Risk Factors and Quality of Life for Post-prostatectomy Vesicourethral Anastomotic Stenoses

Risk Factors and Quality of Life for Post-prostatectomy Vesicourethral Anastomotic Stenoses

Reconstructive Urology Risk Factors and Quality of Life for Post-prostatectomy Vesicourethral Anastomotic Stenoses Rou Wang, David P. Wood, Jr, Brent ...

480KB Sizes 0 Downloads 13 Views

Reconstructive Urology Risk Factors and Quality of Life for Post-prostatectomy Vesicourethral Anastomotic Stenoses Rou Wang, David P. Wood, Jr, Brent K. Hollenbeck, Amy Y. Li, Chang He, James E. Montie, and Jerilyn M. Latini OBJECTIVE

METHODS

RESULTS

CONCLUSION

To evaluate the difference in vesicourethral anastomotic stenosis (VUAS) rates after open radical retropubic prostatectomy (RRP) vs robot-assisted radical prostatectomy (RARP), and to analyze associated factors and effect on quality of life. From 2001 to 2009, a total of 1038 patients underwent RARP and 707 patients underwent open RRP. Perioperative factors and Expanded Prostate Cancer Index Composite (EPIC) quality of life scores were compared between patients who did and did not develop a VUAS. Independent significant predictors of VUAS development were identified using multivariable modeling. The incidence of VUAS in open RRP cases was higher (53/707, 7.5%) than for RARP (22/1038, 2.1%) (P ⬍ .0001). Intervention consisted of dilation in 34 of 75 cases (45.3%), internal urethrotomy in 8 of 75 (10.7%), and multiple procedures in 30 of 75 (40%). Open technique (P ⬍ .0001, odds ratio [OR] ⫽ 3.0, 95% confidence interval [CI] ⫽ 1.8-5.2), prostate-specific antigen (PSA) recurrence (P ⫽ .02, OR ⫽ 2.2, 95% CI ⫽ 1.2-4.1), postoperative hematuria (P ⫽ .02, OR ⫽ 3.7, 95% CI ⫽ 1.2-11.3), urinary leak (P ⫽ .002, OR ⫽ 6.0, 95% CI ⫽ 1.9-19.2), and urinary retention (P ⫽ .004, OR ⫽ 3.5, 95% CI ⫽ 1.5-8.7) were significant independent predictors of VUAS development. EPIC incontinence scores were similar between VUAS and non-VUAS patients, whereas irritative voiding scores were worse initially with VUAS but became similar by 12 months. There is a higher rate of VUAS after open RRP vs RARP. Most cases of VUAS require endoscopic intervention. Predictors include open surgery, PSA recurrence, and postoperative hematuria, urinary leak, and retention. There is no diminution of quality of life scores at 12 months. UROLOGY 79: 449 – 457, 2012. © 2012 Published by Elsevier Inc.

R

obot-assisted radical prostatectomy (RARP) is rapidly becoming the primary surgical treatment for clinically localized prostate cancer. Despite ongoing debate regarding comparative outcomes of RARP vs radical retropubic prostatectomy (RRP), the incidence of vesicourethral anastomotic stenosis (VUAS), more commonly termed a bladder neck contracture, seems to be lower with RARP. A study comparing minimally invasive and open prostatectomy using Surveillance, Epidemiology, and End Results (SEER)–Medicare data reported a decreased incidence of anastomotic stricture in minimally invasive cases (5.8% vs 14.0%, P ⬍ .001).1 This is similar to singleinstitution literature demonstrating VUAS incidence of 2.6%-10% after RRP vs 0%-2% after RARP.2-6 VUAS development may arise from excess anastomotic narrowing or lack of mucosal apposition, although

outcomes also depend on individual anatomy, inherent technical differences between approaches,5 and surgeon volume.7 An anastomotic stenosis (usually characterized as a Clavien grade III complication, requiring surgical, endoscopic, or radiological intervention) can portend recurring complications such as urinary retention or infection, and possible multiple endoscopic surgeries. Studies have not reported routinely on quality of life outcomes after VUAS development. Here, we report our institutional data comparing VUAS rates in open vs robot-assisted prostatectomies, identifying independent predictors of VUAS development, and discussing impact on quality of life.

MATERIAL AND METHODS Patient Selection, Operative Technique, and Postoperative Surveillance

From the Department of Surgery, University of Toledo, Toledo, Ohio; and Department of Urology, University of Michigan, Ann Arbor, Michigan Reprint requests: Rou Wang, M.D., Rou Wang, M.D., GenitoUrinary Surgeons, Inc., 3500 Executive Parkway, Toledo, OH 43606. Email: [email protected] Submitted: April 20, 2011, accepted (with revisions): July 9, 2011

© 2012 Published by Elsevier Inc.

Through our Institutional Review Board (IRB)–approved prostate cancer database, we identified 1038 patients undergoing RARP and 707 patients undergoing open RRP for clinically localized prostate cancer, from September 2001 to December 0090-4295/12/$36.00 doi:10.1016/j.urology.2011.07.1383

449

2009. This period represented a common time frame during which both open and robotic radical prostatectomies were performed at our institution. In addition, we excluded patients who received preoperative radiation therapy and censored patients for the development of VUAS before any adjuvant radiation therapy. Data on consented men were collected from time of diagnosis, and include demographics, clinical and pathologic stage, operative factors, and postoperative complications and outcomes. For patients with a documented VUAS, information regarding treatment was derived from a second IRB-approved reconstructive urology database. A total of 15 surgeons performed prostatectomies over time. Open RRP was performed by 14 surgeons. Neoadjuvant ADT was used in some early cases to decrease prostate volume. In most cases, the bladder neck was reconstructed in a racket handle–type of closure, and the mucosa was everted for better apposition. Usually 6 interrupted 3-0 monofilament sutures were placed circumferentially around the urethra. A drain was removed on postoperative day 1, and patients went home with a catheter. RARP was performed by 9 surgeons. Intraperitoneal or extraperitoneal approach was used per surgeon preference. Bladder neck reconstruction was not recorded in each case, although only 1 robotic surgeon used this routinely. In most cases, no mucosal eversion was performed. The vesicourethral anastomosis was matured using a running suture (two 7- to 8-cm-long sutures of 3-0 monofilament, tied to create a doublearmed suture), and tied together at the 12 o’clock position. Patients were seen back in the office for catheter removal on postoperative day 10. A cystogram was only performed if there was concern for anastomotic disruption. Serum prostate-specific antigen (PSA) levels were usually monitored at 6 weeks postoperatively, then every 3-6 months as indicated. If patients reported symptoms consistent with a possible VUAS (weak stream, incomplete emptying, retention or persistent incontinence), they were evaluated with cystoscopy. Overall, VUAS was diagnosed by clinical evaluation, cystoscopy (during which the documented caliber had to be ⱕ16 French with inability to pass the flexible cystoscope into the bladder), and, as needed, ultrasound postvoid residual or urodynamics.

Expanded Prostate Cancer Index Composite Questionnaire The Expanded Prostate Cancer Index Composite (EPIC) questionnaire is a validated, 50-item prostate cancer health-related quality of life (HRQOL) questionnaire. It assesses function and bother in 4 domains after treatment: urinary, sexual, hormonal, and bowel.8 Response options form a Likert scale, and multiitem scores are transformed into a linear 0-100 scale, with higher scores representing better HRQOL. With regard to urinary function, the EPIC questionnaire investigates urinary incontinence and irritative voiding via the urinary incontinence (UIN) and urinary irritative voiding (UIR) subdomains of the larger urinary HRQOL domain. Since 2000, postprostatectomy patients at our institution have been asked to complete an EPIC questionnaire 1 month preoperatively and at 3, 6, 12, 24, 36, 48, and 60 months post-treatment. Patients were included who had EPIC scores available within the first year after prostatectomy, as this is the time frame during which most cases of VUAS would be expected to develop. 450

Statistical Analysis The Student t test for continuous data and chi-square test for nominal data were used to compare perioperative factors between open RRP and RARP. Then, by fitting simple logistic models separately, we first assessed for difference of VUAS incidence for each clinical characteristic. For the purpose of parsimony, we adjusted the logistic model for only clinical characteristics associated with outcomes and with initial univariable P value ⱕ .3, resulting in the final multivariable model. EPIC scores were compared between patients who developed VUAS and those who did not, using paired t tests.

RESULTS Demographic characteristics of both patients undergoing open RRP and RARP are detailed in Table 1. There were several statistically significant differences. Patients undergoing RARP tended to be slightly older (mean 60.4 vs 59.7 years, P ⫽ .04), have more medical comorbidities, including diabetes mellitus (9.6 vs 6.5%, P ⫽ .02), hypertension (25.9 vs 5.5%, P ⬍ .0001), and congestive heart failure (8 vs 4.1%, P ⫽ .02). They had fewer high-grade (8⫹) biopsy Gleason scores (5.2 vs 10.5%, P ⫽ .0001) and better preoperative EPIC scores on the UIR domain (87.5 vs 84.6, P ⫽ .03). Intraoperatively, RARP patients had longer operative time (236.8 vs 214.3 minutes, P ⬍ .0001), and lower blood loss (155.9 vs 795.9 ml, P ⬍ .0001), and were more likely to undergo nerve sparing (86.9 vs 81.9%, P ⫽ .01). There was no difference in positive margin rate. Final pathology for RARP tended to include less T3 disease (P ⫽ .001). RARP patients had lower rates of PSA recurrence (2 vs 9.9%, P ⬍ .0001) and use of neoadjuvant or adjuvant therapies, including ADT (neoadjuvant 2.7% vs 4.8%, P ⫽ .02, adjuvant 1.2 vs 4.4%, P ⬍ .0001) and EBRT after development of VUAS (2.3% vs 7.8%, P ⬍ .0001), although this is affected by the shorter median follow-up for RARP vs open RRP patients (11.2 vs 39.4 months). Postoperatively, open RRP patients experienced more blood loss anemia (3.1% vs 0.2%, P ⫽ .0001) and more urinary retention (4% vs 1.4%, P ⫽ .0014). RARP patients had a greater incidence of infectious complications (4.4% vs 1.7%, P ⫽ .003), including urinary tract infections and bacteremia. Most strikingly, the incidence of VUAS in the open RRP group was significantly higher than in the RARP group (7.5% vs 2.1%, P ⬍ .0001). Median time to VUAS development was 2.8 (0.7-20.1) months. VUAS was detected in most patients within the first year: 51 of 53 (96.2%) open RRP patients and 22 of 22 (100%) RARP patients. Table 2 demonstrates the clinical characteristics and subsequent treatment for patients developing a VUAS. VUAS was treated already by an outside urologist in 3 of 75 (4%) patients before they presented to our center, and required no additional endoscopic procedures. A total of 42 patients underwent a single endoscopic intervention, either dilation (34/42, 81%) or direct vision internal urethrotomy (DVIU; 8/42, 19%). Of the patients, 40% required multiple endoscopic procedures, including dilaUROLOGY 79 (2), 2012

Table 1. Patients undergoing open RRP versus RARP Factor Age (y), mean ⫾ SE Body mass index Preoperative PSA (ng/mL) Diabetes mellitus, n (%) Yes No Hypertension Yes No Congestive heart failure Yes No Biopsy Gleason score ⱕ6 7 ⱖ8 Gland weight Pathologic stage ⬍T2a T2b T3a ⬎T3b Preoperative EPIC UIN scores Preoperative EPIC UIR scores Operative time (min) Operative blood loss (mL) Nerve-sparing Bilateral Unilateral None Follow-up time (mo) PSA recurrence Yes No Any positive margin Yes No Apical positive margin Yes No Therapy Neoadjuvant ADT, n (%) Yes No Adjuvant ADT Yes No Adjuvant EBRT Yes No Postoperative complications Blood loss anemia (I) Yes No Cardiac (I, II, III) Yes No Fluid collection (III) Yes No Gastrointestinal (I, II, III) Yes No

UROLOGY 79 (2), 2012

Open RRP (n ⫽ 707)

RARP (n ⫽ 1038)

P Value

59.7 ⫾ 0.3 28.5 ⫾ 0.2 3.9 ⫾ 2.1

60.4 ⫾ 0.2 28.5 ⫾ 0.1 2.4 ⫾ 1.2

.04* .92 .52 .02*

46 (6.5) 654 (92.5)

100 (9.6) 936 (90.2)

39 (5.5) 668 (94.5)

269 (25.9) 769 (74.1)

29 (4.1) 671 (94.9)

83 (8.0) 953 (91.8)

311 (44) 322 (45.5) 74 (10.5) 47.2 ⫾ 0.8

510 (49.1) 474 (45.7) 54 (5.2) 48 ⫾ 0.54

<.0001* .002* .0001*

157 (22.2) 417 (59) 83 (11.7) 47 (6.6) 94.6 ⫾ 0.9 84.6 ⫾ 1.3 214.3 ⫾ 4.8 795.9 ⫹ 20.7 497 (70.3) 82 (11.6) 124 (17.5) 39.4 ⫾ 0.9 70 (9.9) 637 (90.1)

223 (21.5) 654 (63) 107 (10.3) 28 (2.7) 93.4 ⫾ 0.5 87.5 ⫾ 0.6 236.8 ⫾ 1.8 155.9 ⫾ 4.4 791 (75.8) 116 (11.1) 136 (13) 11.1 ⫾ 0.46

.35 .001*

.27 .03* <.0001* <.0001* .01*

<.0001* <.0001*

21 (2) 1017 (98) .25

95 (13.4) 611 (86.4)

160 (15.4) 876 (75.7)

41 (5.8) 664 (93.9)

69 (6.6) 965 (93)

34 (4.8) 673 (95.2)

28 (2.7) 1010 (97.3)

31 (4.4) 676 (95.6)

12 (1.2) 1026 (98.8)

55 (7.8) 652 (92.2)

24 (2.3) 1014 (97.7)

22 (3.1) 685 (96.7)

2 (0.2) 1036 (99.8)

2 (0.28) 705 (99.7)

1 (⬍ 0.1) 1037 (99.9)

1 (0.1) 706 (99.9)

0 (0) 1038 (100)

0 (0) 707 (100)

1 (⬍ 0.1) 1037 (99.9)

.47

.02*

<.0001* <.0001*

.0001* .39 .98 .98

451

Table 1. Continued Factor Hematuria (I, III) Yes No Infectious (II, III) Yes No Pelvic hematoma (I, III) Yes No Rectal injury (I, III) Yes No Thrombotic (I, II) Yes No Renal (I) Yes No Urethral stricture (I, III) Yes No VUAS (I, III) Yes No Urinary leak (I, III) Yes No Urinary retention (I, III) Yes No Wound (I, II, III) Yes No

Open RRP (n ⫽ 707)

RARP (n ⫽ 1038)

12 (1.7) 695 (98.3)

22 (2.1) 1016 (97.9)

12 (1.7) 695 (98.3)

46 (4.4) 992 (95.6)

P Value .53 .003* .66

3 (0.4) 704()

6 (0.58) 1032 (99.4)

2 (0.28) 705 (99.7)

0 (0) 1038 (100)

9 (1.3) 698 (99.9)

10 (0.96) 1028 (99.0)

1 (0.1) 706 (99.9)

2 (0.2) 1036 (99.8)

6 (0.8) 701 (99.2)

10 (1) 1028 (99)

53 (7.5) 654 (92.5)

22 (2.1) 1016 (97.9)

8 (1.1) 699 (98.9)

17 (1.6) 1021 (98.4)

28 (4) 679 (96)

15 (1.4) 1023 (98.6)

1 (0.1) 706 (99.9)

1 (⬍ 0.1) 1037 (99.9)

.97 .54 .80 .81

<.0001* .39 .0014* .79

EPIC ⫽ Expanded Prostate Cancer Index Composite; PSA ⫽ prostate-specific antigen; RARP ⫽ robot-assisted radical prostatectomy; RRP ⫽ open radical retropubic prostatectomy; ADT ⫽ androgen deprivation therapy; EBRT ⫽ external beam radiation therapy; UIN ⫽ urinary incontinence; UIR ⫽ urinary irritative voiding. Clavien complication grade is listed next to postoperative complication. Status is unknown if not listed as Yes or No. * Significant P values ⬍0.05 are listed in boldface type.

tion, DVIU, and transurethral resection of the bladder neck. One patient (1.3%) proceeded to posterior urethroplasty. Overall, endoscopic intervention had minimal complications, with 1 of 75 (1.3%) patients experiencing significant hematuria. Documented median caliber on cystoscopy was 10 French6-16 at diagnosis. Median ultrasound postvoid residual was 182.5 mL (0-700) before the intervention and 27 mL (0-200) after the last intervention. There was insufficient data to report uroflow and retrograde urethrogram findings reliably. Of patients who underwent ⱖ2 procedures, 14 of 30 (46.7%) were placed on self-obturation. Ultimately, 4 (5.3%) patients with VUAS underwent a subsequent continence procedure (artificial urinary sphincter and/or male perineal boneanchor sling), with documented stable VUAS caliber ⬎16 French for 1 year before artificial urinary sphincter implantation. We then compared variables between those patients who developed a VUAS (n ⫽ 75) and those who did not (n ⫽ 1670), and noted various differences which were statistically significant on univariable analysis (Table 3). Patients with VUAS were more likely to have undergone 452

open RRP (70.7 vs 39.2% P ⬍ .01), had less hypertension (4 vs 18.3%, P ⫽ .005), experienced longer operating room time (265.8 vs 236 minutes, P ⬍ .01) and greater intraoperative blood loss (810 vs 401.4 mL, P ⬍ .01). The follow-up time was longer for those with VUAS (35.4 vs 22 months, P ⬍ .0001), likely because of subsequent treatments for the stenosis. Postoperatively, they had higher rates of hematuria (5.3% vs 1.8%, P ⫽ .04), urine leak (5.3% vs 1.3%, P ⫽ .008), as well as urinary retention (9.3% vs 2.2%, P ⫽ .0003). With longer follow-up, VUAS patients also had a higher incidence of PSA recurrence (20% vs 4.6%, P ⬍ .0001), any positive margin (25.3% vs 14.1%, P ⫽ .09), as well as adjuvant use of ADT (13.3% vs 2%, P ⬍ .0001) or EBRT after development of VUAS (13.3% vs 4.1%, P ⫽ .0004). A multivariable analysis was then performed, which adjusted for significant factors on univariable analysis: surgery type, age, BMI, hypertension, congestive heart failure, stage, Gleason score, operative time, operative blood loss, follow-up time, PSA recurrence, any and apical positive margin, adjuvant ADT, adjuvant EBRT, postoperative hematuria, rectal injury, urinary leak, uriUROLOGY 79 (2), 2012

Table 2. Characteristics and treatment of patients with VUAS Characteristic/Treatment Type of surgery Open RRP, n (%) RARP No. of interventions None Single Dilation DVIU Multiple 2 Dilation ⫻ 2 Dilation followed by DVIU DVIU ⫻ 2 3 4⫹ Urethroplasty Median PVR at time of diagnosis, mL, n ⫽ 28 Median PVR after last treatment, mL, n ⫽ 33 Median caliber on cystoscopy at time of diagnosis, French, n ⫽ 33 Patients placed on self-obturation Overall After 1 procedure After 2⫹ procedures Complications Hematuria requiring admission (I) Subsequent continence procedure Sling AUS Sling followed by AUS

Patients with VUAS (n ⫽ 75) 53 (70.6%) 22 (29.3%) 3 (4%) 42 (56%) 34 8 12 (16%) 1 4 7 7 (9.3%) 11 (14.7%) 1 (1.3%) 182.5 (0-700) 27 (0-200) 10 (6-16)

20 (26.7%) 6 14 1 (1.3%) 4 (5.3%) 2 1 1

DVIU ⫽ direct vision internal urethrotomy; VUAS ⫽ vesicourethral anastomotic stenosis; AUS ⫽ artificial urinary sphincter.

nary retention, and wound complications. Open technique (P ⬍ .0001, OR ⫽ 3.0, 95% CI ⫽ 1.8-5.2), PSA recurrence (P ⫽ .02, OR ⫽ 2.2, 95% CI ⫽ 1.2-4.1), postoperative hematuria (P ⫽ .02, OR ⫽ 3.7, 95% CI ⫽ 1.2-11.3), urinary leak (P ⫽ .002, OR ⫽ 6.0, 95% CI ⫽ 1.9-19.2), and urinary retention (P ⫽ .004, OR ⫽ 3.5, 95% CI ⫽ 1.5-8.7) were significant independent predictors of VUAS development. With regard to quality of life, EPIC UIN scores were similar between VUAS and non-VUAS patients at 3, 6, and 12 months (Fig. 1A). EPIC UIR scores were worse for patients with VUAS at 3 months (P ⬍ .0001) and 6 months (P ⫽ .02), but became similar to scores for non-VUAS patients by 12 months (Fig. 1B).

COMMENT VUAS incidence is significantly lower after robot-assisted vs open prostatectomy techniques (7.5 vs 2.1% P ⬍ .0001). Independent predictors of VUAS include open surgery, PSA recurrence, and postoperative hematuria, urinary leak and urinary retention. PSA recurrence as an UROLOGY 79 (2), 2012

independent predictor may indicate unrecognized advanced tumor stage with micrometastatic disease, rendering surgery more difficult. Robot assistance is thought to provide enhanced pelvic visualization9 and improved circumferential apical visualization,4,10 potentially leading to improved mucosal apposition. In addition, there is decreased blood loss with the robotic approach, which has been implicated as a factor in VUAS development.9,11 The goal of surgical reconstruction is to create a nonischemic, water-tight vesicourethral anastomosis. Postoperative hematuria with associated hematoma or urine leak may suggest disruption of the vesicourethral anastomosis. Urinary leak has been associated with development of VUAS in some studies,9 but not in others12; likely, the important point is clinical anastomotic disruption. Ischemia may be reduced overall by the running “parachute” anastomosis used in robot-assisted prostatectomies, and possibly by mucosal eversion.5,13 The only study to address this issue randomized patients undergoing open RRP to cohorts with and without bladder neck mucosal eversion.14 Patients with mucosal eversion had increased postoperative radiologic urinary extravasation, but no significant increase in VUAS or incontinence. Because most robotic surgeons (including at our institution) do not use mucosal eversion, it is difficult to parse out the effect of mucosal eversion from the inherent differences between robotic and open surgery. Webb et al discuss the impact of stomatization and end-to-end anastomosis in increasing the incidence of VUAS after open prostatectomy, and the possible protective effect of the robotic “parachute” technique against ischemia.5 Others have discussed using a running vesicourethral anastomotic technique in open prostatectomy,15,16 and other surgical modifications to decrease stenosis development, such as an intussuscepted anastomosis, reconstruction to 28 French diameter, and tractioning the posterior suture.17,18 None of these have been widely adopted, particularly with increasing use of robotics. Inherent patient characteristics may also play a role. This has been investigated by Borboroglu et al, who identified smoking as the largest risk factor for VUAS, as well as hypertension, coronary artery disease, and diabetes.11 Using the Cancer of the Prostate Strategic Urologic Research Endeavor (CaPSURE) database, Elliott et al identified older age, greater body mass index (BMI), and Caucasian race/ethnicity as factors associated with urethral stricture development.19 Interestingly, maximal abdominal scar width after open RRP has also been associated with VUAS,20 with maximal scar width ⬎ 10 mm portending an 8-fold increased likelihood of stricture. These findings imply local anastomotic hypovascularity and hypertrophic scar formation, leading to VUAS formation. Overall, as implicated by another recent study by Sandhu et al investigating risk factors for symptomatic 453

Table 3. Patients with and without VUAS Factor Type of surgery, n (%) Open RRP RARP Age (mean ⫾ SE) Body mass index Preoperative PSA (ng/mL) Diabetes mellitus, n (%) Yes No Hypertension Yes No Congestive heart failure Yes No Biopsy Gleason score ⱕ6 7 ⱖ8 Gland weight Pathologic stage ⬍T2a T2b T3a ⬎T3b Operative time (min) Operative blood loss (mL) Nerve sparing Bilateral Unilateral None Follow-up time (mo) PSA recurrence Yes No Any positive margin Yes No Apical positive margin Yes No Therapy Neoadjuvant ADT, n (%) Yes No Adjuvant ADT Yes No Any EBRT Yes No Postoperative complications Blood loss anemia (I) Yes No Cardiac (I, II, III) Yes No Fluid collection (III) Yes No Gastrointestinal (I, II, III) Yes No

454

VUAS (n ⫽ 75)

No VUAS (n ⫽ 1670)

P Value <.0001*

53 (70.7) 22 (29.3) 58.9 ⫾ 0.8 29.0 ⫾ 0.5 3.3 ⫾ 2.9

654 (39.2) 1016 (60.8) 60.2 ⫾ 0.2 28.5 ⫾ 0.1 3 ⫾ 1.2

5 (6.7) 70 (93.3)

147 (8.8) 1520 (91)

3 (4) 72 (96)

305 (18.3) 1365 (81.7)

2 (2.7) 73 (97.3)

110 (6.6) 1551 (92.9)

.16 .27 .96 .58 .005* .19 .25

29 (38.7) 38 (50.7) 8 (10.7) 50 ⫾ 2.2

792 (47.4) 758 (45.4) 120 (7.2) 47.6 ⫾ 0.5

13 (17.3) 49 (65.3) 9 (12) 4 (5.3) 265.8 ⫾ 12.5 810 ⫾ 84.3

367 (22) 1022 (61.2) 181 (10.8) 71 (4.3) 236 ⫾ 1.7 401.4 ⫾ 11.4

55 (73.3) 6 (8) 14 (18.7) 35.4 ⫾ 2.5

1233 (73.8) 195 (11.7) 241 (14.4) 22 ⫾ 0.6

15 (20) 60 (80)

76 (4.6) 1594 (95.5)

19 (25.3) 56 (74.7)

236 (14.1) 1431 (85.7)

7 (9.3) 68 (90.7)

103 (6.2) 1561 (93.5)

2 (2.7) 73 (97.3)

60 (3.6) 1610 (96.4)

10 (13.3) 65 (86.7)

33 (2) 1637 (98)

10 (13.3) 65 (86.7)

69 (4.1) 1601 (95.9)

1 (1.3) 74 (98.7)

23 (1.4) 1647 (98.6)

0 (0) 75 (100)

3 (1.8) 1667 (99.8)

0 (0) 75 (100)

1 (⬍ 0.01) 1669 (99.9)

0 (0) 75 (100)

1 (⬍ 0.01) 1669 (99.9)

.53 .77

.008* <.0001* .43

<.0001* <.0001* .009* .28

.67

<.0001* .0004*

.97 .99 .99 .99

UROLOGY 79 (2), 2012

Table 3. Continued Factor Hematuria (I, III) Yes No Infectious (II, III) Yes No Pelvic hematoma (I, III) Yes No Rectal injury (I, III) Yes No Thrombotic (I, II) Yes No Renal (I) Yes No Urethral stricture (I, III) Yes No Urinary leak (I, III) Yes No Urinary retention (I, III) Yes No Wound (I, II, III) Yes No

VUAS (n ⫽ 75)

No VUAS (n ⫽ 1670)

P Value .04*

4 (5.3) 71 (94.7)

30 (1.8) 1640 (98.2)

4 (5.3) 71 (94.7)

54 (3.2) 1616 (96.8)

1 (1.3) 74 (98.7)

8 (0.5) 1662 (99.5)

1 (1.3) 74 (98.7)

1 (⬍ 0.01) 1669 (99.9)

1 (1.3) 74 (98.7)

18 (1.1) 1652 (98.9)

0 (0) 75 (100)

3 (0.2) 1667 (99.8)

0 (0) 75 (100)

16 (9.6) 1916 (99)

4 (5.3) 71 (94.7)

21 (1.3) 1649 (98.7)

7 (9.3) 68 (90.7)

36 (2.2) 1634 (97.8)

2 (2.7) 73 (97.3)

11 (0.7) 1659 (99.3)

.33 .33 .03 .83 .99 .99 .008* .0003* .07

Clavien complication grade is listed next to postoperative complication. Status is unknown if not listed as Yes or No. * Significant P values ⬍0.05 are listed in boldface type.

Figure 1. (A) Comparison of Expanded Prostate Cancer Index Composite (EPIC) Urinary Incontinence Subdomain scores in patients with and without VUAS in the first year. (B) Comparison of EPIC Urinary Irritative Voiding Subdomain scores in patients with and without VUAS in the first year.

UROLOGY 79 (2), 2012

455

anastomotic strictures after primarily open radical prostatectomies,21 development of a VUAS is usually multifactorial, including patient-related as well as surgeon- and surgery-related risk factors. These investigators identified patient age, BMI, Charlson comorbidity score, and renal insufficiency as patient-related risk factors on multivariable analysis, and identified individual surgeon and surgical approach as additional risk factors. As in our study, postoperative urinary leak and hematoma were also linked to development of an anastomotic stricture, indicating the importance of meticulous intraoperative technical detail at the anastomosis and appropriate postoperative management. In relation to quality of life, patients with VUAS experience a worsening of urinary irritative voiding scores at 3 and 6 months, although scores become similar by 12 months. There was no difference in urinary continence scores through the first year—this may reflect the subsequent “hypercontinence” with VUAS development. After treatment, some men complain of worsened continence but might be expected to have a decrease in their irritative voiding symptoms, and following quality of life measures, such as the EPIC questionnaire beyond 1 year may demonstrate a waxing and waning of scores on both subdomain measures depending on repeat treatments. Finally, we confirm that treatment of VUAS can become a long-term issue with possible need for repeat endoscopic procedures. The success rate of endoscopic management for stricture decreases with each subsequent procedure, and repeated endoscopic procedures do not represent definitive treatment. Placing patients on selfobturation can maintain patency for recurrent VUAS; we now routinely place patients requiring ⱖ2 procedures on a self-obturation regimen (usually twice a day with a 16 French catheter for 1 month, then daily for 1 month, and with decreasing frequency from that point forward). The literature discusses different approaches for management of recalcitrant VUAS. Most describe a first procedure to establish patency regardless of continence, followed by a staged artificial urinary sphincter for restoration of urinary control. The first procedure can include UroLume urethral stent placement,22 transurethral incision,23 combined transurethral resection/incision and urethral stent placement,24 or posterior urethroplasty.25,26 Placement of multiple urethral stents has also been discussed, with 24% of patients failing to obtain patency.27 Urethral stents pose their own challenges, including stent migration, recurrent tissue ingrowth or stenosis, urinary retention,28 and difficult removal. In clinical practice, we discuss urethral stents and bladder neck reconstruction with patients, but favor simple cystectomy with urinary diversion for end-stage VUAS, especially with a patient history of radiation therapy resulting in a recalcitrant stenosis. Limitations of this study include the possibility of underreporting VUAS rates, given our tertiary care med456

ical system and outside diagnosis and treatment of VUAS. However, we also maintain a prospective prostate cancer database in which a dedicated database manager reviews all records for complications, rather than relying on physician-reported complications. We acknowledge that the open RRP and RARP population demographics are slightly different; however, we compared these surgical modalities during a common time frame, and even within that time frame, RARP patients tended to be older and to have more comorbidities, factors that have been implicated by other studies to predict VUAS development. Even though RARP patients also tended to have less pathologically aggressive disease at baseline, their follow-up was shorter, and they must be followed up to determine whether their disease course is similar. The common time frame also inherently excluded the open RRP learning curve and included the robotic learning curve, and yet surgery type remained a predictor. This cohort study could not adjust for individual surgeon, as this was closely linked to surgery type. Finally, as noted, follow-up was shorter for the robotassisted cases, which is expected, given its more contemporary application; however, most cases of VUAS are diagnosed within the first 12 months,19 and should be detected within the follow-up period. This is also supported by our data, in which nearly all VUAS cases were diagnosed within 12 months.

CONCLUSIONS Overall, the incidence of VUAS is significantly decreased in robot-assisted vs open radical prostatectomies. This may be attributed to a variety of factors, including downward stage migration of prostate cancer over time and the advantages of robotics (improved visualization, a running anastomosis). Surgical expertise is paramount, with precise dissection and creation of a water-tight anastomosis. Here, we attempt to identify significant independent predictors of VUAS development: open technique, PSA recurrence, and postoperative hematuria, urinary leak, and urinary retention. Patients with a VUAS have similar urinary continence as those without, but experience more urinary irritative voiding symptoms within the first 6 months. EPIC scores do normalize by 12 months and need to be followed to document long-term change, given the repeated nature of endoscopic treatments. Treatment of a VUAS is usually required, and consists mainly of endoscopic dilation, internal urethrotomy, and, for many patients, multiple procedures. Attention to technical detail and careful postoperative management will help to further decrease the incidence of VUAS and overall morbidity from prostate cancer care.

References 1. Hu JC, Gu X, Lipsitz SR, et al. Comparative effectiveness of minimally invasive vs open radical prostatectomy. JAMA. 2009; 302:1557-1564.

UROLOGY 79 (2), 2012

2. Krambeck AE, DiMarco DS, Rangel LJ, et al. Radical prostatectomy for prostatic adenocarcinoma: a matched comparison of open retropubic and robot-assisted techniques. BJU Int. 2009;103:448453. 3. Breyer BN, Davis CB, Cowan JE, et al. Incidence of bladder neck contracture after robot-assisted laparoscopic and open radical prostatectomy. BJU Int. 2010;106:1734-1738. 4. Msezane LP, Reynolds WS, Gofrit ON, et al. Bladder neck contracture after robot-assisted laparoscopic radical prostatectomy: evaluation of the incidence, risk factors, and impact on urinary function. J Endourol. 2008;22:377-383. 5. Webb DR, Sethi K, Gee K. An analysis of the causes of bladder neck contracture after open and robot-assisted laparoscopic radical prostatectomy. BJU Int. 2008;103:957-963. 6. Carlsson S, Nilsson AE, Schumacher MC, et al. Surgery-related complications in 1253 robot-assisted and 485 open retropubic radical prostatectomies at the Karolinska University Hospital, Sweden. Urology. 2010;75:1092-1097. 7. Hu JC, Gold KF, Pashos CL, et al. Role of surgeon volume in radical prostatectomy outcomes. J Clin Oncol. 2003;21:401-405. 8. Wei JT, Dunn RL, Litwin MS, et al. Development and validation of the Expanded Prostate Cancer Index Composite (EPIC) for comprehensive assessment of health-related quality of life in men with prostate cancer. Urology. 2000;56:899-905. 9. El-Hakim A, Leung RA, Tewari A. Robotic prostatectomy: a pooled analysis of published literature. Expert Rev Anticancer Ther. 2006;6:11-20. 10. Tewari AK, Srivastava A, Mudaliar K, et al. Anatomical retroapical technique of synchronous (posterior and anterior) urethral transection: a novel approach for ameliorating apical margin positivity during robotic radical prostatectomy. BJU Int. 2010;106: 1364-1373. 11. Borboroglu PG, Sands JP, Roberts JL, et al. Risk factors for vesicourethral anastomotic stricture after radical prostatectomy. Urology. 2000;56:96-100. 12. Schatzl G, Madersbacher S, Hofbauer J, et al. The impact of urinary extravasation after radical retropubic prostatectomy on urinary incontinence and anastomotic strictures. Eur Urol. 1999;36:187190. 13. Erickson BA, Meeks JJ, Roehl KA, et al. Bladder neck contracture after retropubic radical prostatectomy: incidence and risk factors from a large single-surgeon experience. BJU Int. 2009;104:16151619. 14. Srougi M, Paranhos M, Leite KM, et al. The influence of bladder neck mucosal eversion and early urinary extravasation on patient outcome after radical retropubic prostatectomy: a prospective controlled trial. BJU Int. 2005;95:757-760.

UROLOGY 79 (2), 2012

15. Harpster LE, Brien J. Initial results using a running vesicourethral anastomosis following open radical retropubic prostatectomy. J Urol. 2007;177:118-122. 16. Forster JA, Palit V, Myatt A, et al. Technical description and outcomes of a continuous anastomosis in open radical prostatectomy. BJU Int. 2009;104:929-933. 17. Garg T, See WA. Bladder neck contracture after radical retropubic prostatectomy using an intussuscepted vesicourethral anastomosis: incidence with long-term follow-up. BJU Int. 2009;104:925-928. 18. Orvieto MA, Zorn KC, Gofrit ON, et al. Surgical modifications in bladder neck reconstruction and vesicourethral anastomosis during radical retropubic prostatectomy to reduce bladder neck contractures. Can J Urol. 2006;13:3353-3357. 19. Elliott SP, Meng MV, Elkin EP, et al. Incidence of urethral stricture after primary treatment for prostate cancer: data from CaPSURE. J Urol. 2007;178:529-534. 20. Park R, Martin S, Goldberg JD, et al. Anastomotic strictures following radical prostatectomy: insights into incidence, effectiveness of intervention, effect on continence, and factors predisposing to occurrence. Urology. 2001;57:742-746. 21. Sandhu JS, Gotto GT, Herran LA, et al. Age, obesity, medical comorbidities and surgical technique are predictive of symptomatic anastomotic strictures after contemporary radical prostatectomy. J Urol. 2011;185:2148-2152. 22. Elliott DS, Boone TB. Combined stent and artificial urinary sphincter for management of severe recurrent bladder neck contracture and stress incontinence after prostatectomy: a long-term evaluation. J Urol. 2001;165:413-415. 23. Gousse AE, Tunuguntla HS, Leboeuf L. Two-stage management of severe postprostatectomy bladder neck contracture associated with stress incontinence. Urology. 2005;65:316-319. 24. Anger JT, Raj GV, Delvecchio FC, et al. Anastomotic contracture and incontinence after radical prostatectomy: a graded approach to management. J Urol. 2005;173:1143-1146. 25. Wessells H, Morey AF, McAninch JW. Obliterative vesicourethral strictures following radical prostatectomy for prostate cancer: reconstructive armamentarium. J Urol. 1998;160:1373-1375. 26. Simonato A, Gregori A, Lissiani A, et al. Two-stage transperineal management of posterior urethral strictures or bladder neck contractures associated with urinary incontinence after prostate surgery and endoscopic treatment failures. Eur Urol. 2007;52:1499-1504. 27. Magera JS Jr., Inman BA, Elliott DS. Outcome analysis of urethral wall stent insertion with artificial urinary sphincter placement for severe recurrent bladder neck contracture following radical prostatectomy. J Urol. 2009;181:1236-1241. 28. Wilson TS, Lemack GE, Dmochowski RR. UroLume stents: lessons learned. J Urol. 2002;167:2477-2480.

457