Members of the Cancer Registration Committee of the Japanese Urological Association: Tsuneharu Miki, Department of Urology, Graduate School of Medical Sciences, Kyoto Prefectural University of Medicine; Tomoaki Fujioka, Department of Urology, Iwate Medical University; Tomohiko Ichikawa, Department of Urology, Graduate School of Medicine, Chiba University; Seiji Naito, Department of Urology, Faculty of Medical Sciences, Kyushu University; Kenjiro Kohri, Department of Urology, Graduate School of Medical Sciences, Nagoya City University; Hideyuki Akaza, Urology and Andrology, Graduate School of Comprehensive Human Sciences, University of Tsukuba; Hiroyuki Fujimoto, Urology Division, National Cancer Center Hospital; Yoichi Mizutani, Department of Urology, Graduate School of Medical Sciences, Kyoto Prefectural University of Medicine; Tadao Kakizoe, National Cancer Center; Akihiko Okuyama, Department of Urology, Graduate School of Medicine, Osaka University.
Original Article: Clinical Investigation
Clinical outcome of tumor recurrence for Ta, T1 non-muscle invasive bladder cancer from the data on registered bladder cancer patients in Japan: 1999–2001 report from the Japanese Urological Association
Article first published online: 21 JAN 2009
© 2009 The Japanese Urological Association
International Journal of Urology
Volume 16, Issue 3, pages 279–286, March 2009
How to Cite
Kikuchi, E., Fujimoto, H., Mizutani, Y., Okajima, E., Koga, H., Hinotsu, S., Shinohara, N., Oya, M., Miki, T. and the Cancer Registration Committee of the Japanese Urological Association (2009), Clinical outcome of tumor recurrence for Ta, T1 non-muscle invasive bladder cancer from the data on registered bladder cancer patients in Japan: 1999–2001 report from the Japanese Urological Association. International Journal of Urology, 16: 279–286. doi: 10.1111/j.1442-2042.2008.02235.x
- Issue published online: 11 MAR 2009
- Article first published online: 21 JAN 2009
- Received 3 September 2008; accepted 19 November 2008.; Online publication 21 January 2009
Vol. 16, Issue 4, 430, Article first published online: 8 APR 2009
- non-muscle invasive;
- transitional cell
Objective: To characterize the clinical outcome in a large contemporary series of Japanese patients with newly diagnosed Ta, T1 non-muscle invasive bladder cancer who underwent transurethral bladder tumor resection with or without intravesical chemotherapy or Bacillus Calmette-Guérin (BCG) therapy.
Methods: We developed a database incorporating newly diagnosed non-muscle invasive bladder cancer data and outcomes from a Japanese bladder cancer registry between 1999 and 2001 and identified a study population of 3237 consecutive patients who had complete data based on pathological features. Median patient age was 69.9 years.
Results: The 1-year, 3-year, and 5-year overall recurrence-free survival rates were 77.0%, 61.3%, and 52.8%, respectively. In multivariate analyses, the multiplicity of bladder tumors, tumor size greater than 3 cm, pathological stage T1, tumor grade G3, and the absence of adjuvant intravesical instillation were independent risk factors for tumor recurrence. Overall, 1710 patients (52.8%) received intravesical instillation; chemotherapy in 1314 (76.8%) and BCG treatment in 396 (23.2%). In patients treated with intravesical chemotherapy in which an anthracycline chemo-agent was used in 90.5% of the cases, multivariate analyses demonstrated that male gender, multiple bladder tumors, a tumor size greater than 3 cm, and pathological stage T1 were associated with tumor recurrence.
Conclusions: The accumulation and analysis of data from the Japanese National Bladder Cancer Registry made it possible to determine the clinical characteristics, management trends, and survival rates for the period studied. Further study with a dataset created from longer follow-up data would be warranted to analyze tumor progression and disease survival.
Transurethral bladder tumor resection (TUR-BT) with or without intravesical therapy is the primary treatment modality in individuals with non-muscle invasive Ta, T1 bladder cancer. Some patients with risk factors, such as a high grade, lamina propria infiltration, multiplicity, a large tumor, and concomitant carcinoma in situ (CIS) still have a lifelong frequency of multiple recurrences and a probability of stage progression.1–4 To date, several contemporary single center series of patients treated with intravesical chemo- or immunotherapy after TUR-BT for Ta, T1 non-muscle invasive bladder cancer have been published to evaluate the prognostic indicators for tumor recurrence in Japan.5–8 These studies had relatively small patient populations and single center bias. Data compiled from multiple institutions may offer the benefit of understanding and generalizing the diagnosis, patient selection, staging, pathological evaluation, treatment, and follow-up protocol. In addition, a contemporary series of a large number of patients with Ta, T1 non-muscle invasive bladder cancer has the potential benefit of identifying bladder cancer recurrence as an outcome measure.
We developed a database incorporating comprehensive non-muscle invasive bladder cancer data and outcomes from data collected by multiple centers in Japan in retrospective fashion under the program of the Japanese National Bladder Cancer Registry. The aim of the present study was to characterize the clinical outcome in a large contemporary series of Japanese patients with newly diagnosed Ta, T1 non-muscle invasive bladder cancer who underwent TUR-BT with or without intravesical chemotherapy or Bacillus Calmette-Guérin (BCG) therapy for preventing tumor recurrence. In addition, we determined the association of established clinical and pathological characteristics with tumor recurrence.
Data sets from multiple centers in Japan were collected in retrospective fashion under the program of the Japanese National Bladder Cancer Registry. The Registry was sponsored by the Japanese Urological Association. The registration included the background of newly diagnosed bladder cancer patients, their clinical and pathological characteristics, and their 3-year clinical outcome after the initial diagnosis. After combining the data sets, reports were generated for each variable to identify data inconsistencies and other data integrity problems. Ultimately we developed a database incorporating newly diagnosed non-muscle invasive bladder cancer data and outcomes between 1999 and 2001.
A total of 4036 patients with newly diagnosed non-muscle invasive bladder cancer (Ta, T1) N0M0 who were treated with initial TUR-BT from the database were identified. Exclusion criteria included the main histological component of non-urothelial carcinoma (n = 37), a non-curative operation was carried out (n = 703), the presence of a concomitant CIS (n = 32), or incomplete data (n = 27). We identified a study population of 3237 consecutive patients for whom there were complete data sets based on pathological features. After complete tumor resection, intravesical chemotherapy or BCG treatment was carried out in 1314 (40.6%) or in 396 (12.2%) patients, respectively for the purpose of preventing bladder tumor recurrence.
Pathological specimens were reviewed by the genitor-urinary pathologist at each institution. Pathological staging and grading were classified according to the 2nd and 3rd edition of the General Rule for Clinical and Pathological Studies on Bladder Cancer.9,10 In the study we combined the pT1a and pT1b according to the 2nd edition and treated them as pT1. Recurrence was defined as a new tumor appearing in the bladder after initial clearance. Tumor progression was not analyzed in the present study because of the shorter follow-up period of our study. The time until tumor recurrence was calculated as the time from TUR-BT to the date of the first documented bladder tumor recurrence. Patients who had not experienced bladder tumor recurrence were censored at the last follow-up. Patients who died before bladder cancer recurrence were censored at the time of death.
Bladder tumor recurrence-free survival curves were constructed using the Kaplan–Meier method, and were compared with the log-lank test. Cox proportional hazards regression analysis was used to assess the prognostic indicators for recurrence. A P-value less than 0.05 was considered to indicate statistical significance. These analyses were carried out with the STATA version 7.0 statistical software package (Stata Corporation, College Station, TX, USA).
Patient characteristics of overall patient population
Of the 3237 patients (Table 1), 80.3% were men with a median age of 69.9 years (mean; 68.3, range, 18–103). Multiple tumors were found in 1312 cases (40.5%) and a tumor size greater than 3 cm was observed in 374 (11.6%). Disease was grade 1 in 24.2%, 2 in 57.2%, and 3 in 18.7%. Pathological tumor stage was Ta in 51% and T1 in 49%. There were 608 TaG1, 919 TaG2, 124 TaG3, 174 T1G1, 931 T1G2, and 481 T1G3 cases, respectively. Tumor grade was significantly associated with pathological stage (P < 0.001). A total of 1710 patients (52.8%) received intravesical instillation, consisting of chemotherapy in 1314 and BCG treatment in 396.
|No. patients (%)|
|<1 cm||1013 (31.3)|
|1–3 cm||1681 (51.3)|
|>3 cm||374 (11.6)|
|Pathological T category|
|No treatment||1527 (47.2)|
Tumor recurrence in overall patient population
The 1-year, 3-year, and 5-year overall recurrence-free survival rates were calculated to be 77.0%, 61.3%, and 52.8%, respectively. Kaplan–Meier analysis (Table 2) and univariate Cox proportional hazards regression analysis (Table 3) revealed the presence or absence of tumor stalk, tumor multiplicity, tumor size, pathological stage, tumor grade, and with or without adjuvant intravesical therapy to be significant predictors of tumor recurrence. According to the intravesical instillation, the 1- and 3-year recurrence-free survival rates were 73.1% and 56.2%, respectively, in patients without any adjuvant therapy, compared with 80.3% and 64.2% in patients with intravesical chemotherapy, and 81.0% and 70.3% in patients with intravesical BCG therapy (Fig. 1). In multivariate analysis (Table 3), multiple bladder tumors, a tumor size greater than 3 cm, pathological stage T1, tumor grade G3, and the absence of adjuvant intravesical instillation were independent risk factors for tumor recurrence. Recurrence risk was 1.7 times higher for multiple tumors, and 1.5 times higher for tumors greater than 3 cm compared with those equal to or smaller than 3 cm. Intravesical instillations were a protective factor of tumor recurrence since the risk was lower than 1. The risk of recurrence was 0.65 and 0.4 times lower for patients treated with intravesical chemotherapy and BCG instillation, respectively. Patients treated with BCG instillation had significantly less likelihood of papillary type, tumor stalk, solitary, and small size tumor, but greater likelihood of a higher stage and higher grade tumor than those with intravesical chemotherapy or without adjuvant instillation (Table 4). Patients treated with intravesical chemotherapy had significantly less likelihood of a solitary, small size tumor, but greater likelihood of a higher pathological stage and higher grade tumor than those without adjuvant instillation.
|% 1-year (mean ± SE)||% 3-year (mean ± SE)||P-value|
|≤70 years||76.4 ± 0.01||62.8 ± 0.01|
|>70 years||77.7 ± 0.01||59.4 ± 0.01|
|Male||77.3 ± 0.01||61.0 ± 0.01|
|Female||76.0 ± 0.02||62.3 ± 0.02|
|Yes||77.0 ± 0.01||61.4 ± 0.01|
|No||77.6 ± 0.03||60.3 ± 0.03|
|Yes||78.2 ± 0.01||62.8 ± 0.01|
|No||74.3 ± 0.01||57.8 ± 0.02|
|No||82.7 ± 0.01||66.3 ± 0.01|
|Yes||68.8 ± 0.01||54.0 ± 0.02|
|≤3 cm||78.7 ± 0.01||62.8 ± 0.01|
|>3 cm||64.2 ± 0.03||49.8 ± 0.03|
|Pathological T category||<0.001|
|pTa||81.4 ± 0.01||65.6 ± 0.01|
|pT1||72.5 ± 0.01||56.8 ± 0.01|
|G1/2||78.8 ± 0.01||62.7 ± 0.01|
|G3||69.6 ± 0.02||55.2 ± 0.02|
|No Tx||73.1 ± 0.01||56.2 ± 0.01|
|chemoTx||80.3 ± 0.01||64.2 ± 0.01|
|BCG||81.0 ± 0.02||70.3 ± 0.02|
|P-value||Hazards ratio||95% CI||P-value|
|Pathological T category||<0.001||<0.001|
|Patients||IVI chemotherapy No. patients (%)||BCG instillation No. patients (%)||No IVI treatment No. patients (%)||P-value|
|≤70||671 (51.1)||199 (50.3)||790 (51.7)|
|>70 years||643 (48.9)||197 (49.8)||737 (48.3)|
|Male||1038 (79.0)||323 (81.6)||1239 (81.1)|
|Female||276 (21.0)||73 (18.4)||288 (18.9)|
|Yes||1224 (93.2)||325 (82.1)||1389 (91.0)|
|No||69 (5.3)||57 (14.4)||86 (5.6)|
|Unknown||21 (1.6)||14 (3.5)||52 (3.4)|
|Yes||925 (70.4)||225 (56.8)||1101 (72.1)|
|No||316 (24.1)||143 (36.1)||326 (21.4)|
|Unknown||73 (5.6)||28 (7.1)||100 (6.55)|
|Yes||588 (44.8)||242 (61.1)||482 (31.6)|
|No||690 (52.5)||132 (33.3)||988 (64.7)|
|Undetected||14 (1.1)||10 (2.5)||9 (0.6)|
|Unknown||22 (1.7)||12 (3.0)||48 (3.1)|
|<1 cm||369 (28.1)||101 (25.5)||543 (35.6)|
|1–3 cm||716 (54.5)||197 (50.0)||768 (50.3)|
|>3 cm||180 (13.7)||68 (17.2)||126 (8.25)|
|Uncountable||15 (1.1)||13 (3.3)||17 (1.1)|
|unknown||34 (2.6)||17 (4.3)||73 (4.8)|
|Pathological T category||<0.001|
|pTa||672 (51.1)||120 (30.3)||859 (56.3)|
|pT1||642 (48.9)||276 (69.7)||668 (43.8)|
|G1||287 (21.8)||37 (9.3)||458 (30.0)|
|G2||817 (62.2)||178 (45.0)||855 (56.0)|
|G3||210 (16.0)||181 (45.7)||214 (14.0)|
In a subgroup of patients with pT1G3, 164, 154, and 163 patients were treated with intravesical chemotherapy, BCG instillation, and without any adjuvant instillation, respectively. The Kaplan–Meier curve indicated that the differences among the groups were significant (P = 0.039 for intravesical chemotherapy vs BCG instillation, P = 0.007 for intravesical chemotherapy vs no adjuvant therapy, and P < 0.001 for BCG instillation vs no adjuvant therapy) (Fig. 2).
Tumor recurrence in patients treated with intravesical chemotherapy and BCG instillation
We next evaluated the prognostic indicators for tumor recurrence, particularly in patients treated with intravesical chemotherapy, which accounted for 76.8% of the intravesical adjuvant therapy. Intravesical chemotherapy consisted of adriamycin (doxorubicin) in 41, epirubicin in 468, pirarubicin in 680, mitomycin C (MMC) in 120, peplomycin in seven, and tiotepa in one. Overall, an anthracycline chemo-agent was used in 90.5% of the cases of intravesical chemotherapy. When patients treated with an anthracycline chemo-agent were compared with those receiving MMC treatment, Kaplan–Meier curve analysis revealed that the differences among the groups were not significant (Fig. 3).
Kaplan–Meier analysis (Table 5) and univariate Cox proportional hazards regression analysis (Table 6) revealed gender, presence or absence of tumor stalk, tumor multiplicity, tumor size, pathological stage, and tumor grade were significant predictors for tumor recurrence. In multivariate analysis, male gender, multiple bladder tumors, a tumor size greater than 3 cm, and pathological stage T1 were independent risk factors for tumor recurrence. Female patients had a recurrence risk 0.7 times lower than male patients. Recurrence risk was 1.7 times higher for multiple tumors, and 1.3 times higher for tumors greater than 3 cm compared with those equal to or smaller than 3 cm. Patients with T1 tumor had a recurrence risk 1.3 times higher than those with Ta.
|% 1-Year (mean ± SE)||% 3-Year (mean ± SE)||P-value|
|≤70 years||79.2 ± 0.02||64.7 ± 0.02|
|>70 years||81.6 ± 0.02||63.4 ± 0.02|
|Male||79.2 ± 0.01||62.7 ± 0.02|
|Female||84.6 ± 0.02||69.7 ± 0.03|
|Yes||80.7 ± 0.01||64.7 ± 0.02|
|No||74.5 ± 0.05||57.6 ± 0.06|
|Yes||82.0 ± 0.01||66.4 ± 0.02|
|No||76.2 ± 0.02||58.9 ± 0.03|
|No||85.6 ± 0.01||69.6 ± 0.02|
|Yes||73.8 ± 0.02||57.5 ± 0.02|
|≤3 cm||81.8 ± 0.01||65.6 ± 0.02|
|>3 cm||70.7 ± 0.04||55.8 ± 0.04|
|Pathological T category||0.002|
|pTa||84.0 ± 0.02||68.1 ± 0.02|
|pT1||76.4 ± 0.02||60.1 ± 0.02|
|G1/2||82.1 ± 0.01||66.1 ± 0.02|
|G3||71.1 ± 0.03||54.5 ± 0.04|
|P-value||Hazard ratio||95% CI||P-value|
|Pathological T category||0.003||0.004|
In patients treated with BCG instillation, no clinicopathological factors were associated with tumor recurrence in uni- and multivariate analysis (data not shown).
In the present study, we characterized the clinical outcome of newly diagnosed non-muscle invasive bladder cancer in a large contemporary series of patients from a Japanese bladder cancer registry and determined the predictors for tumor recurrence. Overall, bladder tumor multiplicity, a tumor size greater than 3 cm, pathological stage T1, tumor grade G3, and the absence of adjuvant intravesical instillation were found to independently increase the risk of tumor recurrence. Numerous publications have reported the same prognostic indicators as ours for tumor recurrence in non-muscle invasive bladder cancer.1–3 Recently, a combined analysis was carried out using data from 2596 non-muscle invasive bladder cancer patients collected from seven European Organization for Research and Treatment of Cancer (EORTC) trials.4 In the analysis, six clinicopathological risk factors, namely multiplicity, tumor size, prior recurrence rate, pathological stage, concomitant CIS, and tumor grade were determined. Four of the six predictors for tumor recurrence in the present study were shared with their indicators; however, the big difference between their study and ours is that their population has included both primary and recurrent cases. A more homogenous population of patients who initially diagnosed non-muscle invasive bladder tumor was evaluated in our current study.
Overall, 12.2% received BCG instillation in our study. In the subgroup of patients treated with BCG instillation, no clinicopathological factors were associated with tumor recurrence. Kaplan–Meier analysis demonstrated that the recurrence-free survival in the BCG instillation group was significantly higher than that in the intravesical chemotherapy group especially in pT1G3 patients (P = 0.039), which was confirmed by others.11 Furthermore, BCG instillation was significantly selected in patients with multiple, larger, and higher pathological stage tumors, compared with intravesical chemotherapy. These results suggested that BCG instillation was carried out for the prevention of recurrence in a relatively smaller percentage of high risk patients than would have been expected from the current clinical situation.12,13 One reason for the difference in the percentage of BCG instillations carried out between 1999–2001 and the present is that the current clinical management for non-muscle invasive bladder cancer is highly affected by the guidelines.14,15
In our subgroup consisting of the 1314 patients treated with adjuvant intravesical chemotherapy, multivariate analyses demonstrated that male gender, bladder tumor multiplicity, a tumor size greater than 3 cm, and pathological stage T1 were independent risk factors for tumor recurrence. Only about 10% of the patients were treated with MMC intravesical chemotherapy. Au et al. reported that intravesical chemotherapy using a modified 40 mg dose of MMC accompanied by a decrease in urine volume during the procedure and urine alkalinization significantly improved the therapeutic benefit of traditional MMC treatment for the prevention of tumor recurrence.16 Meanwhile, Huncharek et al. have showed that maintenance intravesical chemotherapy reduced tumor recurrence, when compared with a single course of induction chemotherapy.17 Further study is warranted to prove the therapeutic benefit with these modalities, especially in Japanese patients who have these risk factors for recurrence. In our multivariate analyses, gender is an independent predictor for tumor recurrence in patients treated with adjuvant intravesical chemotherapy but not in overall patients. There has been no study to evaluate the influence of gender for tumor recurrence in a large series of bladder cancer patients treated with intravesical chemotherapy. The exact reason why female patients have better outcome for tumor recurrence than male patients has to be elucidated in a future study.
Frydenberg et al. conducted a survey of a population cancer registry that included about 700 newly diagnosed non-muscle invasive bladder cancers between 1990 and 1995 in Victoria of Australia. Logistic regression analysis revealed that tumor grade and pathological T stage were independent factors affecting the risk of recurrence. Less than 10% of the patients received adjuvant intravesical chemotherapy or immunotherapy.18 Gårdmark et al. analyzed the clinical characteristics of about 10 000 newly diagnosed cases of bladder cancer obtained from the Swedish National Bladder Cancer Register between 1997 and 2001.19 A large number of the patients, even in the high risk group, were still undertreated and they concluded that the survival rate of bladder cancer in Sweden during this period seemed to remain at the levels previously reported for the 1980s. The accumulation of data provided by a large cancer registry is of great importance to understanding the trends in the clinical characteristics of the disease and its treatment management, and to providing an opportunity for analysis of the indicators predicting prognosis.20
The present study has several limitations. First, the results were obtained from a dataset created by data only from centers participating in the bladder cancer registry. Since all of the centers in Japan do not participate in the cancer registry, the dataset does not include data for all bladder cancers in Japan. However, approximately 180 institutions participate in the cancer registry in Japan and the dataset contained data for approximately 6000 patients so we believe that the results represent an accurate reflection of the characteristics of patients with newly diagnosed bladder cancer and its clinical outcome in the period from 1999 and 2001.21 Another limitation is that the follow-up period was short. Median follow-up was 24 months and this bias might affect the understanding of true risk factors and the natural course of non-muscle invasive bladder cancer and make us unable to analyze the prediction of tumor progression and survival. In fact recurrence-free survival in our study was somewhat better than that reported in another large series.3 Several papers pointed out the importance of the use of data from long-term follow-up of non-muscle invasive bladder cancer.22,23 Further study would be warranted to accumulate long-term follow-up data in the bladder cancer registry.
In conclusion, patients with multiple tumors, a tumor size greater than 3 cm, tumor grade G3, or pathological T1 tumors were at greater risk, whereas those treated with intravesical BCG instillations had a decreased risk of tumor recurrence in the overall patient population. In patients treated with intravesical chemotherapy, male gender, bladder tumor multiplicity, a tumor size greater than 3 cm, and pathological stage T1 were independent risk factors for tumor recurrence. Further study of datasets created from longer follow-up data is warranted in order to analyze tumor progression and disease survival.
- 8Multivariate evaluation of factors affecting recurrence, progression, and survival in patients with superficial bladder cancer treated with intravesical bacillus Calmette-Guerin (Tokyo 172 strain) therapy: significance of concomitant carcinoma in situ. Int. Urol. Nephrol. 2002; 33: 41–7., , et al.
- 9Japanese Urological Association, The Japanese Society of Pathology. General Rule for Clinical and Pathological Studies on Bladder Cancer, 2nd edn. Kanehara & Co. Ltd, Tokyo, 1993.
- 10Japanese Urological Association, The Japanese Society of Pathology. General Rule for Clinical and Pathological Studies on Bladder Cancer, 3rd edn. Kanehara & Co. Ltd, Tokyo, 2001.
- 21The report of clinical statistical studies on registered bladder cancer patients in Japan 1999–2001. Nippon Hinyoukika Gakkai Zasshi 2006; 97: NP2–31. (In Japanese.), , et al.