1 INTRODUCTION

Urethral strictures have an estimated prevalence of 0.6% in Western countries1 and are caused by medical interventions, infection and inflammation, trauma, or may be idiopathic.2 Strictures can be treated with dilatation, direct visual internal urethrotomy (DVIU), or urethroplasty. Treatment choice is dependent on patient age and comorbidities, stricture etiology, location and length, the number of strictures, risk of recurrence, and patient preference.3 Urethroplasty with (transecting or non-transecting) anastomotic repair or free graft urethroplasty have emerged as the standard management for most strictures, offering high success rates (up to 90%).4 Dilatation or DVIU are less invasive treatment options, but recurrence after these procedures is more common. Recurrence rate at 2 years after first DVIU is 30% to 60% and increases to 50% to 100% after second urethrotomy.4 Therefore, open urethral reconstruction should be offered instead of repeated endoscopic management for recurrent anterior urethral strictures.5 Furthermore, repeat transurethral manipulation of bulbar urethral strictures is associated with increased stricture complexity and prolonged disease duration, due to cumulative tissue injury resulting in amplified spongiofibrosis.6 In addition, repetitive endoscopic treatment may impair the success rate of urethroplasty.7, 8 Also, for the same reason, intermittent self-dilatation (ISD), performed to delay or defer further intervention, might increase the risk of failure.6

Knowledge of factors contributing to treatment success has great importance for preoperative counseling and shared decision-making. The aim of this study is to evaluate the relation between clinically relevant stricture recurrence after first urethroplasty and the number of prior (repeated) endoscopic treatments or ISD.

2 MATERIALS AND METHODS 2.1 Patient cohort

A total of 130 consecutive male patients of 18 years and older received a first anterior urethroplasty between 2011 and April 2019 in a tertiary referral center for urethral strictures. Only patients with penobulbar or bulbar strictures with short penile extension were included. Penobulbar strictures with penile approach during urethroplasty were excluded, as well as patients with a history of hypospadias and pelvic fracture urethral distraction injury. A total of 106 patients were included in this study. Ethical board approval was acquired for this study (18-353). Informed consent was not required.

2.2 Data collection

The collected data consisted of patient age and medical history (including all prior urethral endoscopic procedures and ISD), stricture characteristics (etiology, location, and length), urethroplasty technique, and postoperative outcome parameters. Since 2011, all patients attending our hospital with urethral stricture disease have had a standardized preoperative work-up and postoperative follow-up, allowing prospectively gathered data to be collected from the electronic patient file system.

Preoperative diagnostic evaluation included data extraction on medical history and previous procedures from referral letters and questionnaires on medical history and stricture etiology, International Prostate Symptom Score, uroflowmetry and measurement of postvoid residual, retrograde urethrography (RUG), and 16F flexible cystoscopy. The latter two allowed preoperative identification of stricture location and estimation of stricture length. Stricture length was measured during urethroplasty and included in the operation report. Postoperatively, the transurethral catheter was removed after 3 weeks if voiding cystourethrography (VCUG) showed no leakage of contrast. In case of leakage, VCUG was repeated after 2 to 3 weeks prior to catheter removal (if no leakage).

The standardized follow-up was 1 year and included uroflowmetry at 3, 6, and 12 months, RUG at 6 months, and flexible cystoscopy (16F) at 12 months.

Stricture recurrence was defined as any need for reintervention based on patient symptoms (functional outcomes) combined with diagnostic findings during follow-up (anatomic outcomes), which was also described by the TURNS protocol.9 Reintervention was described as any surgical procedure, including outpatient dilatation, endoscopic reintervention, and redo urethroplasty.

In case there were no signs of treatment failure after 12 months, patients were discharged from follow-up with the instruction to return in case of symptoms suggestive of stricture recurrence.

2.3 Surgical technique

The surgical technique was based on the peroperative clinical judgment and expertise of the three reconstructive urologists of our center. If technically feasible, excision and primary anastomosis (EPA) was performed. In case of a long defect, in general more than 2 cm, dorsal onlay buccal mucosa (BM) graft urethroplasty was performed.

2.4 Statistical analyses

Groups were compared using the χ2 test for categorical data. For a statistically significant χ2 test with three or more groups, post hoc analysis with adjusted z scores and Bonferoni adjusted P value for significance was performed. Our data were not normally distributed. Therefore, the difference in continuous variables was evaluated with the Mann-Whitney U test (two categories) and Kruskal-Wallis test (>2 categories). Multivariable logistic regression analysis was performed with all significant univariate variables to identify independent predictors. A P value <.05 was considered statistically significant. Analyses were performed with IBM SPSS Statistics 25.

3 RESULTS

In total, 106 patients were included in this study. Seventeen (16%) patients had no prior endoscopic procedure, 28 (26%) had one, 25 (24%) had two, and 36 (34%) had three or more. There were 32 patients (30%) who routinely performed ISD before anterior urethroplasty. Median follow-up was 12 (interquartile range [IQR] 8-13) months, median age at surgery was 42 (IQR 31-53) years, and median peroperative stricture length was 2.0 (IQR 1.5-3.5) cm. EPA was performed in 85 (80%) of the included patients with a median peroperative stricture length of 2.0 (IQR 1.2-2.0) cm. The other 21 (20%) were treated with dorsal onlay BM graft urethroplasty, and these patients had a longer median peroperative stricture length of 4.8 (IQR 4.0-5.3) cm (P ≤ .001).

From all 106 included patients, 90 (85%) showed no recurrence during follow-up (Table 1). Sixteen patients had a reintervention after the first urethroplasty, of whom 8 (7.5%) were treated endoscopically (dilatation or DVIU) and 8 (7.5%) with redo urethroplasty.

TABLE 1. Overall results. Patient and stricture characteristics in patients without and with recurrence after first urethroplasty Patient characteristics No recurrence n = 90 (85%) Recurrence n = 16 (15%) P value Median age at surgery, years [IQR] 41 [31-50] 53 [28-67] .202a Median time of follow-up, months [IQR] 12 [9-13] 9 [5-27] .582a Stricture etiology: .019b 48 (94%) 3 (6%) 11 (65%) 6 (35%) 3 (100%) 0 (0%) 28 (80%) 7 (20%) Stricture location: .912b 85 (85%) 15 (15%) 5 (83%) 1 (17%) Prior endoscopic treatments (dilatation, DVIU): .021b 14 (82%) 3 (18%) 28 (100%) 0 (0%) 22 (88%) 3 (12%) 26 (72%) 10 (28%) Prior ISD .014b 23 (72%) 9 (28%) 67 (91%) 7 (9%) Technique .908b 72 (85%) 13 (15%) 18 (86%) 3 (14%) Median stricture length in cm [IQR] 2.0 [1.5-3.0] 1.8 [1.0-3.3] .496a 2.0 [1.5-2.0] 1.5 [1.0-2.0] 5.0 [4.0-5.0] 4.5 [4.0-4.5] Abbreviations: BM, buccal mucosa; DVIU, direct visual internal urethrotomy; EPA, excision and primary anastomosis; IQR, interquartile range; ISD, intermittent self-dilatation. a Kruskal-Wallis test. b Chi-square test.

There was a significant relation between number of prior endoscopic procedures and stricture recurrence after the first anterior urethroplasty (P = .021): Specifically, recurrence was more prevalent after three or more prior endoscopic treatments (28%, P = .009). The prevalence of prior ISD was higher in the stricture recurrence group (56% vs 26%, P = .014). Also, we found a relation between stricture etiology and recurrence. Patients with idiopathic etiology less often had a recurrence (6%, P = .019). No difference was found in median age, stricture length, stricture location, and urethroplasty technique in patients with and without recurrence.

The relation between number of prior endoscopic treatments and other parameters was compared (Table 2). There was no difference in median age at surgery, median peroperative stricture length, stricture etiology and location, and urethroplasty technique between patients with various numbers of prior endoscopic treatments. Patients with repeated transurethral procedures tended to be older (not significant).

TABLE 2. Patient and stricture characteristics in patients with 0, 1, 2, and ≥3 prior endoscopic treatments 0 1 2 >3 n = 17 (16%) n = 28 (26%) n = 25 (24%) n = 36 (34%) P value Median age at surgery, years [IQR] 32 [24-43] 40 [30-50] 42 [37-52] 45 [33-58] .091a Median stricture length, cm [IQR] 1.5 [1.0-3.0] 2.0 [1.5-3.0] 2.0 [1.0-4.3] 2.0 [1.5-4.0] .255a Stricture etiology: .096b 7 (41%) 17 (61%) 12 (48%) 15 (42%) 2 (12%) 2 (7%) 3 (12%) 10 (28%) 0 (0%) 0 (0%) 0 (0%) 3 (8%) 8 (47%) 9 (32%) 10 (40%) 8 (22%) Stricture location: .552b 16 (94%) 26 (93%) 25 (100%) 33 (92%) 1 (6%) 2 (7%) 0 (0%) 3 (8%) Technique .156b 16 (94%) 24 (86%) 20 (80%) 25 (69%) 1 (6%) 4 (14%) 5 (20%) 11 (31%) Abbreviations: BM, buccal mucosa; EPA, excision and primary anastomosis; IQR, interquartile range. a Kruskal-Wallis test. b Chi-square test.

Patients who performed ISD were older than patients who did not perform ISD (48 vs 38 years, P = .009, Table 3). Furthermore, 61% of patients with three or more prior endoscopic procedures performed ISD prior to primary urethroplasty, which is more prevalent than in patients with 0 to 2 prior endoscopic treatments (P < .001).

TABLE 3. Patient and stricture characteristics of patients with and without prior intermittent self-dilatation Prior ISDn = 32 (30%) No prior ISDn = 74 (70%) P value Median age, years [IQR] 48 [38-59] 38 [27-49] .009a Prior endoscopic interventions (%): <.001b 1 (6%) 16 (94%) 4 (14%) 24 (86%) 5 (20%) 20 (80%) 22 (61%) 14 (39%) Stricture etiology (%): .265b 12 (24%) 39 (76%) 7 (41%) 10 (59%) 2 (67%) 1 (33%) 11 (31%) 24 (69%) Stricture location (%): .863b 30 (30%) 70 (70%) 2 (33%) 4 (67%) Technique (%) .052b 22 (26%) 63 (74%) 10 (48%) 11 (52%) Median stricture length, cm [IQR] 2.0 [1.5-4.0] 2.0 [1.5-3.0] .075a Abbreviations: BM, buccal mucosa; EPA, excision and primary anastomosis; IQR, interquartile range; ISD, intermittent self-dilatation. a Mann-Whitney U test. b Chi-square test.

There was no difference in median peroperative stricture length, stricture etiology and location, and urethroplasty technique between patients who did or did not perform preoperative ISD.

The variables etiology, preoperative ISD, and number of prior endoscopic interventions were included in a multivariable logistic regression model. The results are presented in Table 4. In this multivariable model the number of previous endoscopic interventions and ISD did not remain as independent predictors for stricture recurrence after the first anterior urethroplasty. Only iatrogenic strictures (odds ratio [OR] 5.7, P = .035) remained as an independent predictor for stricture recurrence with idiopathic strictures as reference.

TABLE 4. Multivariable logistic regression analysis for the prediction of recurrence after first bulbar urethroplasty Variable OR (95% CI) P value Etiology: Reference - 5.7 (1.1-29.0) .035 NA NA 4.1 (0.9-19.3) .070 Number of prior endoscopic interventions: Reference - NA NA 0.6 (0.1-3.6) .572 1.3 (0.2-7.8) .763 Prior ISD 2.3 (0.6-9.2) .231 Note: All significant (P < .005) univariate variables are included in the analysis. Abbreviations: ISD, intermittent self-dilatation; NA, not applicable (there was no recurrence); OR, odds ratio. 4 DISCUSSION

The overall success rate of urethroplasty in our studied cohort was 85%. An increased risk of recurrence was found in patients with three or more endoscopic interventions and in patients who performed ISD prior to first urethroplasty.

There was no difference in recurrence risk in patients without and with one or two prior endoscopic treatments.

Our results show a difference in recurrence rate in patients who performed ISD prior to urethroplasty (28%) compared to patients who did not (9%). We also found a strong relation between number of previous endoscopic procedures and performance of ISD. It is plausible that patients were advised to start with ISD after repeated endoscopic treatment because of the high risk of recurrence (>50%).10 This may explain why preoperative ISD and number of preoperative endoscopic treatments were not identified as independent risk factors in the multivariable analysis.

The lower recurrence rate in patients with one prior endoscopic procedure compared to no previous endoscopic treatment might be explained by the etiology of the strictures. The recurrence rate in patients with idiopathic strictures was only 6%. Remarkably, 61% of patients with one prior endoscopic procedure had an idiopathic stricture, compared to 41% of patients with no prior endoscopic procedure. More iatrogenic strictures and straddle injuries were found in the latter group. Iatrogenic stricture etiology was identified as an independent risk factor for recurrence in our multivariable analysis (OR 5.7), with idiopathic etiology as reference. A larger study of Kinnaird et al11 identified iatrogenic etiology as independent risk factor in multivariate regression analysis as well. Iatrogenic strictures are mainly found after urethral instrumentation in the context of treatment for benign prostatic hyperplasia, urothelial cell carcinoma, and urolithiasis. It is unknown why urethral instrumentation other than stricture treatment predisposes for stricture recurrence. Factors such as use of electrocoagulation and laser may play a role. Another explanation is that, although not identified in our study, patients with iatrogenic strictures may have other age-related or metabolic characteristics compared with patients with idiopathic or traumatic strictures.

We did not find a longer median stricture length in patients with prior repeated endoscopic interventions or in patients with ISD. Furthermore, stricture length was not associated with stricture recurrence. This is in contrast with some previous studies, which reported increased stricture length in patients with repeated transurethral manipulation6 and increased recurrence risk with longer strictures.11-13

In our series, 31% of the patients with ≥3 prior endoscopic treatments were treated with BM dorsal onlay urethroplasty compared to 6% without prior endoscopic treatment, although these numbers were not significant.

It is widely accepted that recurrence rate after onlay urethroplasty is higher than after EPA, which was not found in our series. In a recent study by Barbagli et al with inclusion of 1242 bulbar urethroplasties, success rate after EPA was 85.9%. The success rate after BM urethroplasty was dependent on the used technique (68%-87.5%).13 A similar success rate with BM and EPA urethroplasty was found in other studies14, 15 and these success rates are comparable to our series. The variation between studies may be explained by differences in experience, used techniques, and definition of success or recurrence.

There is some inconsistency in the literature concerning prior repeated endoscopic interventions as a risk factor for failure after urethroplasty.6-8, 12

In the study of Breyer et al,8 with inclusion of 443 patients, an increased risk of recurrence was found with prior DVIU in a multivariate analysis (hazard ratio [HR] 1.7, P = .04), but previous dilatation was not identified as risk factor (HR 0.8, P = .48). In this study, the number of DVIU was not specified. In the study of Kessler et al,7 with inclusion of 267 patients, two or more previous DVIU was found as risk factor for recurrent strictures after urethroplasty (HR 2.25; 95% CI, 1.05-4.8). In a more recent study by Chapman et al,12 who analyzed 596 patients, the number of previous failed endoscopic procedures did not affect the stricture recurrence rate after urethroplasty. Barbagli et al,13 with inclusion of 1242 bulbar urethral strictures, did not find increased recurrence rates with any previous dilatation or DVIU either.

It should be mentioned that the patients in these studies were less homogenous compared to our cohort due to differences in inclusion criteria, such as stricture location,7, 8 inclusion of redo urethroplasty,7, 8, 12 inclusion of traumatic strictures without further specification,7, 8, 12, 13 and history of hypospadias repair.7, 8, 12

A more similar study by Hudak et al6 with inclusion of 101 bulbar urethroplasties divided their cohort in two groups: 0 to 1 and >2 transurethral manipulations, including any dilatations (operative, in office, or self-dilatation at home). More failures were found during follow-up in patients with >2 transurethral manipulations before urethroplasty compared to 0 to 1 previous treatments, though not statistically significant (12% vs 2%, P = .110). This study concluded that repeat transurethral manipulations were associated with increased stricture complexity and prolonged disease duration.

The results of the OPEN trial have been recently published, a randomized controlled trial where 159 patients with recurrence of bulbar stricture after prior DIVU were allocated for open urethroplasty vs repeated DVIU.16 Patients were followed for up to 4 years: 15 men allocated for urethroplasty needed reintervention compared to 29 allocated for DVIU (HR 0.52 [0.31-0.89], P = .017). An increased risk of recurrence after repeated DVIU was found. However, both intervention groups had a similar improvement in symptom scores at 24 months of follow-up and similar adverse events, despite the increased recurrence rate with repeated DVIU. Therefore, the OPEN trial suggests that both options should be available for recurrent bulbar strictures as both procedures are likely to improve symptoms. Nevertheless, there is a longer duration of benefits with urethroplasty.

The OPEN trial16 did not differentiate in the number of previous endoscopic treatments and did not analyze the risk of recurrence after the first urethroplasty with multiple prior endoscopic treatments. The homogenous patient cohort and the analysis of four groups with different numbers of prior endoscopic treatment in our study provide relevant and new information for the treatment of recurrent bulbar strictures. We found no difference in recurrence rate in patients with one or two prior endoscopic treatments before first urethroplasty. Three or more prior endoscopic treatments, on the other hand, led to an increased risk of recurrence after the first urethroplasty.

The findings from our study and from previous literature suggest that a second endoscopic treatment of a recurrent bulbar stricture could be a valid treatment option, next to urethroplasty. Patients should be well informed about the recurrence risk with second endoscopic treatment. Our data show an increased recurrence risk after the first urethroplasty with ≥3 prior endoscopic treatments. We would therefore advise against a third (or more) endoscopic treatment for recurrent bulbar strictures and would strongly advise a patient to choose urethroplasty if a curative treatment is desired.

Also, a relation between ISD and ≥3 previous endoscopic treatments was found. Multivariate analysis of our data shows that those factors were no independent predictors, suggesting that ISD is often advised in patients with repeated endoscopic procedures.

The retrospective data collection limits the strength of this study although the work-up, treatment, and follow-up in this homogenous cohort were standardized and thorough. It is possible that with a larger cohort more significant differences would have been found.

It is important to note that stricture recurrence was defined as any need for reintervention, based on patient symptoms and diagnostic findings. If cystoscopy showed limited scarring at the operation site, without patient