Bladder Neck Contracture Following Radical Retropubic versus Robotic-Assisted Laparoscopic Prostatectomy

Abstract

Introduction

Radical retropubic prostatectomy (RRP) and robotic-assisted laparoscopic prostatectomy (RALP) are co-standard surgical therapies for localized prostatic adenocarcinoma. These surgical modalities offer similar outcomes; however, lower rate of bladder neck contracture (BNC) is amongst the touted benefits of RALP. The differences between approaches are largely elucidated through multiple-surgeon comparisons, which can be biased by differential experience and practice patterns. We aimed to eliminate inter-surgeon bias through this single-surgeon comparison of BNC rates following RRP and RALP.

Materials and Methods

We retrospectively reviewed all RRPs and RALPs performed by one surgeon over 4 years. We compared clinical characteristics, intraoperative and postoperative outcomes.

Results

RRP patients had more advanced cancer and a higher biochemical recurrence rate. No significant differences were noted between groups in rates of anastomotic leakage, BNC, or 12-month postoperative pad-free continence.

Conclusion

RRP offers similar outcomes to RALP with regard to postoperative urinary extravasation, urinary continence, and BNC.

Introduction

Radical retropubic prostatectomy (RRP) and robotic-assisted laparoscopic prostatectomy (RALP) are currently co-standard surgical therapies for localized prostate cancer. While both surgical approaches offer similar functional and oncologic outcomes [1, 2], benefits of RALP over RRP include reduced blood loss, transfusion rate, and duration of hospital stay [2, 3, 4]. A decreased rate of bladder neck contracture (BNC) is an additional proposed benefit of RALP [5, 6, 7, 8, 9, 10, 11, 12, 13, 14].

The incidence of BNC following RRP varies between 0.48 and 33% [6, 7, 8, 9, 10, 11, 12, 15]. By comparison, the BNC rate following RALP has been reported at 0–6.3% [6, 7, 8, 9, 10, 11, 12, 13, 14]. The superiority of RALP over RRP with regard to BNC rate has been attributed to the improved visual magnification, precision, and dexterity afforded by a robotic surgical approach during creation of a watertight urethrovesical anastomosis [6, 7, 8, 9, 10, 11, 13, 16, 17, 18, 19].

While the precise pathophysiology of BNC development has yet to be elucidated, a possible cause includes prolonged anastomotic leakage resulting in healing by secondary intent with subsequent scarring and bladder neck ischemia causing fibrosis [5, 6, 11, 15]. Identified risk factors for BNC following prostatectomy are age, prostate-specific antigen (PSA) level, cigarette smoking, prior transurethral resection of the prostate (TURP), and prior radiation [5, 6, 8, 9, 11, 14, 15, 20]. Technical factors associated with BNC development include surgical approach, low surgeon volume, poor mucosal apposition, absence of mucosal eversion, excessive intraoperative blood loss, urinary extravasation, and excessive narrowing of the urethrovesical anastomosis [5, 6, 7, 8, 9, 10, 11, 15].

BNC typically presents clinically within 6 months of radical prostatectomy (nearly all cases present within 12 months) with symptoms of poor urinary stream, prolonged postoperative incontinence, urinary frequency, urgency, incomplete voiding, and/or nocturia [5, 6, 8, 9, 10, 15]. Significant morbidities are associated with BNC including urinary retention, urinary tract infection, and the need for a secondary procedure for repair [5, 6, 7, 8, 15].

Though improved BNC rates with RALP compared to RRP have been reported, differences in training, experience, and practice patterns complicate multiple-surgeon comparisons [2, 21]. We aim to diminish these sources of inter-surgeon bias through this single surgeon comparison of BNC rate following RALP versus RRP.

Materials and Methods

Upon Institutional Review Board approval, medical records of 117 men with prostatic adenocarcinoma who underwent prostatectomy (74 consecutive RRPs and 43 consecutive RALPs) from February 2010 through May 2014 were retrospectively reviewed. A single surgeon (J.A.B) performed all procedures with routine assistance by resident physicians. This surgeon received prior laparoscopic and robotic prostatectomy fellowship training and performed over 200 open retropubic, 35 standard laparoscopic, and 25 RALPs as a faculty surgeon prior to initiating this study. When counseling these men preoperatively, the main indication for RRP over RALP was higher-risk disease with the need for more extensive lymphadenectomy; however continence status and patient preference also contributed to the decision of RRP versus RALP.

Patient demographic data, preoperative lower urinary tract dysfunction, PSA level, pathology, continence at 12 months post-operatively, anastomotic leak, and BNC rate were recorded. Continence was defined as the use of no pads per day.

Operative Technique

For RRPs, the urethrovesical anastomosis was performed using a Capio™ RP suture-capturing device (Boston Scientific, Marlborough, MA) to facilitate placement of 6–8 interrupted sutures at the 2, 4, 8, 10, 12, and 6-o'clock positions with 2–0 Monocryl suture. For RALPs, the da Vinci® SI Surgical System (Intuitive Surgical, Inc., Sunnyvale, CA) was used to perform a double-arm suture running urethrovesical anastomosis using two 3–0 V-lock sutures. With both operative techniques, watertight anastomosis was ensured through visual inspection for anastomotic leakage following bladder distention with approximately 100 ml of normal saline. Resident physicians were involved in various aspects of each procedure, including routine performance of a portion of the anastomosis in both cohorts.

Diagnosis of BNC

One RRP patient was diagnosed with BNC by cystoscopy at an outside facility. All other patients with BNC were diagnosed at our institution using flexible cystoscopy. With the exception of one patient, all investigations for BNC were prompted by urinary symptoms, including decreased urinary stream, recurrent urinary tract infection, persistent stress urinary incontinence, and sensation of incomplete voiding with or without elevated post-void residual volume. The one remaining patient was diagnosed incidentally following the inability to pass a Foley catheter during an unrelated emergency department presentation.

Statistical Analysis

Bivariate analysis was performed using the Fisher's exact test for categorical variables and Students t-test. All statistical testing was two-sided and assessed for significance at the 5% level using Prism V 6.05 (GraphPad, San Diego, CA).

Results

Preoperative Baseline Data

Of the 74 men initially reviewed in the RRP group, 70 were included in the final analysis. Four aborted cases were excluded; 2 cases early in the series were aborted due to the finding of grossly positive lymph node metastases, 1 case due to presence of extensive surgical mesh behind and above the symphysis pubis, and 1 case due to the inability to place the patient supine secondary to ankylosing spondylitis. All 43 RALP cases were included in our analysis.

At the time of procedure, the RRP cohort was significantly younger than the RALP cohort and had a larger proportion of current cigarette smokers. There was no difference in rates of prior TURP or preoperative LUTD between the RRP and RALP groups (table 1). The RRP cohort had a significantly higher median preoperative PSA (7.2 vs. 5.4 mg/l, p ≤ 0.01). Additionally, the RRP cohort had higher proportions of patients with preoperative Gleason scores > 7 (20.0 vs. 4.7%, p = 0.07) and high-risk D'Amico categorization [22] (27.1 vs. 9.3%, p = 0.07), however these values did not reach statistical significance.

Table 1.

Baseline demographic data

	RRP	RALP	p
Number of patients	70	43
Mean ± SEM
Age, years	60.3 ± 0.8	64.0 ± 1.0	< 0.01
BMI, kg/m2	28.6 ± 0.5	30.2 ± 0.9	0.19
Number (%)
Current smokers, n (%)	22 (31.4)	4 (9.3)	0.02
Diabetes, n (%)	8 (11.4)	4 (9.3)	1.00
Prior TURP, n (%)	3 (4.3)	0 (0.0)	0.16
Preoperative LUTD, n (%)	5 (7.1)	4 (9.3)	0.71

Intraoperative, Perioperative, and Postoperative Data

When compared to RALPs, RRPs were performed in a significantly shorter mean operating time (table 2). Additionally, the RRP group experienced a higher median estimated blood loss (table 2), and a higher blood transfusion rate (table 3). There was no difference in duration of catheterization postoperatively between groups (table 2). A greater number of RRP specimens were categorized as pathologic stage T3 (30.0 vs. 16.3%, p = 0.11) and Gleason grade > 7 (20.0 vs. 9.3%, p = 0.07), and a greater number of RRP patients experienced biochemical recurrence at last follow-up (24.2 vs. 2.3%, p ≤ 0.01).

Table 2.

Operative outcomes

	RRP	RALP	p
Mean ± SEM
Operative time, minutes	249 ± 5	352 ± 11	< 0.01
Median (range)
EBL, ml	800 (200–3,000)	100 (25–450)	< 0.01
Catheter removal, post operative day	15 (8–36)	13.5 (8–27)	0.17
Number (%)
Presenting for 12 month follow-up	54	30
Pad-free continence*	35 (64.8)	19 (63.3)	0.95

EBL = Estimated blood loss.

^*Percentages reflect the proportion of patients present at 12 month follow-up.

Table 3.

Complications

	RRP	RALP	p
Number (%)
Clinically significant anastomotic leak	3 (4.3)	2 (4.7)	1.00
Blood transfusion	12 (17.1)	0 (0.0)	< 0.01
High-grade complications	17 (24.3)	6 (14.0)	0.23
BNC	2 (2.9)	2 (4.7)	0.63

Complications and Functional Outcomes

There was no difference in overall incidence of clinically significant anastomotic leak between RRP and RALP groups (table 3). Three and 2 (4.3 and 4.7%) patients were diagnosed with a clinically significant anastomotic leak in the RRP and RALP groups, respectively. A clinically significant anastomotic leak was defined as elevated drain creatinine (≥ 0.2 mg/dl above serum creatinine level) measured on postoperative day 1 or 2 that did not resolve prior to hospital discharge, and that necessitated drain placement beyond the duration of hospitalization or catheter placement beyond postoperative day 18. Foley catheters were routinely scheduled for removal at 2 weeks (± 3 days) postoperatively. A single RALP patient without elevated drain creatinine or prolonged drain placement was diagnosed with posterior urine leak by computed tomography after presenting to the emergency department on postoperative day 4 with complaints of lower abdominal pain and gross hematuria; otherwise all patients with clinically significant anastomotic leak were diagnosed by our standard definition. The RALP patient diagnosed with anastomotic leak by computed tomography subsequently developed a urinoma requiring treatment by interventional radiology with drain placement. All other clinically significant anastomotic leaks resolved with conservative management within 36 days (range 13–36 days). None of the men that experienced anastomotic leak were diagnosed with subsequent BNC. There was no difference in high-grade complications (defined as Clavien-Dindo Classification grade III or greater) [23] between the RRP and RALP cohorts (table 3).

There was no difference in overall BNC rate between the 2 cohorts (table 3). Two patients were diagnosed with BNC in each of the RRP and RALP groups (2.9 and 4.7%, respectively). One RRP patient noted weakened urinary stream, beginning approximately 7 months post-operatively, and subsequently presented to a local urologist where he was diagnosed with BNC by cystoscopy. The other presented to the emergency department with acute mental status changes and was incidentally found to have BNC with urinary retention following the inability to pass a Foley catheter at 18 months postoperatively. Both patients were diagnosed via cystoscopy and treated by BNC dilation. In the RALP group 1 patient presented at 15 months postoperatively with symptoms of straining during urination and an elevated post-void residual volume that had progressed over the previous year. He was found to have BNC diagnosed by cystoscopy, but the patient was minimally bothered by his symptoms and no further intervention was pursued. The second RALP patient presented with weak urinary stream and urinary retention at approximately 10 months postoperatively. BNC was diagnosed by cystoscopy. Initially he was treated with dilation with subsequent transurethral incision of the bladder neck after the dilation failed. This patient did well for approximately 2 months before once again noting weakened urinary stream, and cystoscopy revealed near concentric scarring diffusely throughout the bladder neck. A transurethral resection of the bladder neck was then performed.

Among the men who presented for follow-up at 12 months postoperatively (n = 54 for the RRP cohort, n = 30 for the RALP cohort), there was no difference in complete pad-free urinary continence between RRP and RALP groups at that time (table 2). The remaining 16 men in the original RRP cohort and 13 men in the original RALP cohort were lost to follow-up prior to the 12-month postoperative time point.

Discussion

By examining a single surgeon's anastomotic leakage, BNC and pad-free continence rates following RRP and RALP, we sought to eliminate the bias present in multiple surgeon comparison studies. The incidence of anastomotic leakage and BNC following our RRP and RALP procedures fell within previously reported rates. Interestingly, despite the increased presence of known BNC risk factors in the RRP cohort relative to the RALP cohort, including a greater prevalence of cigarette smoking, higher preoperative PSA values, and increased intra-operative blood loss, there was no difference in rate of BNC between the 2 groups. The 3 men with symptomatic BNCs experienced symptom onset within the expected first 12 postoperative months with the typical complaints of LUTD. However, one RRP patient had an asymptomatic BNC discovered incidentally 18 months after surgery. Therefore, the clinically symptomatic BNC rate was lower for the RRP cohort (n = 1, 1.4%) than the RALP cohort (n = 2, 4.7%, p = 0.56). It is possible there are additional patients with minimally symptomatic or asymptomatic BNCs in both cohorts. We did not evaluate asymptomatic patients for the presence of a sub-clinical BNC in either cohort.

We observed the previously demonstrated benefits of RALP over RRP of decreased blood loss and lower transfusion rate. Additionally, our finding of similar pad-free urinary continence rates of approximately 64% between the 2 surgical modalities is consistent with previous reports. However, our finding of equivalent BNC rates following RRP and RALP, despite the presence of a greater number of risk factors associated with BNC in the RRP group, refutes previous suggestions of lower BNC incidence as a benefit of RALP in the management of prostatic adenocarcinoma.

While visual confirmation of the intact bladder neck is an advantage of RALP, there are several potential explanations for our comparable RRP BNC rate. These include: highly accurate placement of interrupted sutures using Capio™ RP suture-capturing device [7], increased ability to preserve an intact bladder neck during RRP compared with RALP, routine performance of mucosal eversion during RRP, and less tension during the approximation of the posterior anastomosis in the supine rather than steep Trendelenburg position.

Conclusion

Open radical retropubic prostatectomy using the Capio™ RP suture-capturing device appears to offer similar outcomes to RALP with regard to the development of postoperative urinary extravasation, BNC, and the recovery of postoperative urinary continence.