Vesicovaginal Fistula |
INTRODUCTION |
Vesicovaginal fistula (VVF) is a subtype of female urogenital fistula (UGF). VVF is an abnormal fistulous tract extending between the bladder and the vagina that allows the continuous involuntary discharge of urine into the vaginal vault. In addition to the medical sequelae from these fistulas, they often have a profound effect on the patient’s emotional well-being. This article reviews the etiology of VVF, the surgical principles of repair, and the techniques developed for their repair.
History of the Procedure: The earliest evidence of a VVF was found in 1923, when Derry examined the mummified body of Queen Henhenit (2050 BC). These dissections revealed a large VVF in a markedly contracted pelvis. The first clear documented reference to genital fistula was reported in the Ebers papyrus in approximately 2000 BC. However, not until 950 AD did Avicenna correlate the combination of pregnancy at a young age and difficult labor with the formation of a vesicovaginal communication. The term fistula (previously called ruptura) was not used until 1597, when Luiz de Mercado first coined the term.
The first basic surgical principles for the repair of VVFs were described in 1663 by Hedrik von Roonhuyse. He stressed the use of a speculum and the lithotomy position to gain adequate exposure and denudation of the margin of the bladder wall, with reapproximation of the edges using sharpened swan quills. Later, using Roonhuyse’s technique, Johann Fatio documented the first successful VVF repair in 1675. However, not until the 19th century did successful repair of VVFs become common. In 1834, Jobert de Lamballe published a report of his VVF repairs in which skin flaps were used in the vagina. Later, he advocated the use of tension-free closures using vaginal-releasing incisions.
James Marion Sims published his famous discourse on the treatment of VVF in 1852. Using leaden or silver wire, as John Peter Mettauer had done successfully in 1838, Sims achieved success on his 30th surgical attempt on a slave named Anarcha. Sims emphasized the importance of good exposure, adequate resection of the fistula and scarred vaginal edges, and the critical importance of continuous postoperative bladder drainage. Sims disparaged the popular technique of coagulation (by the application of silver nitrate), stating it proved entirely ineffective except in the rare case of a very small fistula. In 1861, Maurice Collis was the first to report a layered closure technique, and, in 1893, Schuchardt described a pararectal incision to facilitate improved exposure for the repair of a high VVF.
During the early 20th century, several additional techniques were used to improve outcome for the repair of VVF. In 1914, Latzko published his partial colpocleisis technique for repair of posthysterectomy VVF, in which he employed the resection of scarred vaginal mucosa and a layered horizontal closure. Latzko’s procedure has been cited, with his 95-100% success rates noted, in numerous surgeons?experiences. In 1950, O’Conor and Stovsky popularized the transabdominal approach and also proposed the use of electrocoagulation as an initial treatment modality in women with VVFs of 3.5 mm or less, citing a 73% success rate.
Additionally, numerous surgeons are credited for the development of various flaps for interposition between the bladder and vaginal walls to minimize the failure of VVF repairs. The list includes Garlock in 1928 (pedicled gracilis muscle flap), Martius in 1928 (pedicled bulbocavernosus flap), Ingelman-Sundberg in 1960 (pubococcygeus, bulbocavernosus, rectus abdominis, and gracilis), and Kiricuta and Goldstein in 1972 (pedicled omental flaps).
This historical outline of surgical advances is by no means complete. Countless surgeons not listed above have helped raise awareness of VVF with the public and in the medical community, while making substantial contributions in the research and surgical management of this morbid condition.
Problem: A VVF is an abnormal communication between the urinary bladder and the vagina that results in the continuous involuntary discharge of urine into the vaginal vault. An accurate diagnosis is paramount before consideration of repair. A variety of methods are available to the clinician, and any excessive or suspicious vaginal discharge in a patient who recently underwent pelvic surgery or who has a history of pelvic radiotherapy should be evaluated promptly for a UGF.
¡@
Frequency: In developing countries, the predominant cause (97%) of VVF is prolonged obstructed labor. VVFs are associated with marked pressure necrosis, edema, tissue sloughing, and cicatrization. The frequency of VVF largely is underreported in developing countries. Margolis cites an incidence of 3-4 cases per 1000 deliveries in West Africa, while Smith and Williams estimate that 500,000 cases of untreated VVF have occurred worldwide.
In contrast to developing countries, countries that practice modern obstetrics have a low rate of UGFs and VVF remains the most common type. Less frequently, UGFs may occur (1) between the bladder and cervix or uterus; (2) between the ureter and vagina, uterus, or cervix; and (3) between the urethra and vagina. Of note, a ureteric injury is identified in association with 10-15% of VVFs.
The majority of UGFs in developed countries are a consequence of gynecological surgery. Consequently, the incidence may change as surgical management changes. The incidence of VVF in the United States is debated. Although most authors quote an incidence rate of VVF after total abdominal hysterectomy (TAH) of 0.5-2%, others suggest only a 0.05% incidence rate of injury to either the bladder or ureter.
Lee, in a series of 35,000 hysterectomies, found more than 80% of genitourinary fistulas arose from gynecological surgery for benign disease. Uncomplicated TAH accounted for more than 70% of these surgeries. The indications for these TAH surgeries excluded the more complex diagnoses, such as pelvic inflammatory disease (PID), endometriosis, and carcinoma; instead, they were performed primarily for diagnoses such as abnormal bleeding, fibroids, and prolapse. In approximately 10% of cases of VVF, obstetrical trauma was the associated etiology. Radiotherapy and surgery for malignant gynecologic disease each accounted for 5% of cases.
Notably, a rise in incidence of UGFs paralleled the switch in policy toward the preference of performing a total hysterectomy over a supracervical hysterectomy.
Etiology:
Developing countries
Numerous factors contribute to the development of VVF in developing countries. Commonly, these are areas where the culture encourages marriage and conception at a young age, often before full pelvic growth has been achieved. Chronic malnutrition further limits pelvic dimensions, increasing the risk of cephalopelvic disproportion and malpresentation. In addition, few women are attended by qualified health care professionals or have access to medical facilities during childbirth; their obstructed labor may be protracted for days or weeks.
The effect of prolonged impaction of the fetal presenting part in the pelvis is one of widespread tissue edema, hypoxia, necrosis, and sloughing resulting from prolonged pressure on the soft tissues of the vagina, bladder base, and urethra. Typically in these countries, the UGF is large and involves the bladder, urethra, bladder trigone, and the anterior cervix. Complex neuropathic bladder dysfunction and urethral sphincteric incompetency often result, even if the fistula can be repaired successfully. Other cultural factors that increase the likelihood of obstetrical UGFs include outlet obstruction due to female circumcision, the practice of symphysiotomy, and Gihiri incisions (anterior vaginal walls incisions).
Developed countries
VVFs in developing countries are attributed predominantly to inadvertent bladder injury during pelvic surgery (90%). They involve a relatively limited focal bladder injury leading to smaller VVFs than are observed in developing countries. Numerous authors highlight the risk of various types of bladder trauma during pelvic surgery. Such injuries include unrecognized intraoperative laceration of the bladder, bladder wall injury from electrocautery or mechanical crushing, and the dissection of the bladder into an incorrect plane, causing avascular necrosis.
Suture placement through the bladder wall in itself may not play a significant role in VVF development. However, the risk of formation of a hematoma or avascular necrosis after a suture is placed through the bladder wall can lead to infection, abscess, and subsequent suture erosion through the bladder wall. This wall defect permits the escape of urine into the vagina and may be followed by an eventual epithelialization of the track.
Gynecologic procedures are the most common iatrogenic factor. Symmonds evaluated 800 genitourinary fistulas over a 30-year period at the Mayo Clinic. Of these, 85% of the VVFs were related to pelvic operations and 75% were related to hysterectomy, with more than 50% being secondary to simple uncomplicated total abdominal or vaginal hysterectomy. Symmonds also found that 5% of these VVFs were obstetric and 10% occurred after radiotherapy. Obstetric UGFs in modern centers include vaginal lacerations from forceps rotations, cesarean delivery, hysterectomy, and ruptured uterus.
Other types of pelvic surgery (eg, urologic, gastrointestinal surgery) also contribute to the incidence of VVFs; such surgeries include suburethral sling procedures, surgical repair of urethral diverticulum, electrocautery of bladder papilloma, and surgery for pelvic carcinomas. Other less common causes of VVFs include pelvic infections (eg, tuberculosis, syphilis, lymphogranuloma venereum), vaginal trauma, and vaginal erosion with foreign objects (eg, neglected pessary). Lastly, a congenital urogenital abnormality may exist that includes a VVF.
Risk factors that predispose to VVFs include prior pelvic or vaginal surgery, previous PID, ischemia, diabetes, arteriosclerosis, carcinoma, endometriosis, anatomic distortion by uterine myomas, and infection, particularly postoperative cuff abscess. Tancer found prior cesarean delivery to be the most common factor predisposing to vault fistula after abdominal surgery in his series of 110 cases; here, 29% were associated with prior cesarean delivery. Of interest, Tancer also noted 67% of the VVFs in his series occurred in the absence of any risk factors. He also noted that 24 patients incurred a bladder injury during hysterectomy; the injury was recognized intraoperatively and received immediate intraoperative repair (often by a consulting specialist). Despite prompt identification, treatment, and postoperative continuous bladder drainage for 7-10 days, a VVF could not be averted.
Clinical: The uncontrolled leakage of urine into the vagina is the hallmark symptom of patients with UGFs. Patients may complain of urinary incontinence or an increase in vaginal discharge following pelvic surgery or pelvic radiotherapy with or without antecedent surgery. The drainage may be continuous; however, in the presence of a very small UGF, it may be intermittent. Increased postoperative abdominal, pelvic, or flank pain; prolonged ileus; and fever should alert the physician to possible urinoma or urine ascites and mandates expeditious evaluation. Recurrent cystitis or pyelonephritis, abnormal urinary stream, and hematuria also should initiate a workup for UGF.
The time from initial insult to clinical presentation depends on the etiology of the VVF. A VVF secondary to a bladder laceration typically presents immediately. Approximately 90% of genitourinary fistulas associated with pelvic surgery are symptomatic within 7-30 days postoperatively. An anterior vaginal wall laceration associated with obstetric fistulas typically (75%) presents in the first 24 hours of delivery. In contrast, radiation-induced UGFs are associated with slowly progressive devascularization necrosis and may present 30 days to 30 years later.
INDICATIONS |
Symptomatic VVF merits appropriate treatment. Further details regarding the indications for a specific procedure are described in
Surgical objectives or principles.RELEVANT ANATOMY AND CONTRAINDICATIONS |
Contraindications: In general no absolute contraindications exist for the attempted correction of a VVF in patients who can medically tolerate a surgical procedure.
WORKUP |
Lab Studies:
¡@
Imaging Studies:
¡@
Diagnostic Procedures:
¡@
TREATMENT |
Medical therapy:
Conservative management
If VVF is diagnosed within the first few days of surgery, a transurethral or suprapubic catheter should be placed and maintained for up to 30 days. Small fistulas (<1 cm) may resolve or decrease during this period if caution is used to ensure proper continuous drainage of the catheter.
In 1985, Zimmern concluded that if the fistula is small and the patient's vaginal leakage of urine is cured with Foley placement, the fistula has a high spontaneous cure rate with a 3-week trial of Foley drainage. He also noted that in general, if at the end of 30 days of catheter placement the fistula has diminished in size, a trial of continued catheter drainage for an additional 2-3 weeks may be beneficial. Finally, Zimmern concluded that if no improvement is observed after 30 days, a VVF is not likely to resolve spontaneously. Under these circumstances, prolonged catheterization only increases the risks of infection and offers no increased benefit to fistula cure.
In their series, Davitz and Miranda found complete resolution of 4 VVFs with continuous bladder drainage maintained for 19-54 days. Tancer noted spontaneous closure in 3 of 151 patients (2%). In these 3 patients, continuous bladder catheterization was provided within 3 weeks of index hysterectomy; none had an epithelialized fistula tract, and 2 had transvesical sutures that were removed at the time of the initial cystoscopic examination. The size of the VVFs was not documented.
Elkins and Thompson noted some success with continuous bladder drainage. Unfortunately, the rate of success was unpredictable for the individual patient; the rates ranged from 12-80%. Successful cases were characterized by the following criteria: continuous bladder drainage for up to 4 weeks, the VVFs were diagnosed and treated within 7 days of index surgery, VVFs were less than 1 cm, and they were not associated with carcinoma or radiation.
Surgical therapy:
Surgical objectives or principles
Guidelines to follow intraoperatively to minimize VVF formation are identical to those followed at the time of index surgery to prevent a fistula complication. A summary of these guidelines follows.
Adequate exposure of the operative field should be obtained to avoid inadvertent organ injury and to ensure prompt identification of any injury incurred.
Minimize bleeding and hematoma formation. The closure of dead space at the anterior vaginal wall upon completion of an anterior colporrhaphy will prevent hematoma formation. This technique employs intermittently incorporating pubocervicovaginal fascia with the vaginal mucosal layer as the vaginal wall is sutured.
Widely mobilize the bladder from the vagina during hysterectomy to diminish the risk of suture placement into the bladder wall. A minimum of a 1- to 2-cm margin of dissection of the bladder from the vaginal cuff should be developed prior to cuff closure.
Dissect the pubocervicovaginal endopelvic fascia between the vagina and the bladder in the appropriate plane. Dissection may be easier with a sharp technique compared to a blunt technique; the key is to prevent trauma and separation of bladder wall fibers as the bladder is mobilized off the anterior vaginal wall. The principle of traction and countertraction of the bladder and uterus works well to effect a bloodless dissection at the areolar pubocervicovaginal fascial plane.
If scarring is present at the pubocervicovaginal fascia and dissection is difficult, consider performing an intentional anterior extraperitoneal cystotomy. This technique enables the surgeon to assess the anatomic boundaries of the bladder wall with digital palpation. If scarring is present at the pubocervicovaginal fascia and dissection is difficult, consider employing an intrafascial technique of hysterectomy to best dissect the endopelvic fascial plane.
Intraoperative retrograde filling and emptying of the bladder or mild traction on a temporarily placed small Foley catheter inserted into the fistula itself are helpful to optimally identify anatomical planes and reveal intraoperative bladder lacerations.
Consider supracervical abdominal hysterectomy instead of TAH. The incidence of UGF formation is lower for supracervical versus total hysterectomy.
If an intraoperative bladder injury does occur, Tancer argues strongly for widely mobilizing the bladder from the underlying structures (fascia and vagina, cervix, or uterus). In doing so, the surgeon can effect a VVF closure under no tension.
For VVF at the trigonal area, a transverse closure is preferable over a vertical one. Vertical closure would be more likely to produce ureteral obstruction because the ureteral orifices would be drawn inward toward each other. Ureteral catheters should be considered in repair of a VVF involving or encroaching on ureteric orifices.
Consider performing cystourethroscopy when performing pelvic surgery. Cystourethroscopy to assure bilateral ureteral patency and the absence of suture placement in the bladder or the urethra has been advocated by some authors as a standard for all pelvic surgery.
Preoperative details:
¡@
Timing of repair
The occurrence of a UGF is an anguishing experience for both the patient and the surgeon. The timing of repair should be dictated by the overall medical condition of the patient and the tissue quality surrounding the fistula. While the emotional status of the patient should not be underestimated, it also should not play a dominant role in the decision process of when to repair a VVF.
Controversy surrounds the length of delay between diagnosis and surgical repair of a noninfected VVF in a patient who has not undergone radiation treatment. Complicating the analysis of data is the fact that no definition has been established for "early" and "late" intervals. Traditionally, late referred to waiting for an 8- to 12-week interval between index surgery and repair. Longer intervals are universally accepted as the standard of care in infected or irradiated tissue. A 1-year interval for radiation-induced fistulas is recommended to ensure full resolution of tissue necrosis.
Margolis and Mercer simply recommend delaying surgery until inflamed and infected tissue has been treated and the infection and inflammation have resolved. O'Conor agrees that the exact timing for repair depends on when the tissue health is adequate; most of his patients were brought to surgery approximately 3 months after index surgery. During the waiting period, he discouraged indwelling catheter usage and generally advocated vaginal estrogen therapy. Consideration to adjunctive steroid therapy may be contemplated.
Carr and Webster suggest a strategy of examining the fistula at 2-week intervals and proceeding to surgery when the tissue is pliable, not infected, and not inflamed. In their experience, this typically occurred 4-8 weeks after index surgery.
In Persky's series of 7 patients, a 100% success rate was noted with repair performed between 1 and 10 weeks. All of these patients had an interposition graft of peritoneum and omentum placed.
In a retrospective analysis of 25 patients with VVF referred between 1970 and 1980, Blandy and colleagues noted success with all early and late repairs. Only 12 patients were referred before 6 weeks and, therefore, were candidates for early repair. The remaining 13 were referred after 6 weeks. The surgical technique employed in all cases was midline cystotomy to the level of the fistula, ureteral catheterization, bladder mobilization from vagina, closure of the vaginal defect with 3-0 chromic catgut in interrupted fashion, and placement of an omental interposition graft. Urethral catheterization was employed for 10-12 days. A suprapubic catheter also was placed in approximately half of the patients. Ureteral stents were removed after 5-10 days.
Blaivas documented his philosophy "to repair fistulas as soon as possible and, preferably, by a vaginal approach" in his 1995 article that examined the repair of 24 VVFs between 1989 and 1993. Early repairs were defined as those that occurred within 12 weeks of index surgery. Success rates for early repair were similar to those for late repair as long as general principles of surgery were followed. He concluded that no benefit was noted by delaying surgery once evidence of any inflammation, induration, or infection was resolved.
Lee found a correlation between increased surgical failure and VVF repair performed very early (10-15 d). In his experience, a delay of 8-12 weeks from index surgery or failed repair ensures a full resolution of inflammation and edema and provides an adequate blood supply, thereby optimizing success of VVF repair. However, he exempts certain cases from this general rule. These include fistulas diagnosed within hours of surgery and obstetrical lacerations.
Medications
Estrogen replacement therapy in the postmenopausal patient may assist with optimizing tissue vascularization and healing. Oral hormone replacement therapy/estrogen replacement therapy (HRT/ERT) alone has been found to suboptimally estrogenize urogenital tissue in 40% of patients. Treatment with estrogen vaginal cream is recommended for patients with VVFs who are hypoestrogenic. A 4- to 6-week treatment regimen prior to surgery is commonly recommended. It may be used alone or in combination with oral HRT/ERT. Dosages range from 2-4 g placed vaginally at bedtime once per week. Alternatively, the patient may place 1 g vaginally at bedtime 3 times per week.
Corticosteroid and nonsteroidal anti-inflammatory therapy is theorized to minimize early inflammatory changes at the fistula site. However, its efficacy has not been proven. Because it also carries potential risks for impairment of wound healing, when early repair is planned, cortisone is not recommended for the treatment of VVF.
Acidification of urine to diminish risks of cystitis, mucus production, and formation of bladder calculi may be a consideration, particularly in the interval between the diagnosis and surgical repair of VVF. Vitamin C at 500 mg orally 3 times per day may be used to acidify urine. Alternatively, methenamine mandelate at 550 mg plus sodium acid phosphate at 500 mg 1-4 times per day also can be administered to achieve urine acidification.
Urised is effective for control of postoperative bladder spasms. It is a combination of antiseptics (methenamine, methylene blue, phenyl salicylate, benzoic acid) and parasympatholytics (atropine sulfate, hyoscyamine sulfate).
Sitz baths and barrier ointments, such as zinc oxide preparations, can provide needed relief from local ammoniacal dermatitis.
Intraoperative details:
Antibiotic prophylaxis for VVF repair was the focus of study in a paper from the Benin Republic by Tomlinson and Thorton. In their series of 79 patients who underwent repair of a VVF by a single surgeon, they found intraoperative ampicillin did not reduce the odds of failed repair. However, patients given prophylactic antibiotic therapy did have fewer urinary infections and required less antibiotic therapy postoperatively.
¡@
Lawson position: This position is ideal for proximal urethral and bladder neck fistulas. The patient is placed in a prone position with the knees spread and ankles raised in the air and supported by stirrups. Combining it with reverse Trendelenburg positioning enhances visualization with this technique. Elkins found this technique to work best for him.
Jackknife position: This is ideal for proximal urethral and bladder neck fistulas. The patient is placed in a prone position with the hips abducted and flexed and the table jackknifed.
Dorsal lithotomy position: Dorsal lithotomy position with standard Trendelenburg positioning provides excellent access for repair of a high VVF.
¡@
In their experiences, Raz, Vasavada, Margolis, and Mercer note that routine excision of the fistula tract is not mandatory. They emphasize the risks of increasing the size of the fistula tract with attempts to resect it. Additionally, these surgeons contend that the fibrous ring of the fistula may add to the strength of the repair and prevent postoperative bladder spasms. Cruikshank reported a 100% success rate in his series of 11 patients with fistula repair without tract excision. Elkins and Thompson state that a small fistula may be resected, but large tracts should only be freshened. They warn of the risk of overexcising fistula edges, thereby causing an increase in the size of the fistula. They point out further risks of intracystic bleeding and blood clot formation from the mucosal edge of the bladder with fistula resection. Subsequent blockage of the catheter postoperatively would then increase the risk of failure of the VVF repair.
In contradistinction, Iselin and colleagues strongly feel excision of the fistula tract ensures closure of all layers with viable tissue, thereby optimizing wound healing. In their series of 20 patients who had undergone hysterectomy, a 100% cure rate was obtained with full excision of the fistula tract. They emphasize lack of complications, such as symptomatic vaginal shortening, with their technique.
Deepithelialization of the fistula tract can be accomplished by various techniques. Screw curette is one method. In 1977, Aycinena described the use of a common type of screw to strip away or curet the epithelial lining of small VVFs. He then simply allowed spontaneous healing to occur. Seven patients were reported in this series, all of whom were treated successfully. Experts in the field caution that this procedure is efficacious only in the smallest of VVFs. Other methods used to deepithelialize the fistula tract include electrocoagulation and sharp knife dissection.
The best chance for a surgeon to achieve successful repair is by using the type of surgery with which he or she is most familiar. Techniques of repair include (1) the vaginal approach, (2) the abdominal approach, (3) electrocautery, (4), fibrin glue, (5) electrocautery and endoscopic closure using fibrin glue and bovine collagen, (6) the laparoscopic approach, and (7) using interposition flaps or grafts.
The literature documents excellent success rates for both the vaginal and abdominal approaches if the following general surgical principles are followed: (1) complete preoperative diagnosis, (2) exposure, (3) hemostasis, (4) mobilization of tissue, (5) tissue closure under no tension, (6) watertight closure of bladder with any cystotomy repair, (7) timing to avoid infection and inflammation of tissue, (8) adequate blood supply at area of repair, and (9) continuous catheter drainage postoperatively.
Vaginal approach
Minimal blood loss, low postoperative morbidity, shorter operative time, and shorter postoperative recovery time are characteristics of the vaginal approach, making it an attractive option. Additionally, the vaginal approach obviates bowel manipulation, reducing operative morbidity, particularly in patients with radiation-associated fistulas.
¡@
Suturing of the labial folds to the ipsilateral thigh provides improved visibility of the vaginal vault.
Procedures used to facilitate exposure in the vagina include Dührssen and Schuchardt incisions. As early as 1856, Baker advocated the use of an episiotomy incision to afford greater exposure in the vaginal repair of fistulas that were located high in the vaginal vault. Vaginal incisions used to improve exposure include the Dührssen incision and the Schuchardt incision.
The Dührssen incision is a deep vaginoperineal incision or extended episiotomy initially proposed for usage in other types of vaginal surgery. Its application to fistula surgery was recommended by Mackenrodt in 1894.
In 1893, Schuchardt introduced a parasacral incision as an extension of a Dührssen incision, whereby a deep vaginoperineal incision is carried cephalad to the vault apex and then posteriorly toward the tip of the coccyx. Schuchardt’s paravaginal incision is performed by incising the posterior vaginal wall in a direction angled toward the ischial tuberosity, going through the levator ani and the coccygeus muscle, to ultimately gain access into the ischiorectal fossa.
In 1984, Maisonneuve described the same procedure in his attempts to gain better access in vaginal repair of fistulas. Hemorrhage is an expected complication encountered using this technique. The obstetrician’s mediolateral episiotomy procedure is a modified Schuchardt paravaginal incision.
¡@
Exposure and access to a VVF can be facilitated by catheterization of the
fistula with a bulb catheter, such as a Fogarty catheter. An uninflated catheter
may thread the fistula where the bulb is inflated, then traction is placed on
the catheter to draw the VVF into the field. A small VVF may be probed first
with a lacrimal duct probe and dilated with cervical dilators to permit
placement of a pediatric catheter/ureteral bulb catheter. ¡@
The critical issue of closure of suture lines without any tension is a tenet
of surgical repair of VVF. In an attempt to reduce strain at the site of
anterior vaginal wall closures, surgeons employ several strategies, including
extensive vaginal wall dissection and mobilization from the underlying
vesicovaginal endopelvic fascia. Alternatively, a surgeon may opt to create lateral radial or circumferential
relaxing incisions similar to those first described by Jobert de Lamballe in the
mid 1800s. The relaxing incisions are the full thickness of the vaginal wall
without extension into the endopelvic fascia. The margins are not reapproximated;
instead, they may be sutured in running fashion for desired hemostasis. A significant danger to performing lateral relaxing incisions is further
devascularization of the vaginal tissue. An alternative approach that avoids
this potential complication is to employ vascularized flaps or grafts at the
site of fistula repair, such as a Martius bulbocavernosus fibromuscular pedicle
with or without an intact skin patch. Such grafts are essential in the repair of
large fistulas and radiotherapy-related fistulas, where large areas of
devascularized and scarred vaginal walls commonly are observed (see
Surgical procedures for the vaginal approach
¡@
Numerous authors hold this time-honored procedure, with success rates of
93-100%, to be the standard for repair of simple posthysterectomy VVFs. In 1942, Latzko published his modification of the Simon colpocleisis
procedure designed for repair of obstetric VVFs. The Simon colpocleisis
technique applied a transverse closure of the vagina beneath the fistula defect.
Unfortunately, it often resulted in the formation of a symptomatic diverticulum
between the bladder and cervix. Latzko advocated the prerequisite of total
hysterectomy to obviate such a complication. Additionally, he strongly cautioned
strict adherence to 2 additional prerequisite conditions. First, adequate
preoperative vaginal vault length must be present because the vagina is
shortened by 1.5 cm. Second, the fistula must be located at the vaginal apex "so
that the posterior margin of the fistula and the scar of the vaginal vault
coincide." Advantages of the Latzko procedure include simplicity of technique, high
success rate, low morbidity, no impairment in bladder capacity, and no
compromise of ureteral orifices, even with fistulas lying close to the orifices.
Several surgeons describe symptomatic vaginal vault foreshortening with the
Latzko procedure. However, in the experience of Elkins and Thompson, significant
shortening in vaginal length was not noted unless the patient had antecedent
shortening. In performing the Latzko procedure, Robertson found he never needs to place a
ureteral catheter, even when the fistula margin lies adjacent to a ureteral
orifice, because the ureter is turned into the bladder, preventing ureteral
occlusion. Some authors state that the presence of a cervix is a
contraindication for a Latzko procedure. However, Elkins and Thompson do not
agree. They report that a juxtacervical VVF can be repaired vaginally if the
cervix can be drawn down adequately out of the surgical field of closure. Latzko technique: Vaginal mucosa is sharply denuded in a circular fashion at
a distance of 1.5 cm from the fistula opening. The fistula at the bladder mucosa
is not disturbed. A double row of sagittally oriented sutures is placed in the
raw surfaces on either side of the fistula, with the second row imbricating the
first. Suturing of the vaginal wall is then performed, providing a third layer
of closure. The vaginal wall in contact with the bladder becomes the posterior
vesical wall and eventually is reepithelialized with transitional epithelium.
¡@
In this technique, the vaginal wall is incised circumferentially around the
fistula and widely dissected from the underlying endopelvic fascia in a standard
anterior colporrhaphy technique. Leaving the tract unresected, the bladder is
closed, tension-free, in 2 layers. The surgery is completed with the vaginal
closure over the bladder defect. Elkins, DeLancey, and McGuire published their experiences with repair of VVF
from January 1985 through May 1989. Martius grafts were needed as an adjunctive
technique to a flap-splitting technique in less than 40% of cases. They did not
find adjunctive techniques necessary when the genital tract fistulas were small
( Numerous surgeons, such as Margolis, Mercer, and Raz, have found this
procedure as efficacious as the Latzko technique. It has better applicability
for large VVFs while not foreshortening the vaginal vault. The authors note the
risk of possible ureteral compromise. Technique: The vaginal wall is incised circumferentially around the fistula,
leaving a rim of intact vaginal wall encircling the fistula tract. At the
lateral sides of the fistula incision, the skin incisions are extended toward
the vaginal apex in a parallel fashion. One incision is carried further than the
other, thereby incising a J shape in the vaginal wall. The anterior and posterior flaps are widely dissected from the underlying
endopelvic fascia. The fistula tract is closed with 3-0 chromic or Dexon suture
in a continuous fashion. This closure includes the full thickness of vaginal
skin previously left intact at the fistula tract, along with the partial
thickness of the bladder wall. A second layer of closure in the endopelvic fascia is performed with 3-0
Dexon suture; it is placed perpendicular to the prior suture line. The distal
vaginal flap is trimmed. The proximal flap is advanced beyond the fistula repair
site, reaching the trimmed distal margin, and reapproximated in a running
fashion. Zimmern et al describe their preference for an asymmetric J incision in the
anterior vaginal wall whereby the lower curve of the J loops around the fistula
site. This modification enables the surgeon to advance one flap over the fistula
repair and prevent overlapping suture lines. Martius grafts were added in cases
where fistula closure was tenuous. Tension-free closure of viable tissue,
avoidance of overlapping suture lines, and continuous postoperative bladder
drainage were factors considered crucial to success. Their 1-year success rates
ranged from 90-100%. ¡@
Technique: The patient is placed in dorsal lithotomy position. Cystoscopy is
performed. Traction on the fistula site is obtained by placing a Foley catheter
into the fistula tract from a vaginal approach, inflating the balloon, and
placing traction sutures at 1-cm distances from the fistula. The vaginal mucosa
is denuded circumferentially for a radius of 3-5 mm from the vaginal cuff,
including the fistula. This incision is then extended obliquely to the bladder
wall so as to resect the fistula tract and vaginal cuff scar in a funnel-shaped
specimen. The defect is closed in 4 layers. First, the bladder is closed with
interrupted 4-0 sutures; the subvaginal pubocervicovaginal fascia then is closed
in 2 layers with interrupted 3-0 sutures. This is followed by a vaginal wall
closure. Each of the 4 layers employed polyglycolic acid suture material. Intravenous indigo carmine and cystoscopy is used to ensure bladder and
ureteral integrity. A suprapubic catheter is the preferred method of bladder
drainage and is maintained for approximately 3 weeks postoperatively. The premise on which Iselin and colleagues base their surgical technique is
that scarred tissue margins do not heal well, if at all. In comparison, fresh
viable margins provide for optimal results in the repair of posthysterectomy VVF.
They advocate a total excision of the fistula tract and vaginal cuff scar. In
their series of 20 patients, all were successfully repaired and no symptomatic
vaginal shortening or other complications were encountered. Abdominal approach Exposure: As with the transvaginal approach, exposure with the transabdominal
approach can be augmented with the use of traction sutures and with
catheterization of the fistula with a Fogarty catheter. Similarly, in 1893,
Weinlechner proposed the use of a ball with an attached wire in the combined
transvaginal and transabdominal repair of VVF. The wire was threaded through the
fistula transvaginally and then grasped through the cystotomy. Traction on the
wire elevates the tract into the surgical field. Absolute indications for abdominal approach include (1) the need for
concomitant abdominal surgery, such as augmentation cystoplasty and ureteral
reimplantation; (2) the inability to adequately expose the fistula vaginally;
(3) a complex presentation of VVF involving the ureters, bowel, or other
intraabdominal structures; and (4) involvement of the VVF with ureteric orifices
(Langkilde). The classic positioning of the patient for abdominal procedures is supine,
with Trendelenburg orientation. However, modifying this by flexing the patient's
hips and abducting and supporting her legs in stirrups is wise. Simultaneous
access and examination of the vaginal vault may assist with laparotomy
procedures. The choice of incision may include suprapubic V, Pfannenstiel, or midline
vertical. In 1887, Bardenheuer recommended a transverse suprapubic incision.
Turner-Warwick et al prefer the suprapubic V incision, noting that it provides
superior access to the lower abdomen and pelvis. Several others advocate a
longitudinal suprapubic incision because it allows the surgeon the ability to
more easily obtain an omental graft. ¡@
In 1885, Trendelenburg introduced the first transvesical extraperitoneal
method of vesicovaginal repair. With the patient placed in a steep Trendelenburg
position, a transvesical incision is performed to visualize the fistula. The
bladder mucosa adjacent to the fistula is circumscribed and removed. The bladder
is dissected off the vagina and the bladder, and vaginal defects are closed
separately. ¡@
The transperitoneal technique was developed by von Dittel in 1803 for the
repair of VVFs. In his procedure, a laparotomy was performed. The bladder was
dissected from underlying gynecologic organs involved. The defects in the
bladder and vagina or cervix were closed separately. In 1913, Legueu described his transvesical transperitoneal suprapubic method.
He combined both the Trendelenburg and the von Dittel techniques, whereby the
peritoneal cavity is accessed by laparotomy and a sagittal incision is made in
the bladder. This cystotomy incision is extended to the fistula. The bladder is
mobilized off the vagina, and the bladder and vaginal defects are closed
separately. Unfortunately, transperitoneal surgeries were technically complex and fraught
with high morbidity and mortality at the time. Unless ureteral or coabdominal
surgery was necessary, numerous surgeons (eg, Fritsch, Wertheim, Latzko) favored
the vaginal approach techniques. In the age of modern medicine, surgeons have the benefit of performing
invasive surgeries with sterile fields, antibiotics, and other medical advances.
As such, Margolis, Mercer, O'Conor, and Sokol find the extravesical
transperitoneal procedure of great benefit when the bladder is densely adhered
to the endopelvic fascia and underlying structures (eg, lower uterine segment,
cervix, anterior vaginal wall). ¡@
This method was introduced as an intraperitoneal or transperitoneal
technique. In 1951, O'Conor and Sokol published a Legueu-type technique for the
suprapubic repair of trigonal and supratrigonal VVFs. According to an article
published by O’Conor in 1980, O'Conor and Sokol developed their technique after
they observed Barnes' technique for the resection of adherent bladder
diverticula in 1934. O'Conor then was able to trace Barnes' technique back to a
similar procedure performed by Ward; he disclaims any claim to originality of
their procedure. To their credit, their large studies, with success rates higher
than 85%, did much to popularize the suprapubic technique. Among the successful cases of repaired VVF are patients with complex and
difficult repairs, such as radiation-associated cases. The authors stressed that
it was key to bisect and widely mobilize the bladder from the vagina in order to
produce a closure with separate tension-free layers. The procedure can be performed extraperitoneally; however, in complex cases,
the transperitoneal approach is preferred because it allows for the addition of
interposition grafts. Advocates of their technique cite the advantages of high
success rate, optimum surgical access to the fistula and ureters, and the
ability to add an interposition graft with this procedure. Technique: Using an infraumbilical incision, laparotomy is performed and the
peritoneal cavity is entered. The posterior wall of the bladder is dissected
free as much as possible. The bladder then is bivalved at the dome. This
incision is extended posteriorly to the level of the fistula. Stay sutures are
placed sequentially along the incisional margins every few centimeters to permit
traction and elevation of the bladder wall in order to aid in exposure and
dissection. Ureteral orifices and the location of fistula(s) are identified, and ureteral
catheters are placed if necessary. The fistula tract and scarred and necrotic
tissue are resected. Dissection of the posterior wall of the bladder from the
underlying endopelvic fascia and vagina is completed. The bladder and vagina are
closed in separate layers. The bladder is closed with a 2-0 chromic suture in
continuous running fashion beginning at the apex and extending through the full
muscle layers and imbricated with a second layer with interrupted 1-0 chromic
sutures. Commonly, peritoneal or interposition grafts are added. A suprapubic
catheter is brought out laterally to the sagittal closure. A transurethral
catheter may be placed and discontinued on postoperative day 4 or 5; the
suprapubic catheter is removed on postoperative day 14. ¡@
Gil-Vernet and colleagues presented a bladder wall flap procedure in 1989 as
an alternative technique for the repair of complicated VVF. The approach may be
transvesical, extraperitoneal, or transperitoneovesical. Advantages cited by the
authors are the capability of repairing large VVFs without compromising bladder
capacity, a low-tension closure, direct and easy identification, and
preservation of the submucosal ureteral portion. Technique: The bladder is entered through a transverse incision at the dome.
Catheterization of the ureters is performed. The fistula tract is completely
excised with the assistance of stay sutures secured around the fistula tract.
The bladder wall is carefully mobilized off the endopelvic fascia and vaginal
wall. The vaginal defect is closed with a single-layer closure. A bladder flap
is constructed to close the bladder defect. Incisions are made at the
superolateral angles of the bladder defect and extended cephalad toward the
dome. The anterior margin of the flap is drawn down over the bladder defect to
meet the caudal margin of the bladder defect. It is sutured in place with 3-0
catgut through the submucosal and muscular layers in interrupted fashion with
sutures not less than 10 mm apart. The ureteral catheters are removed, and the
anterior cystotomy is closed in a single extramucosal layer. When a
transperitoneal approach is chosen, Gil-Vernet prefers to also add an
interposition graft. ¡@
The use of autologous bladder mucosa grafts was first introduced in 1947 as a
technique designed for urethral reconstruction. Since that time, research
performed in a canine model by J.W. Coleman and his associates at Cornell
University demonstrated that autographs of bladder mucosa as large as 4 cm could
be used successfully to cover large defects in canine bladder walls. The
application of a free bladder mucosal graft for repair of difficult VVF was
developed by Ostad and his associates. They published their series of 6 patients
with a 100% success rate. The follow-up interval ranged from 2-6 years. All of the VVFs were high, large, multiple, or recurrent and occurred
posthysterectomy; one patient had a history of pelvic irradiation. Three were
repaired early (<3 mo from index surgery), and 3 were repaired late. Simplicity
of technique, high success rates, lack of the need for interposition grafts, and
decreased patient morbidity were notable advantages to this procedure. Exact
fistula sizes were not documented. Reepithelialization of the denuded mucosa
donor site is believed to occur spontaneously over the following 4-6 weeks. Technique: After gaining access to the peritoneal cavity with either a
Pfannenstiel or infraumbilical low vertical midline incision, an extraperitoneal
cystotomy is performed at the anterior bladder wall. Ureteral catheters are
placed. Bladder mucosa is denuded circumferentially at the fistula site at a
distance of 1 cm. The fistula tract and vaginal wall are left undisturbed. A
free bladder mucosal graft is sharply dissected from its underlying muscularis
layer at the edge of the anterior cystotomy margin. This graft of mucosa is then
secured over the fistulous tract with interrupted 4-0 chromic catgut sutures
that are placed into the superficial muscularis at a distance of 2-3 cm. The
anterior cystotomy is closed in 2 layers. A transurethral catheter is used for
24 hours. A suprapubic Malecot drain is left in place for 2-3 weeks, and a
cystogram is obtained prior to its removal. Electrocautery In a series of 15 patients, Stovsky et al reported a 73% cure rate with
electrocoagulation in a highly selected patient group. The fistulas that were
successfully managed with electrocautery as the sole treatment modality were
small in size. Only 4 fistulas were large enough to be cystoscopically
identified with a true fistula opening; the remaining 11 were identified as
either pinhole openings or bladder mucosal dimples. Details of their technique
include both vaginal and cystoscopic routes and fulguration with a Bugbee
electrode and placement of a large Foley catheter for a minimum of 2-3 weeks.
Care was taken to use low-current settings in order to minimize the potential of
thermal damage and enlargement of the fistula. In contradistinction, Margolis and Mercer in 1994 concluded the risk of
destruction of viable tissue with usage of electrocoagulation in the repair of
fistulas is too great to warrant its application. They classified this technique
as worthy of historical interest only. Fibrin glue Cronkite et al introduced fibrin glue in the 1940s. They combined fibrinogen
and thrombin for use as a sealant in skin grafting procedures. Success was
unpredictable, and this technique was largely abandoned until the 1970s. Matras
and associates published their experiences with fibrin glue for interfascicular
nerve repair in animals in 1972. In 1985, Matras reported on the use of a fibrin
sealant in maxillofacial surgery. Occlusion therapy using fibrin glue is considered useful and safe for
intractable fistulas. Fibrin glue facilitates healing by recruiting macrophages
and providing a semisolid support structure rich in growth and angiogenic
factors. This system continues to support the fibroblast to connective tissue
transition. Most of the data are from European investigators because the US Food and Drug
Administration regulated against commercially prepared fibrin agents until
recently. A fibrin-sealant technique has been used to treat a variety of fistula
types, including pancreatic, maxillofacial, enterocutaneous, anorectal,
bronchopleural, and gastrocutaneous. Proponents of its use note that it is
employed as a minimally invasive and technically simple outpatient surgery that
lacks significant morbidity. The commercial fibrin glue presently used in the
United States is Tisseel. Fibrin occlusion of a VVF was first developed by Pettersson and associates in
1979. The VVF was incurred following surgery and radiotherapy and was cured with
the first attempt. Encouraged by this, Hedelin et al performed the technique in
their series of 9 patients with chronic fistulas of 7 months to 10 years in
duration; all had undergone at least one prior surgical repair attempt. They
demonstrated a 50% success rate in treating vesicocutaneous fistulas, with
failure in a single patient with VVF. The VVF was 2 cm long; width was not
reported. In 1998, Venkatesh and Ramanujam published their experiences using autologous
fibrin glue to treat 30 patients with recurrent anorectal and
urethrovesicorectal fistulas. All patients had undergone at least one prior
failed surgery to repair their fistulas. Granulation tissue was removed by
curettage; extensive debridement was not necessary. Cryoprecipitate and thrombin
were dispensed simultaneously into the fistula tract, filling it immediately
with the coagulum. Approximately half of the patients required 2 applications.
The overall success rate was 60%. Failure rates were high if the tract was short
and straight. The 2 urethrovesicorectal fistulas failed to heal, and the authors
presumed urinary contamination played a role in the mechanism. Tsurusaki et al reported on a case in which fibrin glue was used successfully
to heal an intractable kidney transplant ureteral fistula. Urinary leakage
occurred from a ureterocutaneous fistula at the ureteroureterostomy site. The
tract was injected from the cutaneous side on postoperative days 104, 121, and
136 under x-ray fluoroscopic guidance. Success was achieved immediately on the
third injection. The stent and nephrostomy tube were removed, and no recurrence
was noted at the time of publication 11 months later. For optimal success, they
mandate delineating the fistula by radiographic techniques before fibrin
occlusion therapy and caution proper placement of the double-lumen catheter
before injection. Electrocautery and endoscopic closure using fibrin glue and bovine
collagen Morita and Yokue published a case report of successful closure of a
radiation-induced and markedly fibrosed VVF measuring 5 mm. They buttressed the
fibrin glue in the fistula tract between collagen cushions at the proximal and
distal sites of the fistula to prevent its mechanical disruption by the efflux
of urine from the bladder. Technique: After performing electrocoagulation of the fistula, a cystoscope
was introduced transurethrally into the bladder, and 1 mm of bovine collagen was
injected submucosally under direct visualization around the fistula opening.
Fibrin glue was injected transvaginally into the fistula tract. A second
application of 1 mm of bovine collagen was then injected transvaginally into the
vaginal mucosal layer around the fistula tract. A transurethral Foley was
employed for 3 weeks. Laparoscopic approach C.H. Nezhat and colleagues assessed the laparoscopic closure of intentional
and unintentional bladder lacerations in a series of 20 cystotomies. In this
study, the only complication noted was a single VVF that required reoperation.
This fistula was successfully repaired laparoscopically with a single-layer
closure. In another laparoscopic dissection for benign disease, a VVF resulted
postoperatively and was successfully repaired laparoscopically at a later
surgery. Also, see
Omental J flap. Rotated vascularized pedicle flaps are an important adjunct to surgical
techniques employed in the repair of VVF. They increase success by enhancing
granulation tissue formation, increasing neovascularity to the area, and
obliterating dead space. They also provide a barrier layer between the bladder
suture line and the vaginal suture line. Elkins cautions not to expect this
procedure to provide structural periurethral support and not to expect success
if the fistulous space is not completely closed, closed under tension, or closed
with only nonviable tissue. ¡@
¡@ ¡@
¡@
¡@
¡@
¡@
¡@ ¡@
¡@
¡@
¡@
¡@
¡@
¡@
¡@
¡@
¡@ ¡@
¡@
¡@
¡@
Postoperative details: Bladder drainage: The consensus is overwhelming in the literature that
continuous bladder drainage postoperatively is vital for successful UGF repair.
A large-caliber catheter minimizes the potential for catheter blockage by blood
clots, mucus, and calcaneus deposits. However, to date, no prospective
randomized trials have demonstrated the superiority of any single type of
catheter drainage. Type and duration of catheter drainage: For fistulas involving the lower
portion of the bladder trigone, bladder neck, or urethra, transurethral bladder
catheters should not be used. Symmonds finds a large suprapubic catheter for
7-60 days preferable to minimize excess tension on the suture line and to ensure
nonobstructed continuous drainage. In posthysterectomy VVF repairs, both
transurethral and suprapubic catheters may be placed. The urethral catheter may
be discontinued on the fifth to seventh day. If vesical integrity is noted 2
weeks later on a cystogram, the suprapubic catheter may be removed. Surgeries to
repair pelvic radiotherapy-associated VVFs require longer periods of drainage.
Acidification of urine to diminish risks of cystitis, mucus production, and
formation of bladder calculi is a consideration for patients with an indwelling
catheter. Vitamin C at 500 mg orally 3 times per day may be used to acidify
urine. Alternatively, methenamine mandelate at 550 mg plus sodium acid phosphate
at 500 mg 1-4 times daily also can be administered to achieve urine
acidification. Estrogen replacement therapy in the postmenopausal patient may assist with
optimizing tissue vascularization and healing (see Control of postoperative bladder spasms: Urised is effective for control of
postoperative bladder spasms. It is a combination of antiseptics (methenamine,
methylene blue, phenyl salicylate, and benzoic acid) and parasympatholytics
(atropine sulfate, and hyoscyamine sulfate). Antibiotic therapy: The use of antibiotic therapy postoperatively is
controversial. Many physicians administer oral antibiotic prophylaxis to
patients with VVF postoperatively until the Foley catheter is discontinued.
Others check closely for the development of a urinary tract infection and
administer antibiotic therapy when urine cultures are positive for bacterial
growth. Close follow-up and prompt evaluation for any urinary tract infections
and antibiotic therapy, when indicated, are mandatory. Minimizing Valsalva maneuvers: Stool softeners and a high-fiber diet
postoperatively minimize Valsalva maneuvers in the patient. Examinations: Avoid pelvic and speculum vaginal examinations during the first
4-6 weeks postoperatively because the tissue is delicate. Pelvic rest: Prohibit coitus and tampon use for a minimum of 4-6 weeks. Other
authors advocate strict pelvic rest for 3 months.
Integral to all major surgeries are risks of
infection; hemorrhage; injury to other organs, particularly the ureters;
surgical failure of fistula repair; possible new fistula formation;
thromboembolism; and death. Preoperatively, patients should be informed of the
possibilities of sexual dysfunction or dissatisfaction, new-onset incontinence,
and the progression of preexisting urge and/or stress incontinence symptoms.
Authors also mentioned recommendations for cesarean delivery for subsequent
pregnancies. Abdominal approach procedures carry additional risks of abdominal and pelvic
adhesions. Vaginal approach procedures carry increased risks of dyspareunia,
tenderness at the site of the donor Martius graft, and diminished vaginal length
and caliber.
See
COMPLICATIONS
¡@
OUTCOME AND PROGNOSIS
¡@
BIBLIOGRAPHY
¡@
[Medline].