Multi-Institutional Study of Dorsal Onlay Urethroplasty of the Membranous Urethra after Endoscopic Prostate Procedures: Operative Results, Continence, Erectile Function and Patient Reported Outcomes

(1) Background: To critically evaluate dorsal onlay buccal mucosal graft urethroplasty (DOBMGU) for posterior urethral stenosis repair following transurethral resection and other endoscopic prostate procedures. (2) Methods: A retrospective multi-institutional review of patients with membranous or bulbomembranous urethral stenosis for whom treatment with DOBMGU was conducted after receipt of prostate endoscopic procedures. Baseline data, peri-operative care, post-operative care and patient-reported outcomes were analyzed. The primary outcomes were procedural failure and development of de novo stress urinary incontinence (SUI). The secondary outcomes were changes in voiding, sexual function and patient satisfaction. (3) Results: A total of 107 men with a mean age of 69 ± 9.5 years and stenosis length of 3.5 ± 1.8 cm were included. Prior endoscopic procedures among participants were 47 patients (44%) with monopolar TURP, 33 (30.8%) with bipolar TURP, 16 (15%) with Greenlight laser, 9 (8.4%) with Holmium laser enucleation and 2 (1.9%) with bladder neck incision. At a mean follow-up time of 59.3 ± 45.1 months, stenosis recurred in 10 patients (9.35%). Multivariate analysis confirmed that postoperative complications (OR 12.5; p = 0.009), history of radiation (OR 8.3; p = 0.016) and ≥2 dilatations before urethroplasty (OR 8.3; p = 0.032) were independent predictors of recurrence. Only one patient (0.9%) developed de novo SUI. Patients experienced significant improvement in PVR (128 to 60 cc; p = 0.001), Uroflow (6.2 to 16.8 cc/s; p = 0.001), SHIM (11.5 to 11.7; p = 0.028), IPSS (20 to 7.7; p < 0.001) and QoL (4.4 to 1.7; p < 0.001), and 87 cases (81.3%) reported a GRA of + 2 or better. (4) Conclusions: DOBMGU is an effective and safe option for patients with posterior urethral stenosis following TURP and other prostate endoscopic procedures. This non-transecting approach minimizes external urinary sphincter manipulation, thus limiting postoperative risk of SUI or erectile dysfunction.


Introduction
Traditionally, transurethral resection of the prostate (TURP) has been the gold standard surgical treatment for lower urinary tract symptoms due to benign prostatic obstruction (BPO) [1,2]. More recently, other endoscopic approaches, such as KTP laser vaporization and Ho:YAG laser enucleation of the prostate (HoLEP), are proving to be good alternatives to TURP, and transurethral incision of the prostate remains a valid option in the case of small prostate size without a median lobe [3].
None of these approaches are devoid of complications, and the development of urethral stricture along with bladder neck contracture may appear during follow-up [4]. The incidence of urethral stenosis after TURP is reported to be 2.2 to 9.8% and occurs at this rate in either monopolar or bipolar TURP [1,5]. Many etiologic factors are likely involved, including false passage, mechanical trauma from the large caliber of the instruments utilized, longer duration of surgery and insufficient isolation of the electrical current [6,7].
Management of post-TURP urethral stenosis remains a challenge, and many urologists tend to initially employ endoscopic interventions with an elevated risk of recurrence [7]. Stenosis in the membranous or bulbomembranous urethra has special implications due to its proximity to the urethral sphincter and cavernous nerves [8,9]. Short bulbar and/or membranous stenosis can be treated using a tension-free anastomosis technique, but with well-documented risks of sexual dysfunction and stress urinary incontinence (SUI) [10]. The risk of incontinence after posterior excision urethral and primary anastomosis is also especially high in the case of previous radiation treatment [11][12][13].
Non-transecting techniques using dorsal onlay buccal mucosal graft urethroplasty (DOBMGU) might be preferred with long non-obliterative stenosis and seem to better preserve continence in post-TURP sphincteric membranous urethral strictures [14,15]. To date, there have been no direct comparisons between different techniques for post-TURP membranous stenosis. The aim of this study is to critically analyze a multi-institutional series of DOBMGU for membranous and/or bulbomembranous urethral stenosis in patients with previous TURP and related endoscopic prostate procedures for the treatment of BPO. We hypothesize that this procedure is feasible, provides durable patency and satisfactory patient-reported outcomes with minimal deleterious effects on continence and sexual function.

Materials and Methods
After obtaining institutional review board approval (A11/20), patients with a history of TURP or other endoscopic prostate procedures for BPO, who underwent DOBMGU for posterior urethral stenosis at participating institutions between 2010 and 2020, were retrospectively reviewed. In accordance with the preferred SIU/ICUD terminology [16], we reviewed cases with "posterior urethral stenosis" to include patients with stenosis of the membranous urethra, extending from the proximal bulbar urethra up to the distal verumontanum. Patients with a history of radiation before or after TURP were included. Patients with bladder neck contracture and urethral stenosis after simple open prostatectomy were excluded. Only patients with a minimum of six months of follow-up were evaluated.
The primary objective was evaluation of urethral patency and stenosis recurrence along with the incidence of de novo SUI following reconstruction. Secondary objectives included the evaluation of changes in voiding, sexual function and patient satisfaction. Patient demographics, medical history, radiation history, continence status and stenosis length were reviewed. Operative reports were reviewed to identify operative time, complications and estimated blood loss. Perioperative outcomes, stenosis recurrence, preoperative and post-operative uroflowmetry with post-void residual and validated patient reported outcomes (PROs) such as the Global Response Assessment (GRA), IPSS and SHIM questionnaires were assessed. Questionnaires were taken from the most recent follow-up visit during which they were available.

Surgical Technique
Dissection of the posterior urethra is performed dorso-laterally, preserving urethral continuity and sparing the bulbar vessels. The urethra is carefully dissected and rotated laterally to spare the striated muscle of the sphincter and the cavernous nerves, the space between the diverging crura is carefully exposed and the dorsal aspect of the urethra is incised longitudinally through the length of the stenosis, with extension approximately 1 cm proximal and distal into the normal urethra [14].
Prior to initiating the procedure, passage of a size 4-5 French urethral catheter or a guidewire assists in exposing the membranous urethra by allowing a sharp incision along the guide. No cautery is used in proximity to the crura of the corpora cavernosa, membranous urethra or cavernous nerve mesh. Sharp resection of peri-urethral scar tissue is limited to the area between the 11 and 1 o'clock positions, with careful preservation of the omega-shaped rhabdosphincter. At this point, the lumen is widely exposed and three stitches are passed outside-in, through the proximal end of the urethrotomy at the 11, 12 and 1 o'clock positions, to fix the proximal BMG in place. Careful use of an instrument such as a nasal speculum or gorget probe can assist in the placement of these proximal stitches. An appropriately sized oral graft is harvested and secured to the apices of the urethrotomy with the pre-placed sutures. Quilting sutures are used to fix the graft to the corpora cavernosa and the intercrural space.

Follow-Up
Patients were followed-up with at 3-to-6-month intervals postoperatively, and yearly thereafter. Stenosis recurrence was defined as the need for any intervention, including endoscopic treatments or the need to redo urethroplasty during follow-up. During follow-up in all institutions, patients underwent uroflowmetry and post-void residual measurements and responded to questionnaires. Endoscopic and radiographic follow-up protocols varied between sites, but flexible cystoscopy or radiographic studies were performed with a low threshold for patients inclined to recurrence based on post-operative maximum urinary flow deterioration, worsening IPSS and/or any subjective voiding complaint.

Patient-Reported Outcomes
Patient-reported outcomes were measured by validated questionnaires administered prior to a direct patient interview. The International Prostate Symptom Score with qualityof-life domain (IPSS-QOL) evaluated patient perception of urinary symptoms. The Sexual Health Inventory for Men (SHIM) assessed preoperative and postoperative erectile function. The Global Response Assessment (GRA) determined patients' satisfaction with the surgery. The GRA is a single query externally validated questionnaire that assesses patients' impressions of their change in symptoms, and scores range from −3 (markedly worse) to +3 (markedly improved).

Incontinence
Preoperative and postoperative continence status was assessed. Prior to reconstruction, patients were assigned as either preoperatively continent, preoperatively incontinent or unknown. Incontinence was defined as any urine loss. Based on the patient's clinical history and physical examination, incontinence type was assigned as stress incontinence, urge incontinence or mixed. The "unknown" status was assigned to patients in total urinary retention with suprapubic drainage. The same groups were clinically assessed postoperatively. Only one case with early failure retained a suprapubic tube and an unknown incontinence status after surgery.

Statistical Analysis
Paired t-tests compared preoperative and postoperative continuous variables following surgical repair. The Wilcoxon signed-rank test compared the preoperative and postoperative patient-reported survey data. The two-sided Fisher's exact test was per-formed for categorical patient-reported data. Factors affecting recurrence-free intervals were evaluated using the Kaplan-Meier analysis method. A multivariate analysis to evaluate factors for stenosis recurrence was performed by stepwise logistic regression with p = 0.1 entry and p = 0.05 stay criteria. A p-value < 0.05 was considered significant. The statistical analysis was performed using Statistical Analysis System 9.4 (SAS Institute Inc., Cary, NC, USA).

Sexual Function Outcomes
SHIM and other sexual aspects such as penile pain and weak ejaculation were evaluated pre-and postoperatively with data available in 87 patients (81.3%). Mean SHIM value improved from 11.5 ± 6.3 to 11.7 ± 7 (p = 0.028). Of 22 patients with a SHIM score of ≥17 (mild ED or no dysfunction) preoperatively, 2 (9%) scored ≤11 (severe to moderate ED) and 4 (18.2%) scored from 12-16 (mild to moderate ED) postoperatively. Conversely, of 22 patients with a SHIM score of ≤11 (severe to moderate ED), 9 (22%) improved to 12-16 and 2 (4.9%) to ≥17 postoperatively. There was no variation in the percentage of patients with penile pain (9.5% preoperatively vs. 3.8% postoperatively; p = 0.1) but the percentage of patients with postoperative weak ejaculation decreased from 24.5% preoperatively to 3.8% after surgery (p < 0.0001). However, narrative data on how many patients really could ejaculate before urethroplasty, and among them how they ejaculate after the urethral repair, are not available.

Global Response Assessment Outcomes
GRA outcomes were available in all 107 patients, of whom 97 (90.7%) assessed their situation as better than before (GRA + 1 or higher), 87 cases (81.3%) self-defined as much better than before (GRA of + 2 or + 3), and 70 (65.4%) registered themselves as "very much better than before" (GRA + 3). A regression analysis was performed to evaluate factors associated with the best patient perception, and the absence of radiation was the only independent factor associated with a patient definition of GRA + 3 (0R 0.12 (95% CI 0.02-0.62); p = 0.011).

Discussion
Due to the proximity of the rhabdosphincter and cavernous nerves to the membranous urethra, reconstruction of membranous stenosis implies a risk of urinary incontinence and erectile dysfunction, and this concern is greater in patients with previous TURP in which the bladder neck and the internal sphincteric mechanism may be disrupted [17,18]. This could explain the high rates of incontinence observed with excision and primary anastomosis in classic studies facing sphincteric strictures, especially after pelvic fracture and radiation [10,13,19,20].
There is increasing evidence that membranous dorsal onlay urethroplasty, not only in the context of radiation but also after TURP, does not compromise continence or erectile function in most patients [14,21,22]. The multi-institutional experience we present reveals a limited failure rate (9.35%) that is very comparable to intrasphincteric anastomotic urethroplasty after BPO surgery (3.9-10%) [17,23]. However, the mean stricture length of the patients treated with DOBMGU was longer (3.5 cm) and the proportion of de novo incontinence lower (0.9%) compared to intrasphincteric anastomosis (1.5-2.6 cm and 7.8-15%, respectively), and these studies excluded previous radiation and incontinence before urethroplasty [17,23]. Our study includes previous radiation and concomitant need of anterior urethroplasty and gives a better definition of the different types of incontinence both pre-and postoperatively. Additionally, we report favorable SHIM scores, other sexual symptoms (penile pain, ejaculation) and PROMs (both IPSS-QoL and GRA) following repair. Multi-institutional retrospective studies have the advantage of collecting a large number of patients, but at the same time suffer from the limitation of technical heterogeneity This study is also limited by its retrospective non-comparative design, but suggests that DOBMGU is a safe and efficacious technique to treat post-TURP bulbomembranous stenosis.
In fact, the European Guidelines on urethral stricture suggest conducting bulbomembranous urethral stenosis after TURP similar to bulbar strictures, either by excision and primary anastomosis or augmentation urethroplasty, with a graft according to the length and tightness of the stricture, while recognizing the high risk of incontinence when reconstruction takes place in the proximity of the external sphincter after damaging the bladder neck during BPO surgery [24]. However, surgical refinement allows careful sphincteric preservation and removal of the fibrous tissue using a dorsal urethrotomy. Erectile impairment after urethroplasty is largely mediated through cavernous nerve injury [8,25]. Cavernous nerves could be better preserved by a dorsal approach than a ventral or circumferential one, as nerve fibers stand in the 5 and 7 o'clock positions ( Figure 2). Accordingly, Hinata and colleagues have previously described a U-shaped cavernous nerve mesh outside the periprostatic region which would spare the 12 o'clock position [26].
neck during BPO surgery [24]. However, surgical refinement allows careful sphincteric preservation and removal of the fibrous tissue using a dorsal urethrotomy. Erectile impairment after urethroplasty is largely mediated through cavernous nerve injury [8,25]. Cavernous nerves could be better preserved by a dorsal approach than a ventral or circumferential one, as nerve fibers stand in the 5 and 7 o'clock positions (Figure 2). Accordingly, Hinata and colleagues have previously described a U-shaped cavernous nerve mesh outside the periprostatic region which would spare the 12 o'clock position [26].  Anatomic studies regarding the preservation of erectile function after bulboprostatic anastomosis led to better characterization of the cavernous neurovascular bundle paths, and also to the discovery of a delicate sheath of connective tissue between the wall of the membranous urethra and the surrounding circular fibers of the external sphincter [25,27]. This sheath can be used as a surgical plane to separate the urethral wall from the sphincter. The circular muscle fibers of the external sphincter can be reflected from the sheath until the underlying urethral wall is identified and dissected [8]. This allows intra-sphincteric dissection to perform bulboprostatic anastomosis or intrasphincteric anastomotic urethroplasty, as defined by Gómez et al., preserving continence and sparing the bulbar vessels [17]. This technique is optimal in cases with short and occlusive stenosis of the membranous urethra, especially in men after BPO surgery. However, cases with longer stenosis may benefit more from augmentation techniques.
It has been suggested and assumed, without evidence, that ventral onlay graft techniques in proximal bulbar and membranous urethra better preserve the sphincter and cavernous mesh than dorsal onlay grafts [9,28], but membranous dorsal and ventral approaches have not been compared to date. A recent retrospective study including 69 patients with post-TURP stenosis in the area of the distal sphincter confirms that careful ventral onlay graft urethroplasty is a suitable technique after BPO transurethral surgery, demonstrating 84% patency success at a median follow-up of 52 months and a 4.3% incidence of de novo incontinence [15]. The risks of recurrence and of de novo incontinence increase to 28.9% and 10.5%, respectively, when a bulbomembranous ventral onlay graft is performed after radiation [29]. In this sense, a recent multi-institutional assessment of dorsal onlay urethroplasty for post-radiation urethral stenosis revealed a 17.7% risk of recurrence and an 8.1% risk of de novo incontinence [21].
Posterior urethral stenosis repair following surgical endoscopic procedures for BPO is a very challenging situation. This series has been gathered from a retrospective evaluation of patients from nine academic institutions who were operated on and followed-up with over a long period. However, the number of cases per institution per year is not lineal, as not all centers adopted this technique at the same time and the volume of cases from each institution is variable. Our combined experience, including both surgical outcomes and PROs, suggests DOBMG urethroplasty can be a feasible option. However, important limita-tions should be recognized. First of all, the absence of a standardized preoperative protocol in these types of studies limits the accurate evaluation of whether a bulbomembranous stenosis is truly trans-sphincteric or not, especially in cases where urethroscopy does not allow access to the area. Antegrade cystoscopy and urethrogram may add limited information to help solve the issue. The description of the operative report is, although inaccurate, the most reliable observation to categorize the stenosis. Secondly, the definition of failure as the need for re-intervention may overestimate success. Finally, but not less importantly, PROs should be better estimated in prospective studies to avoid any likely bias.
This perplexing condition may be better treated and followed-up with in highly specialized centers to offer the maximum possibility of continence and erectile function preservation to patients. The classic option to treat bulbomembranous stenosis has been excision and primary anastomoses. However, in cases with previous TURP, evidence is accumulating in favor of intrasphincteric anastomotic urethroplasty and onlay graft techniques. The experience we report suggests dorsal BMG urethroplasty is efficacious and safe. Of course, prospective controlled trials comparing different techniques are needed to say which technique is most advantageous over the others. Therefore, further investigation is needed on the subject.

Conclusions
DOBMGU is a safe and efficacious technique to treat post-TURP bulbomembranous stenosis, with long-term durability. Patients at risk of recurrence are those with postoperative complications, a radiation history and ≥2 dilatations prior to urethroplasty. This non-transecting urethral reconstruction spares the bulbar vessels, cavernous nerves and striated sphincter, while further conferring a low risk of de novo SUI in prior TURP recipients. Further work is needed to compare this technique with other surgical approaches, such as ventral onlay, intrasphincteric anastomotic urethroplasty and classical excision and primary anastomosis, used to treat this challenging patient population. Data Availability Statement: Data supporting reported results will be provided by the correspondence author upon reasonable request.