Bladder neck contracture-incidence and management following contemporary robot assisted radical prostatectomy technique.

PURPOSE: Bladder neck contracture (BNC) is a well-recognized complication following robot-assisted radical prostatectomy (RARP) for treatment of localized prostate cancer with a reported incidence of up to 1.4%. In this series, we report our institutional experience and management results.
METHODS: A prospectively collected database of patients who underwent RARP by a single surgeon from 2006 to 2012 was reviewed. Watertight bladder neck to urethral anastomosis was performed over 18-French foley catheter. BNC was diagnosed by flexible cystoscopy in patients who developed symptoms of bladder outlet obstruction. Subsequently, these patients underwent cold knife bladder neck incisions. Patients then followed a strict self regimen of clean intermittent catheterization (CIC). We identify the patient demographics, incidence of BNC, associated risk factors and success of subsequent management.
RESULTS: Total of 930 patients who underwent RARP for localized prostate cancer was identified. BNC was identified in 15 patients, 1.6% incidence. Mean patient age and preoperative prostate-specific antigen was 58.8 years old and 7.83 ng/mL (range, 2.5-14.55 ng/mL) respectively. Mean estimated blood loss was 361±193 mL (range, 50-650 mL). Follow-up was mean of 23.4 months. Average time to BNC diagnosis was 5.5 months. In three patients, a foreign body was identified at bladder neck. On multivariate analysis, estimated blood loss was significantly associated with development of BNC. All patients underwent cystoscopy and bladder neck incision with a 3-month CIC regimen. Out of 15 index patients, none had a BNC recurrence over the follow-up period.
CONCLUSIONS: BNC was identified in 1.6% of patients in our series following RARP. Intraoperative blood loss was a significant risk factor for BNC. In 20% of BNC patients a migrated foreign body was noted at vesicourethral anastomosis. Primary management of patients with BNC following RARP should be bladder neck incision and self CIC regimen.

INTRODUCTION

In the era of robot-assisted radical prostatectomy, the incidence of post-operative bladder neck contracture (BNC) is reported to be up to 1.4% [1-9]. In comparison, a higher variability of range of 5% to 32% has been observed with open retropubic prostatectomy techniques [2,3,6,10,11]. Usually resulting from scar tissue at the anastomosis between bladder neck and membranous urethra, BNC can result in post prostatectomy bladder outlet obstruction. The diagnosis should be suspected in patients with poor urinary stream or prolonged unexplained incontinence. Associated morbidity of BNC may include urinary retention, urinary incontinence, urinary tract infection and need for further surgical interventions.

Although pathophysiology of BNC is multifactorial, preventing anastomotic leak and hematoma is important. Achieving watertight closure and meticulous mucosal apposition during vesicourethral anastomosis is crucial in preventing postoperative urine leakage leading to accumulation of urine [9-13]. Prolonged urine leakage resulting from anastomotic gap may lead to healing by secondary intent, subsequent scarring and contracture at the bladder neck. Hematoma may contribute to development of BNC due to distraction or tissue ischemia of anastomosis [13-15].

Contemporary robot-assisted surgical techniques offer a magnified stereoscopic view of tissue planes and improved precision and dexterity, contributing to a decrease in the anastomotic complications [2]. Various management techniques of BNC have been proposed in open prostatectomy patients; however, ideal treatment of such complications in the minimally invasive era is less well understood. In this large series of single institutional data, we evaluate the patient demographics, perioperative outcomes, incidence of BNC, associated risk factors and our experience of subsequent management.

MATERIALS AND METHODS

Patient cohort

Patients were was prospectively enrolled into a Institutional Review Board approved database consisting of men with clinically localized prostate cancer undergoing robot-assisted radical prostatectomy (RARP) by a single surgeon (I.Y.K) from 2006 to 2012. The results were reviewed retrospectively. Primary and secondary endpoints were the incidence of BNC and associated risk factors respectively. Our protocol and results of management are also reported.

Surgical technique

Using daVinci Sugical Systems (Intuitive Surgical, Sunnyvale, CA, USA), radical prostatectomy was performed in an intraperitoneal antegrade fashion with initial dissection of bladder neck latero-posteriorly followed by seminal vesicles. Nerve sparing when indicated was performed without the use of electrocautery. Hem-o-lok surgical clips, titanium clips, as well as suture ligature with Lapra-Ty were occasionally utilized for hemostatic control of vascular pedicles. Watertight bladder neck to urethral anastomosis was performed over an 18-French catheter using the Van Velthovan technique [16] with a double armed 3-0 monocryl suture in a running fashion. Bladder was stressed with up to 140 mL of saline routinely to ensure anastomotic closure. Intrapelvic drains were rarely placed with majority removed on the following postoperative day. Patients were discharged home on postoperative day one with an 18-French foley catheter to gravity drainage. Catheter was removed at first postoperative visit in seven days.

Postoperative BNC

BNC was diagnosed by flexible cystoscopy in patients who developed signs and symptoms of bladder outlet obstruction. Subsequently, these patients underwent endoscopic bladder neck incisions using cold knife urethrotomy under anesthesia. Incisions were performed at 12, 3, and 9 o’clock position until bleeding of healthy tissue was noted along with easy passage of 21 French rigid cystoscope. No patient received intraoperative steroid instillation. Patients were then prescribed a strict self regimen of clean intermittent catheterization (CIC) using 16-French catheter once daily for one month, then every other day for one month, and finally once a week for one month.

Statistics

Patient demographics, perioperative detail, BNC related complications and follow-up information is analyzed and presented. Statistical measure of central tendency including mean, median, range, and standard deviation of clinical parameters were calculated. Comparison of perioperative variables including age, body mass index (BMI), preoperative prostate-specific antigen (PSA), prostate weight, pathologic stage, grade, estimated blood loss, operative time was performed using the Student t-test. Significant risk factors were analyzed through multivariate logistic regression analysis. P-value of <0.05 was considered statistically significant. SPSS ver. 12.0 (SPSS Inc., Chicago, IL, USA) was utilized.

RESULTS

Total of 930 consecutive patients were identified who underwent robot-assisted radical prostatectomy for localized prostate cancer. Patients were followed clinically for mean of 23.4 months; median follow-up was 24 months. BNC was identified in 15 patients, an incidence rate of 1.6%. Patient demographics are listed in Table 1. The two groups were similar in terms of age, BMI, preoperative PSA, prostate weight, pathologic stage, grade, and operative time. Mean patient age was 58.8 years old with an average preoperative total PSA of 7.83 ng/mL (range, 2.5–14.55 ng/mL). Mean operative time was 200 minutes. Mean estimated blood loss was 361±193 mL (range, 50–650 mL).

Table 1.

Patient demographics

Patient	Age (yr)	BMI (kg/m2)	OR time (min)	Estimated blood loss (mL)	Pre-op PSA (ng/mL)	Pathology	Gleason score	Prostate wt (g)	Margin status	Presentation	Time to BNC (mo)	OR findings
1	51	27	150	500	8.0	T2c	3+4	81	Negative	Urinary retention	2.6
2	58	35	215	650	4.7	T3a	3+4	46	Positive (right posterior)	Urinary retention	14.0
3	63	27	240	500	4.7	T2c	4+3	53	Negative	Hematuria/clot retention	8.6
4	63	26	230	400	9.8	T2c	5+4	42	Negative	Difficulty voiding	2.2
5	51	31	200	500	4.1	T2b	3+3	59	Negative	Difficulty voiding	3.8
6	57	29	205	600	12.8	T2c	3+4	39	Negative	Difficulty voiding	7.0
7	63	31	215	300	9.2	T3a	4+5	42	Negative	Difficulty voiding	4.5
8	54	26	215	200	8.6	T2c	3+4	45	Negative	Difficulty voiding	3.3
9	66	28	160	50	5.1	T2c	3+4	35	Positive (Apex)	Difficulty voiding	12.0
10	61	21	130	100	7.3	T2c	3+4	37	Negative	Difficulty voiding	3.3
11	65	26	240	500	14.6	T2c	3+3	41	Negative	Urinary retention	2.6	Lapra-Ty
12	50	29	215	350	5.4	T2c	3+4	38	Negative	Difficulty voiding	1.5
13	57	28	195	100	13.0	T2c	3+3	59	Negative	Difficulty voiding	4.2
14	66	32	-	-	-	-	-	-		Difficulty voiding	7 Years	Titanium staple
15	57	28	170	300	2.5	T2c	3+3	48	Negative	Difficulty voiding	7.1	Hem-o-lock

BMI, body mass index; OR, operation; Pre-op, preoperative; PSA, prostate-specific antigen; BNC, bladder neck contracture.

Intraoperative blood loss was significantly greater in the cohort who developed BNC (P=0.001) (Table 2). Using logistic regression model, a multivariate analysis of the preoperative variables indicates that intraoperative blood loss is a significant predictor for BNC development (P<0.001) (Table 3). None of the patients who developed BNC had prior history of transurethral resection of prostate (TURP). Excluding one subject who developed BNC after 7 years, the remaining patients presented within 5.5 months (mean) and range up to 14 months in the postoperative interval. As demonstrated in Table 4, there was no difference in BNC rate from the first 500 to the last 430 cases analyzed in this study.

Table 2.

Comparison of baseline characteristics

Variable	BNC (n= 14)	No BNC (n= 916)	P-value
Age (yr)	58.3±5.4	59.5±6.8	0.513
Preoperative PSA (ng/mL)	7.8±3.7	6.8±13.3	0.763
Prostate weight (g)	47.5±12.3	46.2±15.3	0.754
Pathologic T stage			0.748
≤T2	12 (85.7)	704 (76.9)
≥T3	2 (14.3)	212 (23.1)
Gleason score			0.532
≤6	4 (28.6)	397 (43.3)
7	8 (57.1)	426 (46.5)
≥8	2 (14.3)	93 (10.2)
Estimated blood loss (mL)	360.7±193.3	220.1±157.1	0.001
Operation time (min)	198.6±33.9	201.5±47.4	0.818
Body mass index (kg/m2)	27.9±3.3	28.4±5.0	0.705

Values are presented as mean±standard deviation or number (%). PSA, prostate-specific antigen; BNC, bladder neck contracture.

Table 3.

Multivariate logistic regression analysis for prediction of bladder neck contracture

Variable	OR (95% CI)	P-value
Age (yr)	0.968 (0.890–1.052)	0.442
Preoperative PSA (ng/mL)	1.003 (0.975–1.031)	0.829
Prostate weight (g)	1.003 (0.970–1.037)	0.869
Pathologic T stage (≤T2 vs. ≥T3)	0.344 (0.070–1.699)	0.190
Gleason score
≤6	1.000
7	2.548 (0.735–8.833)	0.140
≥8	5.809 (0.866–38.962)	0.070
Estimated blood loss (mL)	1.005 (1.002–1.008)	<0.001
Operation time (min)	0.989 (0.973–1.004)	0.156
Body mass index (kg/m2)	0.951 (0.830–1.089)	0.469

PSA, prostate-specific antigen; OR, odds ratio; CI, confidence interval.

Table 4.

Surgeon volume and BNC rates after RARP

	Surgeon volume		Total	P-value
	Before 500	After 500
Total cases	500	430	930	0.426
No BNC
Cases	491	425	916
% Within surgeon volume	98.2%	98.8%	98.5%
BNC
Cases	9	5	14
% Within surgeon volume	1.8%	1.2%	1.5%

BNC, bladder neck contracture; RARP, robot-assisted radical prostatectomy.

BNC was treated with cystoscopy and bladder neck incision under general anesthesia (Figs. 1, 2). In three patients, migrated foreign body was identified at bladder neck (Fig. 3). First patient had Hem-o-lock clip, second patient had Lapra-Ty and third had a titanium surgical clip. Catheter was placed at the end of endoscopic urethrotomy procedure which was subsequently removed within one week post procedure. Patients were then provided with verbal and written instructions for CIC regimen for the next 3 months. Out of the 15 patients diagnosed with post prostatectomy BNC and further treated as described above, none had a recurrence over the follow-up period.

Fig. 1.

Bladder neck contracture.

Fig. 2.

BNC incised using cold knife urethrotomy.

Fig. 3.

Calcified Hem-o-lok clip identified during cystoscopy.

DISCUSSION

BNC has been encountered less frequently in the era of RARP (Table 5). Improved visual magnification and dexterity of robotic instruments have contributed to the observed difference [2,4]. Both surgical techniques and patient-related factors have been evaluated in contributing to development of BNC. Traditionally described risk factors include prior radiation, absence of mucosal eversion, poor vesicourethral mucosal apposition, urinary extravasation, increased blood loss, ischemia of bladder neck and excessive intraoperative narrowing at site of anastomosis [9,12].

Table 5.

Comparison of contemporary literature

Literature	Sample size	Follow-up period (mo)	Time to BNC (mo)	Surgical approach	BNC incidence	Management	Results
Webb et al. [9]	100	14.3	Median, 2.2	Robotic	None	-	-
Breyer et al. [2]	293	12.0	Median, 4.7; range (1–15)	Robotic	1.4%	Balloon dilation, rarely CKI	48% Required ≥2 treatments
Msezane et al. [3]	634	19.5	Mean, 4.8; range (3–12)	Robotic	1.1%	CKI	No recurrence
Carlsson et al. [4]	1253	19.0	<15	Robotic	0.2%	-	-
Hu et al. [5]	322	-	-	Robotic	0.6%	CKI	-
Menon et al. [6]	1,100	12.0	-	Robotic	0.8%
Coelho et al. [7]	2,500	Median 25	-	Robotic	0.1%	CKI	No recurrence
Zorn et al. [8]	300	17.3	-	Robotic	1.4%	CKI	No recurrence
Gonzalgo et al. [30]	246	13.7	4–6	Lap	1.2%	Dilation, CKI	No recurrence
Hu et al. [5]	358	-	-	Lap	2.2%	CKI
Rassweiler et al. [29]	180	Median 12	-	Lap	3.3%	Laser incision	-

BNC, bladder neck contracture; CKI, cold knife incision.

In our series, out of 930 consecutive patients who underwent RARP for localized prostate cancer, BNC was noted in 15 patients, 1.6% incidence. Patients became symptomatic within mean of 5.5 months, excluding a single patient who became symptomatic of BNC at 7 years. Clinical history often included urinary symptoms of voiding difficulty, hesitancy, frequency, urgency or nocturia. Our findings are consistent with previously reported rates in literature [17] (Table 4).

In a retrospective analysis by Webb et al. [9], total of 200 consecutive patients were followed for mean of 39.8 months in the open radical prostatectomy (ORP) group and 14.3 months in the RARP group. Among the patients who underwent ORP (n=100), 9 developed BNC. In the robot assisted arm (n=100), none of the patients had similar complication. The time to occurrence of BNC in most patients was within 8 weeks.

Our results demonstrate that estimated blood loss is significantly associated with and is prognostic for the development of post robot-assisted radical prostatectomy BNC [17]. Excessive intraoperative bleeding can obscure vision of the operative field and may lead to imprecise dissection of tissue planes. This is particularly of greater consequence in laparoscopic and robot-assisted techniques where impaired visual cues may complicate the watertight mucosal apposition needed for vesicourethral anastomosis. Presence of hematoma or urinoma may lead to anastomotic ischemia and resultant scar formation. Garg and See [18] in their retrospective review of 294 radical retropubic prostatectomy patients reported that intraoperative blood loss greater than 1 L and urine leakage correlated to the formation of anastomotic stricture.

Elliott et al. [19] reviewed CaPSURE database to determine incidence of urethral strictures including BNC following primary treatment of clinically localized prostate cancer. Treatment modalities included watchful waiting, androgen deprivation therapy, radical prostatectomy, brachytherapy, external beam radiotherapy, cryotherapy and combination of these therapies. Among the 6,597 men with localized prostate cancer identified through the database, the overall stricture treatment rate was 5.2% (range, 1.1%–8.4% by prostate cancer treatment type). Radical prostatectomy was associated with the highest rate of stricture of 8.4%. In multivariate analysis, type of treatment (radical prostatectomy vs. watchful waiting; hazard ratio, 10.440 [95% confidence interval, 3.276–33.272]), BMI (obese vs. not overweight; hazard ratio, 2.254 [95% confidence interval, 1.566–3.244]) and age (≥70 years old vs. younger; hazard ratio, 2.206 [95% confidence interval, 1.514–3.215]) were significant predictors of stricture treatment.

Although some series have identified prior TURP as a risk factor for BNC due to resulting fibrosis, it was not found to be the case in a large retrospective review by Borboroglu et al. [12]. Prior TURP, type of anastomotic suture used, size of catheter or duration of catheterization were not related to development of BNC. Interestingly, multivariate analysis of their results suggested current cigarette smoking as a strong predictor of BNC development (P<0.001).

Treatment options must balance the therapies to promote bladder emptying while preserving urinary continence. Management can utilize self-dilation schedule, cystoscopic incision or resection, urethral stent, and bladder neck reconstruction for severe cases. Post prostatectomy BNC with mild degree of stenosis can be successfully treated with transurethral dilation or transurethral incision with a success range of 25% to 73% [12]. More fibrotic contractures may require bladder neck incision or even bladder neck resection. Popken et al. [10] retrospectively analyzed data from 340 radical prostatectomy patients over eight year period and found that 24 cases (7%) required transurethral endoscopic treatment. In their 12- to 72-month follow-up, no significant incontinence was observed. Most commonly, the site of stricture was located below the bladder neck musculature and above the distal urethral sphincter.

There is lack of a standardized protocol for managing patients with BNC. CIC has been popularized as a conservative tool for management of neurogenic bladders, benign prostatic hyperplasia, and BNC [20,21]. To our knowledge, data on specific role of CIC in BNC is scarce. Patel et al. [22] recommended CIC to ensure complete emptying and to maintain the patency after endoscopic incision to correct BNC among patients who had an orthotopic urinary diversion after radical cystectomy. Our experience suggests that endoscopic incision with cold-knife followed by a CIC regimen is associated with a very high success rate. Specifically, during the mean post prostatectomy follow-up period of 23.4 months, no patient experienced a BNC recurrence following our management protocol. Additionally, since no patients required a more aggressive management of BNC such as Urolume stent or radical resection, it is likely that the nature of BNC following RARP is limited and focal.

In three patients in our cohort, a migrated foreign body was identified at bladder neck. Individual case reports and retrospective case reviews exist in literature identifying Weck Hem-o-lok, Lapra-Ty suture clip migration into bladder, bladder neck and even rectum [23-27]. In a cohort of 524 patients who underwent RARP, Blumenthal et al. [28] reported that out of the 4 patients (0.76%) who developed BNC, 2 cases were associated with Hem-o-lok clip migration and erosion into the vesicourethral anastomosis. In addition, 1 patient was discovered to have clip migration into bladder. The authors also refer to similar cases reported to Food and Drug Administration Manufacturer and User Facility Device Experience database. Given our findings and the similar citations in literature, the use of such clips should be used with caution. We ensure and extract any loose clips during the surgery.

The study is limited by the design of a retrospective review of prospectively maintained institutional database. In addition, per institutional protocol, most patients returned to their primary urologist after the stabilization of urinary and sexual function. Thus extended follow-up in these patients were limited to communications with the initial referring urologists and hence BNC may be underrepresented in our results. Lastly, we did not perform any urodynamic procedures. Therefore, we could not investigate changes in urodynamic parameters before and after the treatment of BNC.

In conclusions, as the techniques of robot-assisted radical prostatectomy have improved, the incidence of BNC has declined. In our institutional experience, patients who underwent RARP for localized prostate cancer, the incidence of postoperative BNC was found to be 1.6%. Intraoperative blood loss is a significant predictor for BNC development. Surgical clips should be utilized judiciously as migration is a recognized phenomenon and can be associated with subsequent BNC. Subsequent management of BNC with cold-knife urethrotomy followed by a self CIC schedule leads to successful resolution of the bladder outlet obstruction related symptoms.