Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
Original article
The effect of length of ureteral resection on benign ureterointestinal stricture rate in ileal conduit or ileal neobladder urinary diversion following radical cystectomy Kyle A. Richards, M.D.a,*, Joshua A. Cohn, M.D.a, Michael C. Large, M.D.a,b, Gregory T. Bales, M.D.a, Norm D. Smith, M.D.a, Gary D. Steinberg, M.D.a a
Section of Urology, Department of Surgery, The University of Chicago Medical Center, Chicago, IL b Urology of Indiana, Indianapolis, IN Received 14 March 2014; received in revised form 17 April 2014; accepted 7 May 2014
Abstract Objectives: To assess the effect of the length of the ureter resected and other clinical variables on ureterointestinal anastomotic (UIA) stricture rate following radical cystectomy and ileal segment urinary diversion. Methods and materials: We identified 519 consecutive patients who underwent cystectomy and ileal conduit or ileal orthotopic neobladder diversion from January 2007 to August 2012. The length of the ureter resected was defined as the length of the ureter in the cystectomy specimen plus the length of the distal ureter submitted for pathologic analysis. The primary end point was the risk of UIA stricture formation, assessed by Cox proportional hazards analysis. Results: A total of 463 patients met the inclusion criteria with complete data. Median follow-up was 459 days (interquartile range [IQR]: 211–927). Median time to stricture formation was 235 (IQR: 134–352) and 232 days (IQR: 132–351) on the right and the left ureter, respectively. Overall stricture rate per ureter was 5.9% on the right vs. 10.0% on the left (P ¼ 0.03). There was no difference in demographic, operative, or perioperative variables between patients with and without UIA strictures. On multivariate analysis adjusted for age, sex, anastomosis technique (running vs. interrupted), and length of ureter resected, only a Clavien complication ZIII (hazard ratio ¼ 2.11, 1.01–4.40) and urine leak (hazard ratio ¼ 3.37, 1.08–10.46) significantly predicted for left- and right-sided stricture formation, respectively. The length of the ureter resected did not predict UIA stricture formation on either side. Conclusions: The etiology of benign UIA strictures following ileal urinary diversion is likely multifactorial. Our data suggest that a complicated postoperative course and urine leak are risk factors for UIA stricture formation. The length of the distal ureter resected did not significantly affect stricture rate. r 2014 Elsevier Inc. All rights reserved.
Keywords: Ureteral stricture; Ureterointestinal anastomosis; Urinary bladder neoplasms; Cystectomy; Postoperative complications; Reimplantation
1. Introduction Radical cystectomy (RC) and urinary diversion (UD) are the gold-standard treatments for muscle-invasive bladder cancer (BCa) [1]. Despite advances in anesthesia, surgical techniques, and perioperative care, postoperative morbidity within 3 months of surgery is greater than 60% in the
Corresponding author. Tel.: þ1-773-702-6189; fax: þ1-773-702-1001. E-mail address: [email protected] (K.A. Richards). *
http://dx.doi.org/10.1016/j.urolonc.2014.05.015 1078-1439/r 2014 Elsevier Inc. All rights reserved.
contemporary series [2]. Long-term complications related to the UD can also lead to significant morbidity. For example, the incidence of benign ureterointestinal anastomotic (UIA) strictures following RC and UD is reported at 1% to 13% [3–7]. Although most UIA strictures become clinically evident between 7 and 18 months postoperatively [5,7,8], patients occasionally present after more than 10 years from their initial surgery [4]. Patients with UIA strictures can present with nephrolithiasis, recurrent pyelonephritis, asymptomatic worsening hydronephrosis, or even more insidiously with impaired renal function. The pathophysiology of UIA stricture formation is likely secondary to ischemia and inflammation, resulting in
65.e2
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
fibrosis and ureteral obstruction; however, the exact etiology of UIA strictures remains elusive. Meticulous surgical technique, preservation of periureteral adventitia and blood supply, avoidance of electrocautery, and excision of the distal or compromised ureter before anastomosis are common strategies employed to mitigate the risks of UIA stricture formation [9]. Additionally, a running end-to-side Bricker UIAs [10], a postoperative urinary tract infection (UTI) [10], and an antirefluxing technique [3,4,11] are potential risk factors for UIA stricture formation. By contrast, Anderson et al. [12] were unable to identify any patient- or disease-specific factor to be associated with UIA stricture formation in a cohort of 478 patients from their institution. Although it is a common practice to resect the distal ureter that may or may not be compromised before ureteral anastomosis, the effect that this may have on the future development of UIA strictures is unclear. To the best of our knowledge, no previous studies have evaluated this variable as a risk factor for UIA stricture formation. We evaluated whether the length of the ureter excised at the time of performing an RC affects the future development of UIA strictures.
2. Materials and methods We obtained approval from our institutional review board before initiating this analysis. A prospective database was reviewed for patients undergoing RC with ileal conduit or ileal orthotopic neobladder UD by 2 surgeons (G.D.S. and N.D.S.) from 2007 to 2012. Both the surgeons have extensive experience in urologic oncology and large practices in BCa resection. An orthotopic neobladder was constructed as a Studer pouch or a modified Hautmann pouch with an isoperistaltic afferent limb. A permanent distal ureteral specimen is routinely sent intraoperatively but frozen sections are not. All UIAs were performed in a similar Bricker end-to-side tension-free technique using 4-0 polyglactin sutures on CV-23 needles. The decision to perform a running vs. interrupted UIA or to excise additional ureter was at the discretion of the surgeon. Our technique for running UIA has been previously described [10]. A single 4-0 monofilament traction suture is placed at the 12-o’clock position to avoid unnecessary manipulation and to ensure correct orientation. The left ureter is tunneled through a wide channel created beneath the sigmoid mesentery by grasping the traction suture with a large right-angled clamp. All anastomoses were stented with either 5-F feeding tubes or 7-F single J stents and tested for water tightness upon completion by irrigating the conduit or the afferent limb with a bulb syringe. Stents were removed following tolerance of oral diet (G.D.S) or approximately 2 weeks postoperatively (N.D.S). The length of the ureter resected was defined as the length of the ureter in the cystectomy specimen plus the length of the distal ureter submitted separately for pathologic analysis. It is a routine practice for
our pathologists to quantify the length of the ureter in the specimens and to comment on the presence of dysplasia or neoplasia in the distal ureter on the final pathologic report. The primary end points in our study were occurrence of UIA strictures and time to UIA stricture formation. Patients were followed up according to the National Comprehensive Cancer Network guidelines using serum chemistry studies and axial imaging of the abdomen and the pelvis every 3 to 6 months for 2 years based on the risk of recurrence and then annually or as clinically indicated [13]. All patients with recurrent pyelonephritis, worsening hydronephrosis, or worsening renal function were evaluated with an antegrade study or mercaptoacetyltriglycine renal scintigraphy or both to identify UIA strictures. The variables that were maintained and studied included basic demographics, the length of the ureter resected, age-adjusted Charlson comorbidity index [14], diversion type, clinical/pathologic stage, perioperative chemotherapy or radiation, estimated blood loss, body mass index, length of hospital stay, operative time, running vs. interrupted UIA, and perioperative complications up to 90 days. Complications were graded according to the ClavienDindo classification system [15]. For each period (30 d and 30–90 d), the highest grade of complication for each patient was identified, and the UIA stricture itself was not included as a specific Clavien complication to avoid confusion. UTI was diagnosed based on positive findings on urine culture (4105 organisms) in conjunction with fever or leukocytosis. Urine leak was identified in all cases by elevated drain creatinine levels beyond postoperative day 7. The decision to check drain creatinine levels is not routine and is instead based on clinical factors, such as prolonged ileus or excessive drain outputs. Patients who died within 30 days, had missing ureteral resection data, or developed ureteral obstruction due to recurrent cancer were excluded from this analysis. Statistical analysis was performed using Stata 12 (College Station, TX). A comparison of means was performed using the Student t test, and a comparison of medians was performed using the Mann-Whitney U test. The Fisher exact and the chi-square tests were used for a comparison of categorical variables. A separate analysis was performed for each ureter at risk (left and right), to account for both differential risk of stricture formation by side and the subset of patients with a solitary kidney at the time of cystectomy. A multivariate Cox proportional hazards model was performed for the aforementioned perioperative variables to identify independent predictors of UIA stricture formation. An adjusted analysis was performed by controlling for age, sex, and sum of ureter removed, as well as the following variables that were significant on unadjusted analysis: urine leak, anastomotic technique, and a Clavien complication Zgrade III at 30 days. The Kaplan-Meier survival analysis and the stratified log-rank test were used to compare time to UIA stricture formation per ureter at risk based on the length of the ureter resected. A 2-sided P o 0.05 was considered significant.
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
3. Results The study sample consisted of 519 patients. Of them, 13 patients died within 30 days after undergoing RC and 43 patients had missing data on the length of the ureter resected, leaving 463 in the total study population, of which 439 (95%) underwent open RC. The 24 patients who underwent robotic RC had UDs performed extracorporeally, with only 1 patient developing a UIA stricture. For the entire cohort, 76 patients (16.4%) were evaluated with either mercaptoacetyltriglycine renal scintigraphy (41%) or antegrade nephrostogram (59%) for a rise in creatinine levels (11 patients, 14.5%), worsening hydronephrosis (24 patients, 31.6%), rise in creatinine/worsening hydronephrosis (25 patients, 32.9%), or recurrent pyelonephritis (16 patients, 21.0%). After these evaluations, 58 patients (12.5%) were found to have UIA strictures at an overall median follow-up of 459 days (interquartile range [IQR]: 211–927). UIA strictures developed more commonly in men (14.4%) than in women (5.3%) (P ¼ 0.018). However, there was no difference in other demographic or baseline clinical variables between those with or without UIA strictures (Table 1). UIA stricture formation was more common in patients with organ-confined disease than in patients with non–organ-confined disease on final pathologic
65.e3
analysis (15.9% vs. 8.3%, P ¼ 0.014) and in patients who were node negative than in patients who were node positive (14.8% vs. 5.9%, P ¼ 0.012), although follow-up was longer for patients with more favorable findings on pathologic analysis. For each ureter at risk, the left ureters were more likely than the right ureters to develop UIA strictures (10.0% vs. 5.9%, P ¼ 0.03). For the right ureters at risk, the median follow-up was longer for those patients who developed UIA strictures than for those who did not develop UIA strictures (683 d [IQR: 470–1,384] vs. 427 d [IQR: 169–878], P o 0.01) and the median time to stricture formation was 235 days (IQR: 134–352). For each right ureter at risk, there was no difference in diversion type, anastomotic technique, operative time, estimated blood loss, total length of ureter resected, or length of hospital stay in those patients who developed UIA strictures vs. those free from UIA strictures (Table 2). Similar findings were noted for each left ureter at risk. Regarding perioperative outcomes, 57.0% of the cohort experienced a complication within 30 days of surgery, of which 17.3% were Clavien grade ZIII. Furthermore, 22.5% of the cohort experienced a complication 30 to 90 days after surgery, of which 6.9% were Clavien grade ZIII. For the patients with the right ureters at risk, those who developed right UIA strictures were more likely to have sustained a
Table 1 Cohort characteristics Overall
Stricture
No stricture
Number of patients (%) Age, y (95% CI) Sex, male, n (%) Sex, female, n (%) Body mass index (95% CI) Charlson comorbidity index (95% CI)
463 68.2 369 94 28.2 2.3
58 68.6 53 5 29.1 2.2
405 68.2 316 89 28.0 2.3
Preoperative pathology Clinically muscle invasive, n (%) Non–muscle invasive, n (%)
301 (65.0) 162 (35.0)
32 (10.6) 26 (16.1)
269 (89.4) 136 (83.9)
0.093
64 (13.8) 48 (10.3)
6 (6.3) 7 (14.6)
60 (93.7) 41 (85.4)
0.102 0.649
Neoadjuvant chemotherapy, n (%) Preoperative radiation, n (%)
(67.2–69.3) (79.7) (20.3) (27.6–28.7) (2.1–2.5)
Ureters at risk Both, n (%) Left only, n (%) Right only, n (%)
401 (86.6) 38 (8.2) 24 (5.2)
Stricture laterality Both, n (%) Left only, n (%) Right only, n (%)
11 (19.0) 33 (56.9) 14 (24.1)
Postoperative pathology Organ confined, n (%) Non–organ confined, n (%) Positive soft tissue margins, n (%) Negative soft tissue margins, n (%) Node positive, n (%) Node negative, n (%)
258 205 33 430 118 345
(55.7) (44.3) (7.1) (92.9) (25.5) (74.5)
Note: Bold signifies results that are statistically significant (P o 0.05).
41 17 1 57 7 51
(12.5) (66.2–71.1) (14.4) (5.3) (27.4–30.7) (1.8–2.6)
(15.9) (8.3) (3.0) (13.3) (5.9) (14.8)
217 188 32 373 111 294
(87.5) (67.1–69.3) (85.6) (94.7) (27.5–28.6) (2.2–2.5)
(84.1) (91.7) (97.0) (86.7) (94.1) (85.2)
P value 0.73 0.018 0.239 0.56
0.014 0.087 0.012
65.e4
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
Table 2 Perioperative variables and outcomes Overall
No stricture
Stricture
P value
Right ureter at risk (n ¼ 425) Number of patients (%) Median follow-up, d (IQR) Mean follow-up, d (95% CI) Median time to stricture formation, d (IQR)
463 (–) 459 (211–927) 661 (606–715)
400 (94.1) 427 (169–878) 614 (557–670)
Diversion type Ileal conduit, n (%) Orthotopic neobladder, n (%)
288 (62.2) 175 (37.8)
253 (63.3) 147 (36.8)
12 (48.0) 13 (52.0)
0.13
Anastomosis Interrupted, n (%) Running, n (%)
181 (39.1) 282 (60.9)
158 (39.5) 242 (60.5)
12 (48.0) 13 (52.0)
0.40
Operative time, min (95% CI) Estimated blood loss, ml (95% CI)
304 (298–311) 996 (944–1,047)
304 (297–312) 985 (931–1,038)
295 (276–314) 1,096 (861–1,331)
0.36 0.35
Total length of ureter removed, mm (95% CI) In specimen Resected intraoperatively
26.0 (24.2–27.7) 16.8 (15.2–18.3) 9.2 (8.2–10.2)
22.2 (14.7–29.8) 14.7 (8.6–20.8) 7.5 (4.4–10.7)
0.16 0.51 0.30
Sum of ureter removed, categorized 0–10 mm, n (%) 11–30 mm 31–50 mm 450 mm
98 167 103 32
(24.5) (41.8) (25.8) (8.0)
25 683 950 235
8 11 3 3
(5.9) (470–1,384) (692–1,206) (134–352)
. o0.01 0.01
(32.0) (44.0) (12.0) (12.0)
0.42
11.6 (8.7–14.5)
0.54
Length of stay, d (95% CI)
10.8 (10.2–11.4)
10.7 (10.0–11.4)
Complications, 30 d Any, n (%) Clavien grade 4II, n (%)
264 (57.0) 80 (17.3)
224 (56.8) 65 (16.3)
16 (64.0) 6 (24.0)
0.48 0.31
Complications, 30–90 d Any, n (%) Clavien grade 4II, n (%)
104 (6.9) 32 (13.4)
84 (21.0) 47 (11.8)
9 (36.0) 8 (32.0)
0.08 o0.01
25 48 27 31
5 6 3 1
(20.0) (24.0) (12.0) (3.1)
o0.01 0.08 0.32 0.49
(10.0) (414–1,462) (762–1,179) (132–351)
0.03a 0.01 o0.01 –
30-d 30-d 30-d 30-d
Urine leak, n (%) Urinary tract infection, n (%) Intra-abdominal abscess, n (%) Clostridium difficile infection, n (%)
33 60 36 35
(7.1) (13.0) (7.8) (7.6)
(6.3) (12.0) (6.7) (7.8)
Left ureter at risk (n ¼ 439) Number of patients (%) Median follow-up, d (IQR) Mean follow-up, d (95% CI) Median time to stricture formation, d (IQR)
395 (90.0) 431 (171–878) 617 (560–674)
44 714 970 232
Diversion type Ileal conduit, n (%) Orthotopic neobladder, n (%)
250 (63.3) 145 (36.7)
26 (59.1) 18 (40.9)
0.58
Anastomosis Interrupted, n (%) Running, n (%)
156 (39.5) 239 (60.5)
14 (31.8) 30 (68.2)
0.32
Operative time, min (95% CI) Estimated blood loss, ml (95% CI)
304 (296–311) 974 (921–1,027)
303 (284–321) 1,134 (897–1,371)
0.91 0.19
Total length of ureter removed, mm (95% CI) In specimen Resected intraoperatively
22.8 (21.2–24.4) 16.1 (14.7–17.6) 6.7 (5.8–7.6)
21.5 (16.3–26.6) 14.8 (10.4–19.3) 6.6 (3.2–10.1)
0.61 0.58 0.98
Sum of ureter removed, categorized 0–10 mm, n (%) 11–30 mm 31–50 mm 450 mm
115 176 83 21
(29.1) (44.6) (21.0) (5.3)
13 20 9 2
(29.6) (45.5) (20.5) (4.6)
0.99
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
65.e5
Table 2 Continued Overall
No stricture
Stricture
Length of stay, d (95% CI)
10.7 (10.0–11.4)
Complications, 30 d Any, n (%) Clavien grade 4II, n (%)
222 (56.2) 62 (15.7)
26 (59.1) 12 (27.3)
0.71 0.05
Complications, 30–90 d Any, n (%) Clavien grade 4II, n (%)
84 (21.3) 46 (11.7)
13 (29.6) 9 (20.5)
0.21 0.09
30-d 30-d 30-d 30-d
24 46 27 30
a
Urine leak, n (%) Urinary tract infection, n (%) Intra-abdominal abscess, n (%) Clostridium difficile infection, n (%)
(6.1) (11.7) (6.8) (7.6)
10.7 (8.9–12.4)
P value
5 9 7 3
(11.4) (20.5) (15.9) (6.8)
0.97
0.18 0.09 0.03 0.85
P value refers to left vs. right.
major complication 30 to 90 days from surgery (32.0%) than those patients who were free from right UIA strictures (11.8%) (P o 0.01) were. In addition, the urine leak rate was 20% in those with right UIA strictures vs. 6.3% in those free from right UIA strictures (P o 0.01) (Table 2). For the patients with the left ureters at risk, major complications within 30 days of surgery were more common in those who developed left UIA strictures (27.3%) than in those free of left UIA strictures (15.7%) (P ¼ 0.05). The 30-day intra-abdominal abscess rate was also significantly higher in those with left UIA strictures than in those who were free from left UIA strictures (15.9% vs. 6.8%, P ¼ 0.03). While controlling for sex, age, anastomotic technique, sum of the left ureter resected, and urine leak, major complications within 30 days of surgery were associated with left UIA strictures (hazard ratio [HR] ¼ 2.11; 95% CI: 1.01–4.40; P ¼ 0.046) in our Cox proportional hazards model (Table 3). While controlling for the same variables, urine leak was associated with right UIA strictures (HR ¼ 3.37; 95% CI: 1.08–10.46; P ¼ 0.036). The Kaplan-Meier analysis of time to stricture formation showed no difference in freedom between left and right UIA strictures stratified by the length of the ureter resected (Fig.).
4. Discussion Although several studies have attempted to identify the risk factors for the development of UIA strictures, this is the first study to our knowledge to specifically evaluate the importance of the length of the distal ureter resected at the time of RC and UD. The exact etiology of benign UIA strictures in any individual patient is unclear. Certainly, proper attention to meticulous surgical technique is important. These technical considerations should include careful tissue handling, preservation of ureteral blood supply, minimization of electrocautery, and precise placement of
sutures. The concept of resecting additional distal ureter before performing a UIA (i.e., to “freshen up” the ureter) stems from the idea that the distal ureteral blood supply may have been compromised during its dissection at the time of RC. Although this rationale seems logical, one could also argue that resecting additional ureter could also lead to increased risk of UIA stricture formation if undue tension might be placed on the UIA. It is with these theoretical concerns that we set forth to determine the effect of a resected ureter on the incidence of benign UIA strictures. Anderson et al. [12] evaluated potential risk factors for UIA stricture formation in 478 consecutive patients who underwent RC and UD at their institution from 2007 to 2011. In their cohort, 21.5% underwent robotic RC with extracorporeal UD, and the overall rate of UIA stricture was 9.4%. Furthermore, in their multivariable-adjusted model, none of the perioperative variables (robotic technique, age, diversion type, sex, or the American Society of Anesthesiologists class) were predictors of UIA stricture formation; however, the length of the ureter resected was not studied. Similarly, sex and age were not predictors of UIA stricture formation in our adjusted model, and Charlson comorbidity index and diversion type were not predictors of UIA stricture formation in our unadjusted model. Careful dissection and handling of the ureter, including wide spatulation, remain an art and cannot be accurately measured or quantified. However, there are other components of the UIA technique that have been studied for their effect on UIA stricture development. Large et al. [10] evaluated the effect of running vs. interrupted UIA on benign UIA stricture development in a consecutive series of 266 patients. On multivariable analysis controlling for age, sex, final pathologic analysis, and 30-day Clavien complications Zgrade III, running anastomosis (HR ¼ 1.92, 95% CI: 1.00–3.70) and 30-day perioperative UTI (HR ¼ 2.44, 95% CI: 1.17–5.08) predicted UIA stricture formation. With a larger cohort of patients and longer follow-up, these variables were no longer significant predictors in the present study. However, on
65.e6
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
Table 3 Unadjusted and adjusted Cox proportional hazards model with stricture diagnosis as the primary outcome Variable (referent)
Unadjusted HR (95% CI)
P value
Adjusted HR (95% CI)
P value
0.905 0.081 0.242 0.659 0.536 0.536 0.64 0.065 0.391 0.088
– 2.64 (0.93–7.45) 1.02 (0.99–1.05) – – – – – – 1.68 (0.85–3.33)
– 0.067 0.248 – – – – – – 0.137
Left ureter OR time Sex Age, y Length of stay Preoperative pathology (muscle invasive) Diversion (ONB vs. IC) Non–organ confined Positive nodes Positive margin Running anastomosis
0.99 2.50 1.02 1.01 0.83 0.83 0.86 0.38 0.42 1.79
(0.996–1.004) (0.89–6.99) (0.99–1.04) (0.96–1.06) (0.45–1.51) (0.45–1.51) (0.45–1.63) (0.14–1.06) (0.06–3.05) (0.92–3.49)
Sum L ureter removed (ref: 0–10 mm) 11–30 mm 30–50 mm 450 mm
1.01 (0.50–2.03) 1.02 (0.43–2.41) 0.82 (0.18–3.63)
0.977 0.947 0.792
– – –
– – –
Sum L ureter removed þ Ureteral margin 30-d C. difficile infection Postoperative abscess Dehiscence Wound infection 30-d Urinary tract infection EBL BCG history Urine leak Readmit 30 d Neoadjuvant chemotherapy Radiation history Body mass index Age-adjusted CCI Clavien grade III or more, 30-d complication Any 30–90-d complication
0.99 0.71 0.77 2.81 1.40 1.91 1.84 1.00 0.63 2.10 2.80 0.72 1.24 1.04 1.05 2.6 0.79
0.745 0.64 0.654 0.012 0.196 0.143 0.103 0.082 0.237 0.118 0.001 0.527 0.653 0.154 0.541 0.005 0.5
0.995 (0.98–1.01) – – – – – – – – 1.59 (0.58–4.37) – – – – – 2.11 (1.01–4.40) –
0.64 – – – – – – – – 0.372 – – – – – 0.046 –
0.491 0.147 0.511 0.221 0.645 0.563 0.52 0.983 0.835 0.657
– 2.6 (0.60–11.25) 1.02 (0.98–1.06) – – – – – – 0.88 (0.38–2.04)
– 0.200 0.271 – – – – – – 0.760
(0.98–1.02) (0.17–2.95) (0.24–2.47) (1.25–6.32) (0.84–2.32) (0.80–4.53) (0.88–3.83) (0.99–1.00) (0.29–1.36) (0.83–5.35) (1.52–5.14) (0.26–2.01) (0.49–3.14) (0.99–1.09) (0.89–1.25) (1.33–5.09) (0.39–1.58)
Right ureter OR time Sex Age, y Length of stay Preoperative pathology (muscle invasive) Diversion (ONB vs. IC) Non–organ confined Positive nodes Positive margin Running anastomosis
0.99 2.91 1.01 1.03 0.83 1.26 0.75 0.99 0.81 0.83
(0.992–1.004) (0.69–12.36) (0.98–1.05) (0.98–1.09) (0.38–1.83) (0.57–2.78) (0.31–1.81) (0.37–2.64) (0.11–6.01) (0.38–1.85)
Sum R ureter removed (ref: 0–10 mm) 11–30 mm 30–50 mm 450 mm
0.85 (0.34–2.11) 0.40 (0.11–1.50) 1.17 (0.31–4.42)
0.729 0.173 0.816
– – –
– – –
Sum R ureter removed þ Ureteral margin 30-d C. difficile infection Postoperative abscess Dehiscence Wound infection 30-d Urinary tract infection EBL BCG history Urine leak Readmit 30 d
0.99 1.58 0.43 1.99 1.09 2.16 2.26 1.00 0.55 3.92 1.73
0.326 0.538 0.405 0.265 0.835 0.159 0.082 0.401 0.279 0.006 0.22
0.99 (0.97–1.02) – – – – – – – – 3.37 (1.08–10.46) –
0.515 – – – – – – – – 0.036 –
(0.96–1.01) (0.37–6.69) (0.06–3.16) (0.59–6.65) (0.47–2.52) (0.74–6.32) (0.90–5.66) (0.99–1.00) (0.19–1.61) (1.47–10.47) (0.72–4.14)
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
65.e7
Table 3 Continued Variable (referent)
Unadjusted HR (95% CI)
P value
Adjusted HR (95% CI)
P value
Neoadjuvant chemotherapy Radiation history Body mass index Age-adjusted CCI Clavien grade III or more, 30-d complication Any 30–90-d complication
0.29 1.98 0.99 1.03 2.09 2.53
0.226 0.211 0.861 0.752 0.118 0.026
– – – – 1.44 (0.49–4.22) –
– – – – 0.511 –
(0.04–2.15) (0.68–5.77) (0.92–1.07) (0.83–1.30) (0.83–5.26) (1.12–5.73)
BCG ¼ bacillus Calmette-Guérin; CCI ¼ Charlson comorbidity index; EBL ¼ estimated blood loss; IC ¼ ileal conduit; ONB ¼ orthotopic neobladder; OR ¼ operating room.
unadjusted analysis, the data in the present study did suggest a trend toward increased risk for stricture formation for both the ureters with 30-day UTI and for the left ureter only with a running vs. interrupted anastomosis. Other technical modifications have been studied, including refluxing vs. nonrefluxing techniques and stented vs. nonstented UIA. There is a higher rate of benign UIA stricture in series that have evaluated nonrefluxing techniques (8.3%–16.3%) than in those evaluating refluxing techniques (1.7%–5.4%), without any appreciable benefits for patients with nonrefluxing UIA [3,4,11,16]; however, the type of refluxing UIA (Wallace vs. Bricker) does not seem to matter [8,17]. Stenting the UIA is a common procedure that is thought to attenuate the risk of urine leak. Despite concerns that stents may actually increase the risk of UIA stricture formation (being a foreign body), numerous studies have failed to show an increased rate of UIA stricture in stented than in nonstented UIA [18–20]. All patients in our series underwent refluxing and stented UIA to help control potentially modifiable risk factors for UIA stricture formation. Urine leak was a significant predictor of UIA stricture formation in our right ureter–adjusted multivariable model despite stenting and testing all UIA for water tightness. Therefore, every effort should be made through careful surgical technique of the UIA to minimize the risk of postoperative urine leak. In addition, patients with major (Clavien grade ZIII) 30-day postoperative complications were at an increased risk of UIA stricture formation in the left ureter using the multivariable-adjusted model. Patients with a complicated postoperative course should be followed up closely for insidious development of benign UIA strictures, as they may be at an increased risk for this complication. There are several limitations in this study that deserve specific mention. Ours being a tertiary care medical center with very specialized and complex practices in BCa surgery, these results may not have external validity in the general or community urology practice. With large and widespread referral patterns, follow-up care can be challenging with the potential inability to completely capture all postoperative events once the patient is discharged from the hospital. Our research team contacts patients who may have
been lost to follow-up and reviews outside medical records when needed to ensure complete data collection. In addition, the length of the ureter resected was measured by numerous pathologists during the study period, with the resultant possibility of measurement error or bias. The exact location of the stricture was not available in each patient, and we are therefore unable to comment on other technical factors, including the degree of angulation of the left ureter as it passes beneath the sigmoid mesentery. Finally, other subtle changes in our practice over the 5-year study period (i.e., elimination of mechanical bowel preparation) may
Fig. The Kaplan-Meier curve of the proportion of patients stricture free by the length of the ureter resected for the left ureters at risk (top) and the right ureters at risk (bottom). (Color version of figure is available online.)
65.e8
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
have had an unmeasured bias on the study results for which we were unable to control. Despite the aforementioned limitations, this is the first series to evaluate the effect of the length of the distal ureter resected as a potential risk factor for UIA stricture formation. This study highlights that the exact etiology of UIA strictures remains elusive and is likely multifactorial. Despite technical modifications, a sound surgical technique (which is difficult to quantify) remains critically important.
[7] [8]
[9] [10]
[11]
5. Conclusions The etiology of benign UIA strictures following ileal UD is likely multifactorial. Our data suggest that a complicated postoperative course and urine leak are risk factors for UIA stricture formation. The length of the distal ureter resected did not significantly affect the stricture rate.
[12]
[13]
[14]
References [1] Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol 2001;19:666–75. [2] Shabsigh A, Korets R, Vora KC, et al. Defining early morbidity of radical cystectomy for patients with bladder cancer using standardized reporting methodology. Eur Urol 2009;55:164–76. [3] Studer UE, Burkhard FC, Stumacher M, et al. Twenty years experience with an ileal orthotopic low pressure bladder substitute —lessons to be learned. J Urol 2006;176:161–6. [4] Hautmann RE, de Petriconi RC, Volkmer BG. 25 years of experience with 1,000 neobladders: long-term complications. J Urol 2011;185:2207–12. [5] Tal R, Sivan B, Kedar D, et al. Management of benign ureteral strictures following radical cystectomy and urinary diversion for bladder cancer. J Urol 2007;178:538–42. [6] Katkoori D, Samavedi S, Adiyat KT, et al. Is the incidence of ureterointestinal anastomotic stricture increased in patients undergoing
[15]
[16]
[17] [18]
[19]
[20]
radical cystectomy with previous pelvic radiation? BJU Int 2010;105: 795–8. Shimko MS, Tollefson MK, Umbreit EC, et al. Long-term complications of conduit urinary diversion. J Urol 2011;185:562–7. Kouba E, Sands M, Lentz A, et al. A comparison of the Bricker versus Wallace ureteroileal anastomosis in patients undergoing urinary diversion for bladder cancer. J Urol 2007;178:945–8. Richards KA, Steinberg GD. Perioperative outcomes in radical cystectomy: how to reduce morbidity? Curr Opin Urol 2013;23:456–65. Large MC, Cohn JA, Kiriluk KJ, et al. The impact of running versus interrupted anastomosis on ureterointestinal stricture rate after radical cystectomy. J Urol 2013;190:923–7. Pantuck AJ, Han KR, Perrotti M, et al. Ureteroenteric anastomosis in continent urinary diversion: long-term results and complications of direct versus nonrefluxing techniques. J Urol 2000;163:450–5. Anderson CB, Morgan TM, Kappa S, et al. Ureteroenteric anastomotic strictures after radical cystectomy—does operative approach matter? J Urol 2013;189:541–7. National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (NCCN Guideliness). Bladder Cancer, Version 1.2014. Available at: www.nccn.org; [accessed 17.4.14]. Charlson M, Szatrowski TP, Peterson J, et al. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47:1245–51. Dindo D, Demartines N, Clavien P-A. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205–13. Shaaban AA, Abdel-Latif M, Mosbah A, et al. A randomized study comparing an antireflux system with a direct ureteric anastomosis in patients with orthotopic ileal neobladder. BJU Int 2006;97:1057–62. Evangelidis A, Lee EK, Karellas ME, et al. Evaluation of ureterointestinal anastomosis: Wallace vs. Bricker. J Urol 2006;175:1755–8. Mullins JK, Guzzo TJ, Ball MW, et al. Ureteral stents placed at the time of urinary diversion decreases postoperative morbidity. Urol Int 2012;88:66–70. Mattei A, Birkhaeuser FD, Baermann C, et al. To stent or not to stent perioperatively the ureteroileal anastomosis of ileal orthotopic bladder substitutes and ileal conduits? Results of a prospective randomized trial. J Urol 2008;179:582–6. Regan JB, Barrett DM. Stented versus nonstented ureteroileal anastomoses: is there a difference with regard to leak and stricture? J Urol 1985;134:1101–3.
Original article
The effect of length of ureteral resection on benign ureterointestinal stricture rate in ileal conduit or ileal neobladder urinary diversion following radical cystectomy Kyle A. Richards, M.D.a,*, Joshua A. Cohn, M.D.a, Michael C. Large, M.D.a,b, Gregory T. Bales, M.D.a, Norm D. Smith, M.D.a, Gary D. Steinberg, M.D.a a
Section of Urology, Department of Surgery, The University of Chicago Medical Center, Chicago, IL b Urology of Indiana, Indianapolis, IN Received 14 March 2014; received in revised form 17 April 2014; accepted 7 May 2014
Abstract Objectives: To assess the effect of the length of the ureter resected and other clinical variables on ureterointestinal anastomotic (UIA) stricture rate following radical cystectomy and ileal segment urinary diversion. Methods and materials: We identified 519 consecutive patients who underwent cystectomy and ileal conduit or ileal orthotopic neobladder diversion from January 2007 to August 2012. The length of the ureter resected was defined as the length of the ureter in the cystectomy specimen plus the length of the distal ureter submitted for pathologic analysis. The primary end point was the risk of UIA stricture formation, assessed by Cox proportional hazards analysis. Results: A total of 463 patients met the inclusion criteria with complete data. Median follow-up was 459 days (interquartile range [IQR]: 211–927). Median time to stricture formation was 235 (IQR: 134–352) and 232 days (IQR: 132–351) on the right and the left ureter, respectively. Overall stricture rate per ureter was 5.9% on the right vs. 10.0% on the left (P ¼ 0.03). There was no difference in demographic, operative, or perioperative variables between patients with and without UIA strictures. On multivariate analysis adjusted for age, sex, anastomosis technique (running vs. interrupted), and length of ureter resected, only a Clavien complication ZIII (hazard ratio ¼ 2.11, 1.01–4.40) and urine leak (hazard ratio ¼ 3.37, 1.08–10.46) significantly predicted for left- and right-sided stricture formation, respectively. The length of the ureter resected did not predict UIA stricture formation on either side. Conclusions: The etiology of benign UIA strictures following ileal urinary diversion is likely multifactorial. Our data suggest that a complicated postoperative course and urine leak are risk factors for UIA stricture formation. The length of the distal ureter resected did not significantly affect stricture rate. r 2014 Elsevier Inc. All rights reserved.
Keywords: Ureteral stricture; Ureterointestinal anastomosis; Urinary bladder neoplasms; Cystectomy; Postoperative complications; Reimplantation
1. Introduction Radical cystectomy (RC) and urinary diversion (UD) are the gold-standard treatments for muscle-invasive bladder cancer (BCa) [1]. Despite advances in anesthesia, surgical techniques, and perioperative care, postoperative morbidity within 3 months of surgery is greater than 60% in the
Corresponding author. Tel.: þ1-773-702-6189; fax: þ1-773-702-1001. E-mail address: [email protected] (K.A. Richards). *
http://dx.doi.org/10.1016/j.urolonc.2014.05.015 1078-1439/r 2014 Elsevier Inc. All rights reserved.
contemporary series [2]. Long-term complications related to the UD can also lead to significant morbidity. For example, the incidence of benign ureterointestinal anastomotic (UIA) strictures following RC and UD is reported at 1% to 13% [3–7]. Although most UIA strictures become clinically evident between 7 and 18 months postoperatively [5,7,8], patients occasionally present after more than 10 years from their initial surgery [4]. Patients with UIA strictures can present with nephrolithiasis, recurrent pyelonephritis, asymptomatic worsening hydronephrosis, or even more insidiously with impaired renal function. The pathophysiology of UIA stricture formation is likely secondary to ischemia and inflammation, resulting in
65.e2
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
fibrosis and ureteral obstruction; however, the exact etiology of UIA strictures remains elusive. Meticulous surgical technique, preservation of periureteral adventitia and blood supply, avoidance of electrocautery, and excision of the distal or compromised ureter before anastomosis are common strategies employed to mitigate the risks of UIA stricture formation [9]. Additionally, a running end-to-side Bricker UIAs [10], a postoperative urinary tract infection (UTI) [10], and an antirefluxing technique [3,4,11] are potential risk factors for UIA stricture formation. By contrast, Anderson et al. [12] were unable to identify any patient- or disease-specific factor to be associated with UIA stricture formation in a cohort of 478 patients from their institution. Although it is a common practice to resect the distal ureter that may or may not be compromised before ureteral anastomosis, the effect that this may have on the future development of UIA strictures is unclear. To the best of our knowledge, no previous studies have evaluated this variable as a risk factor for UIA stricture formation. We evaluated whether the length of the ureter excised at the time of performing an RC affects the future development of UIA strictures.
2. Materials and methods We obtained approval from our institutional review board before initiating this analysis. A prospective database was reviewed for patients undergoing RC with ileal conduit or ileal orthotopic neobladder UD by 2 surgeons (G.D.S. and N.D.S.) from 2007 to 2012. Both the surgeons have extensive experience in urologic oncology and large practices in BCa resection. An orthotopic neobladder was constructed as a Studer pouch or a modified Hautmann pouch with an isoperistaltic afferent limb. A permanent distal ureteral specimen is routinely sent intraoperatively but frozen sections are not. All UIAs were performed in a similar Bricker end-to-side tension-free technique using 4-0 polyglactin sutures on CV-23 needles. The decision to perform a running vs. interrupted UIA or to excise additional ureter was at the discretion of the surgeon. Our technique for running UIA has been previously described [10]. A single 4-0 monofilament traction suture is placed at the 12-o’clock position to avoid unnecessary manipulation and to ensure correct orientation. The left ureter is tunneled through a wide channel created beneath the sigmoid mesentery by grasping the traction suture with a large right-angled clamp. All anastomoses were stented with either 5-F feeding tubes or 7-F single J stents and tested for water tightness upon completion by irrigating the conduit or the afferent limb with a bulb syringe. Stents were removed following tolerance of oral diet (G.D.S) or approximately 2 weeks postoperatively (N.D.S). The length of the ureter resected was defined as the length of the ureter in the cystectomy specimen plus the length of the distal ureter submitted separately for pathologic analysis. It is a routine practice for
our pathologists to quantify the length of the ureter in the specimens and to comment on the presence of dysplasia or neoplasia in the distal ureter on the final pathologic report. The primary end points in our study were occurrence of UIA strictures and time to UIA stricture formation. Patients were followed up according to the National Comprehensive Cancer Network guidelines using serum chemistry studies and axial imaging of the abdomen and the pelvis every 3 to 6 months for 2 years based on the risk of recurrence and then annually or as clinically indicated [13]. All patients with recurrent pyelonephritis, worsening hydronephrosis, or worsening renal function were evaluated with an antegrade study or mercaptoacetyltriglycine renal scintigraphy or both to identify UIA strictures. The variables that were maintained and studied included basic demographics, the length of the ureter resected, age-adjusted Charlson comorbidity index [14], diversion type, clinical/pathologic stage, perioperative chemotherapy or radiation, estimated blood loss, body mass index, length of hospital stay, operative time, running vs. interrupted UIA, and perioperative complications up to 90 days. Complications were graded according to the ClavienDindo classification system [15]. For each period (30 d and 30–90 d), the highest grade of complication for each patient was identified, and the UIA stricture itself was not included as a specific Clavien complication to avoid confusion. UTI was diagnosed based on positive findings on urine culture (4105 organisms) in conjunction with fever or leukocytosis. Urine leak was identified in all cases by elevated drain creatinine levels beyond postoperative day 7. The decision to check drain creatinine levels is not routine and is instead based on clinical factors, such as prolonged ileus or excessive drain outputs. Patients who died within 30 days, had missing ureteral resection data, or developed ureteral obstruction due to recurrent cancer were excluded from this analysis. Statistical analysis was performed using Stata 12 (College Station, TX). A comparison of means was performed using the Student t test, and a comparison of medians was performed using the Mann-Whitney U test. The Fisher exact and the chi-square tests were used for a comparison of categorical variables. A separate analysis was performed for each ureter at risk (left and right), to account for both differential risk of stricture formation by side and the subset of patients with a solitary kidney at the time of cystectomy. A multivariate Cox proportional hazards model was performed for the aforementioned perioperative variables to identify independent predictors of UIA stricture formation. An adjusted analysis was performed by controlling for age, sex, and sum of ureter removed, as well as the following variables that were significant on unadjusted analysis: urine leak, anastomotic technique, and a Clavien complication Zgrade III at 30 days. The Kaplan-Meier survival analysis and the stratified log-rank test were used to compare time to UIA stricture formation per ureter at risk based on the length of the ureter resected. A 2-sided P o 0.05 was considered significant.
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
3. Results The study sample consisted of 519 patients. Of them, 13 patients died within 30 days after undergoing RC and 43 patients had missing data on the length of the ureter resected, leaving 463 in the total study population, of which 439 (95%) underwent open RC. The 24 patients who underwent robotic RC had UDs performed extracorporeally, with only 1 patient developing a UIA stricture. For the entire cohort, 76 patients (16.4%) were evaluated with either mercaptoacetyltriglycine renal scintigraphy (41%) or antegrade nephrostogram (59%) for a rise in creatinine levels (11 patients, 14.5%), worsening hydronephrosis (24 patients, 31.6%), rise in creatinine/worsening hydronephrosis (25 patients, 32.9%), or recurrent pyelonephritis (16 patients, 21.0%). After these evaluations, 58 patients (12.5%) were found to have UIA strictures at an overall median follow-up of 459 days (interquartile range [IQR]: 211–927). UIA strictures developed more commonly in men (14.4%) than in women (5.3%) (P ¼ 0.018). However, there was no difference in other demographic or baseline clinical variables between those with or without UIA strictures (Table 1). UIA stricture formation was more common in patients with organ-confined disease than in patients with non–organ-confined disease on final pathologic
65.e3
analysis (15.9% vs. 8.3%, P ¼ 0.014) and in patients who were node negative than in patients who were node positive (14.8% vs. 5.9%, P ¼ 0.012), although follow-up was longer for patients with more favorable findings on pathologic analysis. For each ureter at risk, the left ureters were more likely than the right ureters to develop UIA strictures (10.0% vs. 5.9%, P ¼ 0.03). For the right ureters at risk, the median follow-up was longer for those patients who developed UIA strictures than for those who did not develop UIA strictures (683 d [IQR: 470–1,384] vs. 427 d [IQR: 169–878], P o 0.01) and the median time to stricture formation was 235 days (IQR: 134–352). For each right ureter at risk, there was no difference in diversion type, anastomotic technique, operative time, estimated blood loss, total length of ureter resected, or length of hospital stay in those patients who developed UIA strictures vs. those free from UIA strictures (Table 2). Similar findings were noted for each left ureter at risk. Regarding perioperative outcomes, 57.0% of the cohort experienced a complication within 30 days of surgery, of which 17.3% were Clavien grade ZIII. Furthermore, 22.5% of the cohort experienced a complication 30 to 90 days after surgery, of which 6.9% were Clavien grade ZIII. For the patients with the right ureters at risk, those who developed right UIA strictures were more likely to have sustained a
Table 1 Cohort characteristics Overall
Stricture
No stricture
Number of patients (%) Age, y (95% CI) Sex, male, n (%) Sex, female, n (%) Body mass index (95% CI) Charlson comorbidity index (95% CI)
463 68.2 369 94 28.2 2.3
58 68.6 53 5 29.1 2.2
405 68.2 316 89 28.0 2.3
Preoperative pathology Clinically muscle invasive, n (%) Non–muscle invasive, n (%)
301 (65.0) 162 (35.0)
32 (10.6) 26 (16.1)
269 (89.4) 136 (83.9)
0.093
64 (13.8) 48 (10.3)
6 (6.3) 7 (14.6)
60 (93.7) 41 (85.4)
0.102 0.649
Neoadjuvant chemotherapy, n (%) Preoperative radiation, n (%)
(67.2–69.3) (79.7) (20.3) (27.6–28.7) (2.1–2.5)
Ureters at risk Both, n (%) Left only, n (%) Right only, n (%)
401 (86.6) 38 (8.2) 24 (5.2)
Stricture laterality Both, n (%) Left only, n (%) Right only, n (%)
11 (19.0) 33 (56.9) 14 (24.1)
Postoperative pathology Organ confined, n (%) Non–organ confined, n (%) Positive soft tissue margins, n (%) Negative soft tissue margins, n (%) Node positive, n (%) Node negative, n (%)
258 205 33 430 118 345
(55.7) (44.3) (7.1) (92.9) (25.5) (74.5)
Note: Bold signifies results that are statistically significant (P o 0.05).
41 17 1 57 7 51
(12.5) (66.2–71.1) (14.4) (5.3) (27.4–30.7) (1.8–2.6)
(15.9) (8.3) (3.0) (13.3) (5.9) (14.8)
217 188 32 373 111 294
(87.5) (67.1–69.3) (85.6) (94.7) (27.5–28.6) (2.2–2.5)
(84.1) (91.7) (97.0) (86.7) (94.1) (85.2)
P value 0.73 0.018 0.239 0.56
0.014 0.087 0.012
65.e4
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
Table 2 Perioperative variables and outcomes Overall
No stricture
Stricture
P value
Right ureter at risk (n ¼ 425) Number of patients (%) Median follow-up, d (IQR) Mean follow-up, d (95% CI) Median time to stricture formation, d (IQR)
463 (–) 459 (211–927) 661 (606–715)
400 (94.1) 427 (169–878) 614 (557–670)
Diversion type Ileal conduit, n (%) Orthotopic neobladder, n (%)
288 (62.2) 175 (37.8)
253 (63.3) 147 (36.8)
12 (48.0) 13 (52.0)
0.13
Anastomosis Interrupted, n (%) Running, n (%)
181 (39.1) 282 (60.9)
158 (39.5) 242 (60.5)
12 (48.0) 13 (52.0)
0.40
Operative time, min (95% CI) Estimated blood loss, ml (95% CI)
304 (298–311) 996 (944–1,047)
304 (297–312) 985 (931–1,038)
295 (276–314) 1,096 (861–1,331)
0.36 0.35
Total length of ureter removed, mm (95% CI) In specimen Resected intraoperatively
26.0 (24.2–27.7) 16.8 (15.2–18.3) 9.2 (8.2–10.2)
22.2 (14.7–29.8) 14.7 (8.6–20.8) 7.5 (4.4–10.7)
0.16 0.51 0.30
Sum of ureter removed, categorized 0–10 mm, n (%) 11–30 mm 31–50 mm 450 mm
98 167 103 32
(24.5) (41.8) (25.8) (8.0)
25 683 950 235
8 11 3 3
(5.9) (470–1,384) (692–1,206) (134–352)
. o0.01 0.01
(32.0) (44.0) (12.0) (12.0)
0.42
11.6 (8.7–14.5)
0.54
Length of stay, d (95% CI)
10.8 (10.2–11.4)
10.7 (10.0–11.4)
Complications, 30 d Any, n (%) Clavien grade 4II, n (%)
264 (57.0) 80 (17.3)
224 (56.8) 65 (16.3)
16 (64.0) 6 (24.0)
0.48 0.31
Complications, 30–90 d Any, n (%) Clavien grade 4II, n (%)
104 (6.9) 32 (13.4)
84 (21.0) 47 (11.8)
9 (36.0) 8 (32.0)
0.08 o0.01
25 48 27 31
5 6 3 1
(20.0) (24.0) (12.0) (3.1)
o0.01 0.08 0.32 0.49
(10.0) (414–1,462) (762–1,179) (132–351)
0.03a 0.01 o0.01 –
30-d 30-d 30-d 30-d
Urine leak, n (%) Urinary tract infection, n (%) Intra-abdominal abscess, n (%) Clostridium difficile infection, n (%)
33 60 36 35
(7.1) (13.0) (7.8) (7.6)
(6.3) (12.0) (6.7) (7.8)
Left ureter at risk (n ¼ 439) Number of patients (%) Median follow-up, d (IQR) Mean follow-up, d (95% CI) Median time to stricture formation, d (IQR)
395 (90.0) 431 (171–878) 617 (560–674)
44 714 970 232
Diversion type Ileal conduit, n (%) Orthotopic neobladder, n (%)
250 (63.3) 145 (36.7)
26 (59.1) 18 (40.9)
0.58
Anastomosis Interrupted, n (%) Running, n (%)
156 (39.5) 239 (60.5)
14 (31.8) 30 (68.2)
0.32
Operative time, min (95% CI) Estimated blood loss, ml (95% CI)
304 (296–311) 974 (921–1,027)
303 (284–321) 1,134 (897–1,371)
0.91 0.19
Total length of ureter removed, mm (95% CI) In specimen Resected intraoperatively
22.8 (21.2–24.4) 16.1 (14.7–17.6) 6.7 (5.8–7.6)
21.5 (16.3–26.6) 14.8 (10.4–19.3) 6.6 (3.2–10.1)
0.61 0.58 0.98
Sum of ureter removed, categorized 0–10 mm, n (%) 11–30 mm 31–50 mm 450 mm
115 176 83 21
(29.1) (44.6) (21.0) (5.3)
13 20 9 2
(29.6) (45.5) (20.5) (4.6)
0.99
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
65.e5
Table 2 Continued Overall
No stricture
Stricture
Length of stay, d (95% CI)
10.7 (10.0–11.4)
Complications, 30 d Any, n (%) Clavien grade 4II, n (%)
222 (56.2) 62 (15.7)
26 (59.1) 12 (27.3)
0.71 0.05
Complications, 30–90 d Any, n (%) Clavien grade 4II, n (%)
84 (21.3) 46 (11.7)
13 (29.6) 9 (20.5)
0.21 0.09
30-d 30-d 30-d 30-d
24 46 27 30
a
Urine leak, n (%) Urinary tract infection, n (%) Intra-abdominal abscess, n (%) Clostridium difficile infection, n (%)
(6.1) (11.7) (6.8) (7.6)
10.7 (8.9–12.4)
P value
5 9 7 3
(11.4) (20.5) (15.9) (6.8)
0.97
0.18 0.09 0.03 0.85
P value refers to left vs. right.
major complication 30 to 90 days from surgery (32.0%) than those patients who were free from right UIA strictures (11.8%) (P o 0.01) were. In addition, the urine leak rate was 20% in those with right UIA strictures vs. 6.3% in those free from right UIA strictures (P o 0.01) (Table 2). For the patients with the left ureters at risk, major complications within 30 days of surgery were more common in those who developed left UIA strictures (27.3%) than in those free of left UIA strictures (15.7%) (P ¼ 0.05). The 30-day intra-abdominal abscess rate was also significantly higher in those with left UIA strictures than in those who were free from left UIA strictures (15.9% vs. 6.8%, P ¼ 0.03). While controlling for sex, age, anastomotic technique, sum of the left ureter resected, and urine leak, major complications within 30 days of surgery were associated with left UIA strictures (hazard ratio [HR] ¼ 2.11; 95% CI: 1.01–4.40; P ¼ 0.046) in our Cox proportional hazards model (Table 3). While controlling for the same variables, urine leak was associated with right UIA strictures (HR ¼ 3.37; 95% CI: 1.08–10.46; P ¼ 0.036). The Kaplan-Meier analysis of time to stricture formation showed no difference in freedom between left and right UIA strictures stratified by the length of the ureter resected (Fig.).
4. Discussion Although several studies have attempted to identify the risk factors for the development of UIA strictures, this is the first study to our knowledge to specifically evaluate the importance of the length of the distal ureter resected at the time of RC and UD. The exact etiology of benign UIA strictures in any individual patient is unclear. Certainly, proper attention to meticulous surgical technique is important. These technical considerations should include careful tissue handling, preservation of ureteral blood supply, minimization of electrocautery, and precise placement of
sutures. The concept of resecting additional distal ureter before performing a UIA (i.e., to “freshen up” the ureter) stems from the idea that the distal ureteral blood supply may have been compromised during its dissection at the time of RC. Although this rationale seems logical, one could also argue that resecting additional ureter could also lead to increased risk of UIA stricture formation if undue tension might be placed on the UIA. It is with these theoretical concerns that we set forth to determine the effect of a resected ureter on the incidence of benign UIA strictures. Anderson et al. [12] evaluated potential risk factors for UIA stricture formation in 478 consecutive patients who underwent RC and UD at their institution from 2007 to 2011. In their cohort, 21.5% underwent robotic RC with extracorporeal UD, and the overall rate of UIA stricture was 9.4%. Furthermore, in their multivariable-adjusted model, none of the perioperative variables (robotic technique, age, diversion type, sex, or the American Society of Anesthesiologists class) were predictors of UIA stricture formation; however, the length of the ureter resected was not studied. Similarly, sex and age were not predictors of UIA stricture formation in our adjusted model, and Charlson comorbidity index and diversion type were not predictors of UIA stricture formation in our unadjusted model. Careful dissection and handling of the ureter, including wide spatulation, remain an art and cannot be accurately measured or quantified. However, there are other components of the UIA technique that have been studied for their effect on UIA stricture development. Large et al. [10] evaluated the effect of running vs. interrupted UIA on benign UIA stricture development in a consecutive series of 266 patients. On multivariable analysis controlling for age, sex, final pathologic analysis, and 30-day Clavien complications Zgrade III, running anastomosis (HR ¼ 1.92, 95% CI: 1.00–3.70) and 30-day perioperative UTI (HR ¼ 2.44, 95% CI: 1.17–5.08) predicted UIA stricture formation. With a larger cohort of patients and longer follow-up, these variables were no longer significant predictors in the present study. However, on
65.e6
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
Table 3 Unadjusted and adjusted Cox proportional hazards model with stricture diagnosis as the primary outcome Variable (referent)
Unadjusted HR (95% CI)
P value
Adjusted HR (95% CI)
P value
0.905 0.081 0.242 0.659 0.536 0.536 0.64 0.065 0.391 0.088
– 2.64 (0.93–7.45) 1.02 (0.99–1.05) – – – – – – 1.68 (0.85–3.33)
– 0.067 0.248 – – – – – – 0.137
Left ureter OR time Sex Age, y Length of stay Preoperative pathology (muscle invasive) Diversion (ONB vs. IC) Non–organ confined Positive nodes Positive margin Running anastomosis
0.99 2.50 1.02 1.01 0.83 0.83 0.86 0.38 0.42 1.79
(0.996–1.004) (0.89–6.99) (0.99–1.04) (0.96–1.06) (0.45–1.51) (0.45–1.51) (0.45–1.63) (0.14–1.06) (0.06–3.05) (0.92–3.49)
Sum L ureter removed (ref: 0–10 mm) 11–30 mm 30–50 mm 450 mm
1.01 (0.50–2.03) 1.02 (0.43–2.41) 0.82 (0.18–3.63)
0.977 0.947 0.792
– – –
– – –
Sum L ureter removed þ Ureteral margin 30-d C. difficile infection Postoperative abscess Dehiscence Wound infection 30-d Urinary tract infection EBL BCG history Urine leak Readmit 30 d Neoadjuvant chemotherapy Radiation history Body mass index Age-adjusted CCI Clavien grade III or more, 30-d complication Any 30–90-d complication
0.99 0.71 0.77 2.81 1.40 1.91 1.84 1.00 0.63 2.10 2.80 0.72 1.24 1.04 1.05 2.6 0.79
0.745 0.64 0.654 0.012 0.196 0.143 0.103 0.082 0.237 0.118 0.001 0.527 0.653 0.154 0.541 0.005 0.5
0.995 (0.98–1.01) – – – – – – – – 1.59 (0.58–4.37) – – – – – 2.11 (1.01–4.40) –
0.64 – – – – – – – – 0.372 – – – – – 0.046 –
0.491 0.147 0.511 0.221 0.645 0.563 0.52 0.983 0.835 0.657
– 2.6 (0.60–11.25) 1.02 (0.98–1.06) – – – – – – 0.88 (0.38–2.04)
– 0.200 0.271 – – – – – – 0.760
(0.98–1.02) (0.17–2.95) (0.24–2.47) (1.25–6.32) (0.84–2.32) (0.80–4.53) (0.88–3.83) (0.99–1.00) (0.29–1.36) (0.83–5.35) (1.52–5.14) (0.26–2.01) (0.49–3.14) (0.99–1.09) (0.89–1.25) (1.33–5.09) (0.39–1.58)
Right ureter OR time Sex Age, y Length of stay Preoperative pathology (muscle invasive) Diversion (ONB vs. IC) Non–organ confined Positive nodes Positive margin Running anastomosis
0.99 2.91 1.01 1.03 0.83 1.26 0.75 0.99 0.81 0.83
(0.992–1.004) (0.69–12.36) (0.98–1.05) (0.98–1.09) (0.38–1.83) (0.57–2.78) (0.31–1.81) (0.37–2.64) (0.11–6.01) (0.38–1.85)
Sum R ureter removed (ref: 0–10 mm) 11–30 mm 30–50 mm 450 mm
0.85 (0.34–2.11) 0.40 (0.11–1.50) 1.17 (0.31–4.42)
0.729 0.173 0.816
– – –
– – –
Sum R ureter removed þ Ureteral margin 30-d C. difficile infection Postoperative abscess Dehiscence Wound infection 30-d Urinary tract infection EBL BCG history Urine leak Readmit 30 d
0.99 1.58 0.43 1.99 1.09 2.16 2.26 1.00 0.55 3.92 1.73
0.326 0.538 0.405 0.265 0.835 0.159 0.082 0.401 0.279 0.006 0.22
0.99 (0.97–1.02) – – – – – – – – 3.37 (1.08–10.46) –
0.515 – – – – – – – – 0.036 –
(0.96–1.01) (0.37–6.69) (0.06–3.16) (0.59–6.65) (0.47–2.52) (0.74–6.32) (0.90–5.66) (0.99–1.00) (0.19–1.61) (1.47–10.47) (0.72–4.14)
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
65.e7
Table 3 Continued Variable (referent)
Unadjusted HR (95% CI)
P value
Adjusted HR (95% CI)
P value
Neoadjuvant chemotherapy Radiation history Body mass index Age-adjusted CCI Clavien grade III or more, 30-d complication Any 30–90-d complication
0.29 1.98 0.99 1.03 2.09 2.53
0.226 0.211 0.861 0.752 0.118 0.026
– – – – 1.44 (0.49–4.22) –
– – – – 0.511 –
(0.04–2.15) (0.68–5.77) (0.92–1.07) (0.83–1.30) (0.83–5.26) (1.12–5.73)
BCG ¼ bacillus Calmette-Guérin; CCI ¼ Charlson comorbidity index; EBL ¼ estimated blood loss; IC ¼ ileal conduit; ONB ¼ orthotopic neobladder; OR ¼ operating room.
unadjusted analysis, the data in the present study did suggest a trend toward increased risk for stricture formation for both the ureters with 30-day UTI and for the left ureter only with a running vs. interrupted anastomosis. Other technical modifications have been studied, including refluxing vs. nonrefluxing techniques and stented vs. nonstented UIA. There is a higher rate of benign UIA stricture in series that have evaluated nonrefluxing techniques (8.3%–16.3%) than in those evaluating refluxing techniques (1.7%–5.4%), without any appreciable benefits for patients with nonrefluxing UIA [3,4,11,16]; however, the type of refluxing UIA (Wallace vs. Bricker) does not seem to matter [8,17]. Stenting the UIA is a common procedure that is thought to attenuate the risk of urine leak. Despite concerns that stents may actually increase the risk of UIA stricture formation (being a foreign body), numerous studies have failed to show an increased rate of UIA stricture in stented than in nonstented UIA [18–20]. All patients in our series underwent refluxing and stented UIA to help control potentially modifiable risk factors for UIA stricture formation. Urine leak was a significant predictor of UIA stricture formation in our right ureter–adjusted multivariable model despite stenting and testing all UIA for water tightness. Therefore, every effort should be made through careful surgical technique of the UIA to minimize the risk of postoperative urine leak. In addition, patients with major (Clavien grade ZIII) 30-day postoperative complications were at an increased risk of UIA stricture formation in the left ureter using the multivariable-adjusted model. Patients with a complicated postoperative course should be followed up closely for insidious development of benign UIA strictures, as they may be at an increased risk for this complication. There are several limitations in this study that deserve specific mention. Ours being a tertiary care medical center with very specialized and complex practices in BCa surgery, these results may not have external validity in the general or community urology practice. With large and widespread referral patterns, follow-up care can be challenging with the potential inability to completely capture all postoperative events once the patient is discharged from the hospital. Our research team contacts patients who may have
been lost to follow-up and reviews outside medical records when needed to ensure complete data collection. In addition, the length of the ureter resected was measured by numerous pathologists during the study period, with the resultant possibility of measurement error or bias. The exact location of the stricture was not available in each patient, and we are therefore unable to comment on other technical factors, including the degree of angulation of the left ureter as it passes beneath the sigmoid mesentery. Finally, other subtle changes in our practice over the 5-year study period (i.e., elimination of mechanical bowel preparation) may
Fig. The Kaplan-Meier curve of the proportion of patients stricture free by the length of the ureter resected for the left ureters at risk (top) and the right ureters at risk (bottom). (Color version of figure is available online.)
65.e8
K.A. Richards et al. / Urologic Oncology: Seminars and Original Investigations 33 (2015) 65.e1–65.e8
have had an unmeasured bias on the study results for which we were unable to control. Despite the aforementioned limitations, this is the first series to evaluate the effect of the length of the distal ureter resected as a potential risk factor for UIA stricture formation. This study highlights that the exact etiology of UIA strictures remains elusive and is likely multifactorial. Despite technical modifications, a sound surgical technique (which is difficult to quantify) remains critically important.
[7] [8]
[9] [10]
[11]
5. Conclusions The etiology of benign UIA strictures following ileal UD is likely multifactorial. Our data suggest that a complicated postoperative course and urine leak are risk factors for UIA stricture formation. The length of the distal ureter resected did not significantly affect the stricture rate.
[12]
[13]
[14]
References [1] Stein JP, Lieskovsky G, Cote R, et al. Radical cystectomy in the treatment of invasive bladder cancer: long-term results in 1,054 patients. J Clin Oncol 2001;19:666–75. [2] Shabsigh A, Korets R, Vora KC, et al. Defining early morbidity of radical cystectomy for patients with bladder cancer using standardized reporting methodology. Eur Urol 2009;55:164–76. [3] Studer UE, Burkhard FC, Stumacher M, et al. Twenty years experience with an ileal orthotopic low pressure bladder substitute —lessons to be learned. J Urol 2006;176:161–6. [4] Hautmann RE, de Petriconi RC, Volkmer BG. 25 years of experience with 1,000 neobladders: long-term complications. J Urol 2011;185:2207–12. [5] Tal R, Sivan B, Kedar D, et al. Management of benign ureteral strictures following radical cystectomy and urinary diversion for bladder cancer. J Urol 2007;178:538–42. [6] Katkoori D, Samavedi S, Adiyat KT, et al. Is the incidence of ureterointestinal anastomotic stricture increased in patients undergoing
[15]
[16]
[17] [18]
[19]
[20]
radical cystectomy with previous pelvic radiation? BJU Int 2010;105: 795–8. Shimko MS, Tollefson MK, Umbreit EC, et al. Long-term complications of conduit urinary diversion. J Urol 2011;185:562–7. Kouba E, Sands M, Lentz A, et al. A comparison of the Bricker versus Wallace ureteroileal anastomosis in patients undergoing urinary diversion for bladder cancer. J Urol 2007;178:945–8. Richards KA, Steinberg GD. Perioperative outcomes in radical cystectomy: how to reduce morbidity? Curr Opin Urol 2013;23:456–65. Large MC, Cohn JA, Kiriluk KJ, et al. The impact of running versus interrupted anastomosis on ureterointestinal stricture rate after radical cystectomy. J Urol 2013;190:923–7. Pantuck AJ, Han KR, Perrotti M, et al. Ureteroenteric anastomosis in continent urinary diversion: long-term results and complications of direct versus nonrefluxing techniques. J Urol 2000;163:450–5. Anderson CB, Morgan TM, Kappa S, et al. Ureteroenteric anastomotic strictures after radical cystectomy—does operative approach matter? J Urol 2013;189:541–7. National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology (NCCN Guideliness). Bladder Cancer, Version 1.2014. Available at: www.nccn.org; [accessed 17.4.14]. Charlson M, Szatrowski TP, Peterson J, et al. Validation of a combined comorbidity index. J Clin Epidemiol 1994;47:1245–51. Dindo D, Demartines N, Clavien P-A. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205–13. Shaaban AA, Abdel-Latif M, Mosbah A, et al. A randomized study comparing an antireflux system with a direct ureteric anastomosis in patients with orthotopic ileal neobladder. BJU Int 2006;97:1057–62. Evangelidis A, Lee EK, Karellas ME, et al. Evaluation of ureterointestinal anastomosis: Wallace vs. Bricker. J Urol 2006;175:1755–8. Mullins JK, Guzzo TJ, Ball MW, et al. Ureteral stents placed at the time of urinary diversion decreases postoperative morbidity. Urol Int 2012;88:66–70. Mattei A, Birkhaeuser FD, Baermann C, et al. To stent or not to stent perioperatively the ureteroileal anastomosis of ileal orthotopic bladder substitutes and ileal conduits? Results of a prospective randomized trial. J Urol 2008;179:582–6. Regan JB, Barrett DM. Stented versus nonstented ureteroileal anastomoses: is there a difference with regard to leak and stricture? J Urol 1985;134:1101–3.
