term follow-up.

All patients in this series had iatrogenic injuries leading to UGF. Most common surgery leading to UGF was hysterectomy followed by lower segment caesarean section (LSCS). A recent article has suggested more than 40% increase in the rate of fistula following hysterectomy over the last decade, from about 1 in 1000 in 2000-2001 to 1 in 600 in 2009-2010.[8] World-wide increase in LSCS also has led to an increasing number of iatrogenic UGF.[9]

There are multiple surgical approaches for repair of UGF. Repair of VVF can be done via transvaginal or transabdominal approach. O'Connor's technique is considered as the gold standard for management of supratrigonal VVF.[10] It follows principles of VVF repair given by Turner-Warwick.[11] All steps of open O'Connor's repair are duplicated in laparoscopic VVF repair. However, only few small case series of laparoscopic repair of VVF have been published. The possible explanation can be availability of an alternate route i.e., transvaginal approach for VVF repair. Moreover, laparoscopic repair is a demanding procedure and requires skilled intracorporeal suturing. Laparoscopic repair is particularly suitable for VVF which are not amenable for transvaginal repair or surgeons unfamiliarity with vaginal approach. Surgery for UVF can be done with open, laparoscopic and robot assisted reconstruction. A few case series exist for laparoscopic UVF repair. We have previously published our experience of 18 cases of laparoscopic UVF repair.[12]

The choice of approach for fistula repair is based on the surgeon's preference, location and complexity of fistula. Approach should be sensibly chosen to maximise the chance of successful repair in first go as each subsequent attempt increase the complexity of procedure and decrease the chance of success. There is 10% of failure rate of surgery for recurrent VVF.[13]

We repair most of UGF presenting to our institute using laparoscopic surgery. We were able to repair all five recurrent VVF also using laparoscopy. We have previously published our technique of repair of VVF using mini O'Connor's laparoscopic repair.[7] This approach minimizes the length of cystotomy and decrease the operative time without affecting the outcome. As far as possible we like to interpose the tissue between vaginal and bladder suture lines. The role of interposing a tissue is controversial in literature. A few authors favour it[14] and a few have raised doubts about its utility.[15] We recommend it to be done in all cases. We close the fistula horizontally in two layers ensuring non-overlapping suture lines. It is technically difficult as the direction of needle and intended direction of suturing are parallel to each other because of the position of the ports. During laparoscopic UVF repair we perform vesico-psoas hitch in most of the cases. It obviates the need for excessive mobilization of the ureter to bridge the defect which can compromise ureteral vascularity. Seideman et al. did a refluxing ureteroneocyctotomy in his series of 45 laparoscopic ureteral reimplants for lower ureteral stricture, as the ureteral length was short to obtain an adequate tunnel.[16] We did anti-reflux reimplant in all our patients as most them were sexually active and had a theoretical risk of recurrent UTI. We did not place DJ stent in any of our cases of UVF.

Laparoscopic repair of UGF though technically challenging, has multiple advantages. Apart from all the generic advantages of laparoscopy, this modality is particularly well-suited to work in deep pelvic cavity. Blood loss is typically lower during laparoscopic repair of UGF.[17] It may due to tamponade effect of the pneumoperitoneum, which limits bleeding from perivesical and perivaginal venous plexus. During VVF repair, it provides excellent magnified view of the bloodless surgical field whereby accurate dissection in anatomical planes can be done. It permits wide mobilization of the bladder flaps to achieve precise closure of the fistula and tensionless suturing of the bladder. During UVF repair, it aids wide mobilisation of the bladder to perform vesico-psoas hitch.

Robot assisted reconstruction of these fistulae may overcome the drawbacks of laparoscopy to some extent. It has a shorter learning curve and the endo-wrist technology makes it particularly well-suited to do intra-corporeal suturing at awkward angles. It also provides an excellent 3-D vision. However, it increases the cost of surgical procedure and is not commonly available in developing countries like ours. There are multiple recent reports of successful robot assisted ureteroneocystotomies for UVF[1819] as well robot assisted VVF repairs.[20]

There is a selection bias in this study as laparoscopic repair was opted for smaller and supratriagonal VVF. During the study period, 15 patients underwent open modified O'Connor's repair. These patients had an average VVF size of 3.5 cm. Five of these patients had obstetrical fistulae. Eight of these VVF were located in trigone close to the ureteral orifice. Laparoscopic repair was deferred for these patients as it was deemed difficult. Hence, the good results of laparoscopic repair are partly because of selecting easier cases.

CONCLUSION

Laparoscopic repair of UGF is an excellent alternative to traditional open repair. However, it requires familiarity with pelvic anatomy and experience of free hand intra-corporeal suturing. In addition to the generic advantages of laparoscopy, it offers excellent visualization of the surgical field due to less bleeding, good illumination and magnification. Non-overlapping suture line and interposition of tissue during VVF repair, hitching the bladder to psoas during UVF repair are important steps to ensure successful repair.