- Email: [email protected]
The effect of tumor location on oncologic outcomes in patients with upper urinary tract urothelial carcinoma stratified by pathologic stage
Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
Original article
The effect of tumor location on oncologic outcomes in patients with upper urinary tract urothelial carcinoma stratified by pathologic stage Yi-Sheng Tai, M.D.a, Chung-Hsin Chen, M.D., Ph.D.b,*, Chao-Yuan Huang, M.D., Ph.D.b, Huai-Chin Tai, M.D., Ph.D.b, Sho-Mon Wang, M.D., Ph.D.b, Yeong-Shiau Pu, M.D., Ph.D.b a
b
Department of Urology, National Taiwan University Hospital Yun-Lin Branch, Yun-Lin, Taiwan Department of Urology, National Taiwan University Hospital and College of Medicine, Taipei, Taiwan Received 1 May 2015; received in revised form 9 August 2015; accepted 10 August 2015
Abstract Objectives: The aim of this study was to investigate the effect of tumor location on oncologic outcomes in patients with upper urinary tract urothelial carcinoma (UTUC) stratified according to pathologic stage. Material and methods: Between January 1996 and March 2009, 503 patients with UTUC were enrolled who had undergone radical nephroureterectomies and had no nodal or distal metastases. Preoperative chemotherapy or radiation therapy was not administered to any study patient. Clinicopathologic patient characteristics were obtained and used to analyze recurrence-free survival (RFS), cancer-specific survival, and overall survival with the Cox proportional hazards model. Results: During the median follow-up of 52 months, patients with pathologic tumor (pT) stage 3 ureteral tumors had a shorter duration of RFS compared with those with pT3 renal pelvis tumors (5-y RFS: 50% and 71%, P ¼ 0.047). There was no prognostic relevance to the tumor location in pTa/Tis/T1 and pT2 diseases. RFS and cancer-specific survival were significantly shorter in duration in pT3 ureteral disease compared with pT2 diseases (P o 0.001 and P ¼ 0.028). No differences were found in oncologic outcomes between pT3 renal pelvic and pT2 diseases. The presence of pT3 ureteral tumors actually increased the risks of disease recurrence (hazard ratio [HR] ¼ 7.82, P o 0.001), cancer-specific death (HR ¼ 5.08, P o 0.001), and overall mortality (HR ¼ 3.25, P ¼ 0.031). Conclusions: Patients with UTUC and pT3 ureteral tumors had an increased risk of disease recurrence and cancer-specific death. These results underscore the need for close follow-up and the consideration of adjuvant chemotherapy for patients with pT3 ureteral cancer. r 2016 Elsevier Inc. All rights reserved.
Keywords: Upper urinary tract urothelial carcinoma; Ureteral tumor; Tumor stage; disease recurrence
Introduction Upper urinary tract urothelial carcinoma (UTUC) is a rare disease, accounting for o5% of all urothelial carcinomas worldwide [1]. The incidence of UTUC in Taiwan is unusually higher than that in other countries and comprises 20% to 33% of all urothelial cancers [2,3]. The prognostic effect of primary tumor location on oncologic outcomes in patients with UTUC is still controversial. In some study cohorts, ureteral locations have * Corresponding author. Tel.: þ886-2-2312-3456, ext: 5952; fax: þ8862-2321-9145. E-mail address: [email protected] (C.-H. Chen).
http://dx.doi.org/10.1016/j.urolonc.2015.08.006 1078-1439/r 2016 Elsevier Inc. All rights reserved.
been found to be significantly associated with worse recurrence-free survival (RFS). Ureteral tumors are an independent risk factor that indicates a poor prognosis for cancer-specific survival (CSS) and RFS in patients with pathologic tumor stage (pT) 3 UTUCs [4,5]. Renal pelvic tumors that invade the renal parenchyma have been associated with a better prognosis than tumors invading peripelvic or periureteral fat [6]. Some studies have not reported any differences in oncologic outcomes between patients with ureteral and renal pelvis tumors [7–9]. In this study, we investigated the effect of tumor location on oncologic outcomes in patients with UTUC by stratifying patients according to pT stage.
4.e20
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
Material and methods Patient population and enrollment criteria Between January 1996 and March 2009, 598 consecutive Taiwanese patients with UTUC who underwent radical nephroureterectomies (RNUs) at our institution were enrolled. Of the 598 patients, 95 were excluded for the following reasons: incomplete tumor stage (11), bladder urothelial carcinoma before RNU (33), radical cystectomies with concomitant RNU (7), previous systemic chemotherapy, including neoadjuvant therapy for UTUC (7), previous radiotherapy to the pelvis (2), and nodal metastases (35). This study focused on the remaining 503 patients. All 503 RNUs were performed by either the open or laparoscopic method. Bladder cuff resections were obtained through the open approach. The hilar and regional lymph nodes were resected if enlarged or if palpable nodes were identified preoperatively or intraoperatively. There was no adjuvant chemotherapy performed for pTa-3 diseases postoperatively (which included 495 patients). All of the 8 patients with pT4 disease underwent cisplatin-based chemotherapy after RNU and were excluded from the survival analyses. The Research Ethics Committee of the National Taiwan University Hospital (201304027RINC) approved this study.
6 months for 2 years, and annually afterward. Computed tomography or magnetic resonance imaging or both were performed annually to detect local recurrence, contralateral UTUC occurrence, and distal metastasis. Disease recurrence was defined as tumor relapse in the operative field, regional lymph nodes, or distant metastasis. Cisplatin-based chemotherapy is the primary treatment of all types of recurrent disease. A physician determined each patient's cause of death. All patients who died of UTUC had previous disease recurrence. Statistical methods The chi-square test was used to analyze associations between clinicopathologic parameters. The log-rank test was used to compare RFS, CSS, and overall survival (OS) periods among patient groups. The Cox proportional hazards model was used to evaluate the risk of oncologically adverse events. Prognosticators used in the multivariate model were designated from factors with a P o 0.15 in the univariate analysis. A P o 0.05 was considered statistically significant. Results Population characteristics
Clinical data collection Patient medical records were used to obtain information about age, sex, diabetes mellitus status, hypertension status, serum creatinine levels, dialysis history, herbal medications, smoking status, and pathologic characteristics of UTUC tumors. Estimated glomerular filtration rate was calculated using the Modification in Diet in Renal Disease Study equation [10]. The tumor location was defined as the dominantly advanced lesion at the ureteral or renal pelvis according to pathology reports. In all, 39 patients with equally advanced tumors at both ureteral and renal pelvis locations were excluded from the analyses of the association between the tumor locations and oncologic outcomes. Tumor multifocality was defined as the synchronous appearance of Z2 pathologically confirmed tumors in the urinary tract. Heavy smokers were defined as people who had a cumulative smoking history of 20 pack-years or more. Pathologists at our institution reviewed all surgical specimens. Tumors were graded according to 2004 World Health Organization classifications and staged based on 2002 TNM classifications.
Patient demographics and clinical features stratified by tumor location are listed in Table 1. Our cohort included 249 men (49.5%) and 254 women (50.5%). The median age of all patients was 68 years (interquartile range [IQR]: 60.0– 74.8 y). Overall, 280 patients had renal pelvic tumors (55.7%) and 184 patients had ureteral tumors (36.6%). The remaining 39 patients had the same T stage tumors at both the renal pelvis and ureter (7.7%). Ureteral and combined pelvic and ureteral tumors are more likely associated with high-grade disease compared with renal pelvic tumors alone (64.1%, 60%, and 50.7%, respectively; P ¼ 0.012). It was also more common for patients with ureteral tumors to have advanced chronic kidney disease stages (3, 4, and 5) than patients with renal pelvic and combined tumors (69.1%, 50.3%, and 48.7%, respectively; P o 0.001). The presence of locally advanced disease (pT3/T4) was more frequent in renal pelvic tumors (37.5%) than in the combined locations (30.7%) and ureteral tumors (20.2%) (P o 0.001). No significant difference was found in the rest of the clinicopathologic characteristics among patient groups (Table 1). Patient survival
Follow-up protocol After RNU, all patients were followed regularly with cystoscopic examinations, urine cytologic tests, and renal sonography every 3 months in the first year, then every
In all, 495 patients with pTa-T3N0M0 diseases were enrolled for the survival analyses. Table 2 demonstrates disease recurrence, cancer-specific death, and overall mortality rates stratified by tumor stages and locations.
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25 Table 1 Clinicopathologic characteristics of patients with upper urinary tract urothelial carcinoma Characteristics
Number of patients, n (%)
Tumor location Ureter
Both
280 (55.7%)
184 (36.6%) 39 (7.7%)
Age, y Median (IQR) o70 Z70 Gender, n (%) Female Male
Disease recurrence Renal pelvis Ureter Both 0.623
67 (60–73.8) 69 (60–74) 66 (59–75) 166 (59.3%) 105 (57.1%) 23 (60.0%) 114 (40.7%) 79 (42.9%) 16 (40.0%)
0.891 0.296
133 (47.5%) 147 (52.5%)
101 (54.9%) 20 (51.3%) 83 (45.1%) 19 (48.7%)
Pathologic grade, n (%) Low 135 (48.2%) High 142 (50.7%)
63 (34.2%) 15 (38.5%) 118 (64.1%) 23 (60%)
Pathologic tumor stage Ta/Tis/T1 144 (51.4%) T2 31 (11.1%) T3 101 (36.1%) T4 4 (1.4%)
pTa/Tis/T1 (N ¼ 251)
pT2 (N ¼ 98)
pT3 (N ¼ 146)
15/250 (6%) 8/143 (5.6%) 6/87 (6.9%) 1/20 (5.0%)
17/97 (17.5%) 46/144 (31.9%)a 3/30 (10.0%) 27/100 (27%) 13/60 (21.7%) 16/32 (50.0%) 1/7 (14.3%) 3/12 (25.0%)
Cancer-specific death 13/ 250 (5.2%) Renal pelvis 8/143 (5.6%) Ureter 5/87 (5.7%) Both 0/20
12/95 (12.6%) 29/146 (19.9%) 3/29 (10.3%) 17/101 (16.8%) 9/60 (15%) 9/33 (27.3%) 0/6 3/12 (25.0%)
Overall mortality Renal pelvis Ureter Both
22/95 (23.2%) 36/146 (24.7%) 8/29 (27.6%) 22/101 (21.8%) 13/60 (21.7%) 10/33 (30.3%) 1/6 (16.6%) 4/12 (33.3%)
22/250 (8.8%) 12/143 (8.4%) 8/87 (9.2%) 2/20 (10.0%)
0.012
87 60 33 4
a
Two patients with pT3 disease died perioperatively and were excluded from the analysis of disease recurrence.
(47.3%) 20 (51.3%) o0.001 (32.6%) 7 (18%) (18.0%) 12 (30.7%) (2.2%) 0
LVI, n (%) Absent Present
184 (65.7%) 49 (17.5%)
126 (68.5%) 22 (56.4%) 32 (17.4%) 6 (15.4%)
0.979
Herbal usage, n (%) No Yes
202 (72.1%) 20 (7.1%)
129 (70.1%) 24 (61.5%) 16 (8.7%) 6 (15.4%)
Heavy smoker, n (%) No 234 (83.6%) Yes 42 (15.0%)
154 (83.7%) 29 (74.4%) 29 (15.8%) 9 (23.1%)
Renal function, n (%) CKD stage 0–2a 95 (33.9%) CKD stage 3–4 121 (43.2%) CKD stage 5 20 (7.1%)
38 (20.7%) 9 (23.1%) 98 (53.3%) 11 (28.2%) 29 (15.8%) 8 (20.5%)
0.191 0.409
stages had more disease recurrence. The 5-year RFS rates were 92%, 79%, and 65% in pTa/Tis/T1, pT2, and pT3 disease, respectively (P o 0.001). In the univariate analyses for disease recurrence, highgrade disease (hazard ratio [HR] ¼ 2.44, P o 0.001), advanced tumor stage (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 3.45 and 6.75, P o 0.001 and o0.001, respectively), and lymphovascular invasion (HR ¼ 1.81, P ¼ 0.03) significantly increased the risk of disease recurrence. The multivariate analysis shows that only advanced tumor stage (pT2 and pT3 vs. pTa/Tis/T1, HR ¼ 2.52 and 5.64, P ¼ 0.03 and o0.001, respectively) was an independent risk factor predicting disease recurrence after RNU (Table 3).
o0.001
Diabetes mellitus, n (%) No 228 (81.4%) Yes 52 (18.6%)
144 (78.3%) 34 (87.2%) 39 (21.2%) 5 (12.8%)
Hypertension, n (%) No Yes
108 (58.7%) 26 (66.7%) 74 (40.2%) 13 (33.3%
Cancer-specific survival
0.444
0.582 177 (63.2%) 102 (36.4%)
Table 2 Oncologic outcomes stratified by tumor stage and location
P values
Renal pelvis
4.e21
LVI ¼ lymphovascular invasion; T ¼ tumor stage. CKD stage 0 was defined as normal renal function.
a
Recurrence-free survival During a median follow-up of 52 months (IQR: 23– 77 mo), disease recurrence was discovered in 78 patients (15.8%), consisting of 15 (6%), 17 (17.5%), and 46 (31.9%) patients with pTa/Tis/T1, pT2, and pT3 diseases, respectively. The median interval between RNU and disease recurrence was 9 months (IQR: 4–32 mo) for those with disease recurrence; 73.8% of disease recurrences were identified within 2 years after RNU. Patients with advanced
In this cohort, 54 patients (10.9%) died of UTUCs, including 13 (5.2%), 12 (12.6%), and 29 (19.9%) patients with pTa/Tis/T1, pT2, and pT3 disease, respectively. Advanced tumor stages were associated with poor CSS. The 5-year CSS rates were 94%, 84%, and 75% in patients in the pTa/Tis/T1, pT2, and pT3 groups, respectively (P o 0.001). In the univariate analyses for disease recurrence, highgrade disease (HR ¼ 2.54, P ¼ 0.002) and advanced tumor stage (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 2.76 and 4.72, p ¼ 0.011 and o0.001, respectively) significantly increased the risk for cancer-specific death. In the multivariate analysis, age Z70 years (HR ¼ 3.47, P ¼ 0.003) and pT3 disease (reference: pTa/Tis/T1 tumor, HR ¼ 4.11, P ¼ 0.004) were independent risk factors predicting cancerspecific mortality in patients with UTUC (Table 3). Overall survival During follow-up, a total of 80 patients (16.2%) died of any cause, including 22 (8.8%), 22 (23.2%), and 36 (24.7%) patients with pTa/Tis/T1, pT2, and pT3 disease, respectively.
4.e22
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
Table 3 Cox proportional hazards model for predicting oncologic outcomesa Variables
Univariate analysis
Multivariate analysis
HR
95% CI
P values
HR
95% CI
P values
Disease recurrence Grade Low High
Reference 2.44
– 1.49–3.99
– o0.001
Reference 1.31
– 0.70–2.44
– 0.395
Pathologic T stage Ta/Tis/T1 T2 T3
1.84 Reference 3.46 6.75
1.53–2.20 – 1.72–6.92 3.76–12.09
o0.001 – o0.001 o0.001
Reference 2.52 5.64
– 1.10–5.81 2.74–11.61
– 0.03 o0.001
LVI Absent Present
Reference 1.81
– 1.06–3.08
– 0.030
Reference 1.05
– 0.56–1.95
– 0.886
Tumor location Pelvis Ureter
Reference 1.48
– 0.942.33
– 0.090
Reference 1.33
– 0.74–2.40
– 0.338
Reference 1.62
– 0.96–2.73
– 0.071
Reference 3.47
– 1.53–7.87
– 0.003
Reference 2.54
– 1.41–4.58
– 0.002
Reference 1.93
– 0.81–4.64
– 0.140
Pathologic T stage Ta/Tis/T1 T2 T3
Reference 2.76 4.72
– 1.26–6.07 2.45–9.09
– 0.011 o0.001
1.69 4.11
0.52–5.51 1.59–10.65
0.386 0.004
Tumor location Pelvis Ureter
Reference 1.43
– 0.83–2.45
– 0.196
Reference 0.90
– 0.36–2.25
– 0.828
Overall mortality Age o70 Z70
Reference 1.95
– 1.27–2.99
– 0.002
Reference 2.66
– 1.48–4.75
– 0.001
Grade Low High
Reference 2.33
– 1.44–3.77
– 0.001
Reference 1.67
– 0.87–3.20
– 0.122
Pathologic T stage Ta/Tis/T1 T2 T3
Reference 2.90 3.43
– 1.60–5.24 2.03–5.87
– o0.001 o0.001
Reference 3.55 3.01
– 1.66–7.57 1.42–6.38
– 0.001 0.004
Tumor location Pelvis Ureter
Reference 1.40
– 0.82–2.39
– 0.222
0.76
0.42–1.47
0.449
Cancer-specific death Age o70 Z70 Grade Low High
CKD ¼ chronic kidney disease; LVI, lymphovascular invasion; T stage ¼ tumor stage. Bold values statistically signifies p value o0.05. a Only significant results in univariate or multivariate analyses were listed.
The 5-year CSS rates were 90%, 74%, and 71% for the pTa/ Tis/T1, pT2, and pT3 groups, respectively (P o 0.001). Univariate analyses showed that age Z70 years (HR ¼ 1.95, P ¼ 0.002), high-grade disease (HR ¼ 2.33, P ¼ 0.001), and advanced diseases (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 2.90 and 3.43, P o 0.001 and o0.001,
respectively) significantly increased the risk of overall mortality. In the multivariate analysis, age Z70 years (HR ¼ 2.66, P ¼ 0.001) and advanced diseases (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 3.55 and 3.01, P ¼ 0.001 and 0.004, respectively) were independent risk factors for predicting OS (Table 3).
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
4.e23
Table 4 The 5-year survival rates stratified by tumor stage and location Tumor stage/ location
RFS, %
pTa/Tis/T1 Renal pelvis Ureter pT2 Renal pelvis Ureter pT3 Renal pelvis Ureter
92 92 92 79 87 75 65 71 50
P values 0.645
0.101
0.047
CSS, % 94 93 94 84 88 80 75 80 60
P values 0.872
0.381
0.164
OS, % 90 90 91 74 75 73 71 75 59
P values 0.741
0.908
0.237
Tumor location effect of each stage Table 4 demonstrates the 5-year rates for RFS, CSS, and OS stratified by tumor stage and location. In patients with pTa/Tis/T1 disease, the 5-year RFS, CSS, and OS were greater than 90% despite tumors being located at the renal pelvis or ureter. Our study cohort had a trend toward more advanced tumor stages being associated with the prognostic relevance of the ureteral tumor, although this was significantly different only in pT3 disease recurrence rates. During follow-up, pathologic T3 ureteral disease had a higher frequency of disease recurrence (50% vs. 27%) and a shorter duration of RFS than did renal pelvic tumors (logrank test, P ¼ 0.047). Tumor location had no significant effect on Ta/Tis/T1 and T2 diseases. In the subgroup analysis, the tumor stages and locations were reported as pTa/Tis/T1, pT2, pT3 renal pelvic tumor and pT3 ureteral tumor groups for a new model for survival analyses. Patients with pT3 ureteral tumors had a significantly shorter duration of RFS and CSS compared with patients with pT2 diseases (log-rank test, P o 0.001 and P ¼ 0.028, respectively). There were no differences between pT3 renal pelvic tumors and pT2 diseases in RFS and CSS (log-rank test, P ¼ 0.111 and P ¼ 0.285) (Figs. 1 and 2). Table 5 demonstrates the independent risk factors that predict oncologic outcomes. The pT3 ureteral tumors actually increased the risk of disease recurrence (reference: pTa/Tis/T1 tumor, HR ¼ 7.82, P r 0.001), cancer-specific death (reference: pTa/Tis/T1 tumor, HR ¼ 5.08, P r 0.001), and overall mortality (reference: pTa/Tis/T1 tumor, HR ¼ 3.25, P ¼ 0.031) in the multivariate analysis.
Fig. 1. Kaplan-Meier survival curve for recurrence-free survival stratified by tumor stage (pT3 ureteral vs. pT2 disease, P o 0.001; pT3 renal pelvic vs. pT2 disease, P ¼ 0.111).
demonstrated the results of 5-year RFS, CSS, and OS rates stratified by tumor stage and location. Our findings show that patients with pT3 ureteral tumors have significantly poorer RFS and CSS outcomes compared with outcomes in patients with either pT2 or pTa/Tis/T1 disease. Several studies have demonstrated that pathologic tumor grade is a powerful predictor of cancer-control outcomes [12]. This was not confirmed in the multivariate analyses of the present study, because of the strong correlation found between tumor grades and stages. In our cohort, the highgrade tumor accounted for 44.2%, 63.90%, 64.3%, and 93.80% of pTa/Tis/T1, pT2, pT3 renal pelvic disease, and pT3 ureteral disease, respectively. A few low-grade tumors combined with advanced-stage tumors were included in this study, leading to a relatively low statistical power for
Discussion The incidence of UTUC in Taiwan is the highest in the world, comprising 20% to 33% of all urothelial cancers [2,3]. Aristolochic acid, a component of aristolochia herbal remedies, is thought to be associated with the higher incidence of UTUC in Taiwan [11]. In this study, we investigated the risk factors and oncologic outcomes in patients with UTUC and
Fig. 2. Kaplan-Meier survival curve for cancer-specific survival stratified by tumor stage (pT3 ureteral vs. pT2 disease, P ¼ 0.028; pT3 renal pelvic vs. pT2 disease, P ¼ 0.285).
4.e24
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
Table 5 Prognosticators for oncologic outcomesa Variables
Disease recurrence Tumor stage/location pTa/Tis/T1 pT2 pT3 renal pelvis pT3 ureter Cancer-specific death Age o70 Z70 Tumor stage/location pTa/Tis/T1 pT2 pT3 renal pelvis pT3 ureter Overall mortality Age o70 Z70 Tumor stage/location pTa/Tis/T1 pT2 pT3 renal pelvis pT3 ureter
Multivariate analysis HR
95% CI
P values
Reference 3.28 5.22 7.82
– 1.47–7.29 2.47–11.04 3.17–19.29
–
Reference 3.44
– 1.52–7.79
–
Reference 1.72 4.33 5.08
– 0.56–5.35 1.60–11.70 1.28–20.21
–
Reference 2.46
– 1.40–4.31
–
Reference 3.15 2.98 3.25
– 1.57–6.35 1.38–6.43 1.11–9.45
–
0.004 o0.001 o0.001
0.003
0.346 0.004 0.021
0.002
0.001 0.006 0.031
a The tumor stages and locations refer to pTa/Tis/T1, pT2, pT3 renal pelvic, and pT3 ureteral cancer groups. Only significant results in multivariate analyses were listed.
identifying the difference in RFS, CSS, and OS in patients stratified by tumor grade. To date, the effect of tumor location on the prognosis in patients with UTUC is still controversial. Several studies have demonstrated that ureteral tumors have worse prognoses than renal pelvic tumors for disease recurrence and cancer-specific death. The ureteral tumor was an independent predictive factor for RFS and CSS after adjustment for pathologic variables [4,5]. Park et al. found the ureteral tumor to have prognostic relevance only in patients with pT3 disease. Renal pelvic tumors invading the renal parenchyma were associated with higher RFS and CSS rates than those invading peripelvic or periureteral fat, suggesting that renal parenchyma may play a protective role in preventing local tumors from spreading in patients with a pT3 tumor [6]. On the contrary, 2 population-based studies and 1 singleinstitutional study show that renal pelvic and ureteral tumors have the same RFS and CSS after several clinicopathologic parameters are adjusted, which confirms that tumor location is not an independent prognostic factor in oncologic outcomes [7–9]. In our study, the prognostic relevance of the ureteral tumor was investigated only in pT3 disease. The pT3 ureteral tumor had a higher frequency of disease recurrence
(50% vs. pT3 renal pelvic tumor, 27%) and a shorter duration of RFS than did renal pelvic tumors (log-rank test, P = 0.047). The oncologic outcomes of pT3 ureteral disease demonstrated 5-year rates of 50%, 60%, and 59% in RFS, CSS, and OS, respectively. In addition, pT3 ureteral tumors also had a significantly shorter duration of RFS and CSS compared with tumors in either of the pT2 diseases (logrank test, P o 0.001 and P = 0.028, respectively), but there were no differences between pT3 renal pelvic tumors and pT2 diseases regarding RFS and CSS (log-rank test, P = 0.111 and P = 0.285). Advanced pT, especially pT3 ureteral tumors, was the indicating risk factor for predicting RFS, CSS, and OS in the multivariate analysis. To date, the currently published retrospective studies do not provide a definitive conclusion regarding the effect of tumor location on UTUC prognosis. The conflicting findings support the need for randomized prospective studies to investigate the difference in oncologic outcomes between patients with ureteral and renal pelvis tumors. In our cohort, all of the 8 patients with pathologically proven T4 disease underwent adjuvant chemotherapy after RNU, and only 1 patient (12.5%) developed disease recurrence afterward. Among the 32 chemonaïve patients with pT3 ureteral disease, 16 (50%) had disease recurrence and received systemic chemotherapy. The 5-year RFS rate was 71% in patients with pT4 disease, which was greater than the 50% in those with chemonaïve pT3 ureteral disease (log-rank test, P = 0.593; data not shown). The findings may suggest the need for adjuvant chemotherapy in patients with high-risk disease to improve patient prognosis. A systematic review and meta-analysis analyzed 1 prospective and 9 retrospective studies to investigate the effect of adjuvant chemotherapy in patients with UTUC after RNU. There appears to be an OS and RFS benefit with cisplatinbased adjuvant chemotherapy in UTUC [13]. Kawashima et al. [14] analyzed the results of adjuvant chemotherapy for localized pT3 UTUC in a multi-institutional study, which showed that adjuvant chemotherapy status was significantly associated with CSS (HR ¼ 0.205, P ¼ 0.008). On the contrary, these findings were not confirmed by Hellenthal et al. [15], who used the international collaborative database, which demonstrated that adjuvant chemotherapy after RNU did not significantly correlate with overall or CSS in patients with the individual high-risk diseases (pT3N0, pT4N0, and lymph node–positive disease). The current evidence is limited by the retrospective nature of studies and the absence of subgroup analyses stratified by both tumor stage and location. Adjuvant chemotherapy appears to be reasonable and feasible for the selective and high-risk patients, but more prospective trials are needed to confirm its efficacy. Our current study has several limitations. First, the data were collected in a retrospective manner and several potentially important covariates, such as operative methods, tumor size, and smoking status, were not evaluated. The renal pelvic tumors in pT3 and pT4 disease were not
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
divided into parenchymal and peripelvis invasion. Second, a relatively small sample size of patients was used in the analysis of the association between tumor stage/location and outcomes. Third, the relatively short follow-up duration prevented the observation of long-term survival outcomes. Fourth, the RNUs were performed by several surgeons at our institution, resulting in some variation among cases. No regular lymph node dissection was performed in this study, which may mean that information about certain microscopic metastatic diseases and their effect on oncologic outcomes was not recorded. However, there is a standard protocol for managing bladder cuff and follow-up after RNU at our institution. Therefore, the observational bias of RFS events would not systemically exist in this cohort. Conclusions Patients with UTUC and pT3 ureteral tumors had more frequent disease recurrence and a shorter duration of RFS than did those with pT3 renal pelvic diseases. The 5-year rates for RFS, CSS, and OS were 50%, 60%, and 59%, respectively, in patients with pT3 ureteral cancer. In addition, pT3 ureteral tumors also had a significantly shorter duration of RFS and CSS compared with that in patients with pT2 diseases; but there were no differences between pT3 renal pelvic tumors and pT2 diseases in RFS and CSS. Advanced pT, especially pT3 ureteral tumors, was the indicator of risk factors for oncologic outcomes. These results underscore the need for close follow-up and the consideration of adjuvant chemotherapy for patients with pT3 ureteral cancer. Ethical standards The article does not contain clinical studies or patient data. Acknowledgment The authors have nothing to disclose. References [1] Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1998. CA Cancer J Clin 1998;48:6–29.
4.e25
[2] Chou YH, Huang CH. Unusual clinical presentation of upper urothelial carcinoma in Taiwan. Cancer 1999;85:1342–4. [3] Yang MH, Chen KK, Yen CC, Wang WS, Chang YH, Huang WJ, et al. Unusually high incidence of upper urinary tract urothelial carcinoma in Taiwan. Urology 2002;59:681–7. [4] Akdogan B, Dogan HS, Eskicorapci SY, Sahin A, Erkan I, Ozen H. Prognostic significance of bladder tumor history and tumor location in upper tract transitional cell carcinoma. J Urol 2006;176: 48–52. [5] Ouzzane A, Colin P, Xylinas E, Pignot G, Ariane MM, Saint F, et al. Ureteral and multifocal tumours have worse prognosis than renal pelvic tumours in urothelial carcinoma of the upper urinary tract treated by nephroureterectomy. Eur Urol 2011;60:1258–65. [6] Park J, Ha SH, Min GE, Song C, Hong B, Hong JH, et al. The protective role of renal parenchyma as a barrier to local tumor spread of upper tract transitional cell carcinoma and its impact on patient survival. J Urol 2009;182:894–9. [7] Favaretto RL, Shariat SF, Chade DC, Godoy G, Adamy A, Kaag M, et al. The effect of tumor location on prognosis in patients treated with radical nephroureterectomy at Memorial Sloan-Kettering Cancer Center. Eur Urol 2010;58:574–80. [8] Isbarn H, Jeldres C, Shariat SF, Liberman D, Sun M, Lughezzani G, et al. Location of the primary tumor is not an independent predictor of cancer specific mortality in patients with upper urinary tract urothelial carcinoma. J Urol 2009;182:2177–81. [9] Raman JD, Ng CK, Scherr DS, Margulis V, Lotan Y, Bensalah K, et al. Impact of tumor location on prognosis for patients with upper tract urothelial carcinoma managed by radical nephroureterectomy. Eur Urol 2010;57:1072–9. [10] Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999;130: 461–70. [11] Chen CH, Dickman KG, Moriya M, Zavadil J, Sidorenko VS, Edwards KL, et al. Aristolochic acid-associated urothelial cancer in Taiwan. Proc Natl Acad Sci U S A 2012;109:8241–6. [12] Lughezzani G, Burger M, Margulis V, Matin SF, Novara G, Roupret M, et al. Prognostic factors in upper urinary tract urothelial carcinomas: a comprehensive review of the current literature. Eur Urol 2012;62: 100–14. [13] Leow JJ, Martin-Doyle W, Fay AP, Choueiri TK, Chang SL, Bellmunt J. A systematic review and meta-analysis of adjuvant and neoadjuvant chemotherapy for upper tract urothelial carcinoma. Eur Urol 2014;66:529–41. [14] Kawashima A, Nakai Y, Nakayama M, Ujike T, Tanigawa G, Ono Y, et al. The result of adjuvant chemotherapy for localized pT3 upper urinary tract carcinoma in a multi-institutional study. World J Urol 2012;30:701–6. [15] Hellenthal NJ, Shariat SF, Margulis V, Karakiewicz PI, Roscigno M, Bolenz C, et al. Adjuvant chemotherapy for high risk upper tract urothelial carcinoma: results from the Upper Tract Urothelial Carcinoma Collaboration. J Urol 2009;182:900–6.
Original article
The effect of tumor location on oncologic outcomes in patients with upper urinary tract urothelial carcinoma stratified by pathologic stage Yi-Sheng Tai, M.D.a, Chung-Hsin Chen, M.D., Ph.D.b,*, Chao-Yuan Huang, M.D., Ph.D.b, Huai-Chin Tai, M.D., Ph.D.b, Sho-Mon Wang, M.D., Ph.D.b, Yeong-Shiau Pu, M.D., Ph.D.b a
b
Department of Urology, National Taiwan University Hospital Yun-Lin Branch, Yun-Lin, Taiwan Department of Urology, National Taiwan University Hospital and College of Medicine, Taipei, Taiwan Received 1 May 2015; received in revised form 9 August 2015; accepted 10 August 2015
Abstract Objectives: The aim of this study was to investigate the effect of tumor location on oncologic outcomes in patients with upper urinary tract urothelial carcinoma (UTUC) stratified according to pathologic stage. Material and methods: Between January 1996 and March 2009, 503 patients with UTUC were enrolled who had undergone radical nephroureterectomies and had no nodal or distal metastases. Preoperative chemotherapy or radiation therapy was not administered to any study patient. Clinicopathologic patient characteristics were obtained and used to analyze recurrence-free survival (RFS), cancer-specific survival, and overall survival with the Cox proportional hazards model. Results: During the median follow-up of 52 months, patients with pathologic tumor (pT) stage 3 ureteral tumors had a shorter duration of RFS compared with those with pT3 renal pelvis tumors (5-y RFS: 50% and 71%, P ¼ 0.047). There was no prognostic relevance to the tumor location in pTa/Tis/T1 and pT2 diseases. RFS and cancer-specific survival were significantly shorter in duration in pT3 ureteral disease compared with pT2 diseases (P o 0.001 and P ¼ 0.028). No differences were found in oncologic outcomes between pT3 renal pelvic and pT2 diseases. The presence of pT3 ureteral tumors actually increased the risks of disease recurrence (hazard ratio [HR] ¼ 7.82, P o 0.001), cancer-specific death (HR ¼ 5.08, P o 0.001), and overall mortality (HR ¼ 3.25, P ¼ 0.031). Conclusions: Patients with UTUC and pT3 ureteral tumors had an increased risk of disease recurrence and cancer-specific death. These results underscore the need for close follow-up and the consideration of adjuvant chemotherapy for patients with pT3 ureteral cancer. r 2016 Elsevier Inc. All rights reserved.
Keywords: Upper urinary tract urothelial carcinoma; Ureteral tumor; Tumor stage; disease recurrence
Introduction Upper urinary tract urothelial carcinoma (UTUC) is a rare disease, accounting for o5% of all urothelial carcinomas worldwide [1]. The incidence of UTUC in Taiwan is unusually higher than that in other countries and comprises 20% to 33% of all urothelial cancers [2,3]. The prognostic effect of primary tumor location on oncologic outcomes in patients with UTUC is still controversial. In some study cohorts, ureteral locations have * Corresponding author. Tel.: þ886-2-2312-3456, ext: 5952; fax: þ8862-2321-9145. E-mail address: [email protected] (C.-H. Chen).
http://dx.doi.org/10.1016/j.urolonc.2015.08.006 1078-1439/r 2016 Elsevier Inc. All rights reserved.
been found to be significantly associated with worse recurrence-free survival (RFS). Ureteral tumors are an independent risk factor that indicates a poor prognosis for cancer-specific survival (CSS) and RFS in patients with pathologic tumor stage (pT) 3 UTUCs [4,5]. Renal pelvic tumors that invade the renal parenchyma have been associated with a better prognosis than tumors invading peripelvic or periureteral fat [6]. Some studies have not reported any differences in oncologic outcomes between patients with ureteral and renal pelvis tumors [7–9]. In this study, we investigated the effect of tumor location on oncologic outcomes in patients with UTUC by stratifying patients according to pT stage.
4.e20
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
Material and methods Patient population and enrollment criteria Between January 1996 and March 2009, 598 consecutive Taiwanese patients with UTUC who underwent radical nephroureterectomies (RNUs) at our institution were enrolled. Of the 598 patients, 95 were excluded for the following reasons: incomplete tumor stage (11), bladder urothelial carcinoma before RNU (33), radical cystectomies with concomitant RNU (7), previous systemic chemotherapy, including neoadjuvant therapy for UTUC (7), previous radiotherapy to the pelvis (2), and nodal metastases (35). This study focused on the remaining 503 patients. All 503 RNUs were performed by either the open or laparoscopic method. Bladder cuff resections were obtained through the open approach. The hilar and regional lymph nodes were resected if enlarged or if palpable nodes were identified preoperatively or intraoperatively. There was no adjuvant chemotherapy performed for pTa-3 diseases postoperatively (which included 495 patients). All of the 8 patients with pT4 disease underwent cisplatin-based chemotherapy after RNU and were excluded from the survival analyses. The Research Ethics Committee of the National Taiwan University Hospital (201304027RINC) approved this study.
6 months for 2 years, and annually afterward. Computed tomography or magnetic resonance imaging or both were performed annually to detect local recurrence, contralateral UTUC occurrence, and distal metastasis. Disease recurrence was defined as tumor relapse in the operative field, regional lymph nodes, or distant metastasis. Cisplatin-based chemotherapy is the primary treatment of all types of recurrent disease. A physician determined each patient's cause of death. All patients who died of UTUC had previous disease recurrence. Statistical methods The chi-square test was used to analyze associations between clinicopathologic parameters. The log-rank test was used to compare RFS, CSS, and overall survival (OS) periods among patient groups. The Cox proportional hazards model was used to evaluate the risk of oncologically adverse events. Prognosticators used in the multivariate model were designated from factors with a P o 0.15 in the univariate analysis. A P o 0.05 was considered statistically significant. Results Population characteristics
Clinical data collection Patient medical records were used to obtain information about age, sex, diabetes mellitus status, hypertension status, serum creatinine levels, dialysis history, herbal medications, smoking status, and pathologic characteristics of UTUC tumors. Estimated glomerular filtration rate was calculated using the Modification in Diet in Renal Disease Study equation [10]. The tumor location was defined as the dominantly advanced lesion at the ureteral or renal pelvis according to pathology reports. In all, 39 patients with equally advanced tumors at both ureteral and renal pelvis locations were excluded from the analyses of the association between the tumor locations and oncologic outcomes. Tumor multifocality was defined as the synchronous appearance of Z2 pathologically confirmed tumors in the urinary tract. Heavy smokers were defined as people who had a cumulative smoking history of 20 pack-years or more. Pathologists at our institution reviewed all surgical specimens. Tumors were graded according to 2004 World Health Organization classifications and staged based on 2002 TNM classifications.
Patient demographics and clinical features stratified by tumor location are listed in Table 1. Our cohort included 249 men (49.5%) and 254 women (50.5%). The median age of all patients was 68 years (interquartile range [IQR]: 60.0– 74.8 y). Overall, 280 patients had renal pelvic tumors (55.7%) and 184 patients had ureteral tumors (36.6%). The remaining 39 patients had the same T stage tumors at both the renal pelvis and ureter (7.7%). Ureteral and combined pelvic and ureteral tumors are more likely associated with high-grade disease compared with renal pelvic tumors alone (64.1%, 60%, and 50.7%, respectively; P ¼ 0.012). It was also more common for patients with ureteral tumors to have advanced chronic kidney disease stages (3, 4, and 5) than patients with renal pelvic and combined tumors (69.1%, 50.3%, and 48.7%, respectively; P o 0.001). The presence of locally advanced disease (pT3/T4) was more frequent in renal pelvic tumors (37.5%) than in the combined locations (30.7%) and ureteral tumors (20.2%) (P o 0.001). No significant difference was found in the rest of the clinicopathologic characteristics among patient groups (Table 1). Patient survival
Follow-up protocol After RNU, all patients were followed regularly with cystoscopic examinations, urine cytologic tests, and renal sonography every 3 months in the first year, then every
In all, 495 patients with pTa-T3N0M0 diseases were enrolled for the survival analyses. Table 2 demonstrates disease recurrence, cancer-specific death, and overall mortality rates stratified by tumor stages and locations.
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25 Table 1 Clinicopathologic characteristics of patients with upper urinary tract urothelial carcinoma Characteristics
Number of patients, n (%)
Tumor location Ureter
Both
280 (55.7%)
184 (36.6%) 39 (7.7%)
Age, y Median (IQR) o70 Z70 Gender, n (%) Female Male
Disease recurrence Renal pelvis Ureter Both 0.623
67 (60–73.8) 69 (60–74) 66 (59–75) 166 (59.3%) 105 (57.1%) 23 (60.0%) 114 (40.7%) 79 (42.9%) 16 (40.0%)
0.891 0.296
133 (47.5%) 147 (52.5%)
101 (54.9%) 20 (51.3%) 83 (45.1%) 19 (48.7%)
Pathologic grade, n (%) Low 135 (48.2%) High 142 (50.7%)
63 (34.2%) 15 (38.5%) 118 (64.1%) 23 (60%)
Pathologic tumor stage Ta/Tis/T1 144 (51.4%) T2 31 (11.1%) T3 101 (36.1%) T4 4 (1.4%)
pTa/Tis/T1 (N ¼ 251)
pT2 (N ¼ 98)
pT3 (N ¼ 146)
15/250 (6%) 8/143 (5.6%) 6/87 (6.9%) 1/20 (5.0%)
17/97 (17.5%) 46/144 (31.9%)a 3/30 (10.0%) 27/100 (27%) 13/60 (21.7%) 16/32 (50.0%) 1/7 (14.3%) 3/12 (25.0%)
Cancer-specific death 13/ 250 (5.2%) Renal pelvis 8/143 (5.6%) Ureter 5/87 (5.7%) Both 0/20
12/95 (12.6%) 29/146 (19.9%) 3/29 (10.3%) 17/101 (16.8%) 9/60 (15%) 9/33 (27.3%) 0/6 3/12 (25.0%)
Overall mortality Renal pelvis Ureter Both
22/95 (23.2%) 36/146 (24.7%) 8/29 (27.6%) 22/101 (21.8%) 13/60 (21.7%) 10/33 (30.3%) 1/6 (16.6%) 4/12 (33.3%)
22/250 (8.8%) 12/143 (8.4%) 8/87 (9.2%) 2/20 (10.0%)
0.012
87 60 33 4
a
Two patients with pT3 disease died perioperatively and were excluded from the analysis of disease recurrence.
(47.3%) 20 (51.3%) o0.001 (32.6%) 7 (18%) (18.0%) 12 (30.7%) (2.2%) 0
LVI, n (%) Absent Present
184 (65.7%) 49 (17.5%)
126 (68.5%) 22 (56.4%) 32 (17.4%) 6 (15.4%)
0.979
Herbal usage, n (%) No Yes
202 (72.1%) 20 (7.1%)
129 (70.1%) 24 (61.5%) 16 (8.7%) 6 (15.4%)
Heavy smoker, n (%) No 234 (83.6%) Yes 42 (15.0%)
154 (83.7%) 29 (74.4%) 29 (15.8%) 9 (23.1%)
Renal function, n (%) CKD stage 0–2a 95 (33.9%) CKD stage 3–4 121 (43.2%) CKD stage 5 20 (7.1%)
38 (20.7%) 9 (23.1%) 98 (53.3%) 11 (28.2%) 29 (15.8%) 8 (20.5%)
0.191 0.409
stages had more disease recurrence. The 5-year RFS rates were 92%, 79%, and 65% in pTa/Tis/T1, pT2, and pT3 disease, respectively (P o 0.001). In the univariate analyses for disease recurrence, highgrade disease (hazard ratio [HR] ¼ 2.44, P o 0.001), advanced tumor stage (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 3.45 and 6.75, P o 0.001 and o0.001, respectively), and lymphovascular invasion (HR ¼ 1.81, P ¼ 0.03) significantly increased the risk of disease recurrence. The multivariate analysis shows that only advanced tumor stage (pT2 and pT3 vs. pTa/Tis/T1, HR ¼ 2.52 and 5.64, P ¼ 0.03 and o0.001, respectively) was an independent risk factor predicting disease recurrence after RNU (Table 3).
o0.001
Diabetes mellitus, n (%) No 228 (81.4%) Yes 52 (18.6%)
144 (78.3%) 34 (87.2%) 39 (21.2%) 5 (12.8%)
Hypertension, n (%) No Yes
108 (58.7%) 26 (66.7%) 74 (40.2%) 13 (33.3%
Cancer-specific survival
0.444
0.582 177 (63.2%) 102 (36.4%)
Table 2 Oncologic outcomes stratified by tumor stage and location
P values
Renal pelvis
4.e21
LVI ¼ lymphovascular invasion; T ¼ tumor stage. CKD stage 0 was defined as normal renal function.
a
Recurrence-free survival During a median follow-up of 52 months (IQR: 23– 77 mo), disease recurrence was discovered in 78 patients (15.8%), consisting of 15 (6%), 17 (17.5%), and 46 (31.9%) patients with pTa/Tis/T1, pT2, and pT3 diseases, respectively. The median interval between RNU and disease recurrence was 9 months (IQR: 4–32 mo) for those with disease recurrence; 73.8% of disease recurrences were identified within 2 years after RNU. Patients with advanced
In this cohort, 54 patients (10.9%) died of UTUCs, including 13 (5.2%), 12 (12.6%), and 29 (19.9%) patients with pTa/Tis/T1, pT2, and pT3 disease, respectively. Advanced tumor stages were associated with poor CSS. The 5-year CSS rates were 94%, 84%, and 75% in patients in the pTa/Tis/T1, pT2, and pT3 groups, respectively (P o 0.001). In the univariate analyses for disease recurrence, highgrade disease (HR ¼ 2.54, P ¼ 0.002) and advanced tumor stage (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 2.76 and 4.72, p ¼ 0.011 and o0.001, respectively) significantly increased the risk for cancer-specific death. In the multivariate analysis, age Z70 years (HR ¼ 3.47, P ¼ 0.003) and pT3 disease (reference: pTa/Tis/T1 tumor, HR ¼ 4.11, P ¼ 0.004) were independent risk factors predicting cancerspecific mortality in patients with UTUC (Table 3). Overall survival During follow-up, a total of 80 patients (16.2%) died of any cause, including 22 (8.8%), 22 (23.2%), and 36 (24.7%) patients with pTa/Tis/T1, pT2, and pT3 disease, respectively.
4.e22
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
Table 3 Cox proportional hazards model for predicting oncologic outcomesa Variables
Univariate analysis
Multivariate analysis
HR
95% CI
P values
HR
95% CI
P values
Disease recurrence Grade Low High
Reference 2.44
– 1.49–3.99
– o0.001
Reference 1.31
– 0.70–2.44
– 0.395
Pathologic T stage Ta/Tis/T1 T2 T3
1.84 Reference 3.46 6.75
1.53–2.20 – 1.72–6.92 3.76–12.09
o0.001 – o0.001 o0.001
Reference 2.52 5.64
– 1.10–5.81 2.74–11.61
– 0.03 o0.001
LVI Absent Present
Reference 1.81
– 1.06–3.08
– 0.030
Reference 1.05
– 0.56–1.95
– 0.886
Tumor location Pelvis Ureter
Reference 1.48
– 0.942.33
– 0.090
Reference 1.33
– 0.74–2.40
– 0.338
Reference 1.62
– 0.96–2.73
– 0.071
Reference 3.47
– 1.53–7.87
– 0.003
Reference 2.54
– 1.41–4.58
– 0.002
Reference 1.93
– 0.81–4.64
– 0.140
Pathologic T stage Ta/Tis/T1 T2 T3
Reference 2.76 4.72
– 1.26–6.07 2.45–9.09
– 0.011 o0.001
1.69 4.11
0.52–5.51 1.59–10.65
0.386 0.004
Tumor location Pelvis Ureter
Reference 1.43
– 0.83–2.45
– 0.196
Reference 0.90
– 0.36–2.25
– 0.828
Overall mortality Age o70 Z70
Reference 1.95
– 1.27–2.99
– 0.002
Reference 2.66
– 1.48–4.75
– 0.001
Grade Low High
Reference 2.33
– 1.44–3.77
– 0.001
Reference 1.67
– 0.87–3.20
– 0.122
Pathologic T stage Ta/Tis/T1 T2 T3
Reference 2.90 3.43
– 1.60–5.24 2.03–5.87
– o0.001 o0.001
Reference 3.55 3.01
– 1.66–7.57 1.42–6.38
– 0.001 0.004
Tumor location Pelvis Ureter
Reference 1.40
– 0.82–2.39
– 0.222
0.76
0.42–1.47
0.449
Cancer-specific death Age o70 Z70 Grade Low High
CKD ¼ chronic kidney disease; LVI, lymphovascular invasion; T stage ¼ tumor stage. Bold values statistically signifies p value o0.05. a Only significant results in univariate or multivariate analyses were listed.
The 5-year CSS rates were 90%, 74%, and 71% for the pTa/ Tis/T1, pT2, and pT3 groups, respectively (P o 0.001). Univariate analyses showed that age Z70 years (HR ¼ 1.95, P ¼ 0.002), high-grade disease (HR ¼ 2.33, P ¼ 0.001), and advanced diseases (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 2.90 and 3.43, P o 0.001 and o0.001,
respectively) significantly increased the risk of overall mortality. In the multivariate analysis, age Z70 years (HR ¼ 2.66, P ¼ 0.001) and advanced diseases (pT2 and pT3 vs. pTa/Tis/T1 tumor, HR ¼ 3.55 and 3.01, P ¼ 0.001 and 0.004, respectively) were independent risk factors for predicting OS (Table 3).
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
4.e23
Table 4 The 5-year survival rates stratified by tumor stage and location Tumor stage/ location
RFS, %
pTa/Tis/T1 Renal pelvis Ureter pT2 Renal pelvis Ureter pT3 Renal pelvis Ureter
92 92 92 79 87 75 65 71 50
P values 0.645
0.101
0.047
CSS, % 94 93 94 84 88 80 75 80 60
P values 0.872
0.381
0.164
OS, % 90 90 91 74 75 73 71 75 59
P values 0.741
0.908
0.237
Tumor location effect of each stage Table 4 demonstrates the 5-year rates for RFS, CSS, and OS stratified by tumor stage and location. In patients with pTa/Tis/T1 disease, the 5-year RFS, CSS, and OS were greater than 90% despite tumors being located at the renal pelvis or ureter. Our study cohort had a trend toward more advanced tumor stages being associated with the prognostic relevance of the ureteral tumor, although this was significantly different only in pT3 disease recurrence rates. During follow-up, pathologic T3 ureteral disease had a higher frequency of disease recurrence (50% vs. 27%) and a shorter duration of RFS than did renal pelvic tumors (logrank test, P ¼ 0.047). Tumor location had no significant effect on Ta/Tis/T1 and T2 diseases. In the subgroup analysis, the tumor stages and locations were reported as pTa/Tis/T1, pT2, pT3 renal pelvic tumor and pT3 ureteral tumor groups for a new model for survival analyses. Patients with pT3 ureteral tumors had a significantly shorter duration of RFS and CSS compared with patients with pT2 diseases (log-rank test, P o 0.001 and P ¼ 0.028, respectively). There were no differences between pT3 renal pelvic tumors and pT2 diseases in RFS and CSS (log-rank test, P ¼ 0.111 and P ¼ 0.285) (Figs. 1 and 2). Table 5 demonstrates the independent risk factors that predict oncologic outcomes. The pT3 ureteral tumors actually increased the risk of disease recurrence (reference: pTa/Tis/T1 tumor, HR ¼ 7.82, P r 0.001), cancer-specific death (reference: pTa/Tis/T1 tumor, HR ¼ 5.08, P r 0.001), and overall mortality (reference: pTa/Tis/T1 tumor, HR ¼ 3.25, P ¼ 0.031) in the multivariate analysis.
Fig. 1. Kaplan-Meier survival curve for recurrence-free survival stratified by tumor stage (pT3 ureteral vs. pT2 disease, P o 0.001; pT3 renal pelvic vs. pT2 disease, P ¼ 0.111).
demonstrated the results of 5-year RFS, CSS, and OS rates stratified by tumor stage and location. Our findings show that patients with pT3 ureteral tumors have significantly poorer RFS and CSS outcomes compared with outcomes in patients with either pT2 or pTa/Tis/T1 disease. Several studies have demonstrated that pathologic tumor grade is a powerful predictor of cancer-control outcomes [12]. This was not confirmed in the multivariate analyses of the present study, because of the strong correlation found between tumor grades and stages. In our cohort, the highgrade tumor accounted for 44.2%, 63.90%, 64.3%, and 93.80% of pTa/Tis/T1, pT2, pT3 renal pelvic disease, and pT3 ureteral disease, respectively. A few low-grade tumors combined with advanced-stage tumors were included in this study, leading to a relatively low statistical power for
Discussion The incidence of UTUC in Taiwan is the highest in the world, comprising 20% to 33% of all urothelial cancers [2,3]. Aristolochic acid, a component of aristolochia herbal remedies, is thought to be associated with the higher incidence of UTUC in Taiwan [11]. In this study, we investigated the risk factors and oncologic outcomes in patients with UTUC and
Fig. 2. Kaplan-Meier survival curve for cancer-specific survival stratified by tumor stage (pT3 ureteral vs. pT2 disease, P ¼ 0.028; pT3 renal pelvic vs. pT2 disease, P ¼ 0.285).
4.e24
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
Table 5 Prognosticators for oncologic outcomesa Variables
Disease recurrence Tumor stage/location pTa/Tis/T1 pT2 pT3 renal pelvis pT3 ureter Cancer-specific death Age o70 Z70 Tumor stage/location pTa/Tis/T1 pT2 pT3 renal pelvis pT3 ureter Overall mortality Age o70 Z70 Tumor stage/location pTa/Tis/T1 pT2 pT3 renal pelvis pT3 ureter
Multivariate analysis HR
95% CI
P values
Reference 3.28 5.22 7.82
– 1.47–7.29 2.47–11.04 3.17–19.29
–
Reference 3.44
– 1.52–7.79
–
Reference 1.72 4.33 5.08
– 0.56–5.35 1.60–11.70 1.28–20.21
–
Reference 2.46
– 1.40–4.31
–
Reference 3.15 2.98 3.25
– 1.57–6.35 1.38–6.43 1.11–9.45
–
0.004 o0.001 o0.001
0.003
0.346 0.004 0.021
0.002
0.001 0.006 0.031
a The tumor stages and locations refer to pTa/Tis/T1, pT2, pT3 renal pelvic, and pT3 ureteral cancer groups. Only significant results in multivariate analyses were listed.
identifying the difference in RFS, CSS, and OS in patients stratified by tumor grade. To date, the effect of tumor location on the prognosis in patients with UTUC is still controversial. Several studies have demonstrated that ureteral tumors have worse prognoses than renal pelvic tumors for disease recurrence and cancer-specific death. The ureteral tumor was an independent predictive factor for RFS and CSS after adjustment for pathologic variables [4,5]. Park et al. found the ureteral tumor to have prognostic relevance only in patients with pT3 disease. Renal pelvic tumors invading the renal parenchyma were associated with higher RFS and CSS rates than those invading peripelvic or periureteral fat, suggesting that renal parenchyma may play a protective role in preventing local tumors from spreading in patients with a pT3 tumor [6]. On the contrary, 2 population-based studies and 1 singleinstitutional study show that renal pelvic and ureteral tumors have the same RFS and CSS after several clinicopathologic parameters are adjusted, which confirms that tumor location is not an independent prognostic factor in oncologic outcomes [7–9]. In our study, the prognostic relevance of the ureteral tumor was investigated only in pT3 disease. The pT3 ureteral tumor had a higher frequency of disease recurrence
(50% vs. pT3 renal pelvic tumor, 27%) and a shorter duration of RFS than did renal pelvic tumors (log-rank test, P = 0.047). The oncologic outcomes of pT3 ureteral disease demonstrated 5-year rates of 50%, 60%, and 59% in RFS, CSS, and OS, respectively. In addition, pT3 ureteral tumors also had a significantly shorter duration of RFS and CSS compared with tumors in either of the pT2 diseases (logrank test, P o 0.001 and P = 0.028, respectively), but there were no differences between pT3 renal pelvic tumors and pT2 diseases regarding RFS and CSS (log-rank test, P = 0.111 and P = 0.285). Advanced pT, especially pT3 ureteral tumors, was the indicating risk factor for predicting RFS, CSS, and OS in the multivariate analysis. To date, the currently published retrospective studies do not provide a definitive conclusion regarding the effect of tumor location on UTUC prognosis. The conflicting findings support the need for randomized prospective studies to investigate the difference in oncologic outcomes between patients with ureteral and renal pelvis tumors. In our cohort, all of the 8 patients with pathologically proven T4 disease underwent adjuvant chemotherapy after RNU, and only 1 patient (12.5%) developed disease recurrence afterward. Among the 32 chemonaïve patients with pT3 ureteral disease, 16 (50%) had disease recurrence and received systemic chemotherapy. The 5-year RFS rate was 71% in patients with pT4 disease, which was greater than the 50% in those with chemonaïve pT3 ureteral disease (log-rank test, P = 0.593; data not shown). The findings may suggest the need for adjuvant chemotherapy in patients with high-risk disease to improve patient prognosis. A systematic review and meta-analysis analyzed 1 prospective and 9 retrospective studies to investigate the effect of adjuvant chemotherapy in patients with UTUC after RNU. There appears to be an OS and RFS benefit with cisplatinbased adjuvant chemotherapy in UTUC [13]. Kawashima et al. [14] analyzed the results of adjuvant chemotherapy for localized pT3 UTUC in a multi-institutional study, which showed that adjuvant chemotherapy status was significantly associated with CSS (HR ¼ 0.205, P ¼ 0.008). On the contrary, these findings were not confirmed by Hellenthal et al. [15], who used the international collaborative database, which demonstrated that adjuvant chemotherapy after RNU did not significantly correlate with overall or CSS in patients with the individual high-risk diseases (pT3N0, pT4N0, and lymph node–positive disease). The current evidence is limited by the retrospective nature of studies and the absence of subgroup analyses stratified by both tumor stage and location. Adjuvant chemotherapy appears to be reasonable and feasible for the selective and high-risk patients, but more prospective trials are needed to confirm its efficacy. Our current study has several limitations. First, the data were collected in a retrospective manner and several potentially important covariates, such as operative methods, tumor size, and smoking status, were not evaluated. The renal pelvic tumors in pT3 and pT4 disease were not
Y.-S. Tai et al. / Urologic Oncology: Seminars and Original Investigations 34 (2016) 4.e19–4.e25
divided into parenchymal and peripelvis invasion. Second, a relatively small sample size of patients was used in the analysis of the association between tumor stage/location and outcomes. Third, the relatively short follow-up duration prevented the observation of long-term survival outcomes. Fourth, the RNUs were performed by several surgeons at our institution, resulting in some variation among cases. No regular lymph node dissection was performed in this study, which may mean that information about certain microscopic metastatic diseases and their effect on oncologic outcomes was not recorded. However, there is a standard protocol for managing bladder cuff and follow-up after RNU at our institution. Therefore, the observational bias of RFS events would not systemically exist in this cohort. Conclusions Patients with UTUC and pT3 ureteral tumors had more frequent disease recurrence and a shorter duration of RFS than did those with pT3 renal pelvic diseases. The 5-year rates for RFS, CSS, and OS were 50%, 60%, and 59%, respectively, in patients with pT3 ureteral cancer. In addition, pT3 ureteral tumors also had a significantly shorter duration of RFS and CSS compared with that in patients with pT2 diseases; but there were no differences between pT3 renal pelvic tumors and pT2 diseases in RFS and CSS. Advanced pT, especially pT3 ureteral tumors, was the indicator of risk factors for oncologic outcomes. These results underscore the need for close follow-up and the consideration of adjuvant chemotherapy for patients with pT3 ureteral cancer. Ethical standards The article does not contain clinical studies or patient data. Acknowledgment The authors have nothing to disclose. References [1] Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1998. CA Cancer J Clin 1998;48:6–29.
4.e25
[2] Chou YH, Huang CH. Unusual clinical presentation of upper urothelial carcinoma in Taiwan. Cancer 1999;85:1342–4. [3] Yang MH, Chen KK, Yen CC, Wang WS, Chang YH, Huang WJ, et al. Unusually high incidence of upper urinary tract urothelial carcinoma in Taiwan. Urology 2002;59:681–7. [4] Akdogan B, Dogan HS, Eskicorapci SY, Sahin A, Erkan I, Ozen H. Prognostic significance of bladder tumor history and tumor location in upper tract transitional cell carcinoma. J Urol 2006;176: 48–52. [5] Ouzzane A, Colin P, Xylinas E, Pignot G, Ariane MM, Saint F, et al. Ureteral and multifocal tumours have worse prognosis than renal pelvic tumours in urothelial carcinoma of the upper urinary tract treated by nephroureterectomy. Eur Urol 2011;60:1258–65. [6] Park J, Ha SH, Min GE, Song C, Hong B, Hong JH, et al. The protective role of renal parenchyma as a barrier to local tumor spread of upper tract transitional cell carcinoma and its impact on patient survival. J Urol 2009;182:894–9. [7] Favaretto RL, Shariat SF, Chade DC, Godoy G, Adamy A, Kaag M, et al. The effect of tumor location on prognosis in patients treated with radical nephroureterectomy at Memorial Sloan-Kettering Cancer Center. Eur Urol 2010;58:574–80. [8] Isbarn H, Jeldres C, Shariat SF, Liberman D, Sun M, Lughezzani G, et al. Location of the primary tumor is not an independent predictor of cancer specific mortality in patients with upper urinary tract urothelial carcinoma. J Urol 2009;182:2177–81. [9] Raman JD, Ng CK, Scherr DS, Margulis V, Lotan Y, Bensalah K, et al. Impact of tumor location on prognosis for patients with upper tract urothelial carcinoma managed by radical nephroureterectomy. Eur Urol 2010;57:1072–9. [10] Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999;130: 461–70. [11] Chen CH, Dickman KG, Moriya M, Zavadil J, Sidorenko VS, Edwards KL, et al. Aristolochic acid-associated urothelial cancer in Taiwan. Proc Natl Acad Sci U S A 2012;109:8241–6. [12] Lughezzani G, Burger M, Margulis V, Matin SF, Novara G, Roupret M, et al. Prognostic factors in upper urinary tract urothelial carcinomas: a comprehensive review of the current literature. Eur Urol 2012;62: 100–14. [13] Leow JJ, Martin-Doyle W, Fay AP, Choueiri TK, Chang SL, Bellmunt J. A systematic review and meta-analysis of adjuvant and neoadjuvant chemotherapy for upper tract urothelial carcinoma. Eur Urol 2014;66:529–41. [14] Kawashima A, Nakai Y, Nakayama M, Ujike T, Tanigawa G, Ono Y, et al. The result of adjuvant chemotherapy for localized pT3 upper urinary tract carcinoma in a multi-institutional study. World J Urol 2012;30:701–6. [15] Hellenthal NJ, Shariat SF, Margulis V, Karakiewicz PI, Roscigno M, Bolenz C, et al. Adjuvant chemotherapy for high risk upper tract urothelial carcinoma: results from the Upper Tract Urothelial Carcinoma Collaboration. J Urol 2009;182:900–6.