Introduction

In 1996 Ulmsten et al. (Tension-free Vaginal Tape®, TVT) [1] and in 2001 Delorme (Transobturator Tape®, TOT) [2] developed the concept of mid-urethral support based on the hammock hypothesis proposed by DeLancey [3]. Mid-urethral slings (MUS) drastically changed stress urinary incontinence (SUI) surgery in the mid-1990s, with excellent outcomes and a mini-invasive approach.

Although MUS are the reference treatment of SUI due to their high success rates, 5–20 % of women still experience surgery failure [4]. Several options have been proposed: second MUS, tape shortening, bulking agents, Adjustable Continence Therapy®, and artificial urinary sphincter. To date, there is no consensus on the management of MUS failure. Repeat sling might be an appropriate solution for women showing persistent urethral hypermobility but few studies have demonstrated their effectiveness. The aim of this study was to determine the outcomes of a repeat MUS after a failed primary sling for SUI and if there are any differences in outcome with respect to MUS used.

Materials and methods

Study population

We retrospectively included 112 female patients who received a second sling for recurrent SUI. Patients were operated on in two university hospitals based in Paris (Bichat and Tenon) between 2000 and 2011. Inclusion criteria were failure or recurrence after a first MUS and urethral hypermobility (visually determined by physical examination). Any type of MUS was allowed for the first procedure.

Data for this study were obtained from usual procedures. Biological material and clinical measures were obtained only for standard diagnosis following physicians’ orders (no specific sampling, no modification of the sampling protocol, no supplementary question to the national standardized questionnaire). Data analyses were carried out using an anonymized database. According to the French Law of Public Health (CSP Art L 1121–1.1), such a protocol is exempt from written informed consent.

Evaluation

Preoperative assessment consisted of complete medical history, evaluation of lower urinary tract symptoms, physical examination, and urodynamic evaluation. Intrinsic sphincter deficiency was defined as maximum urethral closure pressure (MUCP) of 20 cmH2O or less. In 24 cases (21.4 %) urodynamic evaluation had been performed outside of the hospital and was summarized in medical records but was not available for statistical analyses.

Patients were postoperatively scheduled for evaluation by their surgeon including interview and physical examination at 1 month, 6 months, and 1 year or more. Patients who did not attend the clinic after 6 months were contacted for a phone interview. They were submitted to a structured questionnaire based upon the Urinary Symptom Profile® [5] and the International Consultation on Incontinence Questionnaire Short Form® (ICIQ-SF) question 3 [6], evaluating stress, urge, frequency, obstructive symptoms, and quality of life. Patients who attended the clinic once within 6 months but could not be contacted afterwards were considered as lost to follow-up.

During follow-up, patients were then classified into three independent groups: (1) subjective cure, defined by the absence of leaks within the last month; (2) improvement, defined as the decrease in number of leaks and better quality of life; or (3) failure. Overall success was defined by the combination of cure and improvement.

Methods, definitions, and units conform to the standards jointly recommended by the International Urogynecological Association (IUGA) and the International Continence Society (ICS) [7, 8].

Statistical analysis

Baseline characteristics were summarized by means (SD) for continuous variables or counts (%) for discrete variables and then compared across MUS types using a Kruskal–Wallis, Pearson χ2, or Fisher’s exact test. Analysis of urodynamic findings was performed on patients whose data was available. Using either success (cure and improvement) or cure after surgery as an endpoint, a multiple logistic regression model was used to estimate odds ratios (OR) between MUS types, while adjusting for MUCP. To address the question of whether any bias was introduced due to patients with short-term follow-up, we performed sensitivity analyses examining the distribution of endpoints while only including patients with information for a minimum of 6, 12, and 24 months after intervention. Additionally, discrete follow-up time was also included in the multiple logistic regression model. All analyses were performed using Stata (v11.0, StataCorp, College Station, TX, USA) and p < 0.05 was considered significant.

Results

Description of the study population

Clinical and urodynamic preoperative characteristics are displayed in Table 1 and show typical values for this type of population. The previous sling was not removed except for 33 patients in whom it was removed or cut due to vaginal or urethral mesh exposure (20), voiding dysfunction (10), or de novo urgency (3).

Table 1 Preoperative characteristics (mean ± SD, range)
Full size table

The second MUS was one of the following: retropubic tape (49 %), transobturator tape (48 %), or a single-incision mini-sling (3 %). The sling type was chosen according to the surgeon’s preference. Among the 12.9 % of patients with sphincter deficiency, there was 67 % of retropubic tapes, 25 % of transobturator tapes, and 8 % of mini-slings. Details pertaining to the initial and second sling types are given in Table 2.

Table 2 Distribution of slings used for the second intervention
Full size table

The population was divided in two groups: retropubic slings (n = 54) and transobturator slings (n = 55). There was no significant difference between retropubic and transobturator groups, except for proportion of patients with hysterectomy (22.9 vs 42.2 %, respectively, p = 0.047) and detrusor overactivity (0 vs 11.4 %, respectively, p = 0.03).

Clinical outcomes during follow-up

The mean follow-up was 21 months (median 12 months, range 1–108 months), with a total of 25 patients (22.4 %) lost to follow-up. In the entire study population, subjective cure rate was 60.7 % and improvement rate was 16.1 %, indicating an overall success of 76.8 %. Among the 20 patients with vaginal or urethral mesh exposure, 17 were cured (85 %), none was improved, and 3 had a second sling failure (15 %). We found that clinical and urodynamic preoperative parameters had no effect on the outcomes in both uni- and multivariable analyses nor did the etiology of failure or the sequence in which the slings were placed (Table 3).

Table 3 Univariable analysis for determinants associated with successful use of suburethral sling
Full size table

Table 4 displays results for patients whose follow-up was greater than or equal to 6, 12, and 24 months, showing that long-term outcomes could be typically established after the first 6 months following surgery.

Table 4 Results for patients with a minimum 6 months of follow-up
Full size table

The cure and overall success rates were 65.4 and 81.8 % in the retropubic group and 57.4 and 72.2 % in the transobturator group, respectively. No significant difference was found between groups (p = 0.5 and p = 0.2, respectively). However, after adjustment for MUCP, patients having a retropubic approach had a 3.02-fold higher odds of cure (p = 0.05) and a 2.86-fold higher odds of overall success (p = 0.02) compared to the transobturator approach (Table 5). When additionally adjusting for follow-up time, the magnitude of effect remained the same for both cure (OR = 2.72, p = 0.046) or overall success (OR = 2.81, p = 0.06).

Table 5 Multiple logistic regression of overall success and cure between MUS
Full size table

Complications associated with operation

There was no reported intraoperative morbidity, especially no bladder perforation or urethral injury. Postoperative complications were as follows: 8.9 % de novo urgency (2.7 % with urgency incontinence), 7.1 % temporary voiding dysfunction (successfully treated with clean intermittent catheterization), 3.6 % pelvic pain, and 2.7 % recurrent urinary tract infections. Five patients (4.5 %) had their sling removed because of three vaginal mesh exposures (2.7 %, none of them had vaginal or urethral mesh exposure after the first sling) and two complete retentions (1.8 %, 1 TOT, 1 TVT). None of the sling types were associated with an increased risk of complications.

Discussion

Currently, MUS are the preferred first-line treatment for female SUI. High cure rates over long periods are a clearly attainable goal, with one prospective evaluation reporting 90 % after 11 years [9]. Nevertheless, there remains 5 to up to 20 % risk of failure depending on the study population. Women with recurrent SUI who still have urethral hypermobility could technically benefit from a second MUS. We report that 76.8 % of patients have a successful second SUI, with the large majority completely resolving SUI. Furthermore, outcomes were typically determined within the first 6 months after surgery and were stable over time. To the best of our knowledge, this study represents the largest number of patients with repeat slings to date.

A literature review of studies since 2007 identified only seven case series of repeat MUS in which more than ten patients were included (Table 6). In three of seven studies, there was a heterogeneous distribution of first and second slings. Cure rates largely varied from 40 to 91.7 % with a weighted average of 64.9 %, while overall success rates similarly varied from 58 to 91.7 % with a weighted average of 76.6 %. In one of the more well-designed, prospective studies, a much higher cure rate (74 %) was reported among 31 women with 18 months of follow-up [10]. Unfortunately, their data can only be extended to a specific population, as numerous exclusion criteria [body mass index (BMI) > 30 kg/m2, associated urge incontinence or voiding dysfunction, concomitant pelvic floor surgery] arguably led to a population at low-risk of failure. With a subjective cure at 60.7 % and overall success at 76.8 %, our findings are similar to what has been previously described in the literature despite the aforementioned discrepancies. Our study comprising 112 patients is the largest performed on repeat sling outcomes to date.

Table 6 Literature review on second MUS for recurrent SUI
Full size table

In our cohort, no clinical or urodynamic parameter was correlated to surgical outcomes. Currently, few studies have included multivariable analysis in which other factors predictive of sling success or failure were observed. Nonetheless, it should be noted that several factors have been previously identified for first sling failure (age, previous incontinence surgery, urge incontinence, BMI > 25 kg/m2, diabetes, intrinsic sphincter deficiency, overactive bladder [11]).

A large number of studies have compared transobturator and retropubic slings. In a recent meta-analysis, TVT appeared to be followed by significantly higher rates of objective continence rates than TOT, with no difference in subjective continence rates [12]. On the contrary, more intraoperative complications and storage lower urinary tract symptoms were observed with TVT than TOT slings. Studies on repeat slings have also aimed to compare the two approaches. Lee et al. found a trend toward higher success for the retropubic approach, although the difference between the two groups was not statistically significant [13]. Only Stav et al. demonstrated a significant difference in favor of the retropubic approach (71 vs 48 %, p = 0.04) [14]. In our cohort there was no significant unadjusted difference between the two groups. On the other hand, after adjusting for closure pressure, retropubic slings were significantly superior to the transobturator approach: the retropubic approach increased in effectiveness as the MUCP decreased. Several studies have already shown that retropubic slings are more effective in woman with intrinsic sphincter deficiency [15, 16]. A possible explanation is the U-shape of the retropubic tape, which may be more supportive and obstructive than the horizontal orientation of the transobturator tape.

One limitation of our study is its retrospective design, considering that prospective randomized controlled trials are difficult to perform for second-line surgery. Tenon and Bichat hospitals represent two reference centers for urogynecology in France, with similar and well-standardized methods of perioperative evaluation and operative procedures. Nonetheless, 21.4 % of patients had missing urodynamic data.

Nineteen patients had missing data on MUCP. It could be argued that significance after adjustment for this factor was more likely due to selection of the patient subpopulation. We performed a sensitivity analysis in which crude OR were calculated among those with MUCP data and found no substantial difference in either the magnitude of effect or significance.

Another limitation is the use of numerous types of slings. Hence, several different kinds of materials were compared (allowing the comparison of retropubic and transobturator approaches). Some have been withdrawn from the market due to their increased risk in vaginal or urethral mesh exposures, fistulas, or cellulitis (ObTape®, Uratape®, IVS®, etc.). Currently, the only recommended type of sling is nonwoven polypropylene.

Despite a 22.4 % loss to follow-up, results were very stable regardless of minimal amount of follow-up time. We can therefore suppose that the outcomes of patients with less than 6, 12, or even 24 months of information would not have significantly altered the results of our study.

Conclusion

With 77 % overall success and low risk of complications, repeat slings are still a worthwhile therapeutic alternative in patients whose symptom is a social handicap. Moreover, we found that retropubic slings are more efficient than transobturator slings after adjusting for low MUCP. These data can be useful when choosing the approach after failure. However, the success of repeat slings is inferior to a first procedure.

Patients requiring a repeat sling should be carefully selected. Tape failure brings to light a great variety of situations that require rigorous assessment. Notwithstanding its simplicity, clinicians should not be driven to repeat MUS automatically when other solutions may be more appropriate (bulking agents, Adjustable Continence Therapy®, artificial sphincter).