Fax: (011) 886 77317123, extension 8004
The development of bladder tumors and contralateral upper urinary tract tumors after primary transitional cell carcinoma of the upper urinary tract
Article first published online: 10 SEP 2003
Copyright © 2003 American Cancer Society
Volume 98, Issue 8, pages 1620–1626, 15 October 2003
How to Cite
Kang, C.-H., Yu, T.-J., Hsieh, H.-H., Yang, J. W., Shu, K., Huang, C.-C., Chiang, P.-H. and Shiue, Y.-L. (2003), The development of bladder tumors and contralateral upper urinary tract tumors after primary transitional cell carcinoma of the upper urinary tract. Cancer, 98: 1620–1626. doi: 10.1002/cncr.11691
- Issue published online: 3 OCT 2003
- Article first published online: 10 SEP 2003
- Manuscript Accepted: 6 JUL 2003
- Manuscript Revised: 1 JUL 2003
- Manuscript Received: 1 APR 2003
- transitional cell carcinoma (TCC);
- field cancerization;
- upper urinary tract (UUT);
- bladder neoplasm
Contralateral, metachronous upper urinary tract (UUT) tumors after primary transitional cell carcinoma (TCC) of the UUT are reported rarely, and to the authors' knowledge the risk factors have not been determined to date. In addition, few reports have described the characteristics of recurrent bladder tumors and contralateral UUT tumors and any relation between theses tumor types.
Statistical analysis of data from 223 patients with documented primary UUT-TCC was undertaken. After excluding bilateral involvement and distant metastases, 12 variables were analyzed by multivariate analysis in 189 patients to determine the risk factors for recurrent urothelial tumors.
The incidence rates of recurrent bladder tumors and contralateral UUT tumors were 31.2% and 5.8%, respectively. Multiplicity was determined as a risk factor for recurrent bladder tumors. Renal insufficiency, uremia, and concurrent bladder tumors significantly predisposed patients to develop contralateral UUT tumors after primary UUT-TCC. The time intervals and stage distributions differed significantly between recurrent bladder tumors and contralateral UTT tumors. Patients who had recurrent bladder tumors had earlier stage tumors and had a shorter time to recur compared with patients who had contralateral, metachronous UUT tumors.
For patients with primary UUT-TCC, regular follow-up by cystoscopy is necessary to detect recurrent bladder tumors. Intravenous urography or retrograde pyelography should be performed for patients who have a high risk of developing contralateral UUT tumors. Cancer 2003. © 2003 American Cancer Society.
Transitional epithelial cells line the inner surface of the urinary tract. Carcinoma associated with these cells is termed transitional cell carcinoma (TCC). An important characteristic of TCC is the formation of tumors in multiple foci throughout the urinary tract simultaneously and subsequently. Thus, the development of bladder tumors or contralateral tumors of the upper urinary tract (UUT) is possible in patients with unilateral primary UUT-TCC.
The prevalence of recurrent urothelial tumors in patients with primary UUT-TCC is equivocal. Incidence rates of 13–47% for recurrent bladder tumors and 0.8–5.8% for contralateral metachronous UUT tumors have been reported.1–3 However, those studies focused on the prognostic factors of UUT-TCC and on risk factors for recurrent bladder tumors. To our knowledge, information is lacking concerning the characteristics of contralateral metachronous UUT tumors and the necessity for routine imaging studies in early diagnosis and treatment.
To clarify the association between primary UUT-TCC and recurrent urothelial tumors, we have taken advantage of the observations of the unusual clinical presentation and high incidence of UUT-TCC in Taiwan.4–7 In the current study, the clinical histories of patients from southern Taiwan with primary UUT-TCC were reviewed to identify risk factors for recurrent bladder tumors and contralateral UUT tumors.
MATERIALS AND METHODS
The surgical logs and tumor registries of patients who underwent surgery for primary UUT-TCC at Chang Gung Memorial Hospital in Kaohsiung from January 1, 1986 through December 31, 1995 were reviewed retrospectively to identify study candidates. A total of 223 patients were identified with histologically confirmed, primary UUT-TCC. The median follow-up for these patients after surgery was 91 months (range, 60–168 months). Follow-up consisted of surveillance cystoscopy every 3 months during the first year, every 6 months during the second year, and annually thereafter. Imaging studies of the UUT were obtained every 1–2 years during follow-up.
Table 1 summarizes the characteristics and clinical features of the patients. The median age was 63 years, and the male:female ratio was approximately 1:1. Ten of the 223 patients studied (4.5%) had bilateral, synchronous UUT-TCC with or without concurrent bladder tumors. There were 41 patients (18.4%) who presented initially with flank or abdominal pain or with irritative voiding symptoms without hematuria. Thirty-two patients (12.3%) had renal insufficiency (plasma creatinine > 1.4 mg%), and 18 patients (8.1%) had uremia with dialysis. At the time of UUT-TCC diagnosis, concurrent bladder tumors were detected in 47 patients (21.1%).
|Painless hematuria||145 (65.0)|
|Flank/abdominal pain with hematuria||37 (16.6)|
|Flank/abdominal pain without hematuria||36 (16.1)|
|From endemic blackfoot disease area||39 (17.5)|
|Preoperative renal function:|
|Insufficient (creatinine > 1.4 mg%)||32 (14.3)|
|Uremic with dialysis||18 (8.1)|
|Renal pelvis||76 (34.1)|
|Renal pelvis and ureter||28 (12.6)|
|Multiplicity (tumor foci ≥ 2)||33 (14.8)|
|<2 cm||85 (38.1)|
|2–5 cm||108 (48.4)|
|>5 cm||30 (13.5)|
|Concurrent bladder tumors||47 (21.1)|
The pathologic stage and grade distributions of the primary tumors in patients with UUT-TCC are listed in Table 2. In 77 of 223 patients (34.5%), the primary UUT-TCC tumor was Ta/Tis and T1. Pathologic staging was T2 in 37 patients (16.6%), T3 in 54 patients (24.2%), and T4 in 55 patients (24.7%). Only approximately 33% of the UUT-TCC were classified as superficial disease. The tumor differentiation was Grade 1 in 13 patients (5.8%), Grade 2 in 95 patients (42.6%), and Grade 3 in 115 patients (51.6%).
|Stage/grade||No. of patients (%)|
|Ta/Tis and T1||77 (34.5)|
The most common treatment was radical nephroureterectomy with bladder cuff removal, which was performed in 191 patients (85.7%). Twenty patients (9.0%) with diminished renal functions, bilateral tumors, solitary kidney or other serious comorbidities were selected for nephron-sparing surgery. The procedures included segmental ureterectomy, distal ureterectomy with reimplantation, open excision, and endoscopic ablation. Unresectable tumors were noted in 12 patients (5.4%); they underwent nephrectomy only or explorative biopsy. Patients with T3–T4,N0,M0 tumors, positive lymph node status, or metastatic disease received 1–6 courses of adjuvant chemotherapy.
To assess the predictive value of prognostic factors with regard to survival, survival analyses with univariate log-rank tests and proportional hazards regression were performed. Kaplan–Meier curves were generated and compared using the log-rank test. Patients with bilateral UUT tumors and distant metastases were excluded to determine which risk factors actually were responsible for recurrent bladder tumors and contralateral UUT tumors.
Multivariate analysis using proportional hazards regression was used to determine the risk factors for recurrent bladder tumors and contralateral UUT tumors. The Fisher exact test was used to assess whether there was a difference in stage and grade distributions between the two types of recurrent urothelial tumors. The difference in time of recurrence was analyzed using a t test.
Prognostic Factors for Survival
The 12 variables listed in Tables 1 and 2 were analyzed for their correlation with patient survival. Unfavorable prognostic factors for overall survival, according to the univariate analysis, included residence in a blackfoot disease area, no presenting hematuria, renal insufficiency, uremia, multiplicity (tumor foci ≥ 2) of UUT-TCC, concurrent bladder tumors, high stage, and high grade, as summarized in Table 3.
|Variable||Univariate analysis||Multivariate analysis|
|Chi-square||P value||Odds ratio||P value|
|Residents in blackfoot disease area||4.07||0.044a||—||0.149|
|No macroscopic hematuria||9.32||0.026a||—||0.933|
|Preoperative renal function||8.24||0.016a||1.72||0.0001b|
|Concurrent bladder tumors||4.85||0.028a||2.01||0.001b|
Kaplan–Meier estimated overall survival curves are illustrated in Figure 1. Only stage, grade, preoperative renal function, and concurrent bladder tumors were statistically significant factors in the proportional hazards regression analysis.
Risk Factors for Recurrent Bladder Tumors and Contralateral UUT Tumors
After excluding the patients with bilateral UUT tumors and distant metastases, the follow-up procedures identified recurrent bladder tumors in 59 of the remaining 189 patients (31.2%) and contralateral metachronous UUT tumors in 11 patients (5.8%). Multivariate analysis indicated that multiplicity was a significant risk factor (P = 0.003; relative risk, 2.9) for recurrent bladder tumors (Table 4). For contralateral metachronous UUT tumors, preoperative renal insufficiency/uremia (P = 0.0007) and concurrent bladder tumors (P = 0.03) were significant risk factors. The relative risks were 3.6 and 4.4, respectively.
|Variable||Recurrent bladder tumors||Contralateral UUT tumors|
|Odds ratio||P value||Odds ratio||P value|
|Residents in blackfoot disease area||—||0.462||—||0.839|
|No macroscopic hematuria||—||0.531||—||0.557|
|Preoperative renal insufficiency/uremia||—||0.055||3.63||0.0007a|
|Multiplicity (tumor foci ≥2)||2.93||0.006a||—||0.157|
|Concurrent bladder tumors||—||0.086||4.42||0.03a|
Of the 11 patients with contralateral metachronous UUT tumors, 2 patients displayed renal insufficiency, 3 patients displayed uremia, and another 5 patients had concurrent bladder tumors preoperatively. Only one patient had normal renal function without concurrent bladder tumors.
The pathologic stage and grade distributions of recurrent urothelial tumors are listed in Table 5. The stage distributions differed significantly between patients with recurrent bladder tumors and patients with contralateral UUT tumors (P = 0.0001; Fisher exact test). However, no significant difference in the grade distributions was evident (P = 0.076).
|Stage/grade||No. of patients (%)|
|Recurrent bladder tumors||Contralateral UUT tumors|
|Ta/Tis and T1||52 (88.1)||5 (45.4)|
|T2||5 (8.5)||3 (27.3)|
|T3||1 (1.7)||3 (27.3)|
|T4||1 (1.7)||0 (0.0)|
|1||7 (11.9)||0 (0.0)|
|2||24 (40.1)||5 (45.4)|
|3||28 (47.5)||6 (54.6)|
|Total||59 (100.0)||11 (100.0)|
Most of recurrent bladder tumors were superficial disease (88.1%). In contrast, 54.6% of contralateral UUT tumors were invasive disease.
During the first 2 years after surgery, recurrent bladder tumors developed in 50 of 59 patients (84.7%), and contralateral UUT tumors developed in 6 of 11 patients (54.5%) (Table 6). The two types of recurrent tumors also differed in their time to recur (P < 0.0001).
|Months after treatment||No. of patients (%)|
|Recurrent bladder tumors||Contralateral UUT tumors|
|3||12 (20.3)||0 (0.0)|
|6||20 (33.9)||1 (9.1)|
|9||27 (45.8)||3 (27.3)|
|12||38 (64.4)||6 (54.5)|
|18||47 (80.1)||6 (54.5)|
|24||50 (84.7)||6 (54.5)|
|36||53 (89.8)||7 (63.6)|
|48||54 (91.5)||8 (72.7)|
|60||56 (94.9)||8 (72.7)|
|72||58 (98.3)||10 (90.9)|
|84||59 (100.0)||10 (90.9)|
|168||59 (100.0)||11 (100.0)|
The locations of bladder tumor recurrences are depicted in Figure 2. Most intravesical tumor recurrences were located on the same side as the previous UUT-TCC and were especially contiguous with the wound from bladder cuff removal.
Primary UUT-TCC is more prevalent and more invasive in Taiwan than in other countries. A number of observations support this statement. The clinical presentations of UUT-TCC in Taiwan display equal gender incidence (male:female ratio, approximately 1:1). There are high ratios of UUT-TCC to all urothelial tumors (20%) and of renal pelvis tumors to all renal tumors (40%). In addition, a high incidence of bilateral involvement (4.5%) has been documented.4–9 Finally, the ratio of superficial tumors to all UUT tumors varied from 43% to 58% in previously reported studies10–15 but was only 34.5% in our study. A suggested (although unsubstantiated) reason for the increased prevalence and invasiveness of UUT-TCC in Taiwan may be the influence of contamination of the local drinking water with arsenic, organic chloride, and ergot alkaloid.5–7
With the exception of age, the poor prognostic factors identified in previous studies,1, 3, 10–16 including high stage, high grade, concurrent bladder tumors, multiplicity of UUT-TCC, preoperative renal insufficiency, uremia, large tumor size, and residence in a blackfoot disease area, were verified statistically in our study by univariate log-rank test. However, only stage, grade, concurrent bladder tumors, and preoperative renal function were found to be statistically significant factors with regard to survival using proportional hazards regression.
An additional, albeit poor, prognostic factor for patients with primary UUT-TCC that emerged from the current study is no macroscopic hematuria upon presentation. Eighteen percent of patients presented initially with flank or abdominal pain or with other voiding symptoms without hematuria in our study. Those patients had more invasive and poorly differentiated tumors that, macroscopically, appeared nonpapillary and infiltrative. Delayed diagnosis, which leads to disease progression and invasion, is a possible reason for the poor prognosis. Another possible reason is that the growth patterns of these tumors are infiltrative and initially involve the invasion of muscle and deeper tissues, causing obstructive uropathy; thus, no macroscopic hematuria would be evident on presentation.
Previous literature has revealed a significant incidence of recurrent bladder tumors (13–47%) after patients primary UUT-TCC.1–3 After excluding the patients with distant metastases and bilateral UUT involvement, the incidence was 31.2% in our study. Therefore, regular cystoscopic examination is necessary for these patients. Hisataki et al. reported that the extent and pathologic stage of UUT-TCC were significant and independent factors for initial intravesical recurrences.17 Our multivariate analysis of the patient data determined that only multiplicity was a significant factor (relative risk, 2.9). Thus, the risk of recurrent bladder tumors is almost three-fold greater for multiple tumors than for single tumors in patients with UUT-TCC. In addition, recurrent bladder tumors had a tendency to be located on the same side as the previous nephroureterectomy and were especially contiguous with the wound from bladder cuff removal (Fig. 2). To prevent the development of bladder tumors, it would be prudent for the surgeon to ligate the ureter downstream from the primary UUT-TCC and to diminish bladder trauma when removing the bladder cuff.
Prior to this study, it was believed that contralateral metachronous UUT tumors were extremely rare,1–3 and there has been controversy over whether it is important to conduct regular examinations of the remaining UUT. However, the current study documented an incidence rate of 5.8%, indicating that contralateral metachronous UUT tumors may be more prevalent than previously supposed.
Our multivariate analysis showed that concurrent bladder tumors and renal insufficiency/uremia are two significant risk factors for the development of contralateral UUT tumors. Thus, intravenous urography (IVU) or retrograde pyelography (RP) should be performed every 6–12 months for patient who are at high risk. In addition, surgeons have the option of performing a prophylactic contralateral nephroureterectomy for uremic patients or for patients with compromised renal function who are ready to receive renal replacement therapy, Furthermore, prophylactic cystectomy should be considered for patients who are not candidates for renal transplantation because of the high incidence of TCC in the useless upper and lower urinary tracts.
Two theories explain the potential pathophysiologic mechanisms of recurrent urothelial tumors: the monoclonality hypothesis (including intraluminal seeding and intraepithelial migration) and the field-cancerization hypothesis.18 Most genetic studies favor a monoclonal origin of recurrent bladder tumors and primary UUT-TCC with intraluminal seeding.19–22 Nakamura et al. also showed that mucosal injury can be a cause of bladder tumor recurrences in vitro.23 Tumor cells could not adhere to intact transitional epithelium but were able to implant on the injured urothelial surface. Our observations showed that most recurrent bladder tumors were located over the site of the nephroureterectomy wound. This was clinical evidence of intraluminal seeding. However, field cancerization may be a minor mechanism of bladder tumor recurrence, because oligoclonal origins were found in a small number of patients with combined tumors of the upper and lower tracts.20, 21
The hypothesis of intraluminal seeding explains only a minority of contralateral metachronous UUT tumors, because most patients with these tumors had no vesicouretreral reflux and did not undergo RP or selective urine cytology of the UUT before the existence of the metachronous tumors. Although genetic studies of a clonal nature are lacking, the mechanism of field cancerization may be a factor in the majority of patients with contralateral UUT tumors. Other evidence in support of the field-cancerization hypothesis includes Chinese-herb nephropathy and analgesic nephropathy, which showed that nephrotoxic and carcinogenic toxins induced renal disease and preneoplastic changes/carcinomas of the whole urothelial field.24, 25 These findings also illustrate why bilateral metachronous UUT-TCC frequently was associated with renal insufficiency or uremia in the current study.
Other interesting findings include the different pathologic stage distributions and times of recurrence between recurrent bladder tumors and contralateral UUT tumors. Patients who had recurrent bladder tumors had more superficial disease (Tis/Ta and T1, 88%) and shorter times of recurrence compared with patients who had contralateral UUT tumors. These phenomena indicate that clinical behavior may differ between recurrent tumors from intraluminal seeding and from field cancerization, although more evidence will be required to confirm this hypothesis. Because of the more invasive nature (≥ T2 disease, 55%) and later occurrence (> 2 years, 45%) of contralateral metachronous UUT tumors, patients with these tumors who are at high risk should be followed for long periods and should be monitored closely so that they may receive early treatment.
Multiplicity is a significant and independent factor for recurrent bladder tumors in patients with primary UUT-TCC. Renal insufficiency or uremia and concurrent bladder tumors are significant risk factors for the development of contralateral metachronous UUT tumors. It is necessary to follow patients closely, using cystoscopy for all patients with primary UUT-TCC and using IVU or RP for patients with a high risk of developing contralateral UUT tumors.
- 6Comparative study on the high prevalence of bladder cancer in the blackfoot disease endemic area in Taiwan. J Formosa Med Assoc. 1998; 87: 1074–1080., , , et al.