Randomised comparison of Burch colposuspension versus anterior colporrhaphy in women with stress urinary incontinence and anterior vaginal wall prolapse


  • Mario Colombo,

    Senior Registrar, Corresponding author
    1. Division of Gynaecology, Department of Obstetrics and Gynaecology, University of Milan, San Gerardo Hospital, Monza, Italy
    Search for more papers by this author
  • Domenico Vitobello,

    Senior House Officer
    1. Division of Gynaecology, Department of Obstetrics and Gynaecology, University of Milan, San Gerardo Hospital, Monza, Italy
    Search for more papers by this author
  • Fabio Proietti,

    Senior House Officer
    1. Division of Gynaecology, Department of Obstetrics and Gynaecology, University of Milan, San Gerardo Hospital, Monza, Italy
    Search for more papers by this author
  • Rodolfo Milani

    Associate Professor
    1. Division of Gynaecology, Department of Obstetrics and Gynaecology, University of Milan, San Gerardo Hospital, Monza, Italy
    Search for more papers by this author

Correspondence: Dr M. Colombo, via Giusti 19/a, 21053 Castellanza VA, Italy. Email: emmecol@tin.net


Objective To compare the Burch colposuspension and the anterior colporrhaphy in women with both stress urinary incontinence and advanced anterior vaginal wall prolapse (cystocele).

Design Prospective randomised study.

Setting Secondary referral centre, Urogynaecology Unit, San Gerardo Hospital, Monza, Italy.

Sample Seventy-one women undergoing surgery for primary genuine stress incontinence and concurrent grade 2 or 3 cystocele (descending at or outside the vaginal introitus).

Methods Full urodynamic investigation performed pre-operatively and repeated six months after surgery. Clinical follow up continued for 8 to 17 years.

Main outcome measures Subjective (patient history) and objective (negative stress test result) cure of stress incontinence. Assessment of cystocele recurrence.

Results Thirty (86%) of the 35 evaluable women who had the Burch colposuspension and 17 (52%) of the 33 evaluable women who had the anterior colporrhaphy were subjectively cured (OR 5.6, 95% CI 1.6 to 21.6; P= 0.005). Objective cure rates were 74% (26 of 35) and 42% (14 of 33), respectively (OR 3.9, 95% CI 1.3 to 12.5; P= 0.02). A recurrent cystocele of grade 2 or 3 with or without prolapse at other vaginal sites was recorded in 34% (12 of 35) and 3% (1 of 33) of women, respectively (OR 16.7, 95% CI 2.0 to 368.1; P= 0.003).

Conclusions The Burch colposuspension was better in controlling stress incontinence but it lead to an unacceptable high rate of prolapse recurrence. The anterior colporrhaphy was more effective in restoring vaginal anatomy but it was accompanied by an unacceptable low cure rate of stress incontinence. Neither of the two operations is recommended for women who are suffering from a combination of stress incontinence and advanced cystocele.


Anterior vaginal wall prolapse (cystocele) and stress urinary incontinence can present together, but are not directly related. Although abdominal urethropexies are better than vaginal repairs in controlling stress incontinence1, the debate on which surgical approach should be used to treat both conditions has always had clinical relevance. A vaginal repair is usually recommended for women with genital prolapse only2. When stress incontinence and significant prolapse present together a retropubic colposuspension combined with abdominal hysterectomy and/or posterior repair is advised2–5. Contrary to this, other authors suggest anterior colporrhaphy6,7 or vaginal retropubic urethropexy8 as effective procedures for treating stress incontinence with or without anterior vaginal wall prolapse. Additionally, an abdominovaginal approach to vaginal repair and simultaneous Burch colposuspension is proposed5,9.

The decision for the route of reconstructive pelvic surgery should be based on scientific data rather than on a surgeons' familiarity with various procedures. Although the Burch colposuspension and the anterior colporrhaphy are widely used procedures, to our knowledge only one randomised trial9 compared the results of these two operations in women with both stress incontinence and advanced cystocele. Unfortunately, the authors reported only the short term functional outcome with no mention of post-operative prolapse recurrence.

We compared in a randomised fashion the procedures of Burch colposuspension and anterior colporrhaphy in women with primary genuine stress urinary incontinence associated with severe cystocele. The two procedures were compared for long term subjective and objective cure rates of stress incontinence and incidence of cystocele recurrence.


Starting in October 1981, all women admitted to our department with the diagnosis of both genuine stress urinary incontinence and advanced anterior vaginal wall prolapse (cystocele) were considered for enrolment. Our standard pre-operative assessment was described in full elsewhere10. It included history, grading of stress incontinence by a clinical score, urine culture, physical examination, stress test, water urethral profilometry at rest, cotton swab test and subtracted provocative water cystometry at a filling rate of 100 mL/min. Urodynamic evaluations were performed with no prolapse reduction.

By the clinical score10 stress incontinence was graded as mild (score 1 to 2), moderate (score 4), or severe (score 8). With grade 2 or 3 cystocele, uterine prolapse or rectocele the lead part descended at or outside the vaginal introitus under maximum strain with the woman in the supine position. Urethrovesical junction hypermobility was defined as a maximum straining angle > 30 degrees (positive cotton swab test result). The diagnosis of genuine stress incontinence was established by visual documentation of urine leakage (positive stress test result) and no evidence of detrusor instability at cystometry.

Inclusion criteria were:

  • 1Genuine stress urinary incontinence.
  • 2Cystocele of grade 2 or 3.
  • 3Positive cotton swab test result (urethrovesical junction hypermobility).

Exclusion criteria were:

  • 1Detrusor instability.
  • 2Previous anti-incontinence or prolapse surgery (retro-pubic or vaginal).
  • 3Concomitant pelvic diseases requiring laparotomy.
  • 4Medical conditions making a patient at high surgical risk for abdominal surgery.

The day before surgery one member of the surgical staff assigned each subsequent enrolled woman to receive either the Burch colposuspension or the anterior colporrhaphy according to an open list of computer-generated random numbers.

In the group of women with Burch colposuspension, total abdominal hysterectomy was performed first and the vaginal vault was anchored to the stumps of the uterosacral ligaments. Next, if a deep cul-de-sac of Douglas was observed, Moschcowitz culdoplasty11 was performed using three to five purse-string sutures of chromic catgut number 0. Our standard technique for Burch colposuspension has previously been described in detail10. It was accomplished with three to four couples of number 0 nonabsorbable polybutylate-coated polyester sutures (Ethibond, Ethicon SpA, Pratica di Mare, Rome, Italy) placed between the paravaginal fascia and the ipsilateral ileopectineal ligament. The most distal pair of suture was inserted at the level of the midurethra, the second at the urethrovesical junction and the remaining one or two pairs were placed cephalad along the bladder base. When necessary, vaginal posterior colporrhaphy with perineorrhaphy12 was performed after the closure of the abdominal wall. A vertical incision was performed in the posterior vaginal wall and the levator ani muscles were exposed and sutured together using interrupted stitches of number 1 chromic catgut. Excess vaginal mucosa was trimmed before closing the vagina.

In women scheduled to receive the anterior colporrhaphy, vaginal hysterectomy was performed first with appropriate shortening of the uterosacral ligaments. Douglas obliteration (high closure of the peritoneal cavity) was accomplished by a single running purse-string suture of chromic catgut number 0. This suture included the anterior peritoneum, the stumps of the uterosacral ligaments and the anterior rectal serosa. The peritoneum of the pouch of Douglas was not removed. Anterior colporrhaphy was performed in the standard manner through a midline vaginal incision extended from just below the external urethral meatus towards the edge of the vaginal vault. The vaginal wall was then separated from the underlying pubocervical fascia and the urethrovesical junction was fully mobilised. Beginning adjacent to the external urethral meatus, a series of interrupted stitches of number 0 chromic catgut were placed in the pubocervical fascia as high and deep as possible on either side of the proximal urethra, the urethrovesical junction, the trigone and the base of the bladder. The sutures were tied to approximate and plicate the pubocervical fascia at the midline. Care was taken to elevate and fixate the urethrovesical junction and proximal urethra to a higher position, giving differentially better support to the posterior aspect of the urethra than to the bladder base8. Excess vaginal mucosa was trimmed and the incision closed. Thereafter, the two sutures (one on each side) used to ligate the uterosacral ligaments during vaginal hysterectomy and the suture used for Douglas obliteration were sewn to the vaginal cuff to provide its closure and to give further support. One arm of each suture was sewn into the anterior portion of the cuff and one into the posterior portion. Posterior colporrhaphy with perineorrhaphy12 was performed in all cases using the same technique described above.

Irrespective of the type of operation, a suprapubic catheter was used to drain the bladder. It was clamped on day two and the women were asked to attempt to void spontaneously. The residual volume was determined three times per day and the catheter was removed once the residual urine was 50 mL or less each determination for two consecutive days (resumption of spontaneous voiding). Women failing at discharge to reach such a condition were sent at home with the suprapubic catheter in place. They were seen subsequently at the outpatient office at intervals of a few days and the catheter was removed when the resumption of spontaneous voiding was achieved.

Six months after surgery the women were evaluated by history, physical examination, stress test, cotton swab test, water profilometry at rest and cystometry. Thereafter they were seen annually for history, physical examination and stress test. When a woman failed to attend her annual visit she was periodically recalled by telephone and invited to return for evaluation. No telephone interviews were conducted. From early 1987, while the follow up of this study was ongoing, we started to perform the stress test with prolapse repositioning. This was done with the prolapse reduced by a posterior Sims' speculum placed in the anterior vaginal fornix without urethral obstruction. This eliminated the possibility that women with recurrence of genital prolapse could have a negative stress test result due to a kinking effect on the urethra.

Cure was defined both subjectively (no incontinence episodes by history) and objectively (negative stress test result). Recurrent prolapse was defined as any descent of grade 2 or 3 (cystocele, vault prolapse or rectocele), even if the patient had no symptoms. Grade 1 prolapse (descending to the level of the midvagina) was not considered as a recurrence because it rarely accounts for significant symptoms and does not warrant operative correction. During physical examinations physicians estimated the vaginal length by comparing the length from the introitus to the vault with the length of their forefinger. Dyspareunia was specifically investigated after surgery as mild (sexual activity maintained) or severe (complete loss of coital function).

Unless otherwise specified all terminology conforms to that recommended by the International Continence Society13. Statistical analysis included either unpaired Student's t test or Mann-Whitney U test for independent groups of data. The parametric Student's t test was used for normally distributed continuous variables. The significance of contingency tables was assessed by χ2 test with Yates' correction and by odds ratios with 95% confidence intervals. The Fisher exact test was substituted for the χ2 test for small sample sizes. All tests were two-tailed. Statistical significance was determined at the 0.05 level.


During the period from October 1981 to November 1986, eighty-nine consecutive women were admitted to undergo surgery for primary genuine stress urinary incontinence and advanced anterior vaginal wall prolapse. Eighteen women met one or more of the exclusion criteria, while 71 were enrolled. Of these 37 women were randomised to receive Burch colposuspension and 34 anterior colporrhaphy. All the enrolled women had pure genuine stress incontinence with concurrent cystocele grade 2 or 3 and a positive cotton swab test result. No woman had a maximum urethral closure pressure < 30 cm H2O at rest profilometry. Two women submitted to the Burch colposuspension and one submitted to the anterior colporrhaphy were lost to follow up and were therefore excluded from the analysis of the results, leaving 68 evaluable women in the trial (35 with Burch colposuspension and 33 with anterior colporrhaphy).

There were no significant differences between the two groups with respect to pre-operative clinical and urodynamic indices (Table 1). The additional surgical procedures performed are summarised in Table 2. In either arm there was no serious intraoperative or early post-operative complication except for a small inadvertent cystotomy during an abdominal hysterectomy which was recognised and sutured at surgery. The mean hospital stay was 6.7 (1.8 days (median 6 days, range 5 to 12 days) after the Burch colposuspension and 6.9 (1.5 days (median 7 days, range 5 to 11 days) after the anterior colporrhaphy (P= 0.42, Mann-Whitney U test). Twenty-six (74%) and 21 (64%) women, respectively, resumed spontaneous voiding before discharge (P= 0.49, χ2 test). The remaining 21 women (9 with Burch colposuspension and 12 with anterior colporrhaphy) resumed spontaneous voiding within two weeks after discharge, except one patient in each group who resumed voiding within four weeks.

Table 1.  Population. Values are given as n (%), mean [SD] or median (range].
 Burch colposuspension (n= 35)Anterior colporrhaphy (n= 33)
Age (years)54.9 [8.6]55.7 [10.3]
Parity3 (0.5)3 [0.61]
Vaginal deliveries ≥ 4000 g9 (26)10 (30)
Forceps deliveries or vacuum extraction6 (17)4 (12)
Body mass index (kg/m2)25.5 [3.7]26.2 [3.4]
Postmenopausal23 (66)22 (67)
Stress incontinence  
 Mild (score 1 to 2)15 (43)14 (43)
 Moderate (score 4)11 (31)12 (36)
Severe (score 8)9 (26)7 (21)
Grade 2 cystocele20 (57)17 (52)
Grade 3 cystocele15 (43)16 (48)
Grade 2 uterine prolapse6 (17)6 (18)
Grade 3 uterine prolapse8 (23)7 (21)
Grade 2 rectocele4 (11)3 (9)
Grade 3 rectocele3 (9)5 (15)
Maximum urethral closure pressure (cm H2O)74.1 [26.9]72.6 [24.5]
Urethral functional length (mm)26.8 [6.2]27.2 [8.3]
Table 2.  Additional procedures. Values are given as n (%).
 Burch colposuspension (n= 35)Anterior colporrhaphy (n= 33)
  1. *Moschcowitz procedure with the Burch colposuspension and high closure of the peritoneal cavity with the anterior colporthaphy.

Total abdominal hysterectomy35 (100)0
Vaginal hysterectomy033 (100)
Anchoring of uterosacral ligaments to the vaginal cuff35 (100)33 (100)
Douglas obliteration*17 (49)33 (100)
Posterior colporrhaphy and perineorrhaphy12 (34)33 (100)

All 68 women underwent follow up for at least eight years after the operation. No woman was lost to follow up except for ten women who died of intercurrent disease 8 to 16 years after surgery (four had the Burch colposuspension and six had the anterior colporrhaphy). All women were expected ten years post-operatively: Ninety-four percent (33 of 35) were evaluated after the Burch colposuspension was performed and 97% (32 of 33) were evaluated after the anterior colporrhaphy was performed. Fifteen years post-operatively 19 (68%) of 28 expected women were evaluated after Burch colposuspension and 17 (68%) of 25 expected women were evaluated after anterior colporrhaphy. The mean follow up was 14.2 [2.7 years, range 9–17 years (Burch colposuspension)] and 13.9 [2.1 years, range 8–17 years (anterior colporrhaphy)].

Subjective and objective cure rates are reported in Table 3. Twenty-four (92%) of 26 women who were objectively cured following Burch colposuspension had a negative cotton swab test result (maximum straining angle ≤ 30°) after surgery. The corresponding figure was 50% (7 of 14) after anterior colporrhaphy (OR 12.0, 95% CI 1.6 to 110.4; P= 0.004, Fisher exact test). Among women who still had a post-operative positive stress test result (objective failures), the cotton swab test had a negative result in 11% (one of nine) of those having received Burch colposuspension compared to 10% (2 of 19) having received anterior colporrhaphy.

Table 3.  Functional results. Values are given as n (%), unless otherwise indicated.
 Subjective CureObjective cure
Burch colposuspension (n= 35)30 (86)26 (74)
Anterior colporrhaphy (n= 33)17 (52)14 (42)
 Odds ratio5.63.9
95% CI1.6–21.61.3–12.5
χ2 testP= 0.005P= 0.02

Among the five women who had a subjective failure following Burch colposuspension, stress urinary incontinence recurred between 1 month and 12 years post-operatively (median three months). In four cases incontinence was mild (score 1 to 2) and the women required no further treatment. One woman had severe (score 8) recurrent incontinence and underwent a second colposuspension elsewhere seven years post-operatively. Among the 16 women who had been classified subjective failures following anterior colporrhaphy, stress incontinence recurred between one month and eight years post-operatively (median three months). In eight of these women incontinence was mild (score 1 to 2) and they required no treatment. The remaining eight women had moderate (score 4) or severe (score 8) stress incontinence and a second surgery was recommended for them. Five women decided against further therapy, while three underwent re-operation (a Marshal-Marchetti-Krantz urethropexy and two suburethral sling procedures were performed between 11 years and 15 years post-operatively).

Nineteen (54%) women had recurrent prolapse at any vaginal site after they underwent Burch colposuspension (Table 4). The median time of recurrence was three years (ranging from 3 months to 13 years). Two of these had also recurrence of mild stress incontinence requiring no treatment per se. Nine (53%) of the 17 women who had a concomitant Moschcowitz culdoplasty had a prolapse recurrence at any vaginal site compared with 10 (56%) of the 18 women who had not (P= 0.85, χ2 test).

Table 4.  Anatomic results. Values are given as n, n (%) or n [women for whom a second surgery was recommended because the prolapse was symptomatic].
 Burch colposuspension (n= 35)Anterior colporrhaphy (n= 33)
  1. *Odds ratio 38.0. 95% CI 4.6–830.1; p < 0., χ2 test.

No evidence of prolapse recurrence16 (46)*32 (97)*
Vault prolapse5 (14)0
 Grade 2  
With grade 2 cystocele10
 With grade 2 rectocele10
Grade 3  
 With grade 3 cystocele1 [1]0
 With grade 2 rectocele1 [1]0
 With grade 3 rectocele1 [1]0
Isolated cystocele10 (29)1 (3)
 Grade 271
 Grade 33[2]0
Isolated rectocele4 (11)0
 Grade 230
 Grade 31 [1]0

When women with prolapse recurrence after Burch colposuspension were divided with respect to pre-operative grade of cystocele, we found that recurrence rate was 35% (7 of 20) among women with a pre-operative cystocele of grade 2 and 80% (12 of 15) among those with a pre-operative cystocele of grade 3 (OR 7.4, 95% CI 1.3 to 49.4; P= 0.02, χ2 test), respectively. A second surgery was recommended for six women (two had pre-operative grade 2 cystocele and four had pre-operative grade 3 cystocele) because the condition was symptomatic (Table 4), but all refused.

Only one (3%) woman had recurrent prolapse (an isolated grade 2 cystocele) eight months after she underwent anterior colporrhaphy (Table 4). She remained asymptomatic (being also continent) and required no further treatment.

With respect to the anterior vaginal segment, 12 (34%) women treated with Burch colposuspension and one (3%) treated with anterior colporrhaphy had recurrent cystocele of grade 2 or 3 with or without prolapse at other vaginal sites (OR 16.7, 95% CI 2.0 to 368.1; P= 0.003, χ2 test). Symptomatic detrusor instability was diagnosed for the first time after surgery in one (3%) woman in each group. These two women were treated successfully with anticholinergics.

Post-operative vaginal lengths in all women, and in the subgroup of those who were sexually active, are listed in Table 5. Women operated on via the vaginal route had a statistically greater incidence of severe dyspareunia after surgery (Table 5) and one of them underwent a dilation plastic operation of the vaginal introitus elsewhere eight months post-operatively. Only two women treated with Burch colposuspension had mild or severe dyspareunia post-operatively and both had a concomitant posterior colporrhaphy with perineorrhaphy at surgery.

Table 5.  Sexual function after surgery. Values are given as n, n (%), mean [SD] or median {range]
 Burch colposuspensionAnterior colporrhaphy
  1. *P≤ 0.001 using Mann-Whitney U test.

  2. Odds ratio 14.3,95% CI 2.3–113.6; P= 0.001, χ2 test.

All women3533
 Post-operative vaginal length (cm)7.7 [1.4]4.5 [1.3]
 8 {5–10)*5 {3–7]*
Sexually active women2423
 Post-operative vaginal length (cm)7.9 [1.4]4.7 [1.2]
 8 (5–10)*5 {3–7]*
Sexual activity22 (92)†10 (44)†
 Maintained with no dyspareunia  
 Maintained with mild dyspareunia1 (4)4 (17)
 Lost for severe dyspareunia1 (4)9 (39)


Reconstructive surgery for genuine stress incontinence combined with severe anterior vaginal wall prolapse remains one of the most challenging aspects of surgical gynaecology. Although the belief that anterior colporrhaphy compares favourably with retropubic urethropexy for the treatment of the sole stress urinary incontinence1 has been dispelled, when stress incontinence and severe genital prolapse present together the optimal surgical approach remains open to discussion.

Our Burch colposuspensions had a long term subjective cure rate of 86% and an objective cure rate of 74%, which are comparable to those reported else-where5,10,14,15. The presence of a pre-operative or post-operative cystocele of grade 2 to 3 appeared to be unrelated to the control of continence. However, among our patients we documented an unacceptable high prolapse recurrence rate at any vaginal site of 54% (19 of 35) following abdominal surgery, including 34% (12 of 35) of women with anterior segment recurrence. We did not perform a Moschcowitz culdoplasty in all women of this group and this may partially explain our results. However, the recurrence of prolapse at any vaginal site was almost the same among women who had (53%) or had not (56%) a concomitant Moschcowitz culdoplasty. The fact that additional surgery performed at the time of Burch colposuspension appears to have had little effect against the recurrence of prolapse had previously been noted by others3. In agreement with Wiskind et al.3, our women with a pre-operative grade 3 cystocele had a statistically higher (P= 0.02) incidence of prolapse recurrence at any vaginal site (80%) when compared with the incidence recorded among women with a pre-operative grade 2 cystocele (35%). In an earlier report10 we had only one case (2.5%) of post-operative genital prolapse two to seven years after Burch colposuspension performed in conjunction with abdominal hysterectomy and Moschcowitz culdoplasty in 40 women who had only minimal pelvic relaxation defects.

Our anterior colporrhaphy had long term subjective and objective cure rates of only 52% and 42%, respectively. Although during surgery care was taken to give a differential better support to the posterior aspect of the urethra than to the bladder base8 the procedure failed in stabilising the urethrovesical junction, as evaluated by post-operative cotton swab test. Subjective cure rates between 31% and 72% were reported with follow up periods varying from 3 years to 13 years5–7,14–16. However, in many centres success rates were partially15,16 or totally5,7 determined by women's response to mailed questionnaires or telephone interviews. Objective cure rates from 31% to 69% were published at follow up times varying from one year to around eight years9,14,15,17, being not specifically indicated in some cases18.

This study confirmed that, when compared with colposuspension, the anterior colporrhaphy had lower functional results1, whereas its ability to correct a severe cystocele was superior. Only one (3%) case of prolapse recurrence (an isolated cystocele) was recorded among our women operated on via the vaginal route. Considering the long history of anterior colporrhaphy, surprisingly few studies have been published on its effectiveness for anterior vaginal wall prolapse. Other authors reported incidences of recurrent cystocele between 11.17 and 30.36 but they did not specify the grade of prolapse. It is possible that women with grade 1 prolapse were not eliminated from the analysis of recurrences.

Our good anatomic outcome with the anterior colporrhaphy was obtained to the prejudice of the vaginal length and width. In fact, even excluding those women who were not sexually active, the post-operative vaginal length was significantly shorter (P < 0.001) after vaginal surgery. Moreover, those women who underwent vaginal surgery had a statistically greater (P= 0.001) incidence of dyspareunia. However, in both groups all instances of dyspareunia (mild or severe) resulted from over-narrowing of the vaginal introitus caused by posterior colporrhaphy and perineorrhaphy, so that inadequate sexual function could not be considered to be due to either type of anterior vaginal wall management. This was in agreement with an earlier finding6. Van Geelen et al.14 had no case of post-operative dyspareunia in women submitted to anterior colporrhaphy (with or without vaginal hysterectomy) when no posterior repair was performed. Post-operative sexual dysfunction rates of 10%6 and 16%19 were recorded among women receiving a posterior colporrhaphy, which was performed with the same technique used by us. We performed a posterior repair on all women operated on via the vaginal route. In light of our current data, we believe that it should be avoided on a routine basis and if unnecessary, particularly among sexually active women.

We used chromic catgut for the anterior colporrhaphy and permanent sutures for the Burch colposuspension. Although other investigators14 reported results similar to ours when the two types of operations were both performed with absorbable sutures (i.e. Burch colposuspension being significantly more effective than anterior vaginal repair for the correction of genuine stress incontinence), we recognise that the difference in suture materials may play a role. In another study20 we recorded an objective continence rate as low as 27% among 15 women followed up for three to eight years after they underwent a vaginal anterior repair performed with permanent material. Subjective cure rates from 31%14 to 47%16 were documented in the literature when anterior colporrhaphy was performed with chromic catgut. The same figures ranged from 69%9 to 72%6 with delayed absorption sutures and from 60%20 to 63%16 with a permanent type of suture material. Beck and McCormick8 reported an objective cure rate increasing from 75% to 91% when polyglycolic acid suture material were substituted for chromic catgut for their vaginal retropubic urethropexy. Unfortunately many other authors did not indicate the type of suture material used7,15,17,18 or did not differentiate between chromic catgut and delayed absorption sutures5. Further research is needed before a definitive advice can be given.

Few authors2,3,9,21 published their results with Burch colposuspension performed in women with both stress incontinence and advanced cystocele. Bergman et al.9 reported their data with Burch colposuspension in women with stress incontinence and coincident pelvic relaxation. The women underwent combined abdominovaginal surgery (i.e. vaginal cystocele and rectocele repair with or without vaginal hysterectomy followed by Burch colposuspension). The authors, however, stated only that all women ‘had pelvic relaxation that necessitated vaginal surgery’ giving no further detail concerning the exact degree of genital prolapse before surgery. Moreover, they did not present a site specific assessment of prolapse. In this randomised trial9 Burch colposuspension was found to be objectively better than the Pereyra procedure and the anterior colporrhaphy at one-year follow up. Nevertheless we were surprised to find no result regarding the anatomic failure. Therefore these data should be considered with caution. It is possible that with a long term follow up the eventual recurrence of genital prolapse could no longer make the combined approach preferable. Stanton et al.2 published their experience with Burch colposuspension performed in 40 women with anterior vaginal wall prolapse. Nine (22%) women still had a diagnosis of cystocele or urethrocystocele 6 to 30 months post-operatively. Again, no grading of prolapse was given, making the interpretation of these results difficult.

Two articles3,21 reported anatomic failure rates as disappointing as ours when Burch colposuspension was performed in women with significant prolapse. Kjølhede et al.21 submitted 21 women with stress incontinence to Burch colposuspension. Despite a pre-operative coexisting ‘moderate’(corresponding to our grade 2) or ‘large’(corresponding to our grade 3) genital prolapse at one or more vaginal sites in 18 women (including 10 cases of cystocele), only one woman had an additional procedure at the time of Burch colposuspension (a ‘supravaginal’ hysterectomy). At a mean follow up of around four years these authors21 diagnosed a moderate or large genital prolapse at any vaginal site in 17 (94%) of the 18 women, including four (22%) with recurrent cystocele. Six (33%) of the 18 women underwent a second surgery for prolapse correction 9 to 42 months post-operatively. Wiskind et al.3 performed prolapse surgery in 27% of 131 women who were initially treated with Burch colposuspension and were followed up for 3 to 14 years. Pre-operatively 27% of women had ‘marked’(corresponding to our grade 2 to 3) cystocele, 14% had marked rectocele and 1% had marked uterine prolapse. Only 27 (21%) women had additional procedures at the time of the Burch colposuspension (abdominal hysterectomy, Moschcowitz culdoplasty or posterior repair).


In light of our current data and of those reported else-where1,3,21, we believe that the Burch colposuspension should no longer be performed in women with advanced anterior vaginal wall prolapse because the operation has an unacceptable high rate of prolapse recurrence. Concomitant abdominal hysterectomy, culdoplasty and vaginal posterior repair seem to be of little value. On the other hand, continued use of the anterior colporrhaphy in women with stress incontinence cannot be justified as its functional results are expected to be very poor.

One possible choice for the treatment of women with both stress incontinence and severe cystocele might be a combined surgery of vaginal repair and simultaneous Burch colposuspension. This abdomino-vaginal operation gave an objective result of 87% at a short term follow up of one year9. Unfortunately the incidence of prolapse recurrence remains to be assessed for this type of approach. Alternatively, the Pereyra bladder neck suspension might be considered. Although this procedure was found to be less effective than the Burch colposuspension in relieving stress incontinence1,9, it has the advantage of being performed vaginally, thus allowing the repair of any coexisting pelvic support defect. We previously recorded a subjective cure rate of 71% three to nine years after the Pereyra procedure was performed, with a prolapse recurrence rate of 15%20.