Urologic Oncology: Seminars and Original Investigations ] (2014) ∎∎∎–∎∎∎

Original article

Risk factors for bladder cancer recurrence after nephroureterectomy for upper tract urothelial tumors: Results from the Canadian Upper Tract Collaboration Vincent Fradet, M.D.1,a, Julian Mauermann, M.D.1,a,b, Wassim Kassouf, M.D.c, Ricardo Rendon, M.D.d, Niels Jacobsen, M.D.e, Adrian Fairey, M.D.e, Jonathan Izawa, M.D.f, Anil Kapoor, M.D.g, Peter Black, M.D.h, Simon Tanguay, M.D.c, Joseph Chin, M.D.f, Alan So, M.D.h, Jean-Baptiste Lattouf, M.D.i, David Bell, M.D.d, Fred Saad, M.D.i, Bobby Sheyegan, M.D.g, Darren Drachenberg, M.D.j, Ilias Cagiannos, M.D.k, Louis Lacombe, M.D., F.R.C.S.C.a,* a

Department of Surgery (Urology), Laval University, Quebec City, Quebec, Canada b Department of Urology, Hietzing Hospital, Vienna, Austria c Department of Surgery (Urology), McGill University, Montreal, Quebec, Canada d Department of Urology, Dalhousie University, Halifax, Nova Scotia, Canada e Department of Surgery (Urology), University of Alberta, Edmonton, Alberta, Canada f Division of Urologic Surgery, University of Western Ontario, London, Ontario, Canada g Division of Urology, McMaster University, Hamilton, Ontario, Canada h Department of Urologic Sciences, University of British Columbia, Vancouver, British Columbia, Canada i Department of Surgery (Urology), University of Montreal, Montreal, Quebec, Canada j Section of Urology, University of Manitoba, Winnipeg, Manitoba, Canada k Division of Urology, University of Ottawa, Ottawa, Ontario, Canada Received 27 February 2014; received in revised form 7 April 2014; accepted 9 April 2014

Abstract Objective: To evaluate risk factors for bladder cancer recurrence in a cohort of patients treated with radical nephroureterectomy (RNU). Patients and methods: At 10 Canadian University Centers, we retrospectively evaluated data, between 1990 and 2010, from 743 patients who were free from bladder cancer and were previously treated with RNU for upper tract urothelial cancer. Results: Of 743 patients, 167 (22.5%) developed bladder tumors after a median time of 17.2 months after RNU. Multivariable analysis detected age (hazard ratio [HR] ¼ 1.028; 95% CI: 1.010–1.046; P ¼ 0.0018), tumor location in both the renal pelvis and the ureter (HR ¼ 2.205; 95% CI: 1.355–3.589; P ¼ 0.0015), the use of adjuvant systemic chemotherapy (HR ¼ 2.309; 95% CI: 1.439–3.705; P ¼ 0.0005), and laparoscopic surgery (HR ¼ 1.876; 95% CI: 1.226–2.87; P ¼ 0.0037) as risk factors for bladder cancer recurrence. Open excision of a bladder cuff (HR ¼ 0.661; 95% CI: 0.453–0.965; P ¼ 0.0319) and transurethral resection of the intramural ureter (HR ¼ 0.548; 95% CI: 0.306– 0.981; P ¼ 0.0429) on comparison with extravesical resection decreased the risk of bladder cancer recurrence significantly. Major limitations were the retrospective design and partially missing data, although the significance of variables did not change in the imputation analysis. Conclusion: Older patients, those with tumor location in both the renal pelvis and the ureter, and those treated with adjuvant systemic chemotherapy were found at higher risk for intravesical recurrence, as were those having undergone extravesical ureterectomy or laparoscopic RNU. r 2014 Elsevier Inc. All rights reserved.

Keywords: Bladder cancer; Radical nephroureterectomy; Recurrence; Risk factors; Upper tract urothelial carcinoma

Corresponding author. Tel.: þ1-418-525-4444, ext: 15068; fax: þ1-418691-5562. E-mail address: [email protected] (L. Lacombe). 1 Co-first authors. *

http://dx.doi.org/10.1016/j.urolonc.2014.04.006 1078-1439/r 2014 Elsevier Inc. All rights reserved.

1. Introduction Urothelial tumors of the upper urinary tract (UTUC) are relatively uncommon cancers, accounting for approximately

2

V. Fradet et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7

5% to 10% of all urothelial tumors and occurring at an annual incidence of 1 to 2 cases per 100,000 inhabitants in the Western countries. Renal pelvic tumors are about 2 times more common than ureteral cancers, with the distal ureter being more frequently involved than the proximal ureter [1]. Although cancer occurrence in the upper tract after primary bladder cancer is rare and develops in approximately only 1% to 5% [2], the incidence of tumor recurrence in the urinary bladder after upper tract urothelial cancer is much more common. Reported rates differ significantly throughout the literature, data range from 22% to 69% [3–10], making the bladder the number one recurrence site for UTUC [11]. In this study, we sought to determine risk factors for bladder cancer recurrence after radical nephroureterectomy (RNU) using a large retrospective population of patients treated for UTUC at 10 Canadian University Centers, as this topic had not yet been addressed in large cohorts.

2. Patients and methods After institutional review board approval, the medical records of patients who had undergone RNU for UTUC at 10 Canadian University Centers between January 1990 and June 2010 were reviewed and combined in 1 central database. Of 1,029 patients, 268 were excluded owing to a prior or simultaneous bladder cancer and 18 because of missing data on bladder cancer recurrence, leaving 743 patients with the necessary information available. Neoadjuvant chemotherapy was not given to any of these patients. Collected clinical and pathological parameters included age, gender, smoking status, presence of hydronephrosis, pT and pN stage, grade, presence of concomitant carcinoma in situ (CIS), tumor location (renal pelvis, ureter, or simultaneous tumors in both the regions), multifocality (multiple tumors in the same anatomical location), previous UTUC, previous administration of abdominal radiotherapy, distal ureter management (extravesical, extravesical and open intravesical, extravesical and endoscopic resection [12] of the ureteral orifice, etc.), tumor histology (pure urothelial vs. mixed type), surgical margin status, use of adjuvant systemic chemotherapy, and type of surgery (open vs. laparoscopic). For grading, the 1973 3-tiered World Health Organization system was used, and the 2002 Union Internationale Contre le Cancer TNM system was applied to assign stage. Bladder recurrence was defined as the development of a pathologically confirmed bladder tumor after RNU. Follow-up after RNU was done according to the treating urologist, diagnostic imaging was carried out as part of follow-up depending on pathological parameters or when clinically indicated. The Student t test and the Pearson chi-square test were used to evaluate the association between continuous and

categorical variables, respectively. To identify risk factors for bladder cancer recurrence after upper tract surgery, univariable and multivariable analyses using the Cox proportional hazards model were performed. Risk factors with a P o 0.25 in univariable analysis were included in the multivariable model. Hazard ratios (HRs) and their 95% CIs were also computed to assess the strength of the individual predictors. Tobacco smoking was regarded as a main variable of interest and, therefore, was forced into the multivariable model; this was based on our previous work that demonstrated a significant association between smoking and outcome in patients with urothelial carcinoma of the lower tract [13]. Patients not reaching the end point of bladder cancer recurrence were censored at time of death or end of follow-up. Verification of a correct Cox proportional hazards assumption was done using an extended Cox model with inclusion of time-dependent covariates generated by creating interactions of the predictors and a function of survival. To control for missing data, a multiple imputation process was performed. Under the assumption of missingness at random, imputations were created using PROC MI (Multiple Imputation using SAS) with the Markov chain Monte Carlo method; data sets were then combined with the MIANALYZE procedure that combines results of the imputations analysis and generates valid statistical inferences. As significances of predictors did not change in the imputed data set, the original model was regarded valid and used for publication. A 2-sided P o 0.05 was regarded statistically significant. Analyses were performed using SAS 9.1.3 statistical software.

3. Results Clinical and pathological characteristics of the patient cohort are shown in Table 1. Tumor location, existence of concomitant CIS, and percentage of patients having received adjuvant chemotherapy were significantly different between patients who experienced bladder cancer recurrence and those who did not. Overall, 56.4% patients underwent laparoscopic RNU (LRNU). Mean (median) follow-up for the entire cohort was 40.4 (24.8, interquartile range: 7.69–56.76) months. Of the patients, 22.5% (n ¼ 167) developed bladder cancer. Mean (median) time to intravesical cancer recurrence after RNU was 33.3 (17.2, interquartile range: 5.4–45.36) months. In total, 217 patients died of UTUC or other causes during follow-up, 43 died after having developed bladder cancer recurrence. The categorization of the bladder tumor was as follows: 26.6% Ta, 2.6% CIS, 53.3% T1, and 17.5% T2-T4. Univariable Cox proportional hazards model detected higher age (HR ¼ 1.017, P ¼ 0.0319), presence of concomitant CIS (HR ¼ 1.430, P ¼ 0.0428), tumor location in the renal pelvis and the ureter (HR ¼ 2.504, P o 0.0001), multifocality (HR ¼ 1.579, P ¼ 0.0089),

V. Fradet et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7 Table 1 Clinical and pathological characteristics of the patient cohort and association with bladder cancer recurrence in patients after RNU for UTUC

Table 1 Continued Characteristics

Characteristics

No. of Bladder recurrence patients (%) Negative Positive (%) (%)

P value

Age, y, mean ⫾ SD 69.7 ⫾ 10.8 69.4 ⫾ 11.0 70.4 ⫾ 10.0

0.3254a

Gender Male Female

0.5223b

Smoking status Never Former or active Hydronephrosis No Yes

438 (59.0) 304 (41.0)

343 (59.7) 232 (40.3)

95 (56.9) 72 (43.1) 0.842b

211 (31.6) 456 (68.4)

160 (31.4) 349 (68.6)

51 (32.3) 107 (67.7) 0.2458b

323 (51.8) 301 (48.2)

Previous abdominal radiotherapy No 597 (93.7) Yes 40 (6.3) Previous upper tract tumor No 611 (96.1) Yes 25 (3.9) Distal ureter management Extravesical 251 Open bladder cuff 267 TUR of orifice 76 Other 8

246 (53.1) 217 (46.9)

77 (47.8) 84 (52.2) 0.3129b

446 (94.3) 27 (5.7)

151 (92.1) 13 (7.9) 0.2674b

453 (95.6) 21 (4.4)

158 (97.5) 4 (2.5) 0.7613b

(41.7) (44.4) (12.6) (1.3)

Pathological T stage rpT1 331 (49.9) pT2 105 (15.8) pT3 182 (27.5) pT4 45 (6.8)

180 197 56 7

(40.9) (44.8) (12.7) (1.6)

71 70 20 1

(43.8) (43.2) (12.3) (0.6) 0.0969b

250 80 152 39

(48.0) (15.3) (29.2) (7.5)

81 25 30 6

(57.0) (17.6) (21.1) (4.2) 0.3139b

Pathological N stage N0 114 (15.3) N1-3 57 (7.7) Nx 571 (77.0)

83 (14.4) 47 (8.2) 445 (77.4)

31 (18.6) 10 (6.0) 126 (75.4)

Grade G1 G2/3

220 (30.4) 503 (69.6)

160 (28.6) 399 (71.4)

60 (36.6) 104 (63.4)

Concomitant CIS No Yes

484 (77.8) 138 (22.2)

367 (80.0) 92 (20.0)

117 (71.8) 46 (28.2)

Surgical margin status Negative 637 (91.7) Positive 58 (8.3)

490 (91.4) 46 (8.6)

147 (92.4) 12 (7.6)

Tumor location Renal pelvis Ureter Pelvis and ureter

346 (61.9) 119 (21.3) 94 (16.8)

74 (45.4) 42 (25.8) 47 (28.8)

Multifocality No Yes Histology Pure urothelial Mixed type

0.0513b

0.0309b

0.6786b

0.0003b 420 (58.2) 161 (22.3) 141 (19.5)

0.0765b 476 (75.7) 153 (24.3)

361 (77.5) 105 (22.5)

115 (70.6) 48 (29.4) 0.5861b

706 (96.4) 26 (3.6)

548 (96.7) 19 (3.3)

158 (95.8) 7 (4.2)

3

No. of Bladder recurrence patients (%) Negative Positive (%) (%)

P value

o0.0001b

Adjuvant chemotherapy No 661 (90.1) Yes 73 (9.9)

527 (92.8) 41 (7.2)

134 (80.7) 32 (19.3)

Type of surgery Open Laparoscopic

205 (45.6) 245 (54.4)

62 (38.3) 100 (61.7)

0.1089b 267 (43.6) 345 (56.4)

SD ¼ standard deviation. a Student t test. b Pearson chi-square test.

use of adjuvant chemotherapy (HR ¼ 2.740, P o 0.0001), and laparoscopic surgery (HR ¼ 1.819, P ¼ 0.0003) as predictors for bladder cancer recurrence (Table 2). On multivariable Cox regression analysis of risk factors as summarized in Table 2, higher age (HR ¼ 1.028, P ¼ 0.0018), tumor location in both the renal pelvis and the ureter (HR ¼ 2.205, P ¼ 0.0015), use of adjuvant chemotherapy (HR ¼ 2.309, P ¼ 0.0005), and laparoscopic surgery (HR ¼ 1.876, P ¼ 0.0037) were identified as statistically significant predictors for bladder cancer recurrence. Resection of the distal ureter with open excision of a bladder cuff (HR ¼ 0.661, P ¼ 0.0319) and ureterectomy with transurethral resection (TUR) of the orifice (HR ¼ 0.548, P ¼ 0.0429) were associated with a significantly decreased risk of bladder cancer recurrence when compared with extravesical-only ureterectomy. The association of surgical technique with clinical and pathological characteristics is shown in Table 3. Those patients having undergone LRNU were significantly less likely to smoke and to develop G2/3 tumors, they more likely to have hydronephrosis and to be treated with adjuvant systemic chemotherapy. Furthermore, ureterectomy with open bladder cuff excision or with TUR of the orifice was more frequently done in LRNU, whereas extravesical ureterectomy was the most frequent procedure in open RNU (ORNU). Positive lymph nodes were significantly more common in the ORNU group. Patients that received adjuvant chemotherapy had significantly more tumors with unfavorable pathology, such as higher stage, higher rate of positive lymph nodes, higher rate of concomitant CIS, and higher rate of positive margins, and had more likely been laparoscopically operated (Table 4). However, excluding patients having received adjuvant chemotherapy from analysis did not change the significance of other predictors (data not shown). The 5- and 10-year overall survival were 74.9% and 58.0%, respectively. The 5- and 10-year overall survival from the time of bladder recurrence were 65.7% and 54.8%, respectively.

V. Fradet et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7

4

Table 2 Univariable and multivariable Cox proportional hazards modeling of bladder cancer recurrence after RNU for UTUC Variable

Multivariablea

Univariable HR

95% CI

P value

Lower

Upper

1.001 0.795 0.724 0.925 0.786 0.216

1.033 1.479 1.428 1.733 2.555 1.574

0.0319 0.6088 0.9216 0.1415 0.2457 0.2871

Distal ureter management (extravesical as reference) Open bladder cuff 0.739 Endourologic resection of orifice 0.797

0.529 0.485

1.032 1.309

0.0760 0.3700

Pathological tumor stage (rpT1 as reference) pT2 1.035 pT3 0.860 pT4 1.030

0.659 0.563 0.448

1.625 1.312 2.371

0.8808 0.4837 0.9437

Pathological N stage (pN0 as reference) pN1-3 Grade (G1 vs. G2/3) Concomitant CIS Surgical margin status

0.930 0.879 1.430 1.342

0.467 0.637 1.012 0.744

1.853 1.213 2.022 2.420

0.8369 0.4325 0.0428 0.3289

Tumor location (renal only as reference) Ureter Renal pelvis and ureter Multifocality Histology Adjuvant chemotherapy Laparoscopic surgery

1.446 2.504 1.579 1.232 2.740 1.819

0.984 1.721 1.121 0.577 1.832 1.313

2.125 3.643 2.222 2.633 4.098 2.521

0.0601 o0.0001 0.0089 0.5894 o0.0001 0.0003

Age, y Gender Smoking status Hydronephrosis Previous abdominal radiotherapy Previous upper tract tumor

1.017 1.084 1.017 1.266 1.417 0.583

HR

95% CI

P value

Lower

Upper

1.028

1.01

1.046

0.0018

1.363 1.075 1.061

0.929 0.750 0.553

2.000 1.541 2.038

0.1128 0.6940 0.8579

0.661 0.548

0.453 0.306

0.965 0.981

0.0319 0.0429

1.133

0.742

1.731

0.5639

1.454 2.205 1.053

0.928 1.355 0.668

2.276 3.589 1.659

0.1023 0.0015 0.8249

2.309 1.876

1.439 1.226

3.705 2.87

0.0005 0.0037

Note: numbers in bold means they are significant (P o 0.05) a Multivariable models included those listed in the Table and year of surgery (stratified).

4. Discussion Prognosticators for the development of subsequent bladder cancer still remain unclear. Publications on this topic are on small series [3–9,14,15], and most are based on populations from the Far East [5,6,8,9,14,15], where risk factors may not match those predominant in Western-style countries [16]. Furthermore, recurrence rates differ significantly, probably owing to geographical differences in patient cohorts, or owing to the inclusion of patients with prior bladder cancer in some studies, among them those with the highest recurrence rates [3,4,7,9]. A recently published review described multifocality as the only repeatedly mentioned significant risk factor for bladder cancer recurrence after UTUC [11]. Nonetheless, only a large recently published contemporary series on oncological outcomes of RNU for UTUC addressed risk factors for bladder cancer recurrence [10]. Two clonal hypotheses tried to explain the development of urothelial tumors at multiple sites. The field-change hypothesis suggests that different clones of cancer cells are induced by mutagens at different locations, whereas the intraluminal seeding and implantation hypothesis proposes

that multiple tumor formation results from the downstream implantation or intraepithelial migration of cancer cells originating from a single clone. In our series, as in others, 50% of UTUCs were invasive and 470% of bladder recurrence were ZpT1 perhaps supporting the monoclonal hypothesis with a higher risk of bladder implantation from more aggressive cancers. Two important features distinguish our work from other published series: the large number of patients and the exclusion of those with a bladder cancer history as primary event. In contrast to other publications [3–5,7,8,14,15,17], we identified increasing age as an independent predictor for intravesical tumor recurrence after UTUC (HR = 1.028, P = 0.002), which was similar to the results of the study conducted by Xylinas et al. [10]. Despite an apparently low HR, it is important to realize that every year of increasing patient age at UTUC diagnosis raises the risk of developing subsequent bladder cancer by 2.8%. Possible explanations for our finding could be a reduced host defense, which might alleviate the implantation of tumor cells that descended during operation or might facilitate the transformation of an already premalignantly modified bladder urothelium.

V. Fradet et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7 Table 3 Significant associations of laparoscopic surgery with clinical and pathological characteristics in patients after RNU for UTUC Characteristics

No. of patients (%)

Laparoscopic surgery No (%)

Yes (%) 0.0036a

Pathological N stage N0 N1-3 NX

103 (16.8) 45 (7.4) 464 (75.8)

39 (14.6) 30 (11.2) 198 (74.2)

64 (18.6) 15 (4.3) 266 (77.1)

Smoking status Never Former or active

179 (32.1) 378 (67.9)

60 (24.1) 189 (75.9)

119 (38.6) 189 (61.4)

Hydronephrosis No Yes

0.0003a

0.0007a 306 (51.5) 288 (38.5)

Distal ureter management Extravesical 251 Open bladder cuff 266 TUR of orifice 76 Other 8 Grade G1 G2/3

157 (59.2) 108 (40.8)

149 (45.3) 180 (54.7)

134 111 9 5

117 155 67 3

o0.0001a (41.8) (44.3) (12.6) (1.3)

(51.7) (42.9) (3.5) (1.9)

(34.2) (45.3) (19.6) (0.9) 0.0434a

201 (33.6) 398 (66.4)

77 (29.2) 187 (70.8)

124 (37.0) 211 (63.0)

Adjuvant chemotherapy No 546 (89.8) Yes 62 (10.2)

245 (92.8) 19 (7.2)

301 (87.5) 43 (12.5)

a

P value

0.0322a

5

of the intramural ureter (HR ¼ 0.548, P ¼ 0.0429) diminished the risk of bladder cancer recurrence. Open distal ureterectomy and TUR have already been shown to be equivalent regarding bladder cancer recurrence [19], and surprisingly, all the 3 methods of distal ureteral resection were reported oncologically equal in a recently published article [20]. However, performing an extravesical ureterectomy does not necessarily translate to leaving the orifice behind, the distal ureter and the bladder cuff can nevertheless be resected completely. To which extent the excision of the orifice had been done in our patient cohort cannot be answered conclusively; hence, this fact could contribute to the different findings. Yet, our results outline the importance of a complete distal ureterectomy to prevent the wellknown risk of recurrence in the ureteral stump [21] and to decrease the probability of subsequent bladder tumors [14]. Laparoscopic surgery was also a significant predictor for bladder cancer recurrence in multivariable analysis (HR ¼ 2.203, P o 0.0001) in the present series, as also seen in other articles [10]. Factors that could be responsible for this difference are tumor cell dissemination throughout the urinary tract owing to the high gas pressure or a longer operating time, resulting in longer tumor exposure [11]. Some authors [6,9] drew a similar conclusion, others did not confirm these results [8,17]. A recently published comparison of ORNU and LRNU by Favaretto et al. [22] did not observe a significant association of surgical approach with intravesical recurrence. However, their series did not exclude patients with prior non–muscle-invasive bladder cancer and included only 53 patients with LRNU,

Pearson chi-square test.

We also found synchronous tumors in the renal pelvis and in the ureter increasing the risk for subsequent bladder cancer by 2.2-fold (HR ¼ 2.252, P ¼ 0.0011). On the contrary, multifocality in terms of multiple single-site cancers did not emerge as a significant predictor (P ¼ 0.7245). These results are in concordance with other articles that described the significant effect of multiple-site tumors on bladder cancer recurrence [5,10,17]. Other publications found multifocality being a more important predictor than tumor location [6,18], but definitions have not always been used in the same context throughout the literature, making comparisons difficult. Novara et al. [7], highlighted the importance of pelvic or concomitant pelvic and ureteral tumors as independent risk factors for bladder cancer recurrence, but in a subgroup analysis of patients without history of bladder tumors, this association was not observed. A drawback of that work is the small sample size of 135 patients, of whom 35.6% relapsed and they were found to have intravesical tumors. Distal ureter management during surgery for UTUC has always been a key point of discussion. When compared with extravesical resection, distal ureterectomy with open bladder cuff excision (HR ¼ 0.661, P ¼ 0.0319) and TUR

Table 4 Significant associations of adjuvant chemotherapy with clinical and pathological characteristics in patients after RNU for UTUC Characteristics

No. of patients (%)

Pathological T stage rpT1 328 (50.0) pT2 104 (15.8) pT3 179 (27.3) pT4 45 (6.9)

Adjuvant chemotherapy No (%)

Yes (%) o0.0001a

319 98 154 30

(53.1) (16.3) (25.6) (5.0)

9 6 25 15

(16.4) (10.9) (45.4) (27.3) o0.0001a

Pathological N stage N0 114 (15.5) N1-3 57 (7.8) NX 563 (76.7)

106 (16.0) 39 (5.9) 516 (78.1)

8 (10.9) 18 (24.7) 47 (64.4)

Concomitant CIS No 482 (78.0) Yes 136 (22.0)

447 (80.8) 106 (19.2)

35 (53.8) 30 (46.2)

Surgical margin status Negative 630 (91.8) Positive 56 (8.2)

574 (93.0) 43 (7.0)

56 (81.2) 13 (18.8)

Laparoscopic surgery No 264 (43.4) Yes 344 (56.6)

245 (44.9) 301 (55.1)

19 (30.6) 43 (69.4)

a

Pearson chi-square test.

P value

o0.0001a

0.0006a

0.0322a

6

V. Fradet et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7

of which 5 had bladder recurrences, limiting the generalizability of their findings. Other large series comparing ORNU and LRNU excluded subsequent bladder cancers from recurrence-free survival analysis [23,24]; but one multicenter analysis on 1,003 patients that underwent LRNU in Japan [25] described a higher rate of intravesical recurrences for patients having undergone hand-assisted LRNU. The authors suggested that hand manipulation might be responsible for subsequent tumor dissemination to the bladder. Just 1 prospective trial comparing ORNU and LRNU had been performed: Simone et al. [26] randomized 80 patients and reported similar intravesical recurrence rates for ORNU and LRNU, but the study lacks power because of the small sample size. The question might arise whether probably other cofactors, as distal ureter management, could be responsible for our results. Patients subjected to LRNU were more likely to undergo open bladder cuff excision or TUR of the orifice. Thus, we would rather expect a selection bias affecting outcomes against ORNU, but that was not the case. Despite the contradictory findings throughout the literature, it might be advisable that during LRNU surgeons consider early ureteral clipping to avoid tumor spillage into the urinary bladder. We also observed that the use of adjuvant systemic chemotherapy is an independent predictor for increased bladder cancer recurrence (HR ¼ 2.533, P o 0.0001). Only 1 article reported similar results [14], whereas others found a reduced risk of recurrence with chemotherapy [15,27] or did not observe any statistically significant difference [5]. Our findings have to be regarded with great caution. As indications and chemotherapy regimen have been subject to change over the years, it is very difficult to draw conclusions. This time trend also appeared concurrently with the introduction of laparoscopy, which could potentially affect the associations observed with laparoscopy. Patients having received systemic adjuvant chemotherapy had less favorable pathology, and especially the higher rate of concomitant CIS and more positive surgical margins might put them at a higher risk for developing subsequent bladder cancers. Thus, our findings might presumably be owing to selection bias. But, even if those patients who received adjuvant chemotherapy were at highest risk for developing systemic metastases, they were still at a high risk for recurring intravesically before dying of the primary UTUC. As far as we know, this is by far the largest study addressing risk factors for bladder cancer recurrence after treatment of UTUC. However, there are several limitations to this work worth noting, first those inherent to any retrospective analysis. We cannot preclude a certain selection bias owing to the fact that some patients without complete follow-up information have not been included in this analysis. Despite accurate data retrieving methods, a considerable amount of pathological parameters in the patient cohort could not be

gathered retrospectively. We did not collect data on the exact location of the intravesical tumor recurrence site or whether it was situated on the ipsilateral ureteral orifice or scar, and information on the exact tumor location within the ureter was not available. Also, the longer time to identification of bladder recurrence relative to modern series may suggest a delay in the bladder recurrence. This fact did not affect the associations found, however, and should not cause bias. We did not differentiate between former and active smokers nor did we assess the amount of pack-years in those, which might add some bias to the comparison of smokers and nonsmokers. Finally, patients had been treated at 10 different university clinics and had been operated by different surgeons over a relatively long time span of 20 years. During that time span, laparoscopy was certainly introduced in some of the centers, and this fact may have contributed to the finding concerning laparoscopic surgery as a significant predictor for bladder cancer recurrence. No standardized regimen was used for follow-up, which could have changed the time to diagnosis between different institutions. Specimen had been examined by several different pathologists, albeit all of them with distinct uropathological expertise. Despite all these aforementioned limitations, multicenter analyses are mandatory to obtain potentially actionable clinical information for the management of this relatively rare tumor. Given the growing evidence of reduced bladder cancer tumors with the administration of intravesical mitomycin C after RNU [28,29], identifying patients at highest risk for recurrence could help selecting those who are most likely to benefit from instillation therapy. 5. Conclusion Older patients, those with tumor location in both the renal pelvis and the ureter and those treated with adjuvant systemic chemotherapy were found to be at a higher risk for intravesical recurrence, as were those patients having undergone extravesical ureterectomy or LRNU. These risk factors for bladder cancer recurrence need consideration when operating and following up patients after RNU. Acknowledgments The authors thank Hassan Behlouli, Ph.D., from the Research Institute of the McGill University Health Centre for assistance with the statistical analysis. References [1] Roupret M, Zigeuner R, Palou J, Boehle A, Kaasinen E, Sylvester R, et al. European guidelines for the diagnosis and management of upper urinary tract urothelial cell carcinomas: 2011 update. Eur Urol 2011;59:584–94. [2] Sanderson KM, Cai J, Miranda G, Skinner DG, Stein JP. Upper tract urothelial recurrence following radical cystectomy for transitional cell

V. Fradet et al. / Urologic Oncology: Seminars and Original Investigations ] (2014) 1–7

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

carcinoma of the bladder: an analysis of 1,069 patients with 10-year followup. J Urol 2007;177:2088–94. Raman JD, Ng CK, Boorjian SA, Vaughan ED Jr., Sosa RE, Scherr DS. Bladder cancer after managing upper urinary tract transitional cell carcinoma: predictive factors and pathology. BJU Int 2005;96:1031–5. Pieras E, Frontera G, Ruiz X, Vicens A, Ozonas M, Piza P. Concomitant carcinoma in situ and tumour size are prognostic factors for bladder recurrence after nephroureterectomy for upper tract transitional cell carcinoma. BJU Int 2010;106:1319–23. Hisataki T, Miyao N, Masumori N, Takahashi A, Sasai M, Yanase M, et al. Risk factors for the development of bladder cancer after upper tract urothelial cancer. Urology 2000;55:663–7. Matsui Y, Utsunomiya N, Ichioka K, Ueda N, Yoshimura K, Terai A, et al. Risk factors for subsequent development of bladder cancer after primary transitional cell carcinoma of the upper urinary tract. Urology 2005;65:279–83. Novara G, De Marco V, Dalpiaz O, Gottardo F, Bouygues V, Galfano A, et al. Independent predictors of metachronous bladder transitional cell carcinoma (TCC) after nephroureterectomy for TCC of the upper urinary tract. BJU Int 2008;101:1368–74. Koda S, Mita K, Shigeta M, Usui T. Risk factors for intravesical recurrence following urothelial carcinoma of the upper urinary tract: no relationship to the mode of surgery. Jpn J Clin Oncol 2007;37:296–301. Kume H, Teramoto S, Tomita K, Nishimatsu H, Takahashi S, Takeuchi T, et al. Bladder recurrence of upper urinary tract cancer after laparoscopic surgery. J Surg Oncol 2006;93:318–22. Xylinas E, Kluth L, Passoni N, Trinh QD, Rieken M, Lee RK, et al. Prediction of intravesical recurrence after radical nephroureterectomy: development of a clinical decision-making tool. Eur Urol 2014;65:650–8. Azemar MD, Comperat E, Richard F, Cussenot O, Roupret M. Bladder recurrence after surgery for upper urinary tract urothelial cell carcinoma: frequency, risk factors, and surveillance. Urol Oncol 2011;29:130–6. Gakis G, Ninkovic M, van Koeveringe GA, Raina S, Sturtz G, Rahnama'i MS, et al. Functional detrusor myoplasty for bladder acontractility: long-term results. J Urol 2011;185:593–9. Yafi FA, Aprikian AG, Chin JL, Fradet Y, Izawa J, Estey E, et al. Contemporary outcomes of 2287 patients with bladder cancer who were treated with radical cystectomy: a Canadian multicentre experience. BJU Int 2011;108:539–45. Koga F, Nagamatsu H, Ishimaru H, Mizuo T, Yoshida K. Risk factors for the development of bladder transitional cell carcinoma following surgery for transitional cell carcinoma of the upper urinary tract. Urol Int 2001;67:135–41. Ku JH, Choi WS, Kwak C, Kim HH. Bladder cancer after nephroureterectomy in patients with urothelial carcinoma of the upper urinary tract. Urol Oncol 2011;29:383–7. Colin P, Koenig P, Ouzzane A, Berthon N, Villers A, Biserte J, et al. Environmental factors involved in carcinogenesis of urothelial cell carcinomas of the upper urinary tract. BJU Int 2009;104:1436–40. Terakawa T, Miyake H, Muramaki M, Takenaka A, Hara I, Fujisawa M. Risk factors for intravesical recurrence after surgical management

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

7

of transitional cell carcinoma of the upper urinary tract. Urology 2008;71:123–7. Milojevic B, Djokic M, Sipetic-Grujicic S, Milenkovic-Petronic D, Vuksanovic A, Dragicevic D, et al. Bladder cancer after managing upper urinary tract transitional cell carcinoma: risk factors and survival. Int Urol Nephrol 2011;43:729–35. Salvador-Bayarri J, Rodriguez-Villamil L, Imperatore V, Palou Redorta J, Villavicencio-Mavrich H, Vicente-Rodriguez J. Bladder neoplasms after nephroureterectomy: does the surgery of the lower ureter, transurethral resection or open surgery, influence the evolution? Eur Urol 2002;41:30–3. Li WM, Shen JT, Li CC, Ke HL, Wei YC, Wu WJ, et al. Oncologic outcomes following three different approaches to the distal ureter and bladder cuff in nephroureterectomy for primary upper urinary tract urothelial carcinoma. Eur Urol 2010;57:963–9. Strong DW, Pearse HD. Recurrent urothelial tumors following surgery for transitional cell carcinoma of the upper urinary tract. Cancer 1976;38:2173–83. Favaretto RL, Shariat SF, Chade DC, Godoy G, Kaag M, Cronin AM, et al. Comparison between laparoscopic and open radical nephroureterectomy in a contemporary group of patients: are recurrence and disease-specific survival associated with surgical technique? Eur Urol 2010;58:645–51. Capitanio U, Shariat SF, Isbarn H, Weizer A, Remzi M, Roscigno M, et al. Comparison of oncologic outcomes for open and laparoscopic nephroureterectomy: a multi-institutional analysis of 1249 cases. Eur Urol 2009;56:1–9. Walton TJ, Novara G, Matsumoto K, Kassouf W, Fritsche HM, Artibani W, et al. Oncological outcomes after laparoscopic and open radical nephroureterectomy: results from an international cohort. BJU Int 2011;108:406–12. Kamihira O, Hattori R, Yamaguchi A, Kawa G, Ogawa O, Habuchi T, et al. Laparoscopic radical nephroureterectomy: a multicenter analysis in Japan. Eur Urol 2009;55:1397–407. Simone G, Papalia R, Guaglianone S, Ferriero M, Leonardo C, Forastiere E, et al. Laparoscopic versus open nephroureterectomy: perioperative and oncologic outcomes from a randomised prospective study. Eur Urol 2009;56:520–6. Soga N, Arima K, Sugimura Y. Adjuvant methotrexate, vinblastine, adriamycin, and cisplatin chemotherapy has potential to prevent recurrence of bladder tumors after surgical removal of upper urinary tract transitional cell carcinoma. Int J Urol 2008;15:800–3. Wu WJ, Ke HL, Yang YH, Li CC, Chou YH, Huang CH. Should patients with primary upper urinary tract cancer receive prophylactic intravesical chemotherapy after nephroureterectomy? J Urol 2010; 183:56–61. O'Brien T, Ray E, Singh R, Coker B, Beard R. Prevention of bladder tumours after nephroureterectomy for primary upper urinary tract urothelial carcinoma: a prospective, multicentre, randomised clinical trial of a single postoperative intravesical dose of mitomycin C (the ODMIT-C Trial). Eur Urol 2011;60:703–10.