Stenting Prior to Cystectomy is an Independent Risk Factor for Upper Urinary Tract Recurrence

Author's Accepted Manuscript Double J stenting prior to cystectomy is an independent risk factor for upper urinary tract recurrence Bernhard Kiss , Marc A. Furrer , Patrick Y. Wuethrich , Fiona C. Burkhard , George N. Thalmann , Beat Roth

PII: DOI: Reference:

S0022-5347(17)74752-0 10.1016/j.juro.2017.06.020 JURO 14798

To appear in: The Journal of Urology Accepted Date: 1 June 2017 Please cite this article as: Kiss B, Furrer MA, Wuethrich PY, Burkhard FC, Thalmann GN, Roth B, Double J stenting prior to cystectomy is an independent risk factor for upper urinary tract recurrence, The Journal of Urology® (2017), doi: 10.1016/j.juro.2017.06.020. DISCLAIMER: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our subscribers we are providing this early version of the article. The paper will be copy edited and typeset, and proof will be reviewed before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to The Journal pertain.

Embargo Policy All article content is under embargo until uncorrected proof of the article becomes available online. We will provide journalists and editors with full-text copies of the articles in question prior to the embargo date so that stories can be adequately researched and written. The standard embargo time is 12:01 AM ET on that date. Questions regarding embargo should be directed to [email protected].

ACCEPTED MANUSCRIPT Double J stenting prior to cystectomy is an independent risk factor for upper urinary tract recurrence Bernhard Kiss1*§, Marc A. Furrer1*, Patrick Y. Wuethrich2, Fiona C. Burkhard1, George N. Thalmann1, Beat Roth1 Department of Urology, University Hospital of Bern, University of Bern, Bern, Switzerland

2

Department of Anaesthesiology and Pain Medicine, University Hospital of Bern, University

RI PT

1

*

These authors equally contributed to this work

§

Corresponding author:

M AN U

Bernhard Kiss M.D.

SC

of Bern, Bern, Switzerland

Department of Urology

University Hospital of Bern University of Bern

TE D

CH-3010 Bern, Switzerland Phone number: +41 31 632 36 21

EP

Fax number: +41 31 632 21 81

AC C

E-email address: [email protected]

Key words:

Bladder cancer, double J, nephrostomy, recurrence, upper urinary tract

Word count:

Abstract: 247; text (not including abstract): 2498 words.

Runninghead:

Double J stenting and upper urinary tract recurrence in bladder cancer

1

ACCEPTED MANUSCRIPT Abstract Purpose: Bladder cancer patients presenting with hydronephrosis may require drainage of the affected kidney before receiving further cancer treatment. Drainage can be done by retrograde stenting or percutaneously. Retrograde stenting, however, bears the risk of tumour cell

RI PT

spillage to the upper urinary tract. The aim of this study was to evaluate whether bladder cancer patients are at higher risk of upper urinary tract recurrence if retrograde stenting has been performed prior to radical cystectomy.

SC

Material and Methods: Retrospective analysis of 1005 consecutive bladder cancer patients who underwent radical cystectomy at our department between 01/2000 and 06/2016. Negative

M AN U

ureteral margins intraoperatively were mandatory. Patients had regular follow-up according to our institutional protocol, including imaging of the upper urinary tract and urine cytology. Results: 114/1005 patients (11%) had preoperative drainage of the upper urinary tract, 53/114 (46%) by double J stent and 61/114 (54%) by percutaneous nephrostomy. Among all patients,

TE D

31/1005 (3%) developed UUT recurrence after a median time from cystectomy of 17 months: 7/53 (13%) in the ‘double J’ group, 0% in the ‘nephrostomy’ group, and 24/891 (3%) in the ‘no drainage’ group. Multivariate regression analysis revealed a higher risk for upper urinary

EP

tract recurrence if patients had double J stenting HR 4.54 [95%CI: 1.43-14.38, p=0.01]) and and preoperative intravesical instillations (HR 2.94 [95%CI: 1.40-6.16, p=0.004]).

AC C

Conclusions: Patients undergoing double J stenting prior to radical cystectomy are at higher risk of upper urinary tract recurrence. If preoperative upper urinary tract drainage is required, percutaneous drainage might be recommended.

2

ACCEPTED MANUSCRIPT Introduction Hydronephrosis at the time of radical cystectomy (RC) is found in up to 54% of bladder cancer (BC) patients [1]. It is not only a sign of advanced tumor stage but also an independent prognostic marker for worse recurrence-free survival (RFS) and cancer-specific survival

RI PT

(CSS) [2,3]. Because prolonged chronic hydronephrosis leads to irreversible deterioration of renal function [4], timely resolution of the obstruction is imperative if the projected interval from diagnosis to RC is too long or if the obstruction is symptomatic (e.g. obstructive

SC

pyelonephritis, pain). If neoadjuvant chemotherapy is planned ˗ as is increasingly the case in daily clinical practice [5] ˗ preservation of every possible nephron is vital because cisplatin-

M AN U

based chemotherapy requires adequate renal function.

Whether drainage of an obstructed kidney in BC patients should be managed by retrograde (usually double J stenting) or antegrade (nephrostomy) drainage is a matter of debate [6]. Studies comparing the two methods in cancer patients show similar morbidity rates in the two

TE D

groups but slightly better decompression rates in patients receiving a nephrostomy [6,7]. These studies, however, were mainly performed in the palliative setting of patients with endstage malignant disease and thus are barely applicable in the pre-operative setting of

EP

potentially curable patients. They also never addressed the impact of upper urinary tract (UUT) drainage on UUT recurrences. The aim of this study was to investigate the impact of

AC C

the type of UUT drainage prior to RC on postoperative UUT recurrence in BC patients.

3

ACCEPTED MANUSCRIPT Materials and Methods Between 01/2000 and 06/2016, 1005 BC patients underwent RC and urinary diversion at our institution. All patients had regular follow-up in accordance with our institutional protocol. Potential factors for postoperative UUT recurrence and the influence of the type of

RI PT

preoperative urinary drainage (double J stent vs nephrostomy vs no drainage) to resolve hydronephrosis were retrospectively evaluated. The study protocol was approved by the local

SC

Ethics Committee (KEK-Be 2016-00660).

Surgical procedures

M AN U

The type of drainage was chosen according to the preference of the urologist. In case of unilateral obstruction, only the obstructed side was drained. The surgical technique for RC was described in detail before [8]. Intraoperatively, ureteral margins were sent for frozen section analysis in all patients and in case of malignant dysplasia - including carcinoma in situ

Staging and follow-up

TE D

(CIS) - further resection of the ureter was performed until the frozen sections were negative.

EP

Preoperative staging included physical examination, blood chemistry and computed

AC C

tomography [CT]. The UUT was evaluated by UUT imaging in every patient; either by intravenous pyelography (IVP) or CT-urography. In case of kidney insufficiency, magnetic resonance tomography or retrograde ureteropyelography was performed. A detailed description of our institutional follow-up protocol was published earlier [9,10]; it included physical examination, blood chemistry, chest x-ray (or chest CT if deemed necessary), and kidney ultrasound. Urethral cytology was performed 3 months after RC, every 6 months until year 5, and annually thereafter. Additional cytology after forced diuresis (with furosemide) for evaluation of the UUT was performed 6, 18, 30, 42 and 54 months postoperatively. Until 2007, the UUT was radiologically assessed using IVP 1, 2, 3, 5, 7 and 4

ACCEPTED MANUSCRIPT 10 years after cystectomy [10]. Thereafter, patients with previous or concomitant CIS at RC or a history of multifocal stage pTa/1 BC received CT urography (or IVP) 1, 2 and 3 years after RC. If patients had a tumor stage ≥ pT3 or pT1–4 pN+, they underwent CT scans of the chest/abdomen/pelvis after 6, 12 and 18 months. Additional imaging was performed if

RI PT

deemed necessary. In case of suspicious findings, selective cytology from the UUT and/or ureterorenoscopy was performed. All recurrences were confirmed by cytology and/or biopsy.

SC

Statistical analyses

Qualitative data were presented as counts and percentages and assessed using the Chi-square

M AN U

test or the Fisher exact test. Quantitative data were presented as median and range or interquartile range (IQR) and assessed by the Kruskal Wallis test or the Wilcoxon rank sum test. The Student’s t test was used for normally distributed, quantitative data. Cox regression analysis (univariate) was performed to evaluate different variables (type of drainage [double

TE D

J, nephrostomy, no drainage], presence of CIS, hydronephrosis, type of urinary diversion, tumor stage and grading, lymphnode stage, [neo]adjuvant chemotherapy, preoperative bladder instillations, patient age and gender) and the incidence of UUT recurrence. For the

EP

multivariate Cox regression model, all variables with a p-value ≤0.05 in univariate regression analysis were selected. Results were presented as hazard ratios (HR) and 95% confidence

AC C

intervals (CI). Survival analysis was performed using the Kaplan–Meier method; the log-rank test was used for univariate comparisons. A two-tailed level of significance was set at 0.05 for all statistical tests. SPSS statistical software package 20.0 (SPSS, Inc., Chicago, IL, USA) was used for all statistical analyses.

5

ACCEPTED MANUSCRIPT Results Of the 1005 RC patients, 23% (226/1005) presented with hydronephrosis prior to RC: 4% (40/1005) bilaterally and 19% (186/1005) unilaterally. Fifty percent (114/226) of patients with preoperative hydronephrosis (comprising 11% [114/1005] of all patients) had

RI PT

preoperative drainage: 53/114 (46%) had preoperative double J stenting, 61/114 (54%) received a nephrostomy tube preoperatively. Eighty-nine percent (891/1005) of patients had no preoperative drainage. Baseline characteristics of all patients are shown in Table 1. Median

SC

follow-up of the complete cohort was 36 months (interquartile range: 14-89 months). Followup was significantly shorter in the ‘double J’ (21.6 months) and the ‘nephrostomy’ (20.4

M AN U

months) groups compared to the ‘no drainage’ group (38.4 months) (p<0.001). Three percent (31/1005) of patients developed an UUT recurrence after a median follow-up from cystectomy of 17 months (range: 3-147 months): 13% (7/53) in the ‘double J’ group, 0%

TE D

(0/61) in the ‘nephrostomy’ group, and 3% (24/891) in the ‘no drainage’ group. All (7/7) recurrences in the ‘double J’ group occurred within the first 24 months and were located on the side that was previously stented. The stent dwell time was longer in the ‘double

EP

J’ group with recurrences (n=7) compared to the ‘double J’ group without recurrences (n=46; 49 days vs 42 days, respectively; p=0.22). The time of double J stenting (before [n=9], during

AC C

[n=39] or after [n=5] TURBT) did not influence the rate of UUT recurrences in the ‘double J’ group (p=0.9).

Six percent (2/31) of UUT recurrences were pTis, 29% (9/31) pTa, 65% (20/31) ≥ pT1 (29% (9/31) low-grade and 71% (22/31) high-grade). In 45% (14/31) of recurrences a radical surgical treatment by either nephroureterectomy or partial ureterectomy had to be performed. Univariate Cox regression analysis revealed a significantly higher risk for UUT recurrence after double J stenting (HR 7.09 [95% CI: 3.03-16.61]; p<0.001), if hydronephrosis was 6

ACCEPTED MANUSCRIPT present (HR 2.91 [95% CI: 1.40-6.06]; p=0.004), and after preoperative intravesical instillations (HR 2.37 [95% CI: 1.15-4.88]; p=0.02). In multivariate analysis, however, only double J stenting (HR 4.54 [95%CI: 1.43-14.38, p=0.01]) and preoperative intravesical instillations (HR 2.94 [95%CI: 1.40-6.16, p=0.004] remained independent variables for UUT

RI PT

recurrences (Table 2 and 3).

Univariate Cox regression analysis in the subgroup of patients with hydronephrosis (n=226) revealed that double J stenting as well as preoperative intravesical instillations remained

SC

significant risk factors for UUT recurrence (Table 4). Upper urinary tract RFS was

M AN U

significantly shorter in the ‘double J’ group compared to the ‘nephrostomy’ and the ‘no drainage’ groups (Figure 1A). CSS was significantly longer in the ‘no drainage’ group than in the ‘double J‘ (p=0.03) and the ‘nephrostomy’ groups (p<0.001; Figure 1B), whereas overall

AC C

EP

TE D

survival (OS) was significantly shorter in the ‘nephrostomy’ group (p<0.001; Figure 1C).

7

ACCEPTED MANUSCRIPT Discussion Literature on the influence of pre-RC urinary drainage on UUT recurrence is lacking. If drainage of the UUT in a hydronephrotic kidney prior to RC is indicated, therefore, selection of the type of drainage (retrograde stenting or percutaneous) will be based primarily on

RI PT

personal preferences rather than on considerations of possible oncological consequences. In the present retrospective analysis of a consecutive single-center series of 1005 BC patients who underwent RC and urinary diversion, we found that double J stenting prior to RC was an

SC

independent risk factor for UUT recurrences in multivariate analysis.

The reason for the significantly higher number of UUT recurrences in the ‘double J’ group is

M AN U

most probably the fact that internal ureteral stenting converts a non-refluxing into a refluxing system, thus facilitating constant tumor cell seeding and implantation from the bladder into the UUT. Already in 1992, Palou et al showed that patients with recurrent bladder tumors and reflux due to resection of the ostium had a significantly higher probability of UUT tumors in

TE D

the refluxing renal unit [11]. This seems to be the same principle at work in patients with an indwelling (double J) stent. The trend towards a longer double J dwell time in the ‘double J’ group with UUT recurrence compared to the ‘double J’ group without UUT recurrence (49

EP

days vs 42 days, respectively) strengthens this hypothesis. Another but probably less substantial reason for the higher number of UUT recurrences is the retrograde manipulation

AC C

during stent insertion, which might actively flush tumor cells into the UUT. Three percent (31/1005) of our patients developed UUT recurrences after RC. This is in line with the literature reporting UUT recurrence rates after RC of between 0.8% and 6.4% [12]. Volkmer et al found in their large single-center study (1420 patients) a slightly lower rate of UUT recurrence (1.8%) after a median post-RC interval of 39 months and a mean follow-up of 58 months [13]. Not surprisingly, rates of UUT recurrence increased with longer follow-up. Thus, UUT recurrences are considered to be a late oncological event occurring a median 24 to 36 months following RC [14-16]. In our study, however, UUT recurrences developed earlier, 8

ACCEPTED MANUSCRIPT namely after a median post-RC interval of 17 months. Because all recurrences in the ‘double J’ group occurred early after RC (<12-24 months), the higher rate of patients with an indwelling double J stent could be the reason for this earlier occurrence of UUT recurrences in our series compared to the series of Volkmer et al. [13].

RI PT

The shortness of the interval to UUT recurrence after RC and the abovementioned longer double J dwell time in the ‘double J’ group with UUT recurrences underlines the importance of the mechanism of reflux of tumor cell-saturated urine prior to RC in the development of

SC

UUT recurrences. Upper urinary tract recurrences occurring late in the postoperative period are more likely to be de novo malignancies, which accords with the concept of urothelial

M AN U

cancer being a pan-urothelial disease and a consequence of the pan-urothelial field defect [1719]. Thus, CIS, recurrent NMIBC, and concurrent tumor involvement of the distal ureter were identified as risk factors for UUT recurrences in a large single-center study and meta-analysis [12,13]. Likewise, multiple intravesical instillations before RC (a surrogate for recurrent

TE D

NMIBC/Cis) was a significant and independent factor for UUT recurrence in our series. While hydronephrosis has been shown to be an independent prognostic marker for worse RFS and worse CSS in cystectomy patients [2,3], in our series it was also associated with higher

EP

risk of UUT recurrences in univariate analysis. In contrast to double J stenting, however, hydronephrosis did not remain a significant variable in multivariate analysis. These

AC C

apparently contradictory results might be explained by the fact that the previously published studies did not analyze preoperative stenting of hydronephrosis, thus inadvertently indicating that hydronephrosis is an independent risk factor for UUT recurrences. Our finding that only double J stenting in response to hydronephrosis is an independent risk factor for UUT recurrences, while hydronephrosis per se is not, was confirmed by an additionally performed subgroup analysis of those 226 RC patients who presented with hydronephrosis preoperatively (Table 4).

9

ACCEPTED MANUSCRIPT In our series, CIS was not an independent risk factor for UUT recurrence; a finding which contrasts with the results of previous studies [12,13]. Interestingly, concomitant CIS in the definitive histology was even significantly less frequent in the ‘double J’ group (p=0.02). Nevertheless, UUT recurrences were more frequent in this group, thus strengthening our

RI PT

findings that CIS per se is not an independent risk factor for UUT recurrence. The previously broadly accepted crucial role of concomitant CIS in the development of UUT recurrences appears to be more complex than once thought and needs to be further investigated regarding

ureteral orifice) and previous intravesical instillations.

SC

its association with urinary drainage, vesicoureterorenal reflux (e.g. after resection of the

M AN U

Patients in the ‘double J’ and ‘nephrostomy’ groups received [neo]adjuvant chemotherapy significantly more often than patients in the ‘no drainage’ group (p<0.0001). In addition, we found that patients who required any urinary drainage before RC showed significantly higher tumor stages and grading, higher Charlson Comorbidity Index (CCI) scores, and lower body

TE D

mass index percentiles compared to patients without drainage. This indicates that these patients have a more consuming disease and advanced tumor burden, probably caused by a more unfavorable tumor biology. In line with these findings, CSS in patients after drainage of

EP

the UUT was significantly worse than in patients who did not need drainage prior to RC (5year CSS rate: 47% in the ‘nephrostomy’ group, 54% in the ‘double J’ group, 71% in the ‘no

AC C

drainage’ group). Because a nephrostomy is performed under local anesthesia and does not require any epidural or general anesthesia, patients with higher CCI scores who needed drainage of the UUT were more likely to have a nephrostomy than double J stenting. This observation together with a significantly worse tumor and lymph node stage explains why overall survival was worse in the ‘nephrostomy group’ than in the ‘double J group’. One may argue that patients in the ‘nephrostomy’ group had fewer UUT recurrences because they had significantly higher tumor stages and died before a UUT recurrence could occur as stated by others who found lower rates of UUT recurrences in patients with higher tumor 10

ACCEPTED MANUSCRIPT grading and staging [14]. As mentioned above, however, most UUT recurrences occurred within the first 24 months after RC (100% in the ‘double J’ group). Therefore, worse OS does not necessarily translate into a lower UUT recurrence rate (Figure 1). Although our study is a large single center study and is the first to address the influence of

RI PT

different types of UUT drainage prior to RC on UUT recurrences, it is not without limitations. Besides the inherent shortcomings of any retrospective study, patients with hydronephrosis were not prospectively randomized to receive either double J or nephrostomy drainage.

SC

Another point of criticism might be that we also included patients with BC other than urothelial carcinoma. Still, the number of such patients was low (<6%), did not differ between

AC C

EP

TE D

M AN U

the different groups, and thus had no relevant influence on our results.

11

ACCEPTED MANUSCRIPT Conclusions Unlike all the other parameters that may cause UUT recurrences after RC, the choice of urinary drainage in patients with hydronephrosis and the need for drainage of the UUT prior to RC can be actively influenced by the urologist. Because double J stenting prior to RC was

RI PT

found to be an independent risk factor for UUT recurrence, retrograde manipulation in patients with BC in situ should be done only reluctantly. If preoperative drainage of the UUT is deemed necessary, percutaneous nephrostomy - if feasible – might be the preferred

SC

technique in these patients.

1

M AN U

References

Leibovitch I, Ben-Chaim J, Ramon J et al: The significance of ureteral obstruction in invasive transitional cell carcinoma of the urinary bladder. J Surg Oncol 1993; 52: 31.

2.

Bartsch GC, Kuefer R, Gschwend JE et al: Hydronephrosis as a prognostic marker in bladder cancer in a cystectomy-only series. Eur Urol 2007; 51: 690. Chapman DM, Pohar KS, Gong MC et al: Preoperative hydronephrosis as an indicator

TE D

3.

of survival after radical cystectomy. Urol Oncol 2009; 27: 491. Leahy AL, Ryan PC, McEntee GM et al: Renal injury and recovery in partial ureteric

EP

4.

obstruction. J Urol 1989; 142: 199. Bajorin DF, Herr HW. Kuhn's paradigms: are those closest to treating bladder cancer

AC C

5.

the last to appreciate the paradigm shift? J Clin Oncol 2011; 29: 2135.

6.

Ku JH, Lee SW, Jeon HG et al: Percutaneous nephrostomy versus indwelling ureteral

stents in the management of extrinsic ureteral obstruction in advanced malignancies:

are there differences? Urology 2004; 64: 895. 7.

Wong LM, Cleeve LK, Milner AD et al: Malignant ureteral obstruction: outcomes after intervention. Have things changed? J Urol 2007; 178: 178.

12

ACCEPTED MANUSCRIPT 8.

Studer UE, Burkhard FC, Schumacher M et al: Twenty years experience with an ileal orthotopic low pressure bladder substitute-lessons to be learned. J Urol 2006; 176: 161.

9.

Giannarini G, Kessler TM, Thoeny HC et al: Do patients benefit from routine follow-

RI PT

up to detect recurrences after radical cystectomy and ileal orthotopic bladder substitution? Eur Urol 2010; 58: 486. 10.

Meissner C, Giannarini G, Schumacher MC et al: The efficiency of excretory

SC

urography to detect upper urinary tract tumors after cystectomy for urothelial cancer. J Urol 2007; 178: 2287.

Palou J, Fariña LA, Villavicencio H et al: Upper tract urothelial tumor after

M AN U

11.

transurethral resection for bladder tumor. Eur Urol 1992; 21: 110. 12.

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:

13.

TE D

2046.

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. Gakis G, Black PC, Bochner BH et al: Systematic review on the fate of the remnant

EP

14.

urothelium after radical cystectomy. Eur Urol 2017; 71: 545. Gakis G, Efstathiou J, Lerner SP. ICUD-EAU International Consultation on Bladder

AC C

15.

Cancer 2012: Radical cystectomy and bladder preservation for muscle-invasive urothelial carcinoma of the bladder. Eur Urol 2013;63:45.

16.

Linder BJ, Boorjian SA, Hudolin T et al: Late recurrence after radical cystectomy: patterns, risk factors and outcomes. J Urol 2014; 191: 1256.

17.

Sidransky D, Frost P, Von Eschenbach A et al: Clonal origin bladder cancer. N Engl J Med 1992; 326: 737.

13

ACCEPTED MANUSCRIPT 18.

Lunec J, Challen C, Wright C et al: C-erbB-2 amplification and identical p53 mutations in concomitant transitional carcinomas of renal pelvis and urinary bladder. Lancet 1992; 339: 439. Habuchi T, Takahashi R, Yamada H et al: Metachronous multifocal development of

EP

TE D

M AN U

SC

RI PT

urothelial cancers by intraluminal seeding. Lancet 1993; 342: 1087.

AC C

19.

14

ACCEPTED MANUSCRIPT Nephrostomy group (n=61)

634 (71) 257 (29) 68.6 (61.6 – 75.8) 38.4 (14.4 – 92.4)

43 (81) 10 (19) 66.0 (58.2 – 75.5) 21.6 (9.4 – 55.0)

37 (61) 24 (39) 72.4 (59.4 – 76.0) 20.4 (9.6 – 43.2)

667 (75) 224 (25) 532 (60) 25.9 (23.0 – 28.9) 112 (13)

44 (64) 19 (36) 31 (58) 25.1 (22.9 – 27.5) 53 (100)

72 (8) 131 (15) 278 (31) 326 (37) 84 (9)

6 (11) 9 (17) 12 (23) 19 (36) 7 (13)

1 (2) 1 (2) 13 (21) 29 (48) 17 (28)

659 (74) 232 (26)

37 (70) 16 (30)

32 (52) 29 (48)

36 (4) 855 (96) 419 (47)

2 (4) 51 (96) 15 (28)

0 (0) 61 (100) 25 (41)

43 (81) 5 (9) 2 (4) 0 (0) 3 (6) 0 (0) 3 (6) 2 (4)

48 (79) 5 (8) 2 (3) 1 (2) 2 (3) 2 (3) 3 (5) 3 (5)

EP

AC C

RI PT

Double J group (n=53)

TE D 767 (86) 66 (7) 37 (4) 16 (2) 13 (1) 13 (1) 33 (4) 12 (1)

34 (56) 27 (44) 40 (66) 24.2 (22.0 – 27.8) 61 (100)

p value 0.05

0.5 0.0002 0.002

0.6 0.05 <0.0001 <0.0001

SC

Gender; n (%) male female Age; yrs, median (IQR) Follow up; mos, median (IQR) Charlson Comorbidity Index score; n (%) ≤2 >2 Smoker (former or current); n (%) BMI; median (IQR) Preoperative hydronephrosis; n (%) Tumor stage; n (%) pTa/pTis pT1 pT2 pT3 pT4 N stage; n (%) N0 N+ Tumor grade; n (%) G1/G2 G3 Concomitant CIS; n (%) Histological type; n (%) Pure urothelial carcinoma Urothelial carcinoma with variant histology Squamous differentiation Sarcomatoid differentiation Other Neuroendocrine carcinoma Pure squamous-cell carcinoma Other histological type Urinary diversion; n (%) Orthotopic bladder substitute Ileal conduit Catheterizable pouch Other Chemotherapy; n (%) Neoadjuvant Adjuvant Prior intravesical instillation; n (%) BCG Mitomycin/Epirubicin

No drainage group (n=891)

M AN U

Table 1 - Patient characteristics

0.001

0.6

0.02 0.3

<0.0001 484 (54) 345 (39) 43 (5) 19 (2)

29 (55) 22 (42) 1 (2) 1 (2)

10 (16) 42 (69) 6 (10) 3 (5)

114 (13) 91 (10)

16 (30) 8 (15)

21 (34) 9 (15)

125 (14) 60 (7)

9 (17) 4 (8)

3 (5) 1 (2)

<0.0001

0.1

IQR: interquartile range; BMI: body mass index; CIS: carcinoma in situ; BCG: Bacille Calmette-Guérin

ACCEPTED MANUSCRIPT Table 2 - Cox regression analysis (univariate) of variables correlating with UUT recurrence (n = 1005 patients) P value 0.8 0.9

0.32 7.09 0.05 1.23 1.43 0.65 1.59 2.91

0.14-0.76 3.03-16.61 0.01-71.87 0.87-1.74 0.95-2.16 0.20-2.18 0.72-3.05 1.40-6.06

0.01 < 0.001 0.4 0.2 0.1 0.5 0.3 0.004

AC C

EP

TE D

UUT: upper urinary tract; CIS: carcinoma in situ

0.36-1.47 0.07-3.82 0.70-2.96 0.00-262.4 0.86-3.60 1.15-4.88

M AN U

0.72 0.52 1.44 0.05 1.76 2.37

RI PT

95% CI 0.97-1.04 0.65-1.44

SC

Age Gender (female) Drainage No drainage Double J Nephrostomy T stage N stage G grading CIS Hydronephrosis Urinary diversion Neobladder Continent pouch Ileal conduit Ureter cutaneostomy Chemotherapy Intravesical instillation(s)

HR 1.00 0.96

0.4 0.5 0.3 0.8 0.1 0.02

ACCEPTED MANUSCRIPT Table 3 - Cox analysis (multivariate) of parameters associated with UUT recurrence (n = 1005 patients) HR

95% CI

P value

Drainage No drainage Double J Hydronephrosis Intravesical instillation(s)

0.63 4.54 1.89 2.94

0.20-1.99 1.43-14.38 0.69-5.18 1.40-6.16

0.4 0.01 0.21 0.004

RI PT

Variable

AC C

EP

TE D

M AN U

SC

UUT: upper urinary tract; CIS: carcinoma in situ

ACCEPTED MANUSCRIPT Table 4 - Cox regression (univariate) analysis of variables correlating with UUT recurrence in patients with hydronephrosis (n= 226 patients) P value 0.8 0.25

0.72 4.77 0.03 1.2 0.97 0.27 2.6

0.23-2.26 1.51-15.04 0.00-6.48 0.67-2.22 0.52-1.84 0.04-2.02 0.83-8.21

0.6 0.008 0.2 0.5 0.9 0.2 0.1

0.31-3.10 0.14-8.37 0.25-2.48 0.03-3.26 0.41-3.94 1.84-18.24

1 0.9 0.7 0.8 0.7 0.003

AC C

EP

TE D

UUT: upper urinary tract; CIS: carcinoma in situ

M AN U

0.99 1.08 0.79 0.05 1.27 5.79

RI PT

95% CI 0.96-1.06 0.30-1.37

SC

Age Gender (female) Drainage No drainage Double J Nephrostomy T stage N stage G grading CIS Urinary diversion Neobladder Continent pouch Ileal conduit Ureter cutaneostomy Chemotherapy Intravesical instillation(s)

HR 1.01 0.64

ACCEPTED MANUSCRIPT Figure legends

Table 1 Baseline characteristics of 1005 patients with bladder cancer who underwent radical cystectomy (RC) and urinary diversion (UD) Table 2 Cox regression analysis (univariate) of variables correlating with UUT

RI PT

recurrence (n = 1005 patients) (signifcant variables are shown in bold letters) Table 3 Cox analysis (multivariate) of parameters associated with UUT recurrence (n = 1005 patients) (signifcant variables are shown in bold letters)

Table 4 Cox regression (univariate) analysis of variables correlating with UUT

SC

recurrence in patients with hydronephrosis (n=226 patients) (signifcant variables are

M AN U

shown in bold letters)

Figure 1A UUT-RFS after radical cystectomy depending on different drainage or nodrainage methods, (Kaplan-Meier method) Figure 1B CSS after radical cystectomy depending on different drainage or no-drainage methods, (Kaplan-Meier method)

AC C

EP

TE D

Figure 1C OS after radical cystectomy depending on different drainage or no-drainage methods, (Kaplan-Meier method)

Figure 1A

p=0.3*

M AN U

100 p<0.001*

80 60 40

no drainage double J nephrostomy

TE D

Percent survival (%)

SC

Upper urinary tract (UUT) recurrence-free survival

RI PT

ACCEPTED MANUSCRIPT

20 0 24

48

72

EP

0

96

120 144 168 192

AC C

months survival

0

6

12

24

36

48

60

120

nephrostomy double J no drainage

61 53 891

54 45 783

42 36 692

26 23 566

19 19 458

13 16 390

13 13 331

3 3 134

*log rank test compared to no drainage (months)

patients at risk (n)

Figure 1B

SC

Cancer-specific survival (CSS)

80 p=0.03*

60 40

M AN U

100

no drainage double J nephrostomy

p<0.001*

20 0 24

48

72

96

120 144 168 192

EP

0

TE D

Percent survival (%)

RI PT

ACCEPTED MANUSCRIPT

months survival

AC C

*log rank test compared to no drainage (months)

patients at risk (n)

nephrostomy double J no drainage

0

6

12

24

36

48

60

120

61 53 891

54 47 786

42 39 697

26 26 572

19 19 465

13 16 395

13 13 337

3 3 137

RI PT

ACCEPTED MANUSCRIPT

Figure 1C

SC

Overall survival (OS)

M AN U

80 60

no drainage double J nephrostomy

p=0.7*

40 p<0.001* 20 0 24

48

72

96

120 144 168 192

EP

0

TE D

Percent survival (%)

100

months survival

AC C

*log rank test compared to no drainage (months)

patients at risk (n)

nephrostomy double J no drainage

0

6

12

24

36

48

60

120

61 53 891

54 47 786

42 39 697

26 26 572

19 19 465

13 16 395

13 13 337

3 3 137

ACCEPTED MANUSCRIPT

AC C

EP

TE D

SC

M AN U

RC: radical cystectomy BC: bladder cancer UUT: upper urinary tract RFS: recurrence-free survival CSS: cancer-specific survival OS: overall survival IQR: interquartile range HR: Hazard ratio CI: confidence interval IVP: intravenous pyelography CT: computed tomography CIS: carcinoma in situ NMIBC: non muscle-invasive bladder cancer CCI: Charlson Comorbidity Index

RI PT

Key of Definitions for Abbreviations