Neurourology and Urodynamics

Surgical Treatment of Neurogenic Stress Urinary Incontinence: A Systematic Review of Quality Assessment and Surgical Outcomes Fawzy Farag,1,2* Martin Koens,1 Karl-Dietrich Sievert,3 Dirk De Ridder,4 Wout Feitz,1 and John Heesakkers1 1

Department of Urology, Radboud University Medical Centre Nijmegen, The Netherlands 2 Department of Urology, Sohag University Hospital, Sohag, Egypt 3 Department of Urology, University of Tuebingen, Tuebingen, Germany 4 Department of Urology, UZ Leuven, Leuven, Belgium

Background: There are many opinions but little firm knowledge about the optimal treatment of neurogenic stress urinary incontinence (NSUI). Objective: To scrutinize the quality and surgical outcomes of the available treatment modalities in the published literature. Evidence Acquisition: A systematic review of the published literature from Pub Med and Web of Science was undertaken for studies describing surgical treatment of NSUI between 1990 and 2013. A checklist of criteria of methodological and reporting quality of interventions in urological publications was applied to assess quality of the retrieved publications. Surgical outcomes of success, failure, and reoperation were calculated. Statistical analyses included one-way ANOVA and post-hoc tests to determine significant differences between groups. Evidence Synthesis: Thirty studies were identified with Level 3 evidence. The quality of reporting was 43–81%, with significantly higher quality noted in studies published after 2002 (64% vs. 45%, P < 0.0001). None of the studies followed a randomized controlled trial (RCT) design. Three primary surgical procedures were used in 29 of 30 studies: artificial urinary sphincter (AUS), urethral slings, and urethral bulking agents. One study used a ProACT device. AUS was considered more successful than urethral bulking agents (77
15% vs. 27
20%, P ¼ 0.002). Urethral bulking agents reported higher failures than urethral sling procedures (49
16% vs. 21
19%, P ¼ 0.016) and AUS (21
19% vs. 10
11%, P < 0.002). Conclusions: The quality of evidence obtained from non-RCTs is modest. Surgeries for NSUI have relatively high success rates but also high complication rates in this highly heterogeneous population. More studies using # 2014 Wiley Periodicals, Inc. modern techniques are required to update our knowledge. Neurourol. Urodynam. Key words: artificial; bulking agents; injection therapy; spinal cord injuries; spinal dysraphism; stress; suburethral slings; urinary incontinence; urinary sphincter

INTRODUCTION

Urinary incontinence (UI) is the complaint of any involuntary leakage of urine. Stress urinary incontinence (SUI) is a subset of UI that occurs upon exertion, sneezing, or coughing.1 Patients with spinal cord lesions below the level of the sacral micturition centre commonly experience weakness of the urinary sphincter leading to the development of neurogenic SUI (NSUI). In addition, NSUI due to sphincteric insufficiency affects 68% of school age children with spina bifida.2 The surgical treatment of non-neurogenic SUI in patients has been intensively investigated. However, there remains a lack of knowledge regarding the optimal treatment modalities for NSUI. Intermittent catheterization (IC) plays a major role in achieving dryness in patients with NSUI. In patients with persistent NSUI despite IC, other procedures have been utilized such as bladder neck closure, urethral lengthening procedures, artificial urinary sphincter (AUS), urethral sling and bladder wrap procedures, and urethral bulking agents.3 The aim of the current review is to systematically analyze the current evidence of NSUI treatment using less invasive surgical modalities. EVIDENCE ACQUISITION

A systematic search was performed using Pub Med and Web of Science for clinical studies with a particular emphasis on surgical techniques for treatment of patients with NSUI from #

2014 Wiley Periodicals, Inc.

1990 to 2013 (Fig. 1). The search terms, inclusion and exclusion criteria, are listed in Supplement Table SI.

Surgical Outcome Assessment

The primary outcome was the success rate of the surgical treatment of NSUI in these studies. The secondary outcome was the individual effectiveness of each surgical modality. The variation of the definition of success used in the 30 studies (Supplement Table SII) is reported in Supplement Table SIII. The percentage of surgical outcome parameters for each individual study were calculated by dividing the number of patients for each parameter by the number of patients who completed the last follow-up assessment.

Christopher Chapple led the peer-review process as the Associate Editor responsible for the paper. Potential conflicts of interest: Dr De Ridder reports grants from American Medical Systems, outside the submitted work.  Correspondence to: Fawzy Farag, Radboud University, Urology postbus 9101 659 urologie, Nijmegen 6500 HB, The Netherlands. E-mail: [email protected] Received 3 March 2014; Accepted 3 September 2014 Published online in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/nau.22682

2

Farag et al. Primary search results

EVIDENCE SYNTHESIS

Pub Med:

13 items

Surgical Outcome Assessment

Web of Science:

46 items

Total:

59 hints

A total of 30 studies were selected for final analysis (Supplement Table SII). These 30 studies included 849 patients (525 males and 324 females) with a median age 21 years old (range 3–80). The etiology of NSUI was spinal dysraphism in 578 (69%) patients, spinal cord injury (SCI) in 191 (22%) patients, and other causes in 80 (9%) patients. The demographics and descriptive data for surgical outcome parameters are shown in Supplement Table SIII. The patients received an AUS in 399 cases (322 males, 77 females), urethral sling procedure in 286 cases (108 males, 178 females) distributed as follows: synthetic male sling ¼ 26, synthetic female sling ¼ 20, autologous fascial sling ¼ 177, wrap procedure ¼ 63. Urethral bulking agents were applied in 126 cases (82 males, 44 females), distributed as follows: collagen ¼ 93, polydimethylsiloxane ¼ 33. Finally, a Pro-ACT device was implanted in 37 cases (13 males, 24 females). A total of 280 patients (35%) underwent bladder augmentation either before or simultaneous with the continence surgery while 46 patients (6%) underwent bladder augmentation surgery during follow-up. The overall median follow-up was 48 months (range 12– 62 months). A total of 821 patients completed the last followup. The total number of successful treatments was 528 cases (64%), 130 case (16%) improvements, whilst 162 (20%) reported no change in their incontinence status. The success rates did not correlate with percentages of males (Pearson correlation test: P ¼ 0.9) nor females (P ¼ 0.9), neurogenic deficit subtypes (dysraphism: P ¼ 0.6; SCI: P ¼ 0.9), nor bladder augmentations prior to or simultaneous with surgery (P ¼ 0.7). The total number of patients who used IC was 396 of 849 (47%) patients (data obtained from 25 studies). A total of 167 patients (20%) had surgical complications, 207 patients (25%) underwent one or more reoperations, 59 patients (7%) underwent a permanent undo, 45 patients (5.5%) received an additional continence procedure to achieve continence: Gitte’s anterior urethropexy in one, AUS in 13, open bladder neck continence in nine, urinary diversions in nine, urethral bulking agents in 11, and urethral slings in two. Twenty patients reported difficulties with IC (data obtained from 20 studies). The total number of female patients who developed postoperative surgical complications was 17 of 170 (10%) patients (data obtained from 20 studies). The total number of male patients who developed postoperative surgical complications was 47 of 234 (20%) patients (data obtained from 20 studies). For statistical reasons, the outcome of the single study that used the Pro-ACT device was excluded from the final analysis. The one-way ANOVA revealed statistically significant differences among the three surgical treatments with regard to each of the following study parameters: percentage of success (F ¼ 9.25, P ¼ 0.001), percentage of failure (F ¼ 9.58, P < 0.0001), percentage of reoperation per patient (F ¼ 19.71, P < 0.0001), mean duration of follow-up (F ¼ 6.81, P ¼ 0.004), percentage of patients who developed complications (F ¼ 4.4, P ¼ 0.02), and the percentage of reoperation (F ¼ 19.8, P < 0.0001). The results of the post-hoc tests were as follows: the analysis revealed a significant difference in the mean follow-up between the AUS and urethral bulking agents (72
18 months vs. 30
13 months, P ¼ 0.01) and between the AUS and urethral sling placement (72
18 months vs. 37
30 months, P ¼ 0.01). However, there were no significant differences

Removal of 5 duplicates 54 items

Exclusion of irrelevant publications 30 articles

28 clinical study articles were added after hand searching the relevant bibliography 58 articles

28 articles with mixed patient population excluded

30 articles included for final analysis

Fig. 1. Flow diagram of study selection.

Quality Assessment

The checklist developed by Tseng et al.4 to evaluate the quality of interventional urological publications was used to assess the quality of the retrieved publications from our search. The Quality-related questions investigated the following components: study background, type of intervention, statistical analysis, results, and discussion. The standardized criteria of the Center for Evidence Based Medicine in Oxford (United Kingdom)5 were applied to determine the level of evidence of the studies. The studies were further divided into two categories. Category 1 included studies performed from 1990 to 2002. Category 2 included studies performed from 2003 to 2012. The difference in quality of reporting between both groups was evaluated. STATISTICAL METHODS

SPSS v.20.0 (SPSS, Chicago, IL) was used for the statistical analysis of data. The Shapiro–Wilk test was used for testing the normality of the data distribution. Related sample Wilcoxon Signed-Rank test, Student’s t-test, the ANOVA, and post-hoc tests were applied to find differences between groups at criteria, a ¼ 0.05. Studies with two different subgroups (gender and/or age), representing values in the majority in the study population, were considered for analyses. Neurourology and Urodynamics DOI 10.1002/nau

Surgical Treatment of Neurogenic Stress Urinary Incontinence in mean follow-up between urethral sling placements and bulking injections. The AUS studies reported highly significant success rates (n ¼ 8 studies, mean ¼ 77
16%) compared to urethral bulking injections (n ¼ 6 studies, mean ¼ 27
20%), P ¼ 0.001. Similarly, urethral sling placements showed significantly higher success rates (n ¼ 15 studies, mean ¼ 58
25%) compared to urethral bulking injections (n ¼ 6 studies, mean ¼ 27
20%), P ¼ 0.02. The difference between AUS and urethral sling placements was not statistically significant. Urethral bulking injections (n ¼ 6 studies, mean ¼ 50
16%) showed higher failure rates compared to the AUS (n ¼ 8 studies, mean ¼ 10
11%), P ¼ 0.001 and urethral sling placements (N ¼ 15 studies, mean ¼ 22
20%), P ¼ 0.01, while the difference between the AUS and urethral sling placements was not statistically significant. The AUS had higher reoperation rates when compared to urethral sling placements (n ¼ 8 studies, mean ¼ 51
25% vs. n ¼ 14 studies, mean ¼ 7
9%; P < 0.003), respectively. The same was true for the AUS when compared to urethral bulking injections (n ¼ 8 studies, mean 51
25% vs. n ¼ 4 studies, mean ¼ 12
14%; P ¼ 0.02). The difference between urethral sling placement and urethral bulking injections was not statistically significant. Post-hoc tests revealed no statistically significant difference in the percentages of patients with complications among the three procedures. The total number of patients who received an AUS or urethral sling at the bladder neck position was 608 patients. In 48 patients, the AUS or sling was implanted at the bulbar urethra. There was no statistically significant difference between the two locations in success, failure or reoperation rates. While the implantation of AUS/sling at the bulbar urethral showed statistically higher rates of complications when compared to implantation at bladder neck (45
14% (n ¼ 3) vs. 16
21% (n ¼ 15), P ¼ 0.04). Table I summarizes the surgical outcomes per individual technique.

TABLE I. Summary of Surgical Outcomes per Individual Technique AUS (n ¼ 8) Follow-up (months) Success% Failure% Complications% Reoperations%

Slings (n ¼ 15)

Bulking agents (n ¼ 6)

P-value (ANOVA)

72
18

37
30

30
13

0.004

77
16% 10
11% 32
27% 51
25%

58
25% 22
20% 14
14% 7
9%

27
20% 50
16% 4
6% 12
14%

0.001 0.001 0.020 0.000

TABLE II. Summary of Quality of Reporting of the 30 Articles Before 2002

Criterion Background Intervention Statistical analysis Results reporting Discussion

After 2002

Total (yes%)

Total (no%)

Total (yes%)

Total (no%)

Overall (yes%)

66.7 46.7 14.2 42.3 80.0

33.4 53.3 85.8 57.7 20.0

78.3 84.5 26.7 58.5 91.4

21.7 15.5 73.3 41.5 8.6

72.5 65.6 20.4 50.4 85.7

Neurourology and Urodynamics DOI 10.1002/nau

3

Quality Assessment

Supplement Table SIV provides the checklist used to qualify the 30 studies. Table II summarizes the overall quality of reporting of the 30 studies and highlights the differences noted between studies published before and after 2002. All studies demonstrated Level 3 evidence and were either cohortprospective or retrospective studies. The overall quality of reporting ranged from 43% to 81%. The median quality of reporting score was significantly higher in studies published 2002–2013 than those published 1990–2002 (64% vs. 45%, P < 0.0001). None of the studies followed a randomized controlled trial (RCT) design. Two of 30 (6.6%) studies 6,7 reported the level of experience of the surgeons. There was no power calculation in any of the 30 studies. The surgical techniques were described adequately in 21 of 30 (70%) studies. Seven of 30 (23%) studies used validated assessment tools, six 7–12 were performed after 2002. The predetermined level of significance was provided in three of 30 (10%) studies; all were performed after 2002. One of 30 studies demonstrated an attempt to blind evaluators. Baseline demographics and clinical data were sufficiently described in 13 of 30 (43%) studies. The number and nature of complications were addressed in 29 of 30 (97%) studies. The exact P-value of significant results was provided in five of 30 (17%) studies, four of them were performed after 2002. Potential sources of bias and study limitations were addressed in 11 of 30 (37%) studies. Table III summarizes the items of the quality assessment checklist that have been unsatisfactorily reported in the 30 studies. DISCUSSION

Most available guidelines for the treatment of neurogenic lower urinary tract dysfunctions (NLUTD) provide an inconsistent and limited indication of the optimal treatment modalities for NSUI. The European Association of Urology (EAU) guidelines13 classify the placement of urethral slings and AUS with a Grade B recommendation while the International Consultation on Incontinence (ICI)14 classify the AUS with a Grade A recommendation. It is apparent that the current urological practice using AUS in treatment of NSUI continues to lack proper clinical evidence, however, it is still considered the gold standard. This implies that those dealing with patients with NSUI consider the AUS or urethral sling as the best option, although the proof is weak. Because of this, it is difficult to standardize such a treatment in such an important group of patients. The scientific facts on which we base our judgment should be improved since many neurogenic patient groups need specialized advice and treatment which is based on current state-of-the-art techniques. For instance, the use of synthetic slings, which is considered the gold standard for nonneurogenic SUI, is regarded as of limited value because of the risk of exposure that could occur due to lack of sensation or an obligation to perform clean intermittent catheterization. Unfortunately, very few studies exist that substantiate this presumption.3 Therefore, more studies with modern techniques are needed to update our knowledge. Surgical Outcome Assessment

The analysis revealed an overall success rate of 64% in a median follow-up of 48 months (range 12–62). This is substantially lower than most series of surgical procedures in non-neurogenic patients. The AUS in non-neurogenic male

4

Farag et al.

TABLE III. Quality Assessment Criteria That Were Reported in the 30 Studies

Criterion Background B3 Can the number of surgeons who participated in the study be determined? B4 Is the experience level of the surgeons performing the procedure specified? Intervention I3 Is there any mention of an attempt to standardize perioperative care? I4 Is the period when all cases were performed clearly stated? I7 Methods for assessing outcomes described? I8 For studies assessing functional outcomes, is it stated whether the assessment tool is validated? I9 Do the authors describe how patients were chosen into each treatment group (e.g., patient choice or specific criteria)? Statistical analysis S1 Calculation to justify sample size? S2 Statistical methods described? S3 Statistical software identified?  (significant PS4 Identification of predetermined a values)? S5 Sidedness of testing reported? S8a Effect size (e.g., HR, RR, NNT) provided for the primary outcome? S9a Precision of effect size (e.g., CI for HR, RR, NNT) provided for the primary outcome? Answer No if no effect size given Results reporting R1 Was any attempt made to blind evaluators during the analysis of data? R2 Was the patient population from which the cases were selected from adequately described or identified (e.g., geographically)? R4 Are relevant baseline demographic and clinical data given for each group? R9 Are exact P-values for significant results provided (