Surgical procedures for urinary stress incontinence have been described for more than a century. Some surgeries, whether for their apparent ingenuity or ease of performance, have seen great periods of popularity, only to later fade from use. Others have seen modifications over decades of use. Still others have stood the proverbial test of time and remain current procedures commonly used by pelvic surgeons around the world. One fact has remained constant: there is little agreement about which procedure is the best. Most surgeons would agree that the type of surgery used should be determined by the specific qualities that make up a particular patient's urinary stress incontinence. There are major academic centers around the United States that recommend a sling procedure as the primary surgery for stress incontinence, whereas other centers prefer a retropubic suspension. This confusing scenario has arisen because randomized controlled trials comparing these procedures have been few and far between. To date, seven such studies have been published. They differ in quality, sample size, and follow-up, but an examination of the results at least permits one to begin to formulate an educated understanding of the strengths and weaknesses of the particular operations.

Since the early 1900s, hundreds of operations have been described to treat genuine stress incontinence (GSI). They can be categorized into six basic surgical approaches: (1) suburethral (Kelly); (2) needle suspension procedures; (3) sling procedures; (4) retropubic procedures; (5) urethral bulking procedures; and (6) artificial sphincter.

A brief review of the literature encompassing these various procedures reveals that the longevity of cure provided by operations in the first two groups makes them unsuitable for primary surgical treatment of GSI.^{1, ,2, ,3, ,4, ,5, ,6} Sling procedures are reviewed in detail in a separate chapter. In this chapter, we concentrate on the retropubic procedures, which represent a significant proportion of operations currently performed for urinary stress incontinence. We also briefly describe the use of the artificial sphincter, although this is generally used in very specific and less common situations.

The avascular potential space behind the pubic symphysis and in front of the bladder is the space of Retzius. Its lateral boundaries are formed by pubic bone and obturator internus muscle. One can easily obtain access to this area by dividing the rectus abdominis muscle along its midline raphe, spreading the bands of muscle laterally and sliding a hand between the rectus muscle and the peritoneum in the direction of the pubic symphysis. The space is further developed with blunt dissection, gently dividing the fine adhesions and fatty connective tissue that normally fill this potential space. This can also be performed with the peritoneum open if another procedure such as hysterectomy is to be performed concomitant to continence operation. As the space is developed, tissue is peeled from the posterior pubic ramus from the superior to inferior aspect and laterally to the extent of the obturator internus muscles.

Once such exposure has been created, landmarks should be noted (Fig. 1). The floor of the space of Retzius is formed by the anterior vagina, which is interwoven to a hammock-like layer of musculofascial tissue (the poorly defined endopelvic and pubocervical fascia) running along the posterior symphysis and from side to side, inserting at the level of the arcus tendineus fasciae pelvis (also called the white line). Sitting on top of the anterior vagina and endopelvic fascia in the midline, beneath the symphysis, is the proximal urethra. Just lateral to it along both sides are the posterior pubourethral ligaments. Proximal to this is the urethrovesical junction and extraperitoneal portion of the bladder. The lateral margins of the bladder can be easily delineated by temporarily filling the bladder through an indwelling Foley catheter. The floor of the space of Retzius contains a large plexus of thin-walled veins called the plexus of Santorini. Bleeding from this plexus may be encountered during Burch and Marshall-Marchetti-Krantz (MMK) procedures and is almost always controlled with direct pressure. The white line is a tendinous arch that originates at the posterior pubic symphysis and runs along the internal aspect of the obturator internus muscle to insert at the ischial spine. The attachment of the endopelvic/pubocervical fascia to the white line maintains the anterolateral support of the vagina. Detachment of this fascia from the white line results in paravaginal defects.

Careful inspection of the pubic ramus reveals other important structures. Lateral to each pubic tubercle, just beneath the superior margin of the pubic ramus, is the iliopectineal line and Cooper's ligament. As one continues to move along the ramus in a lateral and posterior direction, the external iliac artery and vein are encountered as they pass beneath the inguinal ligament into the thigh to become the femoral artery and vein. Still more lateral and posterior is the obturator neurovascular bundle, which exits the pelvis through the obturator foramen. Often, just before the obturator vessels enter the obturator foramina, they give off branches that usually anastomose with the external iliac vessels. These are known as the aberrant or anomalous obturator vessels. Injury to the obturator neurovascular bundle and the aberrant vessels can occur if retractors are placed with excessive traction, tearing the vessels between the instruments and the pubic bone.

Sling procedures

Sling operations are some of the oldest anti-incontinence procedures performed. They evolved from an attempt to both support the urethra and recreate or augment urethral sphincter tone lost to injury or atrophy. These procedures are designed to narrow the urethra, provide urethral support, increase urethral closure pressure (UCP) by external compression, and restore posterior urethrovesical angle.⁷ The concept that sling procedures increase UCP has been called into question because some follow-up studies have failed to show a significant increase.^{8, 9} Although there is no typical patient when it comes to treating urinary incontinence, many urogynecologists reserve sling procedures for patients who have had previous unsuccessful anti-incontinence procedures. These patients are often severely incontinent and display little or no urethral mobility during increases in intra-abdominal pressure. Their UCP is generally low (<20 cmH₂O), and they have low Valsalva leak-point pressures (<60 cmH₂O). These are the generally accepted diagnostic criteria for intrinsic sphincter deficiency.

The modern sling procedure evolved from an operation described by Giordano in 1907, in which gracilis muscle flaps were transplanted near the urethra. In 1917, surgeons Goebell, Frankenheim, and Stoeckel developed a sling procedure using pyramidalis muscle with attached rectus fascia. After the muscle bellies were dissected free to the level of the symphysis, the ends were passed behind the pubic bone and sutured below the urethra. The vesical neck was also plicated. The 1930s saw the decrease of musculofascial slings and the advent of slings composed only of fascia. In 1942, Aldridge described an operation that closely resembles a variant of the sling procedure still performed today. He dissected bilateral strips of rectus fascia from the anterior aspect of the muscle, leaving the medial portions attached to the muscle. The strips of fascia were then tunneled through the muscle, passed behind the symphysis, and sutured below the urethra. Ensuing years have brought the use of synthetic materials for slings and the use of suture bridges and patch slings.¹⁰ Current slings are usually composed of fascia, either cadaveric donor tissue or tissue harvested from the patient herself at the time of operation.

Cure rates with sling procedures are reported to be 70–95%.^{11, 12} The results are similar regardless of type of sling material used. The variability arises from differences in technique, definition of cure, and length of follow-up. Although reports of cure rates are abundant, documentation of early and late complications is poor. In addition to the risks of hemorrhage, infection, and injury to local organs, one must consider the effect of the procedure on voiding. There is a 2–30% risk of severe voiding dysfunction or retention.¹³ This estimation is based primarily on observation and is in need of further study. Detrusor instability and various irritative bladder symptoms such as frequency and urgency occur in anywhere from 2% to 50% of patients.¹⁴ Unfortunately, it is difficult to predict which patients will have these complications. These symptoms often lessen with time and usually can be treated pharmacologically. Less common complications include erosion of the sling material (more common with synthetic slings), fistula or sinus tract formation, nerve injury or entrapment, and abscess formation. As previously mentioned, because of the perceived higher rate of potential complications, many pelvic surgeons continue to perform a retropubic urethropexy as their primary anti-incontinence surgery.

Retropubic operations

Both operations described in this section have as their common goal the identification of strong periurethral tissues near the vesicle neck and the suturing of these tissues to a supportive structure attached to the pubis. This serves to return the bladder neck to an intra-abdominal location so that it sees the same transmural pressures as the bladder. Urethral closure pressure has been shown both to increase and decrease after these procedures and is thus not thought to play a role in their mechanism of achieving continence.^{15, 16}

MARSHALL-MARCHETTI-KRANTZ

In 1949, Marshall reported the empirical observation that suturing the periurethral tissues to the pubic bone alleviated urinary stress incontinence after examining a patient with iatrogenic incontinence after vesical neck resection.¹⁷ The original description called for number-1 chromic suture, but both the MMK and Burch procedures are now typically performed with permanent sutures. Access to the retropubic space is obtained as described. The bladder neck is identified by placing the nondominant hand in the vagina and palpating the Foley bulb with the index and middle fingers. While elevating the vaginal fingers, a Kittner dissector is used to place countertraction on the fatty tissue overlying the periurethral fascia (Fig. 2A inset). A gentle sweeping motion easily cleans the fat away, revealing the white fascia below. This dissection permits the surgeon to take good bites of tissue and fosters adherence of the periurethral tissue to the back of the symphysis. Elevation of the vaginal finger permits the operator to place a figure-eight, full-thickness (excluding the vaginal epithelium if possible) bite of tissue (see inset of Fig. 2B). A single suture is placed on both sides of the urethrovesical junction in this fashion. Each suture is then fixed to the periosteum or fibrocartilage of the pubic bone in such a way that the vesical neck is barely brought in contact with the pubic symphysis (Fig. 3). Injury to the bladder and ureters is ruled out with cystoscopy, suprapubic telescopy, or intentional cystotomy. Because postoperative voiding efficiency is unpredictable, a suprapubic catheter is the preferred method of bladder drainage.

BURCH COLPOSUSPENSION

The Burch procedure was described in 1962 after the originator of the procedure was unable to find adequate periosteum in an elderly patient in whom he was trying to perform an MMK procedure.¹⁸ The retropubic space is entered and prepared as described for the MMK procedure. Two permanent figure-eight sutures are placed on either side of the bladder neck. The proximal sutures are placed 2 cm lateral to the bladder neck, and the distal sutures are placed 2 cm lateral to the proximal third of the urethra (see Fig. 2A). The ends of each suture are then passed through Cooper's ligament either by the use of a curved Mayo needle or by using double-armed suture. Once all sutures are placed, the surgeon elevates the vagina while an assistant ties the sutures down with the knots on top of Cooper's ligament (see Fig. 2B). The distal sutures are tied first. When complete, the surgeon should be able to easily pass two fingers between the pubic bone and the urethra. Suture bridges are not problematic and are commonly present. An intravesical assessment is recommended to ensure that no bladder or ureteral injury has occurred.

Laparoscopic approaches to the Burch procedure have also been described. Retrospective and observational studies suggest cure rates are similar to open procedures.¹⁹ Three prospective trials comparing these two techniques have been published. Burton in 1994 and Su in 1997 found the open approach to be superior (97% vs. 73% and 96% vs. 80%, respectively).^{20, 21} Fatthy and associates reported similar cure rates for the open procedure compared to a modified laparoscopic approach followed-up to 18 months (85% vs. 88%) and found less morbidity and a shorter hospital stay in the laparoscopic group.²² Unfortunately, comparisons are difficult to make between the open and laparoscopic approaches secondary to a multitude of technical variations (aside from the actual approach) from the traditional procedure.

Cure rates for the retropubic procedures are similar, 65–90%, at 1–10 years.^{23, 24} Indeed, the single randomized prospective trial that compared the Burch with the MMK procedures found no significant difference in cure rate.²³ These procedures have stood the test of time, and there are long-term success rate data. This is particularly true for the Burch procedure, which is the more studied of the two operations. It seems that over time, the cure rate of the retropubic suspensions decreases steadily from 90% at 1 year to about 70% by 10 years postoperatively, before reaching a plateau at 65–70% in patients who have been followed-up more than 20 years.²⁴ Complications of retropubic procedures are similar to sling procedures with some differences in incidence. Because more dissection is needed for retropubic procedures compared with sling procedures, one would anticipate a higher incidence of infectious and hemorrhagic complications, but less worry about erosions and sinus tract formation. The risk of de novo detrusor instability is reported from 5% to 27%, but Alcalay and associates have reported on patients with 10-year follow-up with an incidence of 14%. They have also reported voiding dysfunction in 22%.²⁵ One complication unique to retropubic suspensions is the occurrence of osteitis pubis, which occurs in up to 2.5% of patients undergoing the MMK procedure. Long-term studies of the Burch procedure have shown a significant incidence of prolapse formation. Rectocele has been noted in 11–25% and enterocele in 4–10% of patients followed-up 10 to 20 years.²⁴

Paravaginal repair

A discussion of the paravaginal repair is included here because it is a retropubic procedure. It should not be considered a primary anti-incontinence operation. The goal of this operation is to repair a specific anatomic defect: separation of one or both sides of the endopelvic fascial hammock that normally inserts at the arcus tendineus fasciae pelvis (white line) at the pelvic sidewall. In the past, it has been used as a means to treat stress incontinence.²⁶ Although it will make some women continent, presumably by elevation of the bladder neck, it does not produce a lasting result. As Colombo and colleagues have shown, the Burch procedure is clearly superior for the treatment of incontinence.²⁷ In patients who have paravaginal defects with resultant cystocele associated with stress incontinence, a procedure called paravaginal plus has been described. In this procedure, the paravaginal defects are repaired as described in the next paragraph, and Burch colposuspension sutures are placed as described previously (Fig. 4). To perform the abdominal paravaginal repair, one gains access to the retropubic space as described previously. The ischial spine and attached arcus tendineus fascia should be identified. Paravaginal defects typically are readily apparent as a detached portion of the vagina from the white line (Figs. 5A and 5C). Using the nondominant hand, the surgeon elevates the anterolateral vaginal sulcus on the side of the defect. A full-thickness (excluding epithelium) figure-eight bite of vaginal tissue is taken with permanent suture near the vaginal apex and then fixed to the white line or fascia of the obturator internus muscle 1 to 2 cm from the ischial spine. This is tied down. Then, proceeding distally, three or four similar sutures are placed such that the final suture is as close as possible to the pubic ramus (see Fig. 5B).

Artificial sphincter

The use of an artificial sphincter to treat urinary stress incontinence may be appropriate in some cases of severe urine loss. It is an implantable device that occludes the urethra but can be voluntarily opened, permitting the patient to empty her bladder. Because of the technical difficulty encountered in placing such a device and the fairly limited pool of appropriate patients, this means of treatment has not gained wide acceptance.

Artificial urinary sphincters were first used in 1972. Several modifications have resulted in advanced devices consisting of a cuff, a pressure-regulating balloon, and a control pump (Fig. 6). The cuff is placed around the bladder neck, and the balloon is fixed into the retropubic space. The pump is placed subcutaneously into one of the labia majora (Fig. 7). The cuff is normally in the activated state, in which it is inflated, thus squeezing the bladder neck closed. The balloon sees changes in intra-abdominal pressure and incrementally regulates the pressure applied to the cuff. When the patient needs to void, she squeezes the pump located in her labium, which deactivates the cuff. The cuff automatically begins to reinflate, but takes 3 minutes to do so, permitting the patient to empty.

The complex nature of the device makes unmotivated and nondexterous patients poor candidates for this intervention. Other contraindications include bladder overactivity that cannot be controlled with medication or biofeedback and high-grade vesicoureteral reflux. There is also the risk of infection, erosion, and device malfunction. Short-term success rates with the artificial sphincter are reported to be 68–100%, but mechanical complication rates are as high as 21%.^{14, 28} Furthermore, women appear to be more susceptible to erosions with this procedure than men, with up to 56% of women experiencing this complication compared to 23% in men.²⁹ A recent series of 68 women who were followed-up for a median of 12 years reported an overall continence rate of 81%, but only 25 (37%) had the original device still in place, 17% had the device replaced for mechanical failure, and 46% had the device removed for infection of erosion.³⁰

In reviewing the literature for surgical management of stress urinary incontinence, one can quickly make two conclusions: (1) a vast number of different procedures to treat stress incontinence have been described over the past 100 years and (2) the usefulness of the literature is called into question because of the difficulty one has in interpreting the results. Most of the studies are small and retrospective, technique is not clearly described or changes from study to study, follow-up is poor, and the outcome data are difficult to understand. During the past 26 years, there have been seven different groups who have completed nine randomized controlled trials comparing well-established surgical approaches to urinary stress incontinence. As one reviews this important series of articles, the current surgical approach becomes more clearly defined. Because this chapter is intended to concentrate on technique, we only briefly review this literature.

In 1977, Henriksson and Ulmsten conducted a randomized comparison of the MMK and sling procedures.¹⁵ Only 30 patients were studied, and randomization was by alternate selection. In their hands, there was no difference in cure rates: 100%. In addition to the small sample size, the study suffers from a vague description of the authors' preoperative assessment and a follow-up period of only 6 months. One is left with little to conclude except that over a short period of time, these two operations are comparable.

Stanton and Cardozo compared anterior repair with the Burch procedure among 50 patients.² This study marked an improvement in clearly describing preoperative assessment, which was quite thorough, but it also suffers from poor follow-up of only 6 months. Furthermore, some of the patients in this study also underwent reconstructive procedures for correction of prolapse in addition to anti-incontinence surgery. Objective cure rates, however, were significantly different even at 6 months: 84% for the Burch procedure and 36% for anterior repair.

Between 1988 and 1995, Bergman and associates published two articles describing a randomized comparison of Burch, anterior repair, and needle suspension (modified Pereyra) procedures that represent the strongest articles among this group.^{3, 4} These studies include the largest sample size (107 at 1 year and 93 at 5 years) and the longest follow-up of any of the randomized studies of anti-incontinence procedures. At 5 years, the Burch procedure was found to have a significantly higher cure rate (82%) than either anterior colporrhaphy (37%) or the modified Pereyra procedure (43%). Largely because of the preceding studies, it started to become clear that both anterior colporrhaphy and needle suspension procedures did not produce lasting cures. Most urogynecologists agree that neither of these operations should be considered a first-line anti-incontinence procedure except in specific situations (i.e., elderly, medically complicated patient at significant anesthesia or surgical risk precluding more invasive procedure). In 1989, Hilton compared a sling procedure with the Stamey needle suspension in 20 patients with results that would seem to argue against the inadequacy of needle suspension operations.⁵ He found no significant difference in cure rates between the sling and Stamey procedures at 24 months' follow-up. The cure rate was higher in the sling group (90% vs. 80% at 1 year and 90% vs. 70% at 2 years). Perhaps if these patients were followed-up longer or if the sample was larger, a significant difference would have been revealed.

For most pelvic surgeons, the question of which anti-incontinence procedure to perform comes down to retropubic urethropexy (Burch or MMK) or sling procedure. In terms of cure rates, all the procedures have acceptable long-term results. But the question of which procedure among retropubic urethropexy operations to perform has largely been left to surgeon preference. Colombo and coworkers actually compared the Burch procedure to the MMK procedure and found no significant difference in cure rate (Burch 80%, MMK 65%) at a mean follow-up of 3.3 years.²³ Another Italian center compared these two operations and found the MMK to be superior to the Burch procedure (93% vs. 53%), but this was among a group of women with low-pressure urethras (intrinsic sphincteric deficiency).³¹ When Richardson reintroduced the theory of paravaginal defects as an etiology of pelvic relaxation, many surgeons began performing paravaginal repairs to address stress incontinence. This operation is intended to fix a specific anatomic defect and in turn has not emerged as a particularly useful anti-incontinence procedure. Colombo and coworkers compared the Burch procedure with the paravaginal repair among 36 patients and found the Burch procedure to be superior (100% vs. 61% at 2.2 years).²⁷ Kammerer-Doak and associates and compared the Burch procedure to modified anterior colporrhaphy.⁶ As expected, they found the Burch procedure to be superior at 1 year (89% vs. 31% cure rate). Colombo and associates also compared these two procedures in a randomized fashion and reported a higher cure rate in the Burch group (74% vs. 42%) but found a higher percentage of recurrent cystocele (34% vs. 3%).³² These studies provide further evidence of the inadequacy of anterior repair to treat urinary stress incontinence (Table 1).

Table 1. Summary of the randomized comparative trials: rates of objective cure retropubic procedures

MMK, Marshall-Marchetti-Krantz.

Although no pelvic surgeon can say that any particular operation is the gold standard for treating incontinence, one can at least tailor one's approach based on experience and what little evidence exists concerning surgery for urinary incontinence. Using this evidence, one generalized conclusion that can be drawn is that sling procedures and retropubic suspensions currently offer the best hope for curing GSI. We cannot currently conclude that one of these procedures is better than another. The Burch procedure is probably the most studied of these operations, and we can at least offer our patients the benefit of what knowledge we have collected about this procedure. Traditionally, slings have been reserved for patients with intrinsic sphincter deficiency and fixed scarred urethras because they tend to cause more postoperative voiding dysfunction than retropubic procedures. As experience grows and surgeons report their findings, the idea of a sling as a primary incontinence procedure gains support. One thing is clear: we all must work more diligently to carefully follow-up and document our patients' progress and we must honestly report our findings. Such actions will advance the skills and knowledge we offer our patients and permit us to do what is best for them.

As time passes, we may see our surgical armamentarium evolve. The laparoscopic Burch procedure, although poorly studied, is a viable approach. The past 10 years have seen the advent of the tension-free vaginal tape procedure. This quick and minimally invasive operation is quickly becoming very popular. There is a well-designed, ongoing, randomized, controlled trial comparing the tension-free vaginal tape to the open Burch procedure, and early data suggest that the two procedures are comparable.³³ If our experience over the past century has taught us anything, it should be that these questions must encourage us to pursue the study of this complex problem further and that there is not likely to be one operation that is appropriate in all clinical situations. We must always be mindful of the basic tenets of good surgical technique, one of which is that careful selection of both patient and technique yields the best results with the fewest complications.