- Email: [email protected]
Predictive factors for final pathologic ureteral sections on 700 radical cystectomy specimens: Implications for intraoperative frozen section decision-making
Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Original article
Predictive factors for final pathologic ureteral sections on 700 radical cystectomy specimens: Implications for intraoperative frozen section decision-making Alexandra Masson-Lecomte, M.D.a,*, Thomas Francois, M.D.b, Dimitri Vordos, M.D.c, Carole Cordonnier, M.D.d, Yves Allory, M.D.e, Francois Desgrandchamps, Ph.D.a, Alexandre de la Taille, Ph.D.c, Fabien Saint, Ph.D.b,f a
Department of Urology and Transplantation, Saint Louis Academic Hospital, Paris Diderot University, Paris, France b Department of Urology and Transplantation, Picardie Jules Verne University, CHU Amiens, Amiens, France c Department of Urology, Henri Mondor Hospital, Paris Est Creteil University, Créteil, France d Department of Pathology, Picardie Jules Verne University, CHU Amiens, Amiens, France e Department of Pathology, Henri Mondor Hospital, Paris Est Creteil University, Créteil, France f HeRVI EA 3801, Picardie Jules Verne University, CHU Amiens, Amiens, France Received 18 April 2017; received in revised form 16 June 2017; accepted 19 June 2017
Abstract Purpose: To identify preoperative predictive factors for final ureteral section invasion after radical cystectomy (RC) and to validate significant factors on an external independent cohort. Material and methods: We retrospectively reviewed data of all consecutive RC performed for bladder cancer in 2 high-volume institutions. Clinical, pathological, and follow-up data were collected prospectively and reviewed retrospectively. Pathological evaluation was performed by 2 well-trained uropathologists in each center. Logistic regression analyses were performed to identify predictive factors for final ureteral sections involvement. Significant factors in cohort A were validated in cohort B. Receiver operating curve and area under curve were modeled to evaluate predictive accuracy of the markers. Results: A total of 441 RC were performed in center A and 307 RC were performed in center B. Mean follow-ups were 36.2 and 38.1 months, respectively. Invasion of the final ureteral section was observed on 5.5% of patients in cohort A and 4.8% of patients in cohort B. In cohort A, multivariable logistic regression identified preoperative hydronephrosis on computed tomography scan (odds ratio [OR] ¼ 4.9, P ¼ 0.004) and presence of Carcinoma in situ (CIS, OR ¼ 3.9, P ¼ 0.01) as the only factors associated with ureteral sections positivity. In cohort B, hydronephrosis and CIS were both associated with ureteral sections positivity in univariable analysis. In multivariable analysis, only hydronephrosis remained significant (OR ¼ 5.9, P ¼ 0.01). Predictive accuracy of hydronephrosis and CIS combined in 1 variable was 0.72. Conclusion: Hydronephrosis and bladder CIS have good accuracy in predicting ureteral sections positivity after RC. In the presence of those factors, ureteral frozen sections should be performed. r 2017 Elsevier Inc. All rights reserved.
Keywords: Urothelial carcinoma; Radical cystectomy; Ureteral frozen section; Predictive factors
1. Introduction Radical cystectomy (RC) is the standard treatment option for nonmetastatic, resectable muscle-invasive bladder Corresponding author. Tel.: þ33-60-382-8593. E-mail address: [email protected] (A. Masson-Lecomte). *
http://dx.doi.org/10.1016/j.urolonc.2017.06.053 1078-1439/r 2017 Elsevier Inc. All rights reserved.
cancer. Preceded by neoadjuvant chemotherapy, it allows local and distant control of the disease with 5-year survival rates up to 80% in the most favorable cases [1]. However, to maximize oncologic outcome, the surgical procedure has to be standardized [2]. Although it is pretty much consensual that pathologic urethral frozen section (FS) should lead to uretrectomy and
2
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
preclude neobladder reconstruction, intraoperative management of ureteral sections remains a matter of debate. In that regard, the latest European Association of Urology guidelines do not provide any recommendation. Analysis of available literature questions both the ability of intraoperative FSs to properly predict final involvement of ureters and the effect of invaded intraoperative ureteral sections on oncological outcome and upper tract recurrence rates [3–5]. Taking into account the high cost of the procedure [6], FSs are not anymore performed in routine in many centers. However, the risk of upper tract recurrence is probably underestimated because of small cohorts, short follow-ups, and competing risk with dying from bladder cancer [7]. Some patients might benefit from sequential ureteral sections and achievement of healthy final ureteral margins. To better define the patients for whom particular attention should be paid to the distal ureters management, predictive factors for ureteral sections invasion on final pathology report are needed. Using 2 large cohorts of RC, we aimed at (1) identifying preoperative predictive factors for final ureteral section invasion and (2) validating these factors on an external independent cohort.
2. Materials and methods 2.1. Population After obtaining institutional review board approval, we retrospectively reviewed data of all consecutive RC performed for bladder cancer in 2 independent institutions. The procedures were performed between 1980 and 2013 for the exploratory cohort (A) and between 2001 and 2013 for the validation cohort (B). Surgery was performed through either open or laparoscopic approach with an extended pelvic lymph node dissection up to the common iliac bifurcation. Ureters were sectioned at the discretion of each surgeon, generally a few centimeters distal to the iliac bifurcation. If the section was not macroscopically healthy, a new section was performed 1 centimeter above the previous section. In cohort A, most patients did not have intraoperative FS examination although it was performed in some cases depending on the surgeon’s habits (N ¼ 92). All patients had intraoperative FS in the validation cohort B. All cystectomies performed for nononcological purposes were excluded. All patients had nonmetastatic bladder cancer on preoperative computed tomography (CT) scans. Following clinical and pathological data were collected for all patients: sex, age at RC, American Society of Anesthesiologists score, tobacco use, exposure to workrelated hazards, number of transurethral resection on the bladder (TURB), time between TURB and RC, pathology on final TURB, preoperative hydronephrosis on abdominal CT scans (defined by an anteroposterior diameter of the renal pelvis 410 mm), presence of carcinoma in situ (CIS) on TURB and RC specimen, multiplicity, tumor location,
neoadjuvant chemotherapy administration, date of cystectomy, type of urinary diversion, final pathological T stage, N stage, histological subtype, grade, margins status, and final ureteral sections pathology examination. 2.2. Pathological examination Pathological evaluation was performed by 2 well-trained uropathologists in each center (Y.A. and C.C). No central review was performed. In cohort A, both intraoperative and final ureteral sections analyses were available, while only final analysis was available in cohort B. Ureteral sections were considered positive if urothelial carcinoma (UC) (in situ or invasive) was in contact with the edge of the ureteral resection. 2.3. Statistical analysis Descriptive analyses of the 2 cohorts were performed. Continuous variables were compared using a Wilcoxon test and categorical variables using a chi-square test. Unadjusted and adjusted logistic regression analyses were performed to identify predictive factors for final ureteral sections involvement. Variables with a P o 0.2 in univariable analyses were kept for multivariable analysis. Final P o 0.05 were considered significant. Receiver operating curve (ROC) and area under curve (AUC) were modeled/calculated using the pROC package (https://cran.r-project.org/web/packages/ pROC/pROC.pdf). Statistical analyses were performed using “R” version 3.3.2 (2016-10-31).
3. Results 3.1. Characteristics of the cohorts Overall, 441 RC were performed in center A (exploratory cohort) and 307 RC were performed in center B (validation cohort). The mean and median follow-ups were, respectively, 36.2 (1–255) and 14.6 months in cohort A and 38.1 months (3–181) and 27.9 months in cohort B. Neoadjuvant chemotherapy was administered to 14 patients (3.2%) in cohort A and 16 patients (5%) in cohort B. Ninety-two patients had intraoperative examination of the ureters in cohort A (20%) while all patients had FS in cohort B. In cohort A, the most common type of urinary diversion was cutaneous ureterostomy (200; 46%), followed by ileal conduit (115; 26%) and orthotopic neobladder (107; 24%). In cohort B, 173 patients had ileal conduit (46%), 120 had an orthotopic neobladder (39%), and 4 cutaneous ureterostomy (1%). Patients’ characteristics are presented and compared in Table 1. Invasion of the final ureteral section was observed on 5.5% of patients in cohort A and 4.8% of patients in cohort B.
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
3
3.3. Prediction of ureteral section positivity using hydronephrosis and CIS
Table 1 Patients characteristics Cohort A n 441 Sex Male (%) 391 ASA score (%) 1 25 2 200 3 123 4 3 NA 90 Age (mean [sd]) 65.35 Number of TURBT (%) 0 2 1 265 2 73 3 31 Z4 62 NA 8 Hydronephrosis (%) 0 281 1 148 NA 12 pT stage (%) T1 85 T2 95 T3 164 T4 92 NA 5 pN stage (%) Node negative 296 Node positive 115 NA 30 CIS (%) No 387 Yes 53 NA 1 Positive margins, n (%) NA Pathological ureteral sections (%) No 282 Yes 21 NA 138
Cohort B
P
307 (88.7)
270
(87.9)
0.854
(5.7) (45.4) (27.9) (0.7) (20.4) (9.79)
25 153 57 1 71 67.19
(8.1) (49.8) (18.6) (0.3) (23.1) (9.95)
0.04
(0.5) (60.1) (16.6) (7) (14.1) (1.8)
2 154 66 23 32 30
(0.7) (50.2) (21.5) (7.5) (10.4) (9.8)
o0.001
(63.7) (33.6) (2.7)
225 82 0
(73.3) (26.7) 0
0.001
(19.3) (21.5) (37.2) (20.9) (1.1)
95 52 107 53 0
(30.9) (16.9) (34.9) (17.3) 0
0.002
(67.1) (26.1) (6.8)
202 82 23
(65.8) (26.7) (7.5)
0.906
(87.8) (12) (0.2) NA
130 165 12 45
(42.3) (53.7) (3.9) (14)
o0.001
(63.9) (4.8) (31.3)
290 17 0
(94.5) (5.5) 0
o0.001
0.013
In cohort A, sensitivity and specificity of the presence of both hydronephrosis and CIS combined in 1 variable for prediction of ureteral section positivity were 85.7% and 54.3%, respectively. Accuracy of hydronephrosis and CIS combined in the cohort A was 0.72 (Fig. 1). In cohort B, CIS alone was not independently associated with ureteral section positivity. Sensitivity and specificity of hydronephrosis alone for prediction of ureteral positivity were 52.9% and 74.8%, respectively. Accuracy of hydronephrosis alone in the cohort B was 0.64 (Fig. 2). Adding CIS to hydronephrosis increased specificity to 90% while lowering sensitivity to 43% and did not improve AUC (0.64). 4. Discussion
ASA ¼ American Society of Anesthesiologists; NA ¼ not available.
3.2. Factors associated with final ureteral section positivity Logistic regression results are presented in Tables 2 and 3. In cohort A, multivariable logistic regression identified preoperative hydronephrosis on CT scan (odds ratio [OR] ¼ 4.9, P ¼ 0.004) and presence of CIS (OR ¼ 3.9, P ¼ 0.01) as the only factors associated with ureteral sections positivity. Cohort B was explored to validate those results. In cohort B, hydronephrosis and CIS were both associated with ureteral sections positivity in univariable analysis. In multivariable analysis adjusted for T stage and N stage, only hydronephrosis remained significant (OR ¼ 5.9, P ¼ 0.01). OR for CIS followed the same tendency as observed in cohort A but was not significant (OR ¼ 3.4, P ¼ 0.10)
At present, the European guidelines do not mention any recommendation regarding intraoperative management of ureters during RC performed for nonmetastatic UC [8]. Consequently, clinical practice is very heterogeneous, and FSs with sequential ureteral resections are performed at the discretion of the surgeon, based on pre/intraoperative clinical feeling. Controversies regarding the interest of ureteral FSs came from the following: (1) Uneven sensitivity and specificity of FS compared to final pathology [9–12], (2) frequent failure to obtain tumor-free sections after sequential resections, particularly in case of CIS [9,13], (3) unclear effect of tumor presence on ureteral section on upper tract recurrence and patients’ prognosis [3,4,10], (4) and direct and indirect costs [6]. In this series, the overall rate of pathologic final ureteral section was very low and comparable in both cohorts demonstrating the validity in most cases of relying on the macroscopic visual evaluation of the ureters. However, we believe that systematically avoiding intraoperative examination of the ureters because of previously stated limitations is excessive. Reported rates of upper tract recurrences after RC are low probably in part due to lack of/inadequate follow-up, small cohorts, but also competing mortality with the initial bladder tumor itself. In 2008, Tran et al. [7] described recurrence rates after 1,329 RC. The incidence for upper tract recurrence was 4% at 3 years and 7% at 5 years. Three-year risk of upper tract recurrence remained approximately 4% to 6% at any point measured up to 4 years after RC and did not change over time. In a meta-analysis published by Picozzi et al. in 2012 on 13,185 patients, recurrence rates ranged between 0.75 and 6.4% [14]. Even though published studies are contradictory, many suggested that positive ureteral margins were associated with upper tract recurrence [14–17].
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
4
Table 2 Logistic regression for factors associated with ureteral section positivity in the cohort A Univariate analysis
Age ASA score Number of TURB Time TURB—RC Hydronephrosis No Yes Pathological T stage pTVNIM pT2 pT3 pT4 Pathological N stage N0 Nþ Carcinoma in situ No Yes
Multivariate analysis
OR
CI
P value
OR
CI
P value
0.97 0.86 1.06 0.77
[0.93–1.02] [0.33–2.20] [0.83–1.26] [0.48–1.11]
0.26 0.76 0.56 0.25
Ref 5.42
– [2.03–17.08]
– 0.001
Ref 4.93
– [1.72–16.43]
– 0.004
Ref 0.32 1.16 2.34
– [0.01–2.59] [0.32–5.50] [0.79–10.97]
– 0.33 0.82 0.21
Ref 2.15
– [0.66–5.67]
– 0.12
Ref 1.78
– [0.60–4.99]
– 0.27
Ref 3.61
– [1.35–9.14]
– 0.007
Ref 3.99
– [1.25–12.00]
– 0.01
ASA ¼ American Society of Anesthesiologists.
Sequential resection until gaining tumor-free margins seems to allow long-term upper tract recurrence-free survival. In the previously cited meta-analysis on 1,397 patients with 10 to 20 years follow-up, serial resection strategy allowed 12% of positive FS to be converted into only 2.2% of final positive resection margin. Positive initial margin status and final margin status were significantly associated with upper tract recurrence (P o 0.001). Although distal ureteral margin status was not associated with overall- or cancer-specific survival, authors concluded
that conversion to negative margins is possible and associated with decreased recurrence risk during follow-up [17]. Consequently, criteria are needed to identify patients who will most probably benefit from special attention in intraoperative ureteral management. In this study, we used 2 different RC cohorts to identify and validate preoperative factors associated with ureteral section positivity. In the exploratory cohort, hydronephrosis and bladder CIS were very strongly associated with ureteral margin invasion. Only hydronephrosis was validated in the
Table 3 Logistic regression for factors associated with ureteral section positivity in the cohort B Univariate analysis
Age ASA score Number of TURB Time TURB—RC Hydronephrosis No Yes Pathological T stage pTVNIM pT2 pT3 pT4 Pathological N stage N0 Nþ Carcinoma in situ No Yes
Multivariate analysis
OR
CI
P value
OR
CI
P value
1.06 1.77 1.06 0.77
[1.008–1.13] [0.66–4.83] [0.72–1.36] [0.48–1.11]
0.03 0.25 0.68 0.25
Ref 3.34
– [1.23–9.21]
– 0.01
Ref 5.91
– [1.48–27.01]
– 0.01
Ref – 3.25 8.26
– – [0.76–2.22] [1.97–5.62]
– – 0.14 0.009
1.64
[0.19–15.41]
0.64
Ref 4.85
– [1.62–16.25]
– 0.005
3.21
– [0.71–17.85]
– 0.14
Ref 3.32
– [1.02–14.80]
– 0.06
Ref 3.84
– [0.86–27.27]
– 0.10
ASA ¼ American Society of Anesthesiologists.
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Fig. 1. ROC curve for prediction of ureteral section positivity using hydronephrosis and CIS in the cohort A. Area under curve was 0.72. (Color version of figure is available online.)
cohort B possibly owing to heterogeneity in the management of distal ureters in the 2 cohorts. Although there is a lack of published studies looking at factors associated with ureteral invasion at RC, those factors overlap with available literature. In 2006, Schumacher et al. [16] published the results of systematic resection of 805 ureters at the level of the common iliac arteries during RC. Ureteral involvement (UC) or CIS was measured either in the distal ureter or at the proximal section. UC or CIS were observed in 4.8% of FS and 3.6% of permanent sections in the distal ureter, while only in 1.2% of proximal sections. Among patients with CIS on FS, 80% also had CIS in the bladder. After RC, there was tumor recurrence in the upper urinary tract in 3% of patients with negative ureteral FS examination and in 17% with CIS on FS examination. In 2015, Kim et al. looked at risk factors for ureteral involvement on permanent section in 402 RC. Sequential ureteral resections were performed if FS showed UC. Out of 11% of initial positive FS, 8% had final positive ureteral section. Risk factors for ureteral involvement on permanent section in multivariable analysis were the presence of CIS in the bladder (OR ¼ 11, P ¼ 0.003) and intramural invasion of the ureter (OR ¼ 7, P ¼ 0.003) [10].
Fig. 2. ROC curve for prediction of ureteral section positivity using hydronephrosis alone in the cohort B. Area under curve was 0.64. (Color version of figure is available online.)
5
Many other studies identified CIS in the bladder and presence of UC in the distal ureter as associated with upper tract recurrence, supporting the hypothesis that CIS in lower urethral segments is spread intramucously from intravesical tumors (especially bladder CIS): in 2006, Lee et al. [18] demonstrated that all 115 RC patients presenting with upper tract recurrence in their series (2.6%) had pelvic ureteral involvement. Same was demonstrated by Kenworthy et al. [19] in 1996: the only risk factor for upper tract recurrence in their series of 430 patients with 10-year follow-up was the presence of UC in the distal ureter. In 2009, Volkmer et al. [20] identified 4 risk factors for upper urinary tract recurrence in a series of 1,420 patients including history of CIS (relative risk [RR] ¼ 2.3), history of recurrent bladder cancer (RR ¼ 2.6), cystectomy for NMIBC (RR ¼ 3.8), and tumor involvement of the distal ureter in the cystectomy specimen (RR ¼ 2.7). Finally, Moschini et al. [21] recorded that history of CIS before RC is at higher risk to incur positive ureteral margin and should be investigated during RC because achieving negative FSs may be associated with survival benefit. Pointing on statistical association does not provide information about the accuracy of a test/marker and its added value in clinical practice. In this study, we calculated the sensitivity, specificity, and AUC of the combination of hydronephrosis and CIS in the first cohort and of hydronephrosis alone in the second cohort. Accuracy of 74% in the exploratory cohort was good whereas accuracy of hydronephrosis alone in the validation cohort was only fair (64%). Main limitation of this study comes from the strong heterogeneity between patients’ characteristics, particularly rate of CIS (12% in the cohort A while 450% in the cohort B) and intraoperative FS usage. Although most authors, for volume improvement purposes, often decide to pool different series despite heterogeneity between the practices in various centers, we rather decided to use the 2 cohorts independently to perform external validation and improve methodological quality. Identification of the same predictive factors in 2 different cohorts reflects real-life practice and enhances, to our point of view, the robustness of the markers. Other limitation comes from the absence of central review. The difficulty and dedication necessary in determining the presence of concomitant CIS at the morphologic level may have resulted in interobserver variability. Although this limitation can be seen as significant, the purpose of this study was to reflect a realworld practice in which multiple pathologists review tissue specimens leading to clinical decision-making with the patient. Despite the described limitations, hydronephrosis (as a probable surrogate for distal ureteral involvement) and at a lower significance presence of CIS in the bladder were strongly associated with ureteral section positivity on final pathology. Those factors are easily accessible preoperatively. If a subpopulation candidate for FS analysis were to be defined, practitioners should be advised to perform it
6
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
preferentially for patients with preoperative hydronephrosis on CT scans and CIS identified on the TURB. Even though part of this population will demonstrate locally advanced tumor responsible for early mortality, some will likely benefit from sequential resections to try achieving tumorfree ureteral sections.
[8]
[9]
[10]
5. Conclusion Using 2 independent RC cohorts, we identified preoperative hydronephrosis and presence of CIS as predictive of final ureteral sections positivity after RC. In the presence of those factors, we believe ureteral FSs should be assessed to perform sequential resections and obtain final tumor-free ureters. This will likely reduce the risk of upper tract recurrence in patients that will experience long-term survival after RC.
[11]
[12]
[13]
[14]
References [15] [1] Shariat SF, Karakiewicz PI, Palapattu GS, et al. Outcomes of radical cystectomy for transitional cell carcinoma of the bladder: a contemporary series from the Bladder Cancer Research Consortium. J Urol 2006;176:2414–22:[discussion 2422]. [2] Alfred Witjes J, Lebret T, Comperat EM, et al. Updated 2016 EAU guidelines on muscle-invasive and metastatic bladder cancer. Eur Urol 2017 Mar;71:462–75. http://dx.doi.org/10.1016/j.eururo.2016. 06.020. Epub 2016 Jun 30. [3] Schoenberg MP, Carter HB, Epstein JI. Ureteral frozen section analysis during cystectomy: a reassessment. J Urol 1996;155:1218–20. [4] Reder NP, Maxwell SP, Pambuccian SE, Barkan GA. Diagnostic accuracy of intraoperative frozen sections during radical cystectomy does not affect disease-free or overall survival: a study of 364 patients with urothelial carcinoma of the urinary bladder. Ann Diagn Pathol 2015;19:107–12. [5] Donat SM. Argument against frozen section analysis of distal ureters in transitional cell bladder cancer. Nat Clin Pract Urol 2008;5:538–9. [6] Touma N, Izawa JI, Abdelhady M, et al. Ureteral frozen sections at the time of radical cystectomy: reliability and clinical implications. Can Urol Assoc J 2010;4:28–32. [7] Tran W, Serio AM, Raj GV, et al. Longitudinal risk of upper tract recurrence following radical cystectomy for urothelial cancer and the
[16]
[17]
[18]
[19]
[20]
[21]
potential implications for long-term surveillance. J Urol 2008;179:96– 100. Paterakos M, Watkin WG, Edgerton SM, et al. Invasive micropapillary carcinoma of the breast: a prognostic study. Hum Pathol 1999;30:1459–63. Gakis G, Schilling D, Perner S, et al. Sequential resection of malignant ureteral margins at radical cystectomy: a critical assessment of the value of frozen section analysis. World J Urol 2011;29:451–6. Kim HS, Moon KC, Jeong CW, et al. The clinical significance of intraoperative ureteral frozen section analysis at radical cystectomy for urothelial carcinoma of the bladder. World J Urol 2015;33:359–65. Osman Y, El-Tabey N, Abdel-Latif M, et al. The value of frozensection analysis of ureteric margins on surgical decision-making in patients undergoing radical cystectomy for bladder cancer. BJU Int 2007;99:81–4. Raj GV, Tal R, Vickers A, et al. Significance of intraoperative ureteral evaluation at radical cystectomy for urothelial cancer. Cancer 2006;107:2167–72. Silver DA, Stroumbakis N, Russo P, et al. Ureteral carcinoma in situ at radical cystectomy: does the margin matter? J Urol 1997;158: 768–71. Picozzi S, Ricci C, Gaeta M, et al. Upper urinary tract recurrence following radical cystectomy for bladder cancer: a meta-analysis on 13,185 patients. J Urol 2012;188:2046–54. Satkunasivam R, Hu B, Metcalfe C, et al. Utility and significance of ureteric frozen section analysis during radical cystectomy. BJU Int 2016;117:463–8. Schumacher MC, Scholz M, Weise ES, et al. Is there an indication for frozen section examination of the ureteral margins during cystectomy for transitional cell carcinoma of the bladder? J Urol 2006;176:2409– 13:[discussion 2413]. Tollefson MK, Blute ML, Farmer SA, Frank I. Significance of distal ureteral margin at radical cystectomy for urothelial carcinoma. J Urol 2010;183:81–6. Lee SE, Byun SS, Hong SK, et al. Significance of cancer involvement at the ureteral margin detected on routine frozen section analysis during radical cystectomy. Urol Int 2006;77:13–7. Kenworthy P, Tanguay S, Dinney CP. The risk of upper tract recurrence following cystectomy in patients with transitional cell carcinoma involving the distal ureter. J Urol 1996;155:501–3. Volkmer BG, Schnoeller T, Kuefer R, et al. Upper urinary tract recurrence after radical cystectomy for bladder cancer—who is at risk? J Urol 2009;182:2632–7. Moschini M, Gallina A, Freschi M, et al. Effect on postoperative survival of the status of distal ureteral margin: the necessity to achieve negative margins at the time of radical cystectomy. Urol Oncol 2016;34:59.e15-22.
Original article
Predictive factors for final pathologic ureteral sections on 700 radical cystectomy specimens: Implications for intraoperative frozen section decision-making Alexandra Masson-Lecomte, M.D.a,*, Thomas Francois, M.D.b, Dimitri Vordos, M.D.c, Carole Cordonnier, M.D.d, Yves Allory, M.D.e, Francois Desgrandchamps, Ph.D.a, Alexandre de la Taille, Ph.D.c, Fabien Saint, Ph.D.b,f a
Department of Urology and Transplantation, Saint Louis Academic Hospital, Paris Diderot University, Paris, France b Department of Urology and Transplantation, Picardie Jules Verne University, CHU Amiens, Amiens, France c Department of Urology, Henri Mondor Hospital, Paris Est Creteil University, Créteil, France d Department of Pathology, Picardie Jules Verne University, CHU Amiens, Amiens, France e Department of Pathology, Henri Mondor Hospital, Paris Est Creteil University, Créteil, France f HeRVI EA 3801, Picardie Jules Verne University, CHU Amiens, Amiens, France Received 18 April 2017; received in revised form 16 June 2017; accepted 19 June 2017
Abstract Purpose: To identify preoperative predictive factors for final ureteral section invasion after radical cystectomy (RC) and to validate significant factors on an external independent cohort. Material and methods: We retrospectively reviewed data of all consecutive RC performed for bladder cancer in 2 high-volume institutions. Clinical, pathological, and follow-up data were collected prospectively and reviewed retrospectively. Pathological evaluation was performed by 2 well-trained uropathologists in each center. Logistic regression analyses were performed to identify predictive factors for final ureteral sections involvement. Significant factors in cohort A were validated in cohort B. Receiver operating curve and area under curve were modeled to evaluate predictive accuracy of the markers. Results: A total of 441 RC were performed in center A and 307 RC were performed in center B. Mean follow-ups were 36.2 and 38.1 months, respectively. Invasion of the final ureteral section was observed on 5.5% of patients in cohort A and 4.8% of patients in cohort B. In cohort A, multivariable logistic regression identified preoperative hydronephrosis on computed tomography scan (odds ratio [OR] ¼ 4.9, P ¼ 0.004) and presence of Carcinoma in situ (CIS, OR ¼ 3.9, P ¼ 0.01) as the only factors associated with ureteral sections positivity. In cohort B, hydronephrosis and CIS were both associated with ureteral sections positivity in univariable analysis. In multivariable analysis, only hydronephrosis remained significant (OR ¼ 5.9, P ¼ 0.01). Predictive accuracy of hydronephrosis and CIS combined in 1 variable was 0.72. Conclusion: Hydronephrosis and bladder CIS have good accuracy in predicting ureteral sections positivity after RC. In the presence of those factors, ureteral frozen sections should be performed. r 2017 Elsevier Inc. All rights reserved.
Keywords: Urothelial carcinoma; Radical cystectomy; Ureteral frozen section; Predictive factors
1. Introduction Radical cystectomy (RC) is the standard treatment option for nonmetastatic, resectable muscle-invasive bladder Corresponding author. Tel.: þ33-60-382-8593. E-mail address: [email protected] (A. Masson-Lecomte). *
http://dx.doi.org/10.1016/j.urolonc.2017.06.053 1078-1439/r 2017 Elsevier Inc. All rights reserved.
cancer. Preceded by neoadjuvant chemotherapy, it allows local and distant control of the disease with 5-year survival rates up to 80% in the most favorable cases [1]. However, to maximize oncologic outcome, the surgical procedure has to be standardized [2]. Although it is pretty much consensual that pathologic urethral frozen section (FS) should lead to uretrectomy and
2
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
preclude neobladder reconstruction, intraoperative management of ureteral sections remains a matter of debate. In that regard, the latest European Association of Urology guidelines do not provide any recommendation. Analysis of available literature questions both the ability of intraoperative FSs to properly predict final involvement of ureters and the effect of invaded intraoperative ureteral sections on oncological outcome and upper tract recurrence rates [3–5]. Taking into account the high cost of the procedure [6], FSs are not anymore performed in routine in many centers. However, the risk of upper tract recurrence is probably underestimated because of small cohorts, short follow-ups, and competing risk with dying from bladder cancer [7]. Some patients might benefit from sequential ureteral sections and achievement of healthy final ureteral margins. To better define the patients for whom particular attention should be paid to the distal ureters management, predictive factors for ureteral sections invasion on final pathology report are needed. Using 2 large cohorts of RC, we aimed at (1) identifying preoperative predictive factors for final ureteral section invasion and (2) validating these factors on an external independent cohort.
2. Materials and methods 2.1. Population After obtaining institutional review board approval, we retrospectively reviewed data of all consecutive RC performed for bladder cancer in 2 independent institutions. The procedures were performed between 1980 and 2013 for the exploratory cohort (A) and between 2001 and 2013 for the validation cohort (B). Surgery was performed through either open or laparoscopic approach with an extended pelvic lymph node dissection up to the common iliac bifurcation. Ureters were sectioned at the discretion of each surgeon, generally a few centimeters distal to the iliac bifurcation. If the section was not macroscopically healthy, a new section was performed 1 centimeter above the previous section. In cohort A, most patients did not have intraoperative FS examination although it was performed in some cases depending on the surgeon’s habits (N ¼ 92). All patients had intraoperative FS in the validation cohort B. All cystectomies performed for nononcological purposes were excluded. All patients had nonmetastatic bladder cancer on preoperative computed tomography (CT) scans. Following clinical and pathological data were collected for all patients: sex, age at RC, American Society of Anesthesiologists score, tobacco use, exposure to workrelated hazards, number of transurethral resection on the bladder (TURB), time between TURB and RC, pathology on final TURB, preoperative hydronephrosis on abdominal CT scans (defined by an anteroposterior diameter of the renal pelvis 410 mm), presence of carcinoma in situ (CIS) on TURB and RC specimen, multiplicity, tumor location,
neoadjuvant chemotherapy administration, date of cystectomy, type of urinary diversion, final pathological T stage, N stage, histological subtype, grade, margins status, and final ureteral sections pathology examination. 2.2. Pathological examination Pathological evaluation was performed by 2 well-trained uropathologists in each center (Y.A. and C.C). No central review was performed. In cohort A, both intraoperative and final ureteral sections analyses were available, while only final analysis was available in cohort B. Ureteral sections were considered positive if urothelial carcinoma (UC) (in situ or invasive) was in contact with the edge of the ureteral resection. 2.3. Statistical analysis Descriptive analyses of the 2 cohorts were performed. Continuous variables were compared using a Wilcoxon test and categorical variables using a chi-square test. Unadjusted and adjusted logistic regression analyses were performed to identify predictive factors for final ureteral sections involvement. Variables with a P o 0.2 in univariable analyses were kept for multivariable analysis. Final P o 0.05 were considered significant. Receiver operating curve (ROC) and area under curve (AUC) were modeled/calculated using the pROC package (https://cran.r-project.org/web/packages/ pROC/pROC.pdf). Statistical analyses were performed using “R” version 3.3.2 (2016-10-31).
3. Results 3.1. Characteristics of the cohorts Overall, 441 RC were performed in center A (exploratory cohort) and 307 RC were performed in center B (validation cohort). The mean and median follow-ups were, respectively, 36.2 (1–255) and 14.6 months in cohort A and 38.1 months (3–181) and 27.9 months in cohort B. Neoadjuvant chemotherapy was administered to 14 patients (3.2%) in cohort A and 16 patients (5%) in cohort B. Ninety-two patients had intraoperative examination of the ureters in cohort A (20%) while all patients had FS in cohort B. In cohort A, the most common type of urinary diversion was cutaneous ureterostomy (200; 46%), followed by ileal conduit (115; 26%) and orthotopic neobladder (107; 24%). In cohort B, 173 patients had ileal conduit (46%), 120 had an orthotopic neobladder (39%), and 4 cutaneous ureterostomy (1%). Patients’ characteristics are presented and compared in Table 1. Invasion of the final ureteral section was observed on 5.5% of patients in cohort A and 4.8% of patients in cohort B.
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
3
3.3. Prediction of ureteral section positivity using hydronephrosis and CIS
Table 1 Patients characteristics Cohort A n 441 Sex Male (%) 391 ASA score (%) 1 25 2 200 3 123 4 3 NA 90 Age (mean [sd]) 65.35 Number of TURBT (%) 0 2 1 265 2 73 3 31 Z4 62 NA 8 Hydronephrosis (%) 0 281 1 148 NA 12 pT stage (%) T1 85 T2 95 T3 164 T4 92 NA 5 pN stage (%) Node negative 296 Node positive 115 NA 30 CIS (%) No 387 Yes 53 NA 1 Positive margins, n (%) NA Pathological ureteral sections (%) No 282 Yes 21 NA 138
Cohort B
P
307 (88.7)
270
(87.9)
0.854
(5.7) (45.4) (27.9) (0.7) (20.4) (9.79)
25 153 57 1 71 67.19
(8.1) (49.8) (18.6) (0.3) (23.1) (9.95)
0.04
(0.5) (60.1) (16.6) (7) (14.1) (1.8)
2 154 66 23 32 30
(0.7) (50.2) (21.5) (7.5) (10.4) (9.8)
o0.001
(63.7) (33.6) (2.7)
225 82 0
(73.3) (26.7) 0
0.001
(19.3) (21.5) (37.2) (20.9) (1.1)
95 52 107 53 0
(30.9) (16.9) (34.9) (17.3) 0
0.002
(67.1) (26.1) (6.8)
202 82 23
(65.8) (26.7) (7.5)
0.906
(87.8) (12) (0.2) NA
130 165 12 45
(42.3) (53.7) (3.9) (14)
o0.001
(63.9) (4.8) (31.3)
290 17 0
(94.5) (5.5) 0
o0.001
0.013
In cohort A, sensitivity and specificity of the presence of both hydronephrosis and CIS combined in 1 variable for prediction of ureteral section positivity were 85.7% and 54.3%, respectively. Accuracy of hydronephrosis and CIS combined in the cohort A was 0.72 (Fig. 1). In cohort B, CIS alone was not independently associated with ureteral section positivity. Sensitivity and specificity of hydronephrosis alone for prediction of ureteral positivity were 52.9% and 74.8%, respectively. Accuracy of hydronephrosis alone in the cohort B was 0.64 (Fig. 2). Adding CIS to hydronephrosis increased specificity to 90% while lowering sensitivity to 43% and did not improve AUC (0.64). 4. Discussion
ASA ¼ American Society of Anesthesiologists; NA ¼ not available.
3.2. Factors associated with final ureteral section positivity Logistic regression results are presented in Tables 2 and 3. In cohort A, multivariable logistic regression identified preoperative hydronephrosis on CT scan (odds ratio [OR] ¼ 4.9, P ¼ 0.004) and presence of CIS (OR ¼ 3.9, P ¼ 0.01) as the only factors associated with ureteral sections positivity. Cohort B was explored to validate those results. In cohort B, hydronephrosis and CIS were both associated with ureteral sections positivity in univariable analysis. In multivariable analysis adjusted for T stage and N stage, only hydronephrosis remained significant (OR ¼ 5.9, P ¼ 0.01). OR for CIS followed the same tendency as observed in cohort A but was not significant (OR ¼ 3.4, P ¼ 0.10)
At present, the European guidelines do not mention any recommendation regarding intraoperative management of ureters during RC performed for nonmetastatic UC [8]. Consequently, clinical practice is very heterogeneous, and FSs with sequential ureteral resections are performed at the discretion of the surgeon, based on pre/intraoperative clinical feeling. Controversies regarding the interest of ureteral FSs came from the following: (1) Uneven sensitivity and specificity of FS compared to final pathology [9–12], (2) frequent failure to obtain tumor-free sections after sequential resections, particularly in case of CIS [9,13], (3) unclear effect of tumor presence on ureteral section on upper tract recurrence and patients’ prognosis [3,4,10], (4) and direct and indirect costs [6]. In this series, the overall rate of pathologic final ureteral section was very low and comparable in both cohorts demonstrating the validity in most cases of relying on the macroscopic visual evaluation of the ureters. However, we believe that systematically avoiding intraoperative examination of the ureters because of previously stated limitations is excessive. Reported rates of upper tract recurrences after RC are low probably in part due to lack of/inadequate follow-up, small cohorts, but also competing mortality with the initial bladder tumor itself. In 2008, Tran et al. [7] described recurrence rates after 1,329 RC. The incidence for upper tract recurrence was 4% at 3 years and 7% at 5 years. Three-year risk of upper tract recurrence remained approximately 4% to 6% at any point measured up to 4 years after RC and did not change over time. In a meta-analysis published by Picozzi et al. in 2012 on 13,185 patients, recurrence rates ranged between 0.75 and 6.4% [14]. Even though published studies are contradictory, many suggested that positive ureteral margins were associated with upper tract recurrence [14–17].
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
4
Table 2 Logistic regression for factors associated with ureteral section positivity in the cohort A Univariate analysis
Age ASA score Number of TURB Time TURB—RC Hydronephrosis No Yes Pathological T stage pTVNIM pT2 pT3 pT4 Pathological N stage N0 Nþ Carcinoma in situ No Yes
Multivariate analysis
OR
CI
P value
OR
CI
P value
0.97 0.86 1.06 0.77
[0.93–1.02] [0.33–2.20] [0.83–1.26] [0.48–1.11]
0.26 0.76 0.56 0.25
Ref 5.42
– [2.03–17.08]
– 0.001
Ref 4.93
– [1.72–16.43]
– 0.004
Ref 0.32 1.16 2.34
– [0.01–2.59] [0.32–5.50] [0.79–10.97]
– 0.33 0.82 0.21
Ref 2.15
– [0.66–5.67]
– 0.12
Ref 1.78
– [0.60–4.99]
– 0.27
Ref 3.61
– [1.35–9.14]
– 0.007
Ref 3.99
– [1.25–12.00]
– 0.01
ASA ¼ American Society of Anesthesiologists.
Sequential resection until gaining tumor-free margins seems to allow long-term upper tract recurrence-free survival. In the previously cited meta-analysis on 1,397 patients with 10 to 20 years follow-up, serial resection strategy allowed 12% of positive FS to be converted into only 2.2% of final positive resection margin. Positive initial margin status and final margin status were significantly associated with upper tract recurrence (P o 0.001). Although distal ureteral margin status was not associated with overall- or cancer-specific survival, authors concluded
that conversion to negative margins is possible and associated with decreased recurrence risk during follow-up [17]. Consequently, criteria are needed to identify patients who will most probably benefit from special attention in intraoperative ureteral management. In this study, we used 2 different RC cohorts to identify and validate preoperative factors associated with ureteral section positivity. In the exploratory cohort, hydronephrosis and bladder CIS were very strongly associated with ureteral margin invasion. Only hydronephrosis was validated in the
Table 3 Logistic regression for factors associated with ureteral section positivity in the cohort B Univariate analysis
Age ASA score Number of TURB Time TURB—RC Hydronephrosis No Yes Pathological T stage pTVNIM pT2 pT3 pT4 Pathological N stage N0 Nþ Carcinoma in situ No Yes
Multivariate analysis
OR
CI
P value
OR
CI
P value
1.06 1.77 1.06 0.77
[1.008–1.13] [0.66–4.83] [0.72–1.36] [0.48–1.11]
0.03 0.25 0.68 0.25
Ref 3.34
– [1.23–9.21]
– 0.01
Ref 5.91
– [1.48–27.01]
– 0.01
Ref – 3.25 8.26
– – [0.76–2.22] [1.97–5.62]
– – 0.14 0.009
1.64
[0.19–15.41]
0.64
Ref 4.85
– [1.62–16.25]
– 0.005
3.21
– [0.71–17.85]
– 0.14
Ref 3.32
– [1.02–14.80]
– 0.06
Ref 3.84
– [0.86–27.27]
– 0.10
ASA ¼ American Society of Anesthesiologists.
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
Fig. 1. ROC curve for prediction of ureteral section positivity using hydronephrosis and CIS in the cohort A. Area under curve was 0.72. (Color version of figure is available online.)
cohort B possibly owing to heterogeneity in the management of distal ureters in the 2 cohorts. Although there is a lack of published studies looking at factors associated with ureteral invasion at RC, those factors overlap with available literature. In 2006, Schumacher et al. [16] published the results of systematic resection of 805 ureters at the level of the common iliac arteries during RC. Ureteral involvement (UC) or CIS was measured either in the distal ureter or at the proximal section. UC or CIS were observed in 4.8% of FS and 3.6% of permanent sections in the distal ureter, while only in 1.2% of proximal sections. Among patients with CIS on FS, 80% also had CIS in the bladder. After RC, there was tumor recurrence in the upper urinary tract in 3% of patients with negative ureteral FS examination and in 17% with CIS on FS examination. In 2015, Kim et al. looked at risk factors for ureteral involvement on permanent section in 402 RC. Sequential ureteral resections were performed if FS showed UC. Out of 11% of initial positive FS, 8% had final positive ureteral section. Risk factors for ureteral involvement on permanent section in multivariable analysis were the presence of CIS in the bladder (OR ¼ 11, P ¼ 0.003) and intramural invasion of the ureter (OR ¼ 7, P ¼ 0.003) [10].
Fig. 2. ROC curve for prediction of ureteral section positivity using hydronephrosis alone in the cohort B. Area under curve was 0.64. (Color version of figure is available online.)
5
Many other studies identified CIS in the bladder and presence of UC in the distal ureter as associated with upper tract recurrence, supporting the hypothesis that CIS in lower urethral segments is spread intramucously from intravesical tumors (especially bladder CIS): in 2006, Lee et al. [18] demonstrated that all 115 RC patients presenting with upper tract recurrence in their series (2.6%) had pelvic ureteral involvement. Same was demonstrated by Kenworthy et al. [19] in 1996: the only risk factor for upper tract recurrence in their series of 430 patients with 10-year follow-up was the presence of UC in the distal ureter. In 2009, Volkmer et al. [20] identified 4 risk factors for upper urinary tract recurrence in a series of 1,420 patients including history of CIS (relative risk [RR] ¼ 2.3), history of recurrent bladder cancer (RR ¼ 2.6), cystectomy for NMIBC (RR ¼ 3.8), and tumor involvement of the distal ureter in the cystectomy specimen (RR ¼ 2.7). Finally, Moschini et al. [21] recorded that history of CIS before RC is at higher risk to incur positive ureteral margin and should be investigated during RC because achieving negative FSs may be associated with survival benefit. Pointing on statistical association does not provide information about the accuracy of a test/marker and its added value in clinical practice. In this study, we calculated the sensitivity, specificity, and AUC of the combination of hydronephrosis and CIS in the first cohort and of hydronephrosis alone in the second cohort. Accuracy of 74% in the exploratory cohort was good whereas accuracy of hydronephrosis alone in the validation cohort was only fair (64%). Main limitation of this study comes from the strong heterogeneity between patients’ characteristics, particularly rate of CIS (12% in the cohort A while 450% in the cohort B) and intraoperative FS usage. Although most authors, for volume improvement purposes, often decide to pool different series despite heterogeneity between the practices in various centers, we rather decided to use the 2 cohorts independently to perform external validation and improve methodological quality. Identification of the same predictive factors in 2 different cohorts reflects real-life practice and enhances, to our point of view, the robustness of the markers. Other limitation comes from the absence of central review. The difficulty and dedication necessary in determining the presence of concomitant CIS at the morphologic level may have resulted in interobserver variability. Although this limitation can be seen as significant, the purpose of this study was to reflect a realworld practice in which multiple pathologists review tissue specimens leading to clinical decision-making with the patient. Despite the described limitations, hydronephrosis (as a probable surrogate for distal ureteral involvement) and at a lower significance presence of CIS in the bladder were strongly associated with ureteral section positivity on final pathology. Those factors are easily accessible preoperatively. If a subpopulation candidate for FS analysis were to be defined, practitioners should be advised to perform it
6
A. Masson-Lecomte et al. / Urologic Oncology: Seminars and Original Investigations ] (2017) ∎∎∎–∎∎∎
preferentially for patients with preoperative hydronephrosis on CT scans and CIS identified on the TURB. Even though part of this population will demonstrate locally advanced tumor responsible for early mortality, some will likely benefit from sequential resections to try achieving tumorfree ureteral sections.
[8]
[9]
[10]
5. Conclusion Using 2 independent RC cohorts, we identified preoperative hydronephrosis and presence of CIS as predictive of final ureteral sections positivity after RC. In the presence of those factors, we believe ureteral FSs should be assessed to perform sequential resections and obtain final tumor-free ureters. This will likely reduce the risk of upper tract recurrence in patients that will experience long-term survival after RC.
[11]
[12]
[13]
[14]
References [15] [1] Shariat SF, Karakiewicz PI, Palapattu GS, et al. Outcomes of radical cystectomy for transitional cell carcinoma of the bladder: a contemporary series from the Bladder Cancer Research Consortium. J Urol 2006;176:2414–22:[discussion 2422]. [2] Alfred Witjes J, Lebret T, Comperat EM, et al. Updated 2016 EAU guidelines on muscle-invasive and metastatic bladder cancer. Eur Urol 2017 Mar;71:462–75. http://dx.doi.org/10.1016/j.eururo.2016. 06.020. Epub 2016 Jun 30. [3] Schoenberg MP, Carter HB, Epstein JI. Ureteral frozen section analysis during cystectomy: a reassessment. J Urol 1996;155:1218–20. [4] Reder NP, Maxwell SP, Pambuccian SE, Barkan GA. Diagnostic accuracy of intraoperative frozen sections during radical cystectomy does not affect disease-free or overall survival: a study of 364 patients with urothelial carcinoma of the urinary bladder. Ann Diagn Pathol 2015;19:107–12. [5] Donat SM. Argument against frozen section analysis of distal ureters in transitional cell bladder cancer. Nat Clin Pract Urol 2008;5:538–9. [6] Touma N, Izawa JI, Abdelhady M, et al. Ureteral frozen sections at the time of radical cystectomy: reliability and clinical implications. Can Urol Assoc J 2010;4:28–32. [7] Tran W, Serio AM, Raj GV, et al. Longitudinal risk of upper tract recurrence following radical cystectomy for urothelial cancer and the
[16]
[17]
[18]
[19]
[20]
[21]
potential implications for long-term surveillance. J Urol 2008;179:96– 100. Paterakos M, Watkin WG, Edgerton SM, et al. Invasive micropapillary carcinoma of the breast: a prognostic study. Hum Pathol 1999;30:1459–63. Gakis G, Schilling D, Perner S, et al. Sequential resection of malignant ureteral margins at radical cystectomy: a critical assessment of the value of frozen section analysis. World J Urol 2011;29:451–6. Kim HS, Moon KC, Jeong CW, et al. The clinical significance of intraoperative ureteral frozen section analysis at radical cystectomy for urothelial carcinoma of the bladder. World J Urol 2015;33:359–65. Osman Y, El-Tabey N, Abdel-Latif M, et al. The value of frozensection analysis of ureteric margins on surgical decision-making in patients undergoing radical cystectomy for bladder cancer. BJU Int 2007;99:81–4. Raj GV, Tal R, Vickers A, et al. Significance of intraoperative ureteral evaluation at radical cystectomy for urothelial cancer. Cancer 2006;107:2167–72. Silver DA, Stroumbakis N, Russo P, et al. Ureteral carcinoma in situ at radical cystectomy: does the margin matter? J Urol 1997;158: 768–71. Picozzi S, Ricci C, Gaeta M, et al. Upper urinary tract recurrence following radical cystectomy for bladder cancer: a meta-analysis on 13,185 patients. J Urol 2012;188:2046–54. Satkunasivam R, Hu B, Metcalfe C, et al. Utility and significance of ureteric frozen section analysis during radical cystectomy. BJU Int 2016;117:463–8. Schumacher MC, Scholz M, Weise ES, et al. Is there an indication for frozen section examination of the ureteral margins during cystectomy for transitional cell carcinoma of the bladder? J Urol 2006;176:2409– 13:[discussion 2413]. Tollefson MK, Blute ML, Farmer SA, Frank I. Significance of distal ureteral margin at radical cystectomy for urothelial carcinoma. J Urol 2010;183:81–6. Lee SE, Byun SS, Hong SK, et al. Significance of cancer involvement at the ureteral margin detected on routine frozen section analysis during radical cystectomy. Urol Int 2006;77:13–7. Kenworthy P, Tanguay S, Dinney CP. The risk of upper tract recurrence following cystectomy in patients with transitional cell carcinoma involving the distal ureter. J Urol 1996;155:501–3. Volkmer BG, Schnoeller T, Kuefer R, et al. Upper urinary tract recurrence after radical cystectomy for bladder cancer—who is at risk? J Urol 2009;182:2632–7. Moschini M, Gallina A, Freschi M, et al. Effect on postoperative survival of the status of distal ureteral margin: the necessity to achieve negative margins at the time of radical cystectomy. Urol Oncol 2016;34:59.e15-22.