- Email: [email protected]
T1 Transitional Cell Carcinoma of the Bladder
0022-5347/04/1711-0153/0 THE JOURNAL OF UROLOGY® Copyright © 2004 by AMERICAN UROLOGICAL ASSOCIATION
Vol. 171, 153–157, January 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000100386.07370.0a
A RANDOMIZED CONTROLLED TRIAL OF SHORT-TERM VERSUS LONG-TERM PROPHYLACTIC INTRAVESICAL INSTILLATION CHEMOTHERAPY FOR RECURRENCE AFTER TRANSURETHRAL RESECTION OF TA/T1 TRANSITIONAL CELL CARCINOMA OF THE BLADDER HIROFUMI KOGA, KENTARO KUROIWA, AKITO YAMAGUCHI, YUKIO OSADA, MASAZUMI TSUNEYOSHI AND SEIJI NAITO* From the Departments of Urology, Graduate School of Medical Sciences, Kyushu University (HK, KK, SN) and Harasanshin General Hospital (AY), Kyushu University Urological Oncology Group (HK, KK, AY, YO, SN), Department of Anatomic Pathology, Graduate School of Medical Sciences, Kyushu University (MT), Fukuoka and Department of Urology, Miyazaki Medical College, Miyazaki, Japan (YO)
ABSTRACT
Purpose: In a prospective randomized controlled study, we investigated the optimal schedule for intravesical instillation of epirubicin for maximizing its effect on prophylaxis and disease progression after transurethral resection of newly diagnosed Ta/T1 bladder cancer. Materials and Methods: The patients were instilled with epirubicin (30 mg/30 ml in normal saline) within 24 hours after transurethral resection and then randomized into 2 groups after a definite histopathological diagnosis of Ta/T1 bladder cancer. One group of 77 patients received 19 intravesical instillations of epirubicin in the year after transurethral resection (group 1). The second group of 73 patients received 9 intravesical instillations of epirubicin during the 3 months after transurethral resection (group 2). Nonrecurrence rates and toxicity were compared. Results: In the followup period, 10 group 1 patients (13.0%) and 23 group 2 patients (31.5%) had recurrent disease. The 3-year nonrecurrence rate was 85.2% in group 1, whereas it was 63.9% in group 2. The nonrecurrence rate of group 1 was significantly higher than that of group 2 throughout the observation period (p ⫽ 0.005). The incidence and severity of toxicity were not significantly different between the 2 groups. Conclusions: Our study indicates that long-term instillation of epirubicin is more effective than short-term instillation in preventing recurrence after transurethral resection of Ta/T1 bladder cancer. KEY WORDS: randomized controlled trial, bladder neoplasms; administration, intravesical; drug therapy, epirubicin
In bladder cancer the most common histological phenotype is urothelial carcinoma. About 70% of newly diagnosed cases are confined to the mucosa or lamina propria (Ta and T1). Almost all cases of Ta/T1 disease macroscopically exhibit papillary tumors, and transurethral resection (TUR) is the gold standard treatment for them. However, TUR alone is not adequate therapy for superficial bladder cancer. The recurrence rate of Ta/T1 bladder cancer within 5 years after TUR is about 50% to 70%. In addition, 5% to 20% of cases show pathologically progressive disease.1 The purpose of intravesical chemotherapy after TUR (adjuvant therapy) is to eradicate existing cancer cells and to prevent recurrence and progression. To date, intravesical chemotherapy in an adjuvant setting for TUR of Ta/T1 disease decreases the short-term (1 to 3-year) recurrence rate by approximately 20% and long-term rate (8-year) by 8.2%.2 Mitomycin C, doxorubicin and epirubicin are commonly used chemotherapeutic agents. The intravesical instillation of bacillus Calmette-Guerin has been shown to be more effective than those chemotherapeutic agents, including doxorubicin, in preventing recurrence after TUR.3 However, bacillus Calmette-Guerin has the disadvantage of causing various
and frequent local and/or systemic side effects. Recently, epirubicin (4-epi-doxorubicin) has shown an in vitro cytotoxic effect on transitional cell carcinoma lines, and in clinical use is as or more effective but less toxic than its parent compound, doxorubicin.4, 5 The timing and duration of intravesical chemotherapy have been extensively studied. A single intravesical instillation of epirubicin,6 – 8 mitomycin C9 and doxorubicin2 immediately after TUR significantly decreased the tumor recurrence rate compared with no adjuvant intravesical therapy. It has been reported that long-term maintenance instillation therapy does not further increase the recurrence-free interval or the long-term recurrence rate compared with immediate short-term instillation therapy after surgery.6, 10 –13 Solsona et al reported that the beneficial effect of long-term instillation was limited to early (1-year) recurrence and not maintained during long-term followup use of mitomycin C.9 On the other hand, Tolley et al reported that 5 instillations of mitomycin C offered a slight advantage over 1 instillation.13 The discrepancies in these reports are due to the differences in the dose of chemotherapeutic agents and enrolled patient backgrounds. Thus, the timing and duration of intravesical chemotherapy remain controversial. In this randomized controlled study we investigated the optimal time of intravesical instillation of epirubicin for its effect on prophylaxis after TUR of newly diagnosed Ta/T1 bladder cancer.
Accepted for publication July 11, 2003. * Correspondence and requests for reprints: Department of Urology, Graduate School of Medical Sciences, 3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan (telephone: 81-92-642-5603; FAX: 81-92-642-5618; e-mail: [email protected]). 153
154
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION MATERIALS AND METHODS
This study was a multicenter, prospective, randomized controlled trial. The criteria of eligibility included new, untreated transitional cell carcinoma of the bladder, Ta or T1 disease, no other active neoplasms and no serious complications. Patients who had a history of urothelial carcinoma of the renal pelvis or ureter were excluded from study. Before treatment patient history, physical examination, urinalysis, urine cytology, complete blood count, blood urea nitrogen, serum creatinine, liver function and electrocardiography were performed. Chest plain x-rays and excretory urography were also performed. At entry in this trial, no patients had evidence of a residual tumor as judged by endoscopic examination and urine cytology. Registration was performed by facsimile at Kyushu University Urological Oncology Group Center. All cases were randomly allocated to a long-term instillation group (group 1) or a short-term instillation group (group 2) by a central computer. All patients were fully informed about this trial and signed the informed consent form. In both groups 30 mg epirubicin dissolved in 30 ml of normal saline was instilled into the bladder through a sterile catheter after emptying the bladder. Patients were instructed to refrain from voiding for 2 hours after instillation. The intravesical instillation schedule of each group is diagrammed (fig. 1). In group 1 instillations 1 and 2 were given within 24 hours and 2 or 3 days after TUR, respectively. The instillation was subsequently given weekly for 2 weeks, and every 2 weeks for an additional 14 weeks. After 4 months the instillation was given every month for 8 months. Thus, a total of 19 instillations were administered in 1 year in group 1. On the other hand, in group 2 a total of 9 instillations were administered in 12 weeks according to the group 1 schedule. Pathological slides stained with hematoxylin and eosin were reviewed by 1 pathologist (MT). All tumors were classified into 3 grades (G1, G2 or G3) according to 1973 WHO classifications.14 This trial enrolled only patients with superficial transitional carcinoma of the bladder (Ta and T1). Patients who were histopathologically diagnosed with muscle invasive carcinoma or concomitant carcinoma in situ were excluded from this trial. In both groups urinalysis and cytologic examination of urine samples were performed every month after TUR. Cystoscopy was performed every 3 months for 2 years and at 6-month intervals thereafter. Local and systemic toxicities
were also monitored. Routine laboratory tests (hematology and biochemistry) were performed before and at 1, 3, 6, 9 and 12 months after TUR, and then every 6 months thereafter. All patients were followed until the first bladder recurrence, and treatment after recurrence was left to the discretion of each investigator. In the present trial we hypothesized that the 3-year nonrecurrence rates of groups 1 and 2 would be 80% and 60%, respectively. Using a power of 75% and a 2-sided test at 5% we calculated the sample size required to detect a difference of 20%. As a result we decided that recruitment of 76 patients in each group, 172 total, was needed. Considering ineligible cases we set the number of patients to be recruited as 85 in each group, 170 total. Statistical analysis of the data was performed using the Statistical Analysis System Package (SAS Institute, Japan, Tokyo). The significance of any differences in patient characteristics, and the incidence and severity of toxicity between the 2 groups was tested by the chi-square test or Fisher’s exact test. The difference in time to first recurrence was assessed by the nonrecurrence curve calculated by the Kaplan-Meier method. Statistical significance was analyzed by the generalized Wilcoxon test. The Cox proportional hazards model was used to adjust for any possible bias as in background factors. Differences were considered statistically significant at p ⬍0.05. Hazard of recurrence was defined as the fraction of patients who had recurrence in every 6-month period. The hazard ratios in each time period were calculated using maximum likelihood estimates from a piecewise exponential model.15 RESULTS
Between March 1993 and July 1995 a total of 171 patients were enrolled in this study at 18 collaborating hospitals. A total of 21 patients were diagnosed as ineligible by histopathological review. As a result 150 patients (77 in group 1 and 73 in group 2) were eligible. No significant differences were seen between each group in terms of age, sex, number of tumors, tumor size, pathological grade or stage (table 1). Average patient age was 66.2 years in group 1 and 63.9 in group 2 (chi-square test p ⫽ 0.2855). Median followup was 30.6 months (range 0.2 to 65.2), 33.6 months (range 0.2 to 61.4) in group 1 and 29.9 months (range 1.1 to 65.2) in group 2. Of the 150 patients all except 16 (7 in group 1 and 9 in group 2) received the first intravesical instillation immediately after TUR, and there was no statistical difference between the 2 groups in regard to the time of the first instillation. In the clinical course 22 patients were not regarded as evaluable. Of those 22 patients, 13 (10 in group 1 and 3 in group 2) were lost to followup with inadequate treatment (less than 13
TABLE 1. Patient characteristics
FIG. 1. Treatment schedule of group 1 (A) and group 2 (B). EPI-ADM, epirubicin-doxorubicin.
Sex: Males Females Multiplicity: 1 2 or More Tumor size (cm): Less than 1 1–3 3–5 Greater than 5 Tumor grade: G1 G2 G3 Stage: Ta T1
No. Group 1
No. Group 2
55 22
55 18
47 30
44 29
26 47 4 0
27 40 4 2
16 50 11
21 46 6
61 16
62 11
p Value (chi-square test) 0.7210 0.9430 0.4794
0.3316
0.3055
155
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION
and 6 instillations in groups 1 and 2, respectively), 1 in group 1 withdrew due to side effects, 2 in group 2 had inadequate examination after TUR, and 6 (3 in group 1 and 3 in group 2) had other protocol violations. In the present study we performed statistical analysis on an intent to treat design. Then we analyzed the nonrecurrence rate for all 150 eligible patients. In followup 10 group 1 patients (13.0%) and 23 group 2 patients (31.5%) had recurrent disease. The overall 3-year nonrecurrence rate was 85.2% in group 1 and 63.9% in group 2. The nonrecurrence rate of group 1 was always higher than that of group 2 throughout the observation period (fig. 2), and there is a significant difference between the 2 groups (the generalized Wilcoxon test p ⫽ 0.005). The Cox proportional hazards model was used to adjust for any possible bias due to the imbalance of patient backgrounds. The findings of univariate analysis are shown in table 2. Of 7 background factors 3 (group, multiplicity and tumor stage) were found to influence the nonrecurrence rate significantly. In a multivariate study the significance of the difference in the nonrecurrence rate between groups 1 and 2 was adjusted for 2 other factors, ie, multiplicity and tumor stage. As shown in table 3 significance was also detected by multivariate analysis. Figure 3 demonstrates each 6-month hazard of recurrence after TUR. In group 2 the peak hazard of recurrence occurred at about 1.5 and 4 years after TUR. In group 1 the hazard of recurrence was suppressed throughout the observation period, especially in the first 3 years. Side effects were assessed in all 150 eligible patients. The most common local toxicities were irritative bladder symptoms (micturition pain and frequency of urination) and hematuria (table 4). Severe local toxicity was observed in 4 of the 77 group 1 patients and 6 of 73 group 2 patients. Instillation was stopped in 1 group 1 patient because of severe micturition pain. No significant difference in frequency of local toxicity was observed between the 2 groups. Systemic toxicity was seen in 2 group 2 patients (1 with general fatigue and the other with low grade fever). However, severe systemic toxicity did not develop in patients in either group. DISCUSSION
Epirubicin, a derivative of doxorubicin, is an anthracycline antibiotic and undergoes minimal transurothelial absorption.16 It was reported that epirubicin has antitumor efficacy which decreases the nonrecurrence rate as an intravesical chemotherapeutic agent after TUR for superficial bladder cancer.6 – 8 In a large controlled study performed by the Eu-
TABLE 2. Univariate analysis of factors influencing recurrence Variable
HR (95% CI)
p Value
Group Age Sex Multiplicity Tumor size Tumor grade Tumor stage
0.387 (0.1821–0.8243) 0.431 (0.1927–0.9643) 0.746 (0.3336–1.6681) 0.338 (0.1636–0.6968) 0.367 (0.0691–1.9437) 0.473 (0.1419–1.5785) 0.0769 (0.0769–0.8184)
0.0138 0.4050 0.4754 0.0033 0.2837 0.2235 0.0219
TABLE 3. Multivariate adjusted hazard ratios for recurrence Variable ⫹ Models
HR (95% CI)
p Value
Univariate Adjusted for multiplicity Adjusted for tumor stage Adjusted for multiplicity ⫹ tumor stage
0.387 (0.1821–0.8243) 0.393 (0.1843–0.8367) 0.362 (0.1688–0.7743) 0.361 (0.1679–0.7759)
0.0138 0.0154 0.0088 0.0091
ropean Organization for the Research and Treatment of Cancer, Oosterlinck et al reported that a single instillation of 80 mg epirubicin given immediately (within 6 hours) after TUR decreased the recurrence rate by nearly 50% compared to the water instilled group.7 Epirubicin was also reported to be less toxic than doxorubicin.4, 5 We have performed randomized controlled trials of intravesical adjuvant chemotherapy using doxorubicin to determine its impact on recurrence after TUR in superficial bladder cancer.17, 18 In consideration of the previously described characteristics of epirubicin, we chose epirubicin as an instillation drug and decided that the optimal dose was 30 mg in 30 ml of normal saline. There are many reports regarding the timing and duration of instillation therapy including epirubicin.6 – 8, 10, 11 Instillation of epirubicin immediately after TUR of superficial bladder cancer significantly decreased disease recurrence compared with instillation of water.6 – 8 It was reported that 1 of the mechanisms of intravesical recurrence after TUR of bladder cancer is implantation of tumor cells in the bladder wall,19 and instillation immediately after TUR seems to be effective in preventing free tumor cells from implanting in the bladder wall during and/or after TUR. Although immediate intravesical instillation after TUR is effective, we were afraid that few hospitals would be able to perform this immediate intravesical instillation at the start of this clinical trial. Therefore, we defined first instillation as within 24 hours after TUR in consideration of feasibility. In actual practice only 16 patients (7 in group 1 and 9 in group 2)
FIG. 2. Nonrecurrence curve after TUR in 2 groups. Nonrecurrence rate of group 1 (A) was significantly higher than that of group 2 (B) (p ⫽ 0.005).
156
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION
FIG. 3. Hazard of tumor recurrence for groups 1 (A) and 2 (B). Hazard of recurrence in group 1 was suppressed throughout observation period, especially in first 3 years, compared to that in group 2.
TABLE 4. Local toxicity No. Group 1
No. Group 2
p Value (chi-square test)
Macrohematuria: None 54 61 Mild 18 9 0.263 Moderate 4 2 Severe 1 1 Micturition pain: None 48 46 Mild 27 22 0.619 Moderate 1 3 Severe 1 2 Frequency: None 52 51 Mild 17 17 0.551 Moderate 6 2 Severe 2 3 In regard to systemic toxicity there were no cases of general fatigue or low grade fever in group 1. In group 2 there was 1 case of general fatigue and 1 of low grade fever.
received the first intravesical instillation 6 to 24 hours after TUR. Thus, all except those 16 received the first instillation immediately after TUR. There was no statistical difference between the 2 groups in terms of time to first intravesical instillation after TUR. In this study the nonrecurrence rate of bladder tumors in the long-term instillation group was much higher than that in the short-term instillation group throughout the observation period, especially in the first 3 years. This significance was maintained even after adjustment for other significant factors by multivariate analysis. In regard to epirubicin, it was reported that prospective studies did not show any difference in the recurrence-free interval or long-term recurrence rate compared to no maintenance.6, 10, 11 Ali-el-Dein et al reported that single dose immediate epirubicin therapy was as effective as delayed maintenance therapy.6 However, their trial differed in regard to patient selection and instillation schedule. Moreover, G3 tumors were excluded from study and immediate instillation was not performed in the long-term group in their trial. Okamura et al performed randomized trials comparing 6 and 17 instillations of 40 mg/40 ml epirubicin in normal saline in patients with new or recurrent bladder cancer.10 They reported no significant difference between the 2 groups in 2-year nonrecurrence rates, which were 77.2% and 75.1%, respectively. In our trial the
3-year nonrecurrence rates were 63.9% in the short-term group and 85.2% in the long-term group using 30 mg/30 ml epirubicin in normal saline. Moreover, our trial included only new, untreated cases. Nomata et al also reported no significant difference between 12 and 19 instillations using epirubicin at 30 mg/30 ml.11 Their 3-year nonrecurrence rates were 55.1% in the short-term group and 48.5% in the longterm group. Their trial included recurrent tumors but excluded G3 tumors, whereas our trial excluded recurrent tumors and included G3 tumors in eligible cases. Furthermore, their treatment schedule was different from ours in that the short-term arm was 12 instillations and there was no perioperative instillation. These findings suggest that the optimal duration of instillation as intravesical chemotherapy of epirubicin after TUR for superficial bladder cancer remains unclear because of the differences in patient selection and drug concentration among the trials. Using a smoothed hazard function Hinotsu et al reported that a peak of tumor recurrence after TUR of Ta/T1 bladder cancer was detected during the first 500 days after operation, and that a prophylactic effect was achieved during the same time.20 In our study the nonrecurrence rate of bladder tumors in the long-term group was much higher than that in the short-term group throughout the observation period, especially in the first 3 years. It was surmised that the peak hazard of recurrence at around 1.5 years could not be suppressed because of completion of the instillations after only 12 weeks in the short-term group. In regard to toxicity no significant differences in the frequency or degree of local toxicity were observed between the 2 treatment groups. Slight systemic toxicity (general fatigue and low grade fever) was observed in 2 patients in the shortterm instillation group. However, these adverse effects were not definitely determined to be due to epirubicin, and the instillation could be continued. Epirubicin toxicity was minimal as previously reported.4, 5, 16 Therefore, the instillation of epirubicin in the present study is as tolerable as adjuvant intravesical chemotherapy. CONCLUSIONS
Long-term instillation of epirubicin is considered safe and more effective than short-term instillation in preventing recurrence of superficial bladder cancer after TUR. Considering the high 3-year nonrecurrence rate in the long-term
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION
group, long-term instillation still seems to be useful. Further study is needed to clarify characteristics which need longterm instillation of epirubicin after TUR in preventing tumor recurrence.
11.
REFERENCES
1. Soloway, M. S.: Intravesical therapy for bladder cancer. Urol Clin N Amer, 15: 661, 1988 2. Nilsson, S., Ragnhammar, P., Glimelius, B., Nygen, P. and SBUgroup. Swedish Council of Technology Assessment in Health Care: A systematic overview of chemotherapy effects in urothelial bladder cancer. Acta Oncol, 40: 371, 2001 3. Lamm, D. L., Blumenstein, B. A., Crawford, E. D., Montie, J. E., Scardino, P., Grossman, H. B. et al: A randomized trial of intravesical doxorubicin and immunotherapy with bacille Calmette-Guerin for transitional-cell carcinoma of the bladder. N Engl J Med, 325: 1205, 1991 4. Badalament, R. A. and Farah, R. N.: Treatment of superficial bladder cancer with intravesical chemotherapy. Semin Surg Oncol, 13: 335, 1997 5. Ali-El-Dein, B., El-Baz, M., Aly, A. N. M., Shamaa, S. and Ashamallah, A.: Intravesical epirubicin versus doxorubicin for superficial bladder tumors (stages pTa and pT1): a randomized prospective study. J Urol, 158: 68, 1997 6. Ali-el-Dein, B., Nabeeh, A., el-Baz, M., Shamaa, S. and Ashamallah, A.: Single-dose versus multiple instillations of epirubicin as prophylaxis for recurrence after transurethral resection of pTa and pT1 transitional-cell bladder tumours: a prospective, randomized controlled study. Br J Urol, 79: 731, 1997 7. Oosterlinck, W., Kurth, K. H., Schro¨ der, F., Bultinck, J., Hammond, B., Sylvester, R. et al: A prospective European Organization for Research and Treatment of Cancer Genitourinary Group randomized trial comparing transurethral resection followed by a single intravesical instillation of epirubicin or water in single stage Ta, T1 papillary carcinoma of the bladder. J Urol, 149: 749, 1993 8. Melekos, M. D., Dauaher, H., Fokaefs, E. and Barbalias, G.: Intravesical instillations of 4-epi-doxorubicin (epirubicin) in the prophylactic treatment of superficial bladder cancer: results of a controlled prospective study. J Urol, 147: 371, 1992 9. Solsona, E., Iborra, I., Rico´ s, J. V., Monro´ s, J. L., Casanova, J. and Dumont, R.: Effectiveness of a single immediate mitomycin C instillation in patients with low risk superficial bladder cancer: short- and long-term followup. J Urol, 161: 1120, 1999 10. Okamura, K., Kinukawa, T., Tsumura, Y., Otani, T., Itoh, H., Kobayashi, H. et al: A randomized study of short-versus longterm intravesical epirubicin instillation for superficial bladder
12.
13.
14.
15. 16.
17.
18.
19.
20.
157
cancer. Nagoya University Urological Oncology Group. Eur Urol, 33: 285, 1998 Nomata, K., Noguchi, M., Kanetake, H., Tsuda, N., Hayashi, M., Yamashita, S. et al: Intravesical adjuvant chemotherapy for superficial transitional cell bladder carcinoma: results of a randomized trial with epirubicin comparing short-term versus long-term maintenance treatment. Cancer Chemother Pharmacol, 50: 266, 2002 Flamm, J.: Long-term versus short-term doxorubicin hydrochloride instillation after transurethral resection of superficial bladder cancer. Eur Urol, 17: 119, 1990 Tolley, D. A., Parmar, M. K. B., Grigor, K. M., Lallemand, G. and the Medical Research Council Superficial Bladder Cancer Working Party: The effect of intravesical mitomycin C on recurrence of newly diagnosed superficial bladder cancer: a further report with 7 years of followup. J Urol, 155: 1233, 1996 Mostofi, F. K., Sobin, H. L. and Torloni, H.: Histological typing of urinary bladder tumours. International Classification of Tumours, No. 10. Geneva, Switzerland: World Health Organization, 1973 Holford, T. R.: Life tables with concomitant information. Biometrics, 32: 587, 1976 Cersosimo, R. J. and Hong, W. K.: Epirubicin; a review of the pharmacology, clinical activity, and adverse effects of an adriamycin analogue. J Clin Oncol, 4: 425, 1986 Ueda, T., Naito, S., Iguchi, A., Sagiyama, K., Osada, Y., Ariyoshi, A. et al: Adjuvant chemotherapy with early intravesical instillation of adriamycin and long-term oral administration of 5-fluorouracil in superficial bladder cancer. The Kyushu University Urological Oncology Group. Cancer Chemother Pharmacol, suppl., 30: S31, 1992 Naito, S., Kotoh, S., Omoto, T., Osada, Y., Sagiyama, K., Iguchi, A. et al: Prophylactic intravesical instillation chemotherapy against recurrence after a transurethral resection of superficial bladder cancer: a randomized controlled trial of doxorubicin plus verapamil versus doxorubicin alone. The Kyushu University Urological Oncology Group. Cancer Chemother Pharmacol, 42: 367, 1998 Akaza, H.: New strategy of bio-chemoprevention on recurrence of superficial bladder cancer based on a hypothesis of the mechanism of recurrence. Gan To Kagaku Ryoho, suppl., 24: 253, 1997 Hinotsu, S., Akaza, H., Ohashi, Y. and Kotake, K.: Intravesical chemotherapy for maximum prophylaxis of new early phase superficial bladder carcinoma treated by transurethral resection: a combined analysis of trials by Japanese Urological Cancer Research Group using smoothed hazard function. Cancer, 86: 1818, 1999
Vol. 171, 153–157, January 2004 Printed in U.S.A.
DOI: 10.1097/01.ju.0000100386.07370.0a
A RANDOMIZED CONTROLLED TRIAL OF SHORT-TERM VERSUS LONG-TERM PROPHYLACTIC INTRAVESICAL INSTILLATION CHEMOTHERAPY FOR RECURRENCE AFTER TRANSURETHRAL RESECTION OF TA/T1 TRANSITIONAL CELL CARCINOMA OF THE BLADDER HIROFUMI KOGA, KENTARO KUROIWA, AKITO YAMAGUCHI, YUKIO OSADA, MASAZUMI TSUNEYOSHI AND SEIJI NAITO* From the Departments of Urology, Graduate School of Medical Sciences, Kyushu University (HK, KK, SN) and Harasanshin General Hospital (AY), Kyushu University Urological Oncology Group (HK, KK, AY, YO, SN), Department of Anatomic Pathology, Graduate School of Medical Sciences, Kyushu University (MT), Fukuoka and Department of Urology, Miyazaki Medical College, Miyazaki, Japan (YO)
ABSTRACT
Purpose: In a prospective randomized controlled study, we investigated the optimal schedule for intravesical instillation of epirubicin for maximizing its effect on prophylaxis and disease progression after transurethral resection of newly diagnosed Ta/T1 bladder cancer. Materials and Methods: The patients were instilled with epirubicin (30 mg/30 ml in normal saline) within 24 hours after transurethral resection and then randomized into 2 groups after a definite histopathological diagnosis of Ta/T1 bladder cancer. One group of 77 patients received 19 intravesical instillations of epirubicin in the year after transurethral resection (group 1). The second group of 73 patients received 9 intravesical instillations of epirubicin during the 3 months after transurethral resection (group 2). Nonrecurrence rates and toxicity were compared. Results: In the followup period, 10 group 1 patients (13.0%) and 23 group 2 patients (31.5%) had recurrent disease. The 3-year nonrecurrence rate was 85.2% in group 1, whereas it was 63.9% in group 2. The nonrecurrence rate of group 1 was significantly higher than that of group 2 throughout the observation period (p ⫽ 0.005). The incidence and severity of toxicity were not significantly different between the 2 groups. Conclusions: Our study indicates that long-term instillation of epirubicin is more effective than short-term instillation in preventing recurrence after transurethral resection of Ta/T1 bladder cancer. KEY WORDS: randomized controlled trial, bladder neoplasms; administration, intravesical; drug therapy, epirubicin
In bladder cancer the most common histological phenotype is urothelial carcinoma. About 70% of newly diagnosed cases are confined to the mucosa or lamina propria (Ta and T1). Almost all cases of Ta/T1 disease macroscopically exhibit papillary tumors, and transurethral resection (TUR) is the gold standard treatment for them. However, TUR alone is not adequate therapy for superficial bladder cancer. The recurrence rate of Ta/T1 bladder cancer within 5 years after TUR is about 50% to 70%. In addition, 5% to 20% of cases show pathologically progressive disease.1 The purpose of intravesical chemotherapy after TUR (adjuvant therapy) is to eradicate existing cancer cells and to prevent recurrence and progression. To date, intravesical chemotherapy in an adjuvant setting for TUR of Ta/T1 disease decreases the short-term (1 to 3-year) recurrence rate by approximately 20% and long-term rate (8-year) by 8.2%.2 Mitomycin C, doxorubicin and epirubicin are commonly used chemotherapeutic agents. The intravesical instillation of bacillus Calmette-Guerin has been shown to be more effective than those chemotherapeutic agents, including doxorubicin, in preventing recurrence after TUR.3 However, bacillus Calmette-Guerin has the disadvantage of causing various
and frequent local and/or systemic side effects. Recently, epirubicin (4-epi-doxorubicin) has shown an in vitro cytotoxic effect on transitional cell carcinoma lines, and in clinical use is as or more effective but less toxic than its parent compound, doxorubicin.4, 5 The timing and duration of intravesical chemotherapy have been extensively studied. A single intravesical instillation of epirubicin,6 – 8 mitomycin C9 and doxorubicin2 immediately after TUR significantly decreased the tumor recurrence rate compared with no adjuvant intravesical therapy. It has been reported that long-term maintenance instillation therapy does not further increase the recurrence-free interval or the long-term recurrence rate compared with immediate short-term instillation therapy after surgery.6, 10 –13 Solsona et al reported that the beneficial effect of long-term instillation was limited to early (1-year) recurrence and not maintained during long-term followup use of mitomycin C.9 On the other hand, Tolley et al reported that 5 instillations of mitomycin C offered a slight advantage over 1 instillation.13 The discrepancies in these reports are due to the differences in the dose of chemotherapeutic agents and enrolled patient backgrounds. Thus, the timing and duration of intravesical chemotherapy remain controversial. In this randomized controlled study we investigated the optimal time of intravesical instillation of epirubicin for its effect on prophylaxis after TUR of newly diagnosed Ta/T1 bladder cancer.
Accepted for publication July 11, 2003. * Correspondence and requests for reprints: Department of Urology, Graduate School of Medical Sciences, 3-1-1 Maidashi, Higashi-ku, Fukuoka, 812-8582, Japan (telephone: 81-92-642-5603; FAX: 81-92-642-5618; e-mail: [email protected]). 153
154
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION MATERIALS AND METHODS
This study was a multicenter, prospective, randomized controlled trial. The criteria of eligibility included new, untreated transitional cell carcinoma of the bladder, Ta or T1 disease, no other active neoplasms and no serious complications. Patients who had a history of urothelial carcinoma of the renal pelvis or ureter were excluded from study. Before treatment patient history, physical examination, urinalysis, urine cytology, complete blood count, blood urea nitrogen, serum creatinine, liver function and electrocardiography were performed. Chest plain x-rays and excretory urography were also performed. At entry in this trial, no patients had evidence of a residual tumor as judged by endoscopic examination and urine cytology. Registration was performed by facsimile at Kyushu University Urological Oncology Group Center. All cases were randomly allocated to a long-term instillation group (group 1) or a short-term instillation group (group 2) by a central computer. All patients were fully informed about this trial and signed the informed consent form. In both groups 30 mg epirubicin dissolved in 30 ml of normal saline was instilled into the bladder through a sterile catheter after emptying the bladder. Patients were instructed to refrain from voiding for 2 hours after instillation. The intravesical instillation schedule of each group is diagrammed (fig. 1). In group 1 instillations 1 and 2 were given within 24 hours and 2 or 3 days after TUR, respectively. The instillation was subsequently given weekly for 2 weeks, and every 2 weeks for an additional 14 weeks. After 4 months the instillation was given every month for 8 months. Thus, a total of 19 instillations were administered in 1 year in group 1. On the other hand, in group 2 a total of 9 instillations were administered in 12 weeks according to the group 1 schedule. Pathological slides stained with hematoxylin and eosin were reviewed by 1 pathologist (MT). All tumors were classified into 3 grades (G1, G2 or G3) according to 1973 WHO classifications.14 This trial enrolled only patients with superficial transitional carcinoma of the bladder (Ta and T1). Patients who were histopathologically diagnosed with muscle invasive carcinoma or concomitant carcinoma in situ were excluded from this trial. In both groups urinalysis and cytologic examination of urine samples were performed every month after TUR. Cystoscopy was performed every 3 months for 2 years and at 6-month intervals thereafter. Local and systemic toxicities
were also monitored. Routine laboratory tests (hematology and biochemistry) were performed before and at 1, 3, 6, 9 and 12 months after TUR, and then every 6 months thereafter. All patients were followed until the first bladder recurrence, and treatment after recurrence was left to the discretion of each investigator. In the present trial we hypothesized that the 3-year nonrecurrence rates of groups 1 and 2 would be 80% and 60%, respectively. Using a power of 75% and a 2-sided test at 5% we calculated the sample size required to detect a difference of 20%. As a result we decided that recruitment of 76 patients in each group, 172 total, was needed. Considering ineligible cases we set the number of patients to be recruited as 85 in each group, 170 total. Statistical analysis of the data was performed using the Statistical Analysis System Package (SAS Institute, Japan, Tokyo). The significance of any differences in patient characteristics, and the incidence and severity of toxicity between the 2 groups was tested by the chi-square test or Fisher’s exact test. The difference in time to first recurrence was assessed by the nonrecurrence curve calculated by the Kaplan-Meier method. Statistical significance was analyzed by the generalized Wilcoxon test. The Cox proportional hazards model was used to adjust for any possible bias as in background factors. Differences were considered statistically significant at p ⬍0.05. Hazard of recurrence was defined as the fraction of patients who had recurrence in every 6-month period. The hazard ratios in each time period were calculated using maximum likelihood estimates from a piecewise exponential model.15 RESULTS
Between March 1993 and July 1995 a total of 171 patients were enrolled in this study at 18 collaborating hospitals. A total of 21 patients were diagnosed as ineligible by histopathological review. As a result 150 patients (77 in group 1 and 73 in group 2) were eligible. No significant differences were seen between each group in terms of age, sex, number of tumors, tumor size, pathological grade or stage (table 1). Average patient age was 66.2 years in group 1 and 63.9 in group 2 (chi-square test p ⫽ 0.2855). Median followup was 30.6 months (range 0.2 to 65.2), 33.6 months (range 0.2 to 61.4) in group 1 and 29.9 months (range 1.1 to 65.2) in group 2. Of the 150 patients all except 16 (7 in group 1 and 9 in group 2) received the first intravesical instillation immediately after TUR, and there was no statistical difference between the 2 groups in regard to the time of the first instillation. In the clinical course 22 patients were not regarded as evaluable. Of those 22 patients, 13 (10 in group 1 and 3 in group 2) were lost to followup with inadequate treatment (less than 13
TABLE 1. Patient characteristics
FIG. 1. Treatment schedule of group 1 (A) and group 2 (B). EPI-ADM, epirubicin-doxorubicin.
Sex: Males Females Multiplicity: 1 2 or More Tumor size (cm): Less than 1 1–3 3–5 Greater than 5 Tumor grade: G1 G2 G3 Stage: Ta T1
No. Group 1
No. Group 2
55 22
55 18
47 30
44 29
26 47 4 0
27 40 4 2
16 50 11
21 46 6
61 16
62 11
p Value (chi-square test) 0.7210 0.9430 0.4794
0.3316
0.3055
155
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION
and 6 instillations in groups 1 and 2, respectively), 1 in group 1 withdrew due to side effects, 2 in group 2 had inadequate examination after TUR, and 6 (3 in group 1 and 3 in group 2) had other protocol violations. In the present study we performed statistical analysis on an intent to treat design. Then we analyzed the nonrecurrence rate for all 150 eligible patients. In followup 10 group 1 patients (13.0%) and 23 group 2 patients (31.5%) had recurrent disease. The overall 3-year nonrecurrence rate was 85.2% in group 1 and 63.9% in group 2. The nonrecurrence rate of group 1 was always higher than that of group 2 throughout the observation period (fig. 2), and there is a significant difference between the 2 groups (the generalized Wilcoxon test p ⫽ 0.005). The Cox proportional hazards model was used to adjust for any possible bias due to the imbalance of patient backgrounds. The findings of univariate analysis are shown in table 2. Of 7 background factors 3 (group, multiplicity and tumor stage) were found to influence the nonrecurrence rate significantly. In a multivariate study the significance of the difference in the nonrecurrence rate between groups 1 and 2 was adjusted for 2 other factors, ie, multiplicity and tumor stage. As shown in table 3 significance was also detected by multivariate analysis. Figure 3 demonstrates each 6-month hazard of recurrence after TUR. In group 2 the peak hazard of recurrence occurred at about 1.5 and 4 years after TUR. In group 1 the hazard of recurrence was suppressed throughout the observation period, especially in the first 3 years. Side effects were assessed in all 150 eligible patients. The most common local toxicities were irritative bladder symptoms (micturition pain and frequency of urination) and hematuria (table 4). Severe local toxicity was observed in 4 of the 77 group 1 patients and 6 of 73 group 2 patients. Instillation was stopped in 1 group 1 patient because of severe micturition pain. No significant difference in frequency of local toxicity was observed between the 2 groups. Systemic toxicity was seen in 2 group 2 patients (1 with general fatigue and the other with low grade fever). However, severe systemic toxicity did not develop in patients in either group. DISCUSSION
Epirubicin, a derivative of doxorubicin, is an anthracycline antibiotic and undergoes minimal transurothelial absorption.16 It was reported that epirubicin has antitumor efficacy which decreases the nonrecurrence rate as an intravesical chemotherapeutic agent after TUR for superficial bladder cancer.6 – 8 In a large controlled study performed by the Eu-
TABLE 2. Univariate analysis of factors influencing recurrence Variable
HR (95% CI)
p Value
Group Age Sex Multiplicity Tumor size Tumor grade Tumor stage
0.387 (0.1821–0.8243) 0.431 (0.1927–0.9643) 0.746 (0.3336–1.6681) 0.338 (0.1636–0.6968) 0.367 (0.0691–1.9437) 0.473 (0.1419–1.5785) 0.0769 (0.0769–0.8184)
0.0138 0.4050 0.4754 0.0033 0.2837 0.2235 0.0219
TABLE 3. Multivariate adjusted hazard ratios for recurrence Variable ⫹ Models
HR (95% CI)
p Value
Univariate Adjusted for multiplicity Adjusted for tumor stage Adjusted for multiplicity ⫹ tumor stage
0.387 (0.1821–0.8243) 0.393 (0.1843–0.8367) 0.362 (0.1688–0.7743) 0.361 (0.1679–0.7759)
0.0138 0.0154 0.0088 0.0091
ropean Organization for the Research and Treatment of Cancer, Oosterlinck et al reported that a single instillation of 80 mg epirubicin given immediately (within 6 hours) after TUR decreased the recurrence rate by nearly 50% compared to the water instilled group.7 Epirubicin was also reported to be less toxic than doxorubicin.4, 5 We have performed randomized controlled trials of intravesical adjuvant chemotherapy using doxorubicin to determine its impact on recurrence after TUR in superficial bladder cancer.17, 18 In consideration of the previously described characteristics of epirubicin, we chose epirubicin as an instillation drug and decided that the optimal dose was 30 mg in 30 ml of normal saline. There are many reports regarding the timing and duration of instillation therapy including epirubicin.6 – 8, 10, 11 Instillation of epirubicin immediately after TUR of superficial bladder cancer significantly decreased disease recurrence compared with instillation of water.6 – 8 It was reported that 1 of the mechanisms of intravesical recurrence after TUR of bladder cancer is implantation of tumor cells in the bladder wall,19 and instillation immediately after TUR seems to be effective in preventing free tumor cells from implanting in the bladder wall during and/or after TUR. Although immediate intravesical instillation after TUR is effective, we were afraid that few hospitals would be able to perform this immediate intravesical instillation at the start of this clinical trial. Therefore, we defined first instillation as within 24 hours after TUR in consideration of feasibility. In actual practice only 16 patients (7 in group 1 and 9 in group 2)
FIG. 2. Nonrecurrence curve after TUR in 2 groups. Nonrecurrence rate of group 1 (A) was significantly higher than that of group 2 (B) (p ⫽ 0.005).
156
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION
FIG. 3. Hazard of tumor recurrence for groups 1 (A) and 2 (B). Hazard of recurrence in group 1 was suppressed throughout observation period, especially in first 3 years, compared to that in group 2.
TABLE 4. Local toxicity No. Group 1
No. Group 2
p Value (chi-square test)
Macrohematuria: None 54 61 Mild 18 9 0.263 Moderate 4 2 Severe 1 1 Micturition pain: None 48 46 Mild 27 22 0.619 Moderate 1 3 Severe 1 2 Frequency: None 52 51 Mild 17 17 0.551 Moderate 6 2 Severe 2 3 In regard to systemic toxicity there were no cases of general fatigue or low grade fever in group 1. In group 2 there was 1 case of general fatigue and 1 of low grade fever.
received the first intravesical instillation 6 to 24 hours after TUR. Thus, all except those 16 received the first instillation immediately after TUR. There was no statistical difference between the 2 groups in terms of time to first intravesical instillation after TUR. In this study the nonrecurrence rate of bladder tumors in the long-term instillation group was much higher than that in the short-term instillation group throughout the observation period, especially in the first 3 years. This significance was maintained even after adjustment for other significant factors by multivariate analysis. In regard to epirubicin, it was reported that prospective studies did not show any difference in the recurrence-free interval or long-term recurrence rate compared to no maintenance.6, 10, 11 Ali-el-Dein et al reported that single dose immediate epirubicin therapy was as effective as delayed maintenance therapy.6 However, their trial differed in regard to patient selection and instillation schedule. Moreover, G3 tumors were excluded from study and immediate instillation was not performed in the long-term group in their trial. Okamura et al performed randomized trials comparing 6 and 17 instillations of 40 mg/40 ml epirubicin in normal saline in patients with new or recurrent bladder cancer.10 They reported no significant difference between the 2 groups in 2-year nonrecurrence rates, which were 77.2% and 75.1%, respectively. In our trial the
3-year nonrecurrence rates were 63.9% in the short-term group and 85.2% in the long-term group using 30 mg/30 ml epirubicin in normal saline. Moreover, our trial included only new, untreated cases. Nomata et al also reported no significant difference between 12 and 19 instillations using epirubicin at 30 mg/30 ml.11 Their 3-year nonrecurrence rates were 55.1% in the short-term group and 48.5% in the longterm group. Their trial included recurrent tumors but excluded G3 tumors, whereas our trial excluded recurrent tumors and included G3 tumors in eligible cases. Furthermore, their treatment schedule was different from ours in that the short-term arm was 12 instillations and there was no perioperative instillation. These findings suggest that the optimal duration of instillation as intravesical chemotherapy of epirubicin after TUR for superficial bladder cancer remains unclear because of the differences in patient selection and drug concentration among the trials. Using a smoothed hazard function Hinotsu et al reported that a peak of tumor recurrence after TUR of Ta/T1 bladder cancer was detected during the first 500 days after operation, and that a prophylactic effect was achieved during the same time.20 In our study the nonrecurrence rate of bladder tumors in the long-term group was much higher than that in the short-term group throughout the observation period, especially in the first 3 years. It was surmised that the peak hazard of recurrence at around 1.5 years could not be suppressed because of completion of the instillations after only 12 weeks in the short-term group. In regard to toxicity no significant differences in the frequency or degree of local toxicity were observed between the 2 treatment groups. Slight systemic toxicity (general fatigue and low grade fever) was observed in 2 patients in the shortterm instillation group. However, these adverse effects were not definitely determined to be due to epirubicin, and the instillation could be continued. Epirubicin toxicity was minimal as previously reported.4, 5, 16 Therefore, the instillation of epirubicin in the present study is as tolerable as adjuvant intravesical chemotherapy. CONCLUSIONS
Long-term instillation of epirubicin is considered safe and more effective than short-term instillation in preventing recurrence of superficial bladder cancer after TUR. Considering the high 3-year nonrecurrence rate in the long-term
INTRAVESICAL CHEMOTHERAPY AFTER BLADDER CANCER RESECTION
group, long-term instillation still seems to be useful. Further study is needed to clarify characteristics which need longterm instillation of epirubicin after TUR in preventing tumor recurrence.
11.
REFERENCES
1. Soloway, M. S.: Intravesical therapy for bladder cancer. Urol Clin N Amer, 15: 661, 1988 2. Nilsson, S., Ragnhammar, P., Glimelius, B., Nygen, P. and SBUgroup. Swedish Council of Technology Assessment in Health Care: A systematic overview of chemotherapy effects in urothelial bladder cancer. Acta Oncol, 40: 371, 2001 3. Lamm, D. L., Blumenstein, B. A., Crawford, E. D., Montie, J. E., Scardino, P., Grossman, H. B. et al: A randomized trial of intravesical doxorubicin and immunotherapy with bacille Calmette-Guerin for transitional-cell carcinoma of the bladder. N Engl J Med, 325: 1205, 1991 4. Badalament, R. A. and Farah, R. N.: Treatment of superficial bladder cancer with intravesical chemotherapy. Semin Surg Oncol, 13: 335, 1997 5. Ali-El-Dein, B., El-Baz, M., Aly, A. N. M., Shamaa, S. and Ashamallah, A.: Intravesical epirubicin versus doxorubicin for superficial bladder tumors (stages pTa and pT1): a randomized prospective study. J Urol, 158: 68, 1997 6. Ali-el-Dein, B., Nabeeh, A., el-Baz, M., Shamaa, S. and Ashamallah, A.: Single-dose versus multiple instillations of epirubicin as prophylaxis for recurrence after transurethral resection of pTa and pT1 transitional-cell bladder tumours: a prospective, randomized controlled study. Br J Urol, 79: 731, 1997 7. Oosterlinck, W., Kurth, K. H., Schro¨ der, F., Bultinck, J., Hammond, B., Sylvester, R. et al: A prospective European Organization for Research and Treatment of Cancer Genitourinary Group randomized trial comparing transurethral resection followed by a single intravesical instillation of epirubicin or water in single stage Ta, T1 papillary carcinoma of the bladder. J Urol, 149: 749, 1993 8. Melekos, M. D., Dauaher, H., Fokaefs, E. and Barbalias, G.: Intravesical instillations of 4-epi-doxorubicin (epirubicin) in the prophylactic treatment of superficial bladder cancer: results of a controlled prospective study. J Urol, 147: 371, 1992 9. Solsona, E., Iborra, I., Rico´ s, J. V., Monro´ s, J. L., Casanova, J. and Dumont, R.: Effectiveness of a single immediate mitomycin C instillation in patients with low risk superficial bladder cancer: short- and long-term followup. J Urol, 161: 1120, 1999 10. Okamura, K., Kinukawa, T., Tsumura, Y., Otani, T., Itoh, H., Kobayashi, H. et al: A randomized study of short-versus longterm intravesical epirubicin instillation for superficial bladder
12.
13.
14.
15. 16.
17.
18.
19.
20.
157
cancer. Nagoya University Urological Oncology Group. Eur Urol, 33: 285, 1998 Nomata, K., Noguchi, M., Kanetake, H., Tsuda, N., Hayashi, M., Yamashita, S. et al: Intravesical adjuvant chemotherapy for superficial transitional cell bladder carcinoma: results of a randomized trial with epirubicin comparing short-term versus long-term maintenance treatment. Cancer Chemother Pharmacol, 50: 266, 2002 Flamm, J.: Long-term versus short-term doxorubicin hydrochloride instillation after transurethral resection of superficial bladder cancer. Eur Urol, 17: 119, 1990 Tolley, D. A., Parmar, M. K. B., Grigor, K. M., Lallemand, G. and the Medical Research Council Superficial Bladder Cancer Working Party: The effect of intravesical mitomycin C on recurrence of newly diagnosed superficial bladder cancer: a further report with 7 years of followup. J Urol, 155: 1233, 1996 Mostofi, F. K., Sobin, H. L. and Torloni, H.: Histological typing of urinary bladder tumours. International Classification of Tumours, No. 10. Geneva, Switzerland: World Health Organization, 1973 Holford, T. R.: Life tables with concomitant information. Biometrics, 32: 587, 1976 Cersosimo, R. J. and Hong, W. K.: Epirubicin; a review of the pharmacology, clinical activity, and adverse effects of an adriamycin analogue. J Clin Oncol, 4: 425, 1986 Ueda, T., Naito, S., Iguchi, A., Sagiyama, K., Osada, Y., Ariyoshi, A. et al: Adjuvant chemotherapy with early intravesical instillation of adriamycin and long-term oral administration of 5-fluorouracil in superficial bladder cancer. The Kyushu University Urological Oncology Group. Cancer Chemother Pharmacol, suppl., 30: S31, 1992 Naito, S., Kotoh, S., Omoto, T., Osada, Y., Sagiyama, K., Iguchi, A. et al: Prophylactic intravesical instillation chemotherapy against recurrence after a transurethral resection of superficial bladder cancer: a randomized controlled trial of doxorubicin plus verapamil versus doxorubicin alone. The Kyushu University Urological Oncology Group. Cancer Chemother Pharmacol, 42: 367, 1998 Akaza, H.: New strategy of bio-chemoprevention on recurrence of superficial bladder cancer based on a hypothesis of the mechanism of recurrence. Gan To Kagaku Ryoho, suppl., 24: 253, 1997 Hinotsu, S., Akaza, H., Ohashi, Y. and Kotake, K.: Intravesical chemotherapy for maximum prophylaxis of new early phase superficial bladder carcinoma treated by transurethral resection: a combined analysis of trials by Japanese Urological Cancer Research Group using smoothed hazard function. Cancer, 86: 1818, 1999