- Email: [email protected]
Invasive Bladder Cancer: Our Experience With Bladder Sparing Approach
Int. J. Radiation Oncology Biol. Phys., Vol. 41, No. 2, pp. 273–278, 1998 Copyright © 1998 Elsevier Science Inc. Printed in the USA. All rights reserved 0360-3016/98 $19.00 1 .00
PII S0360-3016(98)00013-3
●
Clinical Investigation INVASIVE BLADDER CANCER: OUR EXPERIENCE WITH BLADDER SPARING APPROACH JOZˇ ICA CˇERVEK, M.D.,* TANJA CˇUFER, PH.D., M.D.,* BRANKO ZAKOTNIK, M.D.,* BORUT KRAGELJ, M.D.,* SIMONA BORSˇ TNAR, M.D.,* TADEJA MATOS, M.D.* AND MIRJANA ZˇUMER-PREGELJ, M.D.† *Institute of Oncology, Ljubljana, Slovenia; and †Department of Urology, University Medical Center, Ljubljana, Slovenia Purpose: Muscle-invasive bladder cancer (MIBC) is a disease associated with several unresolved therapeutic questions. Radical cystectomy still represents the most frequent treatment approach. The aim of our study was to evaluate the effect and feasibility of bladder-sparing treatment by transurethral resection (TUR) and sequential chemoradiotherapy in patients with biopsy-proven invasive bladder cancer. Methods and Materials: After maximal TUR, 105 patients were treated with two to four cycles of methotrexate, cisplatinum, and vinblastine polychemotherapy. In complete responders, the treatment was continued by radiotherapy (50 Gy to the bladder and 40 Gy to the regional lymph nodes), whereas in nonresponders, cystectomy was performed when feasible. Results: Complete response after TUR and chemotherapy was achieved in 52% of patients. After a median follow-up of 42 months, 52 of 75 patients (69%) selected for bladder preservation were without evidence of disease in the bladder. Freedom from local failure in complete responders to chemotherapy was 80% [95% confidence interval (CI), 69 –91%) at 4 years. The actuarial survival of the entire group was 58% (95% CI, 47– 69%), whereas the survival rate with the bladder intact was 45% (95% CI, 34 –56%) at 4 years. Survival was significantly better in patients who responded to chemotherapy (79%) than in nonresponders (35%, p < 0.0001). There was no significant difference in survival between nonresponders who underwent cystectomy and nonresponders who completed treatment with radiotherapy (approximately 30% at 3 years). Conclusion: The present study confirms that MIBC is a heterogeneous disease, and that in more than half of patients who are affected, a bladder-sparing approach is safe. Our study has also demonstrated that in nonresponders, radical cystectomy as the treatment of choice is questionable. © 1998 Elsevier Science Inc. Invasive bladder cancer, Bladder preservation, Combined modality treatment.
abolic disorders, and sexual dysfunction. The other drawback of this treatment approach lies in the fact that over half of the patients will die with distant metastases (3). In view of these problems, several clinical studies were conducted using a bladder-sparing approach to the treatment of this disease (4 –11). In the past decade, the most promising advance has been achieved using transurethral surgery (TUR) and combined chemoradiotherapy regimens (4, 7–10). This treatment approach uses the advantages of the favorable effects of cis-platinum– based chemotherapy as well as the synergistic effects of chemotherapy and radiotherapy (12). The aim of our study was to evaluate the effect and feasibility of combined modality treatment in patients with invasive bladder cancer. The preliminary results of this study demonstrated a high response rate of 53% to chemotherapy, and about three quarters of patients had their bladder preserved with a 2-year survival rate of 73% (13). Here, we report the final results of this study.
INTRODUCTION Bladder cancer is a relatively frequent malignant disease. According to the data of the Cancer Registry of Slovenia, its incidence is increasing; in the period 1988 –1990, it was 11.1/100,000 males and 4/100,000 females. In that period, the observed 5-year survival rate was about 40% (1). The prognosis and treatment outcome depend on the depth of bladder-wall invasion. Superficial disease has an excellent prognosis, whereas the survival rate decreases significantly for deeply invasive tumors (2). Muscle-invasive bladder cancer (MIBC) is a disease associated with a relatively low cure rate and several yet unresolved therapeutic questions. The optimal management of MIBC has been a continuous subject of controversy. Radical cystectomy represents the most frequent treatment approach. Despite great advances in surgical techniques and better perioperative support even a neovesica cannot substitute for the patient’s original bladder without entailing a high risk of infection and consequential renal failure, metReprint requests to: Jozˇica Cˇervek, M.D., Institute of Oncology, Zaloska 2, 61105 Ljubljana, Slovenia.
Accepted for publication 29 December 1997. 273
274
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 41, Number 2, 1998
Table 1. Pretreatment characteristics of evaluable patients Characteristics
N
All patients Age Median Range Sex Male Female Clinical stage cT1 cT23 cT4 Histology Transitional cell ca G 2 Trasitional cell ca G 3 Transitional cell ca with metaplasia Anaplastic ca Transurethral resection Radical Nonradical Unknown Hydronephrosis Yes No
105 62 34–77 83 22 7 78 20 25 59 17 4 27 68 10 27 78
PATIENTS AND METHODS Patients Between December 1988 and June 1995, 105 patients with biopsy-proven invasive bladder cancer were entered into the study. There were 83 males and 22 females, with ages ranging from 34 to 77 years (median 62). The distribution of patients by clinical stage was as follows: cT1 in 7, cT2–3 in 78, and cT4 in 20 patients. All patients with Stage T1 had grade 3 tumors nonresectable by TUR. Of the 105 patients, 84 had pure transitional cell carcinoma; of these, 25 were grade 2 and 59 grade 3. The remaining 21 patients had either transitional cell carcinoma with metaplasia or anaplastic carcinoma. TUR of the tumor was judged by the urological surgeon to be complete in 27 patients. Hydronephrosis requiring nephrostomy to improve renal function was found in 27 patients. Patient characteristics are listed in Table 1. Pretreatment evaluation included medical history, physical examination, chest X ray, excretory urogram, computed tomography (CT) of the abdomen and pelvis, complete blood count, blood urea nitrogen, serum creatinine, creatinine clearance, and liver function tests. Most patients had a bone scan. On entry, all patients had as thorough as possible TUR of the tumor with bimanual palpation under anesthesia. Patients with clinical Stage T1– 4NXM0 according to American Joint Committee 1983 criteria, performance status # 2 (ECOG), normal bone marrow function (peripheral blood leukocyte values $ 4.0 3 109/l, thrombocytes $ 100 3 109/l), and normal renal function (24-h creatinine clearance . 70 ml/min) were considered eligible for the study.
Fig. 1. Treatment schema. Number of patients treated at each step is shown. *Includes 27 operable patients with residual tumor after chemotherapy, and three patients with residual tumor after chemotherapy and radiotherapy. **Includes 55 patients with complete response, four patients with incomplete evaluation, and 19 nonresponders unsuitable for surgery.
Treatment The treatment schedule is shown in Fig. 1. After maximal TUR of the tumor, the patients were treated with two to four cycles of polychemotherapy according to the MCV schedule [methotrexate (M) 30 mg/m2, intravenously (i.v.), on days 1 and 14; cisplatinum (C) 100 mg/m2, i.v., on day 2; and vinblastine (V) 3 mg/m2, i.v., on days 1 and 14]. The cycles were repeated after 21 days. Cystoscopic evaluation was performed after three to four cycles of MCV. Only in patients suspected upon clinical and laboratory examinations of having progression was restaging cystoscopy performed after two cycles of chemotherapy. The mean number of chemotherapy cycles received was 3.3. The average dose of cisplatinum received per cycle was 80 mg/m2 and was the same in patients with and without hydronephrosis before treatment. Restaging included examination under anesthesia, cystoscopy with tumor-site biopsy, and urine cytology. Patients were considered to have complete response if there was no evidence of tumor upon all of the above investigations. In patients with complete response, treatment was continued by radiotherapy, whereas in all other patients, cystectomy was performed whenever feasible. Radiotherapy was started within 2–3 weeks after com-
Bladder sparing in invasive bladder cancer
Table 2. Patients with complete response (CR) Tumor stage
No. of patients (CR/total)
%
55/105 5/7 43/78 7/20
52 71 55 35
67/105 6/7 51/78 10/20
64 85 65 50
After TUR and ChT All patients Stage cT1 Stages cT2T3 Stage cT4 After TUR, ChT, and RT All patients Stage cT1 Stages cT2T3 Stage cT4
TUR 5 transurethral resection; ChT 5 chemotherapy; RT 5 radiotherapy.
pleted chemotherapy. Patients were treated in a supine position, using an 8- or 10-MV linear accelerator and the following technique: a four-field arrangement to the pelvis and a three-field arrangement coned down on the urinary bladder. In patients with complete response, the total dose to the urinary bladder was 50 Gy, and to the regional lymph nodes, 40 Gy, given in 5 2-Gy fractions/week. Patients who did not achieve complete response were irradiated with a total dose of 60 Gy. Three months after completed chemotherapy and radiotherapy, cystoscopic reevaluation, chest X ray, and CT scan were done. Thereafter, follow-up examinations (clinical and laboratory diagnostic studies, cystoscopy, urine cytology, chest X ray, and CT scan) were performed every 3 months for 2 years, and every 6 months thereafter. Toxicity during chemotherapy and radiotherapy was recorded according to World Health Organization criteria. Hematologic side effects were calculated on the basis of day 1 blood count values. Methotrexate dose was reduced 25% in the case of grade 3 or 4 stomatitis. Methotrexate and cisplatinum reductions for 25–50% were required when creatinine clearance fell below the normal value. Long-term toxicity after combined treatment was defined according to Abrahamsen and Fossa (14). Statistical methods Actuarial survival was calculated according to the method of Kaplan and Meier (15). For comparison of survival curves, the log rank test was used. Survival was measured from the date of diagnosis. The comparison of survival of complete responders versus non– complete responders was calculated using the landmark method proposed by Anderson et al. (16, 17). RESULTS The results obtained with the analyses of the first 45 evaluable patients entered into our study were previously reported with a shorter follow-up (13). The present report includes 105 patients with a median follow-up of 42 months (range 4 –96). The rate of complete responses is shown in
●
J. CˇERVEK et al.
275
Table 2. After TUR and chemotherapy, complete response was achieved in 52% of patients; after TUR, chemotherapy, and radiotherapy, no residual tumor was found in 64% of patients. The response rates were higher in patients with a lower T stage. All patients with complete response after TUR and chemotherapy were irradiated. Four patients who were not completely evaluated for response after chemotherapy were treated with radiotherapy, and complete response was obtained in three. A total of 46 patients with residual disease at restaging were assigned to cystectomy; 27 of them underwent recommended cystectomy. Nineteen did not undergo cystectomy per protocol because of medical reasons (n 5 7), inoperable tumor (n 5 2), development of metastases (n 5 1), or patient refusal (n 5 9); all continued treatment with radiotherapy. With subsequent radiotherapy, complete response was achieved in 12 of these patients. Three patients underwent cystectomy after completing radiotherapy; in one, radiotherapy downstaged inoperable disease to operable, and two patients with residual disease after radiotherapy changed their opinion and accepted cystectomy. Treatments given to particular groups of patients are outlined in Fig. 1. The pattern of first failure in 105 patients is presented in Table 3. After a median follow-up of 42 months, locoregional failure was found in 29 of 105 patients (28%) and distant metastases occurred in 23 of 105 patients (22%). In the cystectomy group, locoregional failure occurred in six patients (16%), whereas distant metastases occurred in 13 of 30 patients (43%). Among 75 patients who had their bladder preserved, 23 (31%) had locoregional failures, and 10 (13%) developed distant metastases. Among 23 locoregional failures, there were eight treatment failures (i.e., patients who did not achieve complete response after treatment), one patient with regional nodes metastases, and 14 patients with bladder recurrence. Eleven patients, complete responders after TUR and chemotherapy, developed bladder
Table 3. Patterns of first failure in 105 patients
Treatment Cystectomy After ChT or ChT plus RT Radiotherapy After ChT CR after ChT Not CR after ChT Not evaluated after ChT All patients
Total (n)
Locoregional Distant failure metastases [n (%)] [n (%)]
30
6 (20)
13 (43)
75 55 16 4 105
23 (31) 12 (22)* 10 (63)† 1‡ 29 (28)
10 (13) 6 (11) 3 (19) 1 23 (22)
CR 5 complete response; ChT 5 chemotherapy; RT 5 radiotherapy. * Includes 11 patients with bladder relapse and one with regional nodes metastases. † Includes seven patients with induction failure and three with bladder relapse. ‡ Includes one patient with induction failure.
276
I. J. Radiation Oncology
●
Biology
●
Physics
Fig. 2. Overall survival (OS) and survival with bladder intact (SWBI).
recurrences which were found to be invasive in seven patients and noninvasive in four. Three more patients, complete responders after TUR, chemotherapy, and radiotherapy, developed bladder recurrence which was found to be invasive in two patients and noninvasive in one. The rate of local recurrence was found to be much higher in complete responders with cT4 [3 of 10 (30%)] disease than in patients with cT2,T3 [9 of 51 (18%)]. Noninvasive recurrences were managed by TUR, which had to be repeated in two patients; in one patient, invasive cancer occurred subsequently. Salvage cystectomy was performed in seven patients with invasive recurrences: Four developed distant metastases only a few months after cystectomy and died, one patient died owing to septic shock in postoperative period, and two are free of disease 6 and 21 months after salvage cystectomy. At a median follow-up of 42 months, 52 of 75 patients (69%) selected for bladder preservation were without evidence of disease in the bladder. Freedom from local failure in complete responders to chemotherapy was 80% (95% CI, 69 –91%) at 4 years. The survival curve for all 105 eligible and evaluable patients is shown in Fig. 2. The 4-year actuarial survival rate was 58% (95% CI, 47– 69%). The 4-year survival with the bladder intact was 45% (95% CI, 34 –56%). When cT1 patients are excluded, the 4-year actuarial survival was 50% (95% CI, 36 – 64%) and 4-year survival with the bladder intact was 36% (95% CI, 23– 49%). As shown in Fig. 3, 4-year actuarial survival was significantly better in patients who responded to chemotherapy (79%) than in nonresponders (35%, p , 0.0001). There was no significant difference in survival between nonresponders who underwent cystectomy and nonresponders who completed treatment with radiotherapy. The 3-year actuarial survival of both groups was found to be around 30% (Fig. 4). Figure 5
Volume 41, Number 2, 1998
Fig. 3. Overall survival, according to response to transurethral resection and chemotherapy. CR 5 complete responders; NCR 5 nonresponders.
shows that survival was better in patients without hydronephrosis than in those with hydronephrosis at presentation (p 5 0.012). Table 4 shows the worst reported toxicity during chemotherapy and radiation treatment per patient. The most frequent adverse reactions observed during chemotherapy in almost all patients were myelosuppression and nausea. Leukopenia grade 4 was registered in two patients. One of these patients also presented with severe stomatitis, and another
Fig. 4. Overall survival of nonresponders after transurethral resection and chemotherapy, according to local therapy. RT 5 radiotherapy.
Bladder sparing in invasive bladder cancer
●
J. CˇERVEK et al.
277
Table 5. Late radiation-related morbidity in 63 evaluable patients Grade
Diarrhea Frequency Hematuria Dysuria
0
1
2
3
4
56 45 58 57
6 12 2 2
1 4 3 4
0 2 0 0
0 0 0 0
DISCUSSION
Fig. 5. Overall survival, according to absence or presence of hydronephrosis. OU 5 obstructive uropathy.
patient with septicemia which was found to be the direct cause of death. There was one case of severe uroinfection with septicemia which resolved to antibiotic therapy. Other chemotherapy-related toxic side effects were generally mild and transient. During radiotherapy, the most pronounced toxic side effects were dysuria and diarrhea. Radiation therapy was interrupted in 11 patients because of uncontrolled diarrhea. Twelve patients had mild myelosupression during radiation therapy without clinical consequences. Myelosupression was mainly registered in the beginning of radiotherapy and was probably due to the effects of previous chemotherapy. Among the late sequelae of radiation treatment (Table 5), only the increased frequency of micturition is worth mentioning. No patient required cystectomy for treatment-related bladder morbidity.
Table 4. Chemotherapy- and radiotherapy-related acute toxicity Grade
ChT-related toxicity (105 patients) Myelosuppression Stomatitis Nausea Alopecia Neuropathy Infection RT-related toxicity (78 patients) Myelosuppression Diarrhea Dysuria/frequency
0
1
2
3
4
63 95 23 83 99 70
29 6 50 14 3 16
8 1 25 7 2 14
3 2 7 1 1 3
2 1 0 0 0 2
66 27 28
11 18 24
1 22 24
0 11 2
0 0 0
ChT 5 chemotherapy; RT 5 radiotherapy.
Our results of combined modality treatment of invasive bladder cancer are encouraging. After completing treatment, 75 of 105 patients (71%) had their bladder preserved, and at a median follow-up of 42 months, 52 of 75 patients (69%) with a preserved bladder have an intact, functioning bladder. The 4-year actuarial survival rate of the entire group is 58%, whereas survival with the bladder intact is 45%. To avoid complications associated with cystectomy after irradiation, in contrast to all other published studies, a strictly sequential regime of chemotherapy and radiotherapy was used in our protocol. However, the survival and bladder preservation rates in our patients are very similar to results obtained in all other studies with partially (8, 9) or completely (4, 7, 10, 18) concomitant chemoradiotherapy. Those authors reported bladder preservation in 58 –79% of patients; in our case, it was 71%. The authors cited reported overall survival rates of 47– 62% at 4 –5 years; in our case, it was 58% at 4 years. The same authors reported freedom from local failure from 73% to 89% at 4 –5 years, whereas in our study, it was 80% at 4 years. In contrast, we had a higher proportion of invasive bladder recurrences compared to noninvasive ones, which was not the case in other studies (9, 18). The reason for this could be that biopsies were not always performed at follow-up cystoscopy. As reported by others using a similar chemotherapy and radiotherapy protocol (8, 9), the tolerance of our protocol was also acceptable. Among acute side effects, myelosupression prevailed. Grade 4 neutropenia with accompanying septicemia was found to be the direct cause of death in one patient. Radiation therapy had to be interrupted in 11 patients owing to uncontrollable diarrhea. Among the late sequelae of treatment, only an increased frequency of micturition is worth mentioning, but it was very rarely expressed at such a scale as to interfere with patients’ daily activities or social life. However, treatment results obtained by chemoradiotherapy and selective bladder preservation have to be compared with results obtained by standard treatment. The survival of patients in our study as well as in other studies with combined modality treatment and bladder sparing is similar, if not superior, to that of patients treated with only local therapy— either cystectomy (2, 19) or radiotherapy (5, 6). The exact answer regarding the superiority of a combined treatment over cystectomy will be known when the results
278
I. J. Radiation Oncology
●
Biology
●
Physics
of the randomized, ongoing EORTC/MRC study are published (20). As in other studies (4, 7), we found that patients with complete response to chemotherapy have a better prognosis than those who do not respond to chemotherapy. Nonresponders have a dismal prognosis whatever treatment they receive afterward. Actually, in our patients, the overall survival rate of nonresponders who underwent cystectomy according to the protocol was as poor as the overall survival rate of nonresponders who were irradiated (approximately 30% at 3 years). Complete response after TUR and chemotherapy seems to be a very important prognostic factor for the treatment outcome. Nonresponders presented with an equally high percentage of disseminated disease, regardless of the treatment approach. These results suggest that the response of the primary disease to chemotherapy is indicative of the response of occult metastases to the same agents. The present study confirms that MIBC is a heterogeneous disease, and that among the patients affected, in more than half a bladder-sparing approach is safe and does not com-
Volume 41, Number 2, 1998
promise the survival outcome. Our study also indicates that in nonresponders, radical cystectomy as the treatment of choice is questionable, since it cannot ensure a cure, despite the radical surgical procedure. It is obvious that these patients have biologically more aggressive tumors that are not manageable with either cystectomy or the currently used combined modality treatment. Apparently, for these patients, a new, more aggressive treatment approach should be considered. This could be either concomitant radiochemotherapy with new, more potent chemotherapeutic agents that have demonstrated evidence of radiation enhancement (21), or high-dose chemotherapy (12). The development of biologic markers which could predict the responsiveness of chemoradiotherapy might be extremely useful in the selection of patients for bladder preservation. Some of these biological markers, such as DNA ploidy, various oncogenes, markers of drug resistance, and markers of angiogenesis, are already under study (12); it is hoped that according to them, better treatment selection for the individual patient will be possible in the future.
REFERENCES: 1. Pompe-Kirn, V.; Zakotnik, B.; Volk, N.; Benulicˇ, T.; Sˇkrk, J. Cancer patients in Slovenia 1963–1990. Ljubljana: Onkolosˇki Insˇtitut; 1995:56 – 61. 2. Scher, H.; Shipley, W. U.; Herr, H. Cancer of the bladder. In: DeVita, V. T., Jr.; Helman, S.; Rosenberg, S. A., eds. Cancer: Principles and practice in oncology. 5th ed. Philadelphia, PA: Lippincott; 1997:1300 –1322. 3. Raghavan, D.; Huben, R. Management of bladder cancer. Curr. Probl. Cancer 19:3– 63; 1995. 4. Dunst, J.; Sauer, R.; Schrot, K. M.; Wittekind, C.; AltendorfHoffman, A. Organ-sparing treatment of advanced bladder cancer: A 10-year experience. Int. J. Radiat. Oncol. Biol. Phys. 30:261–266; 1994. 5. Gospodarowicz, M. K.; Hawkins, N. V.; Rawlings, G. A.; Connolly, J. G.; Jewett, M. A. S.; Thomas, G. M.; Herman, J. G.; Garrett, P. G.; Chua, T.; Duncan, W.; Buckspan, M.; Sugar, L.; Rider, W. D. Radical radiotherapy for muscle invasive transitional cell carcinoma of the bladder: Failure analysis. J. Urol. 142:1448 –1454; 1989. 6. Gospodarowicz, M. K.; Rider, W. D.; Keen, C. W.; Connoly, J. G.; Jewett, M. A. S.; Cummings, B. J.; Duncan, W.; Warde, P.; Chua, T. Bladder Cancer: Long-term follow-up results of patients treated with radical radiation. Clin. Oncol. 3:155–161; 1991. 7. Housset, M.; Maulard, C.; Chretien, Y. C.; Dufour, B.; Delanian, S.; Huart, J.; Colardelle, F.; Brunel, P.; Baellet, F. Combined radiation and chemotherapy for invasive transitional-cell carcinoma of the bladder: A prospective study. J. Clin. Oncol. 11:2150 –2157; 1993. 8. Kaufman, D. S.; Shipley, W. U.; Griffin, P. P.; Henery, N. M.; Althausen, A. F.; Efird, J. T. Selective preservation by combination treatment of invasive bladder cancer. N. Engl. J. Med. 329:1377–1382; 1993. 9. Tester, W.; Caplan, R.; Heaney, J.; Venner, P.; Whittington, R.; Byhardt, R.; True, L.; Shipley, W. Neoadjuvant combined modality program with selective organ preservation for invasive bladder cancer: Results of Radiation Therapy Oncology Group Phase II Trial 8802. J. Clin. Oncol. 14:119 –126; 1996.
10. Tester, W.; Porter, A.; Asbell, S. Combined modality program with possible organ preservation for invasive bladder carcinoma: Results of RTOG Protocol 85-12. Int. J. Radiat. Oncol. Biol. Phys. 25:738 –790; 1993. 11. Zietman, A. L.; Shipley, W. U. Organ sparing treatment for bladder cancer: Time to beat the drum. Int. J. Radiat. Oncol. Biol. Phys. 30:741–742; 1994. 12. Sternberg, C. N. Neoadjuvant and adjuvant chemotherapy in locally advanced bladder cancer. Semin. Oncol. 23:621– 632; 1996. 13. Cˇervek, J.; Cˇufer, T.; Kragelj, B.; Zakotnik, B.; Stanonik, M. Sequential transurethral surgery, multiple drug chemotherapy and radiation therapy for invasive bladder carcinoma: Initial report. Int. J. Radiat. Oncol. Biol. Phys. 25:777–782; 1993. 14. Foss Abrahamsen, J.; Fossa, S. D. Long-term morbidity after curative radiotherapy for carcinoma of the bladder: A retrospective study. Strahlenther. Oncol. 166:580 –583; 1990. 15. Kaplan, E. L.; Meier, P. Non-parrametric estimation from incomplete obervations. J. Am. Stat. Assoc. 53:457– 481; 1958. 16. Anderson, J. A.; Cain, K. C.; Gelber, R. D. Analysis of survival by tumor response. J. Clin. Oncol. 1:710 –719; 1983. 17. Anderson, J. A.; Cain, K. C.; Gelber, R. D.; Gelmson, K. S. Analysis and interpretation of the comparisation of survival by treatment outcome variables in cancer clinical trials. Cancer Treat. Rep. 69:1139 –1144; 1985. 18. Rifkin, M. N.; Oblon, D.; Parsons, J.; McCarley, D.; Wajsman, Z. Bladder-sparing treatment for muscle invasive bladder cancer with systemic chemotherapy followed by radiation therapy and adjuvant cisplatin: A 3-year follow-up. Proc. Am. Soc. Clin. Oncol. 11:208; 1992. 19. Bales, G. T.; Kim, H.; Steinberg, G. D. Surgical therapy for locally advanced bladder cancer. Semin. Oncol. 23:605– 613; 1996. 20. Sternberg, C. N. The treatment of advanced bladder cancer. Ann. Oncol. 6:113–126; 1995. 21. Roth, B. J. Preliminary experience with Paclitaxel in advanced bladder cancer. Semin. Oncol. 22:1–5; 1995.
PII S0360-3016(98)00013-3
●
Clinical Investigation INVASIVE BLADDER CANCER: OUR EXPERIENCE WITH BLADDER SPARING APPROACH JOZˇ ICA CˇERVEK, M.D.,* TANJA CˇUFER, PH.D., M.D.,* BRANKO ZAKOTNIK, M.D.,* BORUT KRAGELJ, M.D.,* SIMONA BORSˇ TNAR, M.D.,* TADEJA MATOS, M.D.* AND MIRJANA ZˇUMER-PREGELJ, M.D.† *Institute of Oncology, Ljubljana, Slovenia; and †Department of Urology, University Medical Center, Ljubljana, Slovenia Purpose: Muscle-invasive bladder cancer (MIBC) is a disease associated with several unresolved therapeutic questions. Radical cystectomy still represents the most frequent treatment approach. The aim of our study was to evaluate the effect and feasibility of bladder-sparing treatment by transurethral resection (TUR) and sequential chemoradiotherapy in patients with biopsy-proven invasive bladder cancer. Methods and Materials: After maximal TUR, 105 patients were treated with two to four cycles of methotrexate, cisplatinum, and vinblastine polychemotherapy. In complete responders, the treatment was continued by radiotherapy (50 Gy to the bladder and 40 Gy to the regional lymph nodes), whereas in nonresponders, cystectomy was performed when feasible. Results: Complete response after TUR and chemotherapy was achieved in 52% of patients. After a median follow-up of 42 months, 52 of 75 patients (69%) selected for bladder preservation were without evidence of disease in the bladder. Freedom from local failure in complete responders to chemotherapy was 80% [95% confidence interval (CI), 69 –91%) at 4 years. The actuarial survival of the entire group was 58% (95% CI, 47– 69%), whereas the survival rate with the bladder intact was 45% (95% CI, 34 –56%) at 4 years. Survival was significantly better in patients who responded to chemotherapy (79%) than in nonresponders (35%, p < 0.0001). There was no significant difference in survival between nonresponders who underwent cystectomy and nonresponders who completed treatment with radiotherapy (approximately 30% at 3 years). Conclusion: The present study confirms that MIBC is a heterogeneous disease, and that in more than half of patients who are affected, a bladder-sparing approach is safe. Our study has also demonstrated that in nonresponders, radical cystectomy as the treatment of choice is questionable. © 1998 Elsevier Science Inc. Invasive bladder cancer, Bladder preservation, Combined modality treatment.
abolic disorders, and sexual dysfunction. The other drawback of this treatment approach lies in the fact that over half of the patients will die with distant metastases (3). In view of these problems, several clinical studies were conducted using a bladder-sparing approach to the treatment of this disease (4 –11). In the past decade, the most promising advance has been achieved using transurethral surgery (TUR) and combined chemoradiotherapy regimens (4, 7–10). This treatment approach uses the advantages of the favorable effects of cis-platinum– based chemotherapy as well as the synergistic effects of chemotherapy and radiotherapy (12). The aim of our study was to evaluate the effect and feasibility of combined modality treatment in patients with invasive bladder cancer. The preliminary results of this study demonstrated a high response rate of 53% to chemotherapy, and about three quarters of patients had their bladder preserved with a 2-year survival rate of 73% (13). Here, we report the final results of this study.
INTRODUCTION Bladder cancer is a relatively frequent malignant disease. According to the data of the Cancer Registry of Slovenia, its incidence is increasing; in the period 1988 –1990, it was 11.1/100,000 males and 4/100,000 females. In that period, the observed 5-year survival rate was about 40% (1). The prognosis and treatment outcome depend on the depth of bladder-wall invasion. Superficial disease has an excellent prognosis, whereas the survival rate decreases significantly for deeply invasive tumors (2). Muscle-invasive bladder cancer (MIBC) is a disease associated with a relatively low cure rate and several yet unresolved therapeutic questions. The optimal management of MIBC has been a continuous subject of controversy. Radical cystectomy represents the most frequent treatment approach. Despite great advances in surgical techniques and better perioperative support even a neovesica cannot substitute for the patient’s original bladder without entailing a high risk of infection and consequential renal failure, metReprint requests to: Jozˇica Cˇervek, M.D., Institute of Oncology, Zaloska 2, 61105 Ljubljana, Slovenia.
Accepted for publication 29 December 1997. 273
274
I. J. Radiation Oncology
●
Biology
●
Physics
Volume 41, Number 2, 1998
Table 1. Pretreatment characteristics of evaluable patients Characteristics
N
All patients Age Median Range Sex Male Female Clinical stage cT1 cT23 cT4 Histology Transitional cell ca G 2 Trasitional cell ca G 3 Transitional cell ca with metaplasia Anaplastic ca Transurethral resection Radical Nonradical Unknown Hydronephrosis Yes No
105 62 34–77 83 22 7 78 20 25 59 17 4 27 68 10 27 78
PATIENTS AND METHODS Patients Between December 1988 and June 1995, 105 patients with biopsy-proven invasive bladder cancer were entered into the study. There were 83 males and 22 females, with ages ranging from 34 to 77 years (median 62). The distribution of patients by clinical stage was as follows: cT1 in 7, cT2–3 in 78, and cT4 in 20 patients. All patients with Stage T1 had grade 3 tumors nonresectable by TUR. Of the 105 patients, 84 had pure transitional cell carcinoma; of these, 25 were grade 2 and 59 grade 3. The remaining 21 patients had either transitional cell carcinoma with metaplasia or anaplastic carcinoma. TUR of the tumor was judged by the urological surgeon to be complete in 27 patients. Hydronephrosis requiring nephrostomy to improve renal function was found in 27 patients. Patient characteristics are listed in Table 1. Pretreatment evaluation included medical history, physical examination, chest X ray, excretory urogram, computed tomography (CT) of the abdomen and pelvis, complete blood count, blood urea nitrogen, serum creatinine, creatinine clearance, and liver function tests. Most patients had a bone scan. On entry, all patients had as thorough as possible TUR of the tumor with bimanual palpation under anesthesia. Patients with clinical Stage T1– 4NXM0 according to American Joint Committee 1983 criteria, performance status # 2 (ECOG), normal bone marrow function (peripheral blood leukocyte values $ 4.0 3 109/l, thrombocytes $ 100 3 109/l), and normal renal function (24-h creatinine clearance . 70 ml/min) were considered eligible for the study.
Fig. 1. Treatment schema. Number of patients treated at each step is shown. *Includes 27 operable patients with residual tumor after chemotherapy, and three patients with residual tumor after chemotherapy and radiotherapy. **Includes 55 patients with complete response, four patients with incomplete evaluation, and 19 nonresponders unsuitable for surgery.
Treatment The treatment schedule is shown in Fig. 1. After maximal TUR of the tumor, the patients were treated with two to four cycles of polychemotherapy according to the MCV schedule [methotrexate (M) 30 mg/m2, intravenously (i.v.), on days 1 and 14; cisplatinum (C) 100 mg/m2, i.v., on day 2; and vinblastine (V) 3 mg/m2, i.v., on days 1 and 14]. The cycles were repeated after 21 days. Cystoscopic evaluation was performed after three to four cycles of MCV. Only in patients suspected upon clinical and laboratory examinations of having progression was restaging cystoscopy performed after two cycles of chemotherapy. The mean number of chemotherapy cycles received was 3.3. The average dose of cisplatinum received per cycle was 80 mg/m2 and was the same in patients with and without hydronephrosis before treatment. Restaging included examination under anesthesia, cystoscopy with tumor-site biopsy, and urine cytology. Patients were considered to have complete response if there was no evidence of tumor upon all of the above investigations. In patients with complete response, treatment was continued by radiotherapy, whereas in all other patients, cystectomy was performed whenever feasible. Radiotherapy was started within 2–3 weeks after com-
Bladder sparing in invasive bladder cancer
Table 2. Patients with complete response (CR) Tumor stage
No. of patients (CR/total)
%
55/105 5/7 43/78 7/20
52 71 55 35
67/105 6/7 51/78 10/20
64 85 65 50
After TUR and ChT All patients Stage cT1 Stages cT2T3 Stage cT4 After TUR, ChT, and RT All patients Stage cT1 Stages cT2T3 Stage cT4
TUR 5 transurethral resection; ChT 5 chemotherapy; RT 5 radiotherapy.
pleted chemotherapy. Patients were treated in a supine position, using an 8- or 10-MV linear accelerator and the following technique: a four-field arrangement to the pelvis and a three-field arrangement coned down on the urinary bladder. In patients with complete response, the total dose to the urinary bladder was 50 Gy, and to the regional lymph nodes, 40 Gy, given in 5 2-Gy fractions/week. Patients who did not achieve complete response were irradiated with a total dose of 60 Gy. Three months after completed chemotherapy and radiotherapy, cystoscopic reevaluation, chest X ray, and CT scan were done. Thereafter, follow-up examinations (clinical and laboratory diagnostic studies, cystoscopy, urine cytology, chest X ray, and CT scan) were performed every 3 months for 2 years, and every 6 months thereafter. Toxicity during chemotherapy and radiotherapy was recorded according to World Health Organization criteria. Hematologic side effects were calculated on the basis of day 1 blood count values. Methotrexate dose was reduced 25% in the case of grade 3 or 4 stomatitis. Methotrexate and cisplatinum reductions for 25–50% were required when creatinine clearance fell below the normal value. Long-term toxicity after combined treatment was defined according to Abrahamsen and Fossa (14). Statistical methods Actuarial survival was calculated according to the method of Kaplan and Meier (15). For comparison of survival curves, the log rank test was used. Survival was measured from the date of diagnosis. The comparison of survival of complete responders versus non– complete responders was calculated using the landmark method proposed by Anderson et al. (16, 17). RESULTS The results obtained with the analyses of the first 45 evaluable patients entered into our study were previously reported with a shorter follow-up (13). The present report includes 105 patients with a median follow-up of 42 months (range 4 –96). The rate of complete responses is shown in
●
J. CˇERVEK et al.
275
Table 2. After TUR and chemotherapy, complete response was achieved in 52% of patients; after TUR, chemotherapy, and radiotherapy, no residual tumor was found in 64% of patients. The response rates were higher in patients with a lower T stage. All patients with complete response after TUR and chemotherapy were irradiated. Four patients who were not completely evaluated for response after chemotherapy were treated with radiotherapy, and complete response was obtained in three. A total of 46 patients with residual disease at restaging were assigned to cystectomy; 27 of them underwent recommended cystectomy. Nineteen did not undergo cystectomy per protocol because of medical reasons (n 5 7), inoperable tumor (n 5 2), development of metastases (n 5 1), or patient refusal (n 5 9); all continued treatment with radiotherapy. With subsequent radiotherapy, complete response was achieved in 12 of these patients. Three patients underwent cystectomy after completing radiotherapy; in one, radiotherapy downstaged inoperable disease to operable, and two patients with residual disease after radiotherapy changed their opinion and accepted cystectomy. Treatments given to particular groups of patients are outlined in Fig. 1. The pattern of first failure in 105 patients is presented in Table 3. After a median follow-up of 42 months, locoregional failure was found in 29 of 105 patients (28%) and distant metastases occurred in 23 of 105 patients (22%). In the cystectomy group, locoregional failure occurred in six patients (16%), whereas distant metastases occurred in 13 of 30 patients (43%). Among 75 patients who had their bladder preserved, 23 (31%) had locoregional failures, and 10 (13%) developed distant metastases. Among 23 locoregional failures, there were eight treatment failures (i.e., patients who did not achieve complete response after treatment), one patient with regional nodes metastases, and 14 patients with bladder recurrence. Eleven patients, complete responders after TUR and chemotherapy, developed bladder
Table 3. Patterns of first failure in 105 patients
Treatment Cystectomy After ChT or ChT plus RT Radiotherapy After ChT CR after ChT Not CR after ChT Not evaluated after ChT All patients
Total (n)
Locoregional Distant failure metastases [n (%)] [n (%)]
30
6 (20)
13 (43)
75 55 16 4 105
23 (31) 12 (22)* 10 (63)† 1‡ 29 (28)
10 (13) 6 (11) 3 (19) 1 23 (22)
CR 5 complete response; ChT 5 chemotherapy; RT 5 radiotherapy. * Includes 11 patients with bladder relapse and one with regional nodes metastases. † Includes seven patients with induction failure and three with bladder relapse. ‡ Includes one patient with induction failure.
276
I. J. Radiation Oncology
●
Biology
●
Physics
Fig. 2. Overall survival (OS) and survival with bladder intact (SWBI).
recurrences which were found to be invasive in seven patients and noninvasive in four. Three more patients, complete responders after TUR, chemotherapy, and radiotherapy, developed bladder recurrence which was found to be invasive in two patients and noninvasive in one. The rate of local recurrence was found to be much higher in complete responders with cT4 [3 of 10 (30%)] disease than in patients with cT2,T3 [9 of 51 (18%)]. Noninvasive recurrences were managed by TUR, which had to be repeated in two patients; in one patient, invasive cancer occurred subsequently. Salvage cystectomy was performed in seven patients with invasive recurrences: Four developed distant metastases only a few months after cystectomy and died, one patient died owing to septic shock in postoperative period, and two are free of disease 6 and 21 months after salvage cystectomy. At a median follow-up of 42 months, 52 of 75 patients (69%) selected for bladder preservation were without evidence of disease in the bladder. Freedom from local failure in complete responders to chemotherapy was 80% (95% CI, 69 –91%) at 4 years. The survival curve for all 105 eligible and evaluable patients is shown in Fig. 2. The 4-year actuarial survival rate was 58% (95% CI, 47– 69%). The 4-year survival with the bladder intact was 45% (95% CI, 34 –56%). When cT1 patients are excluded, the 4-year actuarial survival was 50% (95% CI, 36 – 64%) and 4-year survival with the bladder intact was 36% (95% CI, 23– 49%). As shown in Fig. 3, 4-year actuarial survival was significantly better in patients who responded to chemotherapy (79%) than in nonresponders (35%, p , 0.0001). There was no significant difference in survival between nonresponders who underwent cystectomy and nonresponders who completed treatment with radiotherapy. The 3-year actuarial survival of both groups was found to be around 30% (Fig. 4). Figure 5
Volume 41, Number 2, 1998
Fig. 3. Overall survival, according to response to transurethral resection and chemotherapy. CR 5 complete responders; NCR 5 nonresponders.
shows that survival was better in patients without hydronephrosis than in those with hydronephrosis at presentation (p 5 0.012). Table 4 shows the worst reported toxicity during chemotherapy and radiation treatment per patient. The most frequent adverse reactions observed during chemotherapy in almost all patients were myelosuppression and nausea. Leukopenia grade 4 was registered in two patients. One of these patients also presented with severe stomatitis, and another
Fig. 4. Overall survival of nonresponders after transurethral resection and chemotherapy, according to local therapy. RT 5 radiotherapy.
Bladder sparing in invasive bladder cancer
●
J. CˇERVEK et al.
277
Table 5. Late radiation-related morbidity in 63 evaluable patients Grade
Diarrhea Frequency Hematuria Dysuria
0
1
2
3
4
56 45 58 57
6 12 2 2
1 4 3 4
0 2 0 0
0 0 0 0
DISCUSSION
Fig. 5. Overall survival, according to absence or presence of hydronephrosis. OU 5 obstructive uropathy.
patient with septicemia which was found to be the direct cause of death. There was one case of severe uroinfection with septicemia which resolved to antibiotic therapy. Other chemotherapy-related toxic side effects were generally mild and transient. During radiotherapy, the most pronounced toxic side effects were dysuria and diarrhea. Radiation therapy was interrupted in 11 patients because of uncontrolled diarrhea. Twelve patients had mild myelosupression during radiation therapy without clinical consequences. Myelosupression was mainly registered in the beginning of radiotherapy and was probably due to the effects of previous chemotherapy. Among the late sequelae of radiation treatment (Table 5), only the increased frequency of micturition is worth mentioning. No patient required cystectomy for treatment-related bladder morbidity.
Table 4. Chemotherapy- and radiotherapy-related acute toxicity Grade
ChT-related toxicity (105 patients) Myelosuppression Stomatitis Nausea Alopecia Neuropathy Infection RT-related toxicity (78 patients) Myelosuppression Diarrhea Dysuria/frequency
0
1
2
3
4
63 95 23 83 99 70
29 6 50 14 3 16
8 1 25 7 2 14
3 2 7 1 1 3
2 1 0 0 0 2
66 27 28
11 18 24
1 22 24
0 11 2
0 0 0
ChT 5 chemotherapy; RT 5 radiotherapy.
Our results of combined modality treatment of invasive bladder cancer are encouraging. After completing treatment, 75 of 105 patients (71%) had their bladder preserved, and at a median follow-up of 42 months, 52 of 75 patients (69%) with a preserved bladder have an intact, functioning bladder. The 4-year actuarial survival rate of the entire group is 58%, whereas survival with the bladder intact is 45%. To avoid complications associated with cystectomy after irradiation, in contrast to all other published studies, a strictly sequential regime of chemotherapy and radiotherapy was used in our protocol. However, the survival and bladder preservation rates in our patients are very similar to results obtained in all other studies with partially (8, 9) or completely (4, 7, 10, 18) concomitant chemoradiotherapy. Those authors reported bladder preservation in 58 –79% of patients; in our case, it was 71%. The authors cited reported overall survival rates of 47– 62% at 4 –5 years; in our case, it was 58% at 4 years. The same authors reported freedom from local failure from 73% to 89% at 4 –5 years, whereas in our study, it was 80% at 4 years. In contrast, we had a higher proportion of invasive bladder recurrences compared to noninvasive ones, which was not the case in other studies (9, 18). The reason for this could be that biopsies were not always performed at follow-up cystoscopy. As reported by others using a similar chemotherapy and radiotherapy protocol (8, 9), the tolerance of our protocol was also acceptable. Among acute side effects, myelosupression prevailed. Grade 4 neutropenia with accompanying septicemia was found to be the direct cause of death in one patient. Radiation therapy had to be interrupted in 11 patients owing to uncontrollable diarrhea. Among the late sequelae of treatment, only an increased frequency of micturition is worth mentioning, but it was very rarely expressed at such a scale as to interfere with patients’ daily activities or social life. However, treatment results obtained by chemoradiotherapy and selective bladder preservation have to be compared with results obtained by standard treatment. The survival of patients in our study as well as in other studies with combined modality treatment and bladder sparing is similar, if not superior, to that of patients treated with only local therapy— either cystectomy (2, 19) or radiotherapy (5, 6). The exact answer regarding the superiority of a combined treatment over cystectomy will be known when the results
278
I. J. Radiation Oncology
●
Biology
●
Physics
of the randomized, ongoing EORTC/MRC study are published (20). As in other studies (4, 7), we found that patients with complete response to chemotherapy have a better prognosis than those who do not respond to chemotherapy. Nonresponders have a dismal prognosis whatever treatment they receive afterward. Actually, in our patients, the overall survival rate of nonresponders who underwent cystectomy according to the protocol was as poor as the overall survival rate of nonresponders who were irradiated (approximately 30% at 3 years). Complete response after TUR and chemotherapy seems to be a very important prognostic factor for the treatment outcome. Nonresponders presented with an equally high percentage of disseminated disease, regardless of the treatment approach. These results suggest that the response of the primary disease to chemotherapy is indicative of the response of occult metastases to the same agents. The present study confirms that MIBC is a heterogeneous disease, and that among the patients affected, in more than half a bladder-sparing approach is safe and does not com-
Volume 41, Number 2, 1998
promise the survival outcome. Our study also indicates that in nonresponders, radical cystectomy as the treatment of choice is questionable, since it cannot ensure a cure, despite the radical surgical procedure. It is obvious that these patients have biologically more aggressive tumors that are not manageable with either cystectomy or the currently used combined modality treatment. Apparently, for these patients, a new, more aggressive treatment approach should be considered. This could be either concomitant radiochemotherapy with new, more potent chemotherapeutic agents that have demonstrated evidence of radiation enhancement (21), or high-dose chemotherapy (12). The development of biologic markers which could predict the responsiveness of chemoradiotherapy might be extremely useful in the selection of patients for bladder preservation. Some of these biological markers, such as DNA ploidy, various oncogenes, markers of drug resistance, and markers of angiogenesis, are already under study (12); it is hoped that according to them, better treatment selection for the individual patient will be possible in the future.
REFERENCES: 1. Pompe-Kirn, V.; Zakotnik, B.; Volk, N.; Benulicˇ, T.; Sˇkrk, J. Cancer patients in Slovenia 1963–1990. Ljubljana: Onkolosˇki Insˇtitut; 1995:56 – 61. 2. Scher, H.; Shipley, W. U.; Herr, H. Cancer of the bladder. In: DeVita, V. T., Jr.; Helman, S.; Rosenberg, S. A., eds. Cancer: Principles and practice in oncology. 5th ed. Philadelphia, PA: Lippincott; 1997:1300 –1322. 3. Raghavan, D.; Huben, R. Management of bladder cancer. Curr. Probl. Cancer 19:3– 63; 1995. 4. Dunst, J.; Sauer, R.; Schrot, K. M.; Wittekind, C.; AltendorfHoffman, A. Organ-sparing treatment of advanced bladder cancer: A 10-year experience. Int. J. Radiat. Oncol. Biol. Phys. 30:261–266; 1994. 5. Gospodarowicz, M. K.; Hawkins, N. V.; Rawlings, G. A.; Connolly, J. G.; Jewett, M. A. S.; Thomas, G. M.; Herman, J. G.; Garrett, P. G.; Chua, T.; Duncan, W.; Buckspan, M.; Sugar, L.; Rider, W. D. Radical radiotherapy for muscle invasive transitional cell carcinoma of the bladder: Failure analysis. J. Urol. 142:1448 –1454; 1989. 6. Gospodarowicz, M. K.; Rider, W. D.; Keen, C. W.; Connoly, J. G.; Jewett, M. A. S.; Cummings, B. J.; Duncan, W.; Warde, P.; Chua, T. Bladder Cancer: Long-term follow-up results of patients treated with radical radiation. Clin. Oncol. 3:155–161; 1991. 7. Housset, M.; Maulard, C.; Chretien, Y. C.; Dufour, B.; Delanian, S.; Huart, J.; Colardelle, F.; Brunel, P.; Baellet, F. Combined radiation and chemotherapy for invasive transitional-cell carcinoma of the bladder: A prospective study. J. Clin. Oncol. 11:2150 –2157; 1993. 8. Kaufman, D. S.; Shipley, W. U.; Griffin, P. P.; Henery, N. M.; Althausen, A. F.; Efird, J. T. Selective preservation by combination treatment of invasive bladder cancer. N. Engl. J. Med. 329:1377–1382; 1993. 9. Tester, W.; Caplan, R.; Heaney, J.; Venner, P.; Whittington, R.; Byhardt, R.; True, L.; Shipley, W. Neoadjuvant combined modality program with selective organ preservation for invasive bladder cancer: Results of Radiation Therapy Oncology Group Phase II Trial 8802. J. Clin. Oncol. 14:119 –126; 1996.
10. Tester, W.; Porter, A.; Asbell, S. Combined modality program with possible organ preservation for invasive bladder carcinoma: Results of RTOG Protocol 85-12. Int. J. Radiat. Oncol. Biol. Phys. 25:738 –790; 1993. 11. Zietman, A. L.; Shipley, W. U. Organ sparing treatment for bladder cancer: Time to beat the drum. Int. J. Radiat. Oncol. Biol. Phys. 30:741–742; 1994. 12. Sternberg, C. N. Neoadjuvant and adjuvant chemotherapy in locally advanced bladder cancer. Semin. Oncol. 23:621– 632; 1996. 13. Cˇervek, J.; Cˇufer, T.; Kragelj, B.; Zakotnik, B.; Stanonik, M. Sequential transurethral surgery, multiple drug chemotherapy and radiation therapy for invasive bladder carcinoma: Initial report. Int. J. Radiat. Oncol. Biol. Phys. 25:777–782; 1993. 14. Foss Abrahamsen, J.; Fossa, S. D. Long-term morbidity after curative radiotherapy for carcinoma of the bladder: A retrospective study. Strahlenther. Oncol. 166:580 –583; 1990. 15. Kaplan, E. L.; Meier, P. Non-parrametric estimation from incomplete obervations. J. Am. Stat. Assoc. 53:457– 481; 1958. 16. Anderson, J. A.; Cain, K. C.; Gelber, R. D. Analysis of survival by tumor response. J. Clin. Oncol. 1:710 –719; 1983. 17. Anderson, J. A.; Cain, K. C.; Gelber, R. D.; Gelmson, K. S. Analysis and interpretation of the comparisation of survival by treatment outcome variables in cancer clinical trials. Cancer Treat. Rep. 69:1139 –1144; 1985. 18. Rifkin, M. N.; Oblon, D.; Parsons, J.; McCarley, D.; Wajsman, Z. Bladder-sparing treatment for muscle invasive bladder cancer with systemic chemotherapy followed by radiation therapy and adjuvant cisplatin: A 3-year follow-up. Proc. Am. Soc. Clin. Oncol. 11:208; 1992. 19. Bales, G. T.; Kim, H.; Steinberg, G. D. Surgical therapy for locally advanced bladder cancer. Semin. Oncol. 23:605– 613; 1996. 20. Sternberg, C. N. The treatment of advanced bladder cancer. Ann. Oncol. 6:113–126; 1995. 21. Roth, B. J. Preliminary experience with Paclitaxel in advanced bladder cancer. Semin. Oncol. 22:1–5; 1995.