TREATMENT OF INVASIVE bladder cancer remains a

Transcription

1 Combined-Modality Treatment and Selective Organ Preservation in Invasive Bladder Cancer: Long-Term Results By Claus Rödel, Gerhard G. Grabenbauer, Reinhard Kühn, Thomas Papadopoulos, Jürgen Dunst, Martin Meyer, Karl M. Schrott, and Rolf Sauer Purpose: To evaluate our long-term experience with combined modality treatment and selective bladder preservation and to identify factors that may predict treatment response, risk of relapse, and survival. Patients and Methods: Between 1982 and 2000, 415 patients with bladder cancer (high-risk T1, n 89; T2 to T4, n 326) were treated with radiotherapy (RT; n 126) or radiochemotherapy (RCT; n 289) after transurethral resection (TUR) of the tumor. Six weeks after RT/RCT, response was evaluated by restaging- TUR. In case of complete response (CR), patients were observed at regular intervals. In case of persistent or recurrent invasive tumor, salvage-cystectomy was recommended. Median follow-up was 60 months (range, 6 to 199 months). Results: CR was achieved in 72% of patients. Local control after CR without muscle-invasive relapse was TREATMENT OF INVASIVE bladder cancer remains a triple challenge: first, the eradication of local disease, second, the elimination of potential micro-metastases, and third, the maintenance of the best quality of life possible without compromising survival. In the United States as well as in Europe, the gold standard for the treatment of muscle-invasive disease is radical cystectomy with removal of the bladder, prostate, and seminal vesicles in men or the uterus, tubes, ovaries, anterior vaginal wall, and urethra in women. Sophisticated techniques for urinary diversion have been developed in recent years to improve patients quality of life, and considerable progress has been made in this regard. However, even a continent urinary diversion with orthotopic lower urinary tract reconstruction cannot substitute for the patient s original bladder. Over the past 10 to 20 years, multimodality organ-sparing treatment has become the standard of care for many malignancies, including breast cancer, anal cancer, and soft tissue sarcoma among others. Therefore, the question has been asked as to whether primary cystectomy in invasive bladder cancer could also be replaced by an organ-sparing treatment option. However, attempts to obtain bladder preservation are only justified when they have a high likelihood of achieving local cure with no compromise in survival rates. When used alone, neither transurethral resection (TUR) of the bladder tumor nor chemotherapy or radiation alone results in significant local control. Several groups have reported the value of combined-modality therapy, including maintained in 64% of patients at 10 years. Distant metastases were diagnosed in 98 patients with an actuarial rate of 35% at 10 years. Ten-year diseasespecific survival was 42%, and more than 80% of survivors preserved their bladder. Early tumor stage and a complete TUR were the most important factors predicting CR and survival. RCT was more effective than RT alone in terms of CR and survival. Salvage cystectomy for local failure was associated with a 45% disease-specific survival rate at 10 years. Cystectomy because of a contracted bladder was restricted to 2% of patients. Conclusion: TUR with RCT is a reasonable option for patients seeking an alternative to radical cystectomy. Ideal candidates are those with early-stage and unifocal tumors, in whom a complete TUR is accomplished. J Clin Oncol 20: by American Society of Clinical Oncology. TUR, radiation therapy, and systemic chemotherapy With these programs, cystectomy has been reserved for patients with incomplete response or local relapse after combined-modality treatment. Five-year survival rates in the range of 50% to 60% have been published in these series, and approximately three quarters of the surviving patients maintained their own bladder. 13 However, in the absence of any prospective, randomized trial directly comparing radical cystectomy versus trimodality treatment, the value of the bladder-preserving approach in terms of longterm cure as well as the optimal selection of patients for the respective treatment alternatives remain to be established. Bladder cancers constitute a heterogeneous group of tumors. Response to treatment cannot be tailored by absolute predictors of success. As a consequence, physicians recommending selective bladder preservation must exercise considerable judgment, and the patient must accept some From the Departments of Radiation Oncology, Urology, and the Institute of Pathology, University of Erlangen; Department of Radiation Oncology, University of Halle; and Population Based Cancer Registry, Bavaria, Germany. Submitted November 7, 2001; accepted April 16, Address reprint requests to Claus Rödel, MD, Department of Radiation Oncology, University of Erlangen, Universitätsstr 27, Erlangen, Germany; claus.roedel@strahlen.med.uni-erlangen.de by American Society of Clinical Oncology X/02/ /$20.00 Journal of Clinical Oncology, Vol 20, No 14 (July 15), 2002: pp DOI: /JCO

2 3062 RÖDEL ET AL risk in using such an approach. Thus, selection criteria helpful in determining appropriate patients for organ preservation are sorely required. It has been the ongoing policy at the Departments of Urology and Radiation Oncology at the University of Erlangen, Germany, to use definitive radiotherapy (RT) with or without concurrent chemotherapy after conservative surgery for invasive bladder cancer since ,7,12 We now present our 18-year experience with the bladder-sparing approach in a large group of 415 patients and report on clinicopathologic and treatment-related factors that may predict treatment response, risk of relapse, and long-term survival. Patient Characteristics PATIENTS AND METHODS Between May 1982 and December 2000, 428 patients suffering from muscle-invasive (T2 to T4) or high-risk T1-bladder cancer were treated with either RT (n 126) alone or with concomitant radiochemotherapy (RCT; n 302) after initial TUR of the tumor. Risk factors for T1-cancer were defined as tumor grade 3/4, associated carcinoma-insitu (Tis), multifocal tumors, or recurrent tumors refractory to repeated TUR with or without intravesical therapy. Excluded from analysis were 13 patients in which treatment was regarded to be palliative because of concomitant distant disease or in which radiation dose to the bladder was insufficient (less than 45 Gy). All of the remaining 415 patients were free of distant metastases at the time of RT/RCT. Pelvic lymph node metastases (detected by computed tomography or ultrasound), multiple TURs before RT/RCT, or poor general condition with contraindications for radical cystectomy were not considered exclusion criteria. Patient and tumor characteristics are listed in Table 1. Treatment Protocol Treatment was commenced by bimanual palpation and cystoscopic evaluation under general anesthesia with tumor mapping and TUR as thoroughly as was judged safely possible. Residual tumor was assessed histologically by representative biopsies from all resection margins: R0 indicated microscopically complete TUR, R1 microscopic tumor residual, and R2 macroscopic tumor residual. T category was assessed according to the tumor-node-metastasis classification of 1987 (International Union Against Cancer) 14, thus grouping all subepithelial tissue invading tumors under stage T1, all superficial muscle invasive tumors under stage T2, all deep muscle invasive tumors as stage T3, and any tumor with evidence of invasion of surrounding organs under stage T4. Since 1997, tumor staging was modified according to the 1997 tumor-node-metastasis classification by the American Joint Committee on Cancer. 15 For the purpose of this present analysis, all stages were redefined according to the former 1987 tumor-node-metastasis system to make comparisons with other contemporary series possible. RT/RCT was initiated 4 to 8 weeks after initial TUR using 6- to 10-mV photons and a 4-field box technique with individually shaped portals and daily fractions of 1.8 to 2.0 Gy on 5 consecutive days. A median total dose of 54 Gy (range, 45 to 69.4 Gy) was applied to the bladder, and the pelvis was irradiated with a median total dose of 45 Gy (range, 40 to 59.4 Gy). Seventy-nine patients with G3-tumors, macroscopically incomplete (R2) TUR, or evidence of pelvic lymph node metastases additionally received a median total dose of 45 Gy (range, 16.2 to 54 Table 1. Patient and Tumor Characteristics No. of Patients Total no. 415 Sex Male 327 Female 88 Age, years Median 67 Range T category T1, high-risk 89 G2/G3-4 39/50 R0/R1/R2 49/29/11 Unifocal/multifocal/unknown 42/42/5 Tis absent/tis present/unknown 47/32/10 Primary/recurrent tumor 54/35 T2 100 T3 195 T4 31 Grading G1/2 197 G3/4 218 Pelvic lymph node metastases cn0 331 cn 28 Unknown 56 Resection-status after TUR R0 118 R1 135 R2 149 Unknown 13 Multifocality Present 127 Absent 251 Unknown 37 Tis Present 80 Absent 281 Unknown 54 Lymph vessel involvement Present (L1) 175 Absent (L0) 187 Unknown 53 Gy) to the para-aortic lymph nodes up to the level of the third lumbar vertebra. A total of 126 patients were treated by RT alone. Since October 1985, chemotherapy has been administered simultaneously with RT. Chemotherapy was applied during the first and fifth week of RT and consisted of cisplatin (25 mg/m 2 /d; 30-minute infusion on 5 consecutive days) in 145 patients or carboplatin (65 mg/m 2 /d; 30- minutes infusion on 5 consecutive days) in 95 patients with decreased creatinine clearance ( 60 ml/min) or congestive heart disease. Since 1993, a combination of cisplatin (20 mg/m 2 /d; 30-minute infusion on 5 consecutive days) and 5-fluorouracil (5-FU) (600 mg/m 2 /d; 120-hour continuous infusion) was applied to 49 patients. Full-dose chemotherapy was received by 198 patients (68%), ie, the prescribed doses of cisplatin, carboplatin, and 5-FU were administered in the first and at least 75% in the second cycle. In 92 patients (32%), the doses had to be

3 ORGAN PRESERVATION IN BLADDER CANCER 3063 reduced because of hematotoxicity, gastrointestinal toxicity, or nephrotoxicity (graded according to National Cancer Institute-Common Toxicity Criteria 16 ). However, for analysis of the impact of chemotherapy on the different end points, all patients were included (intent-totreat analysis). Criteria for Response, Follow-Up, and Late Toxicity Six weeks after completion of RT/RCT, response quality was evaluated by cystoscopy and deep TUR of the former tumor bed. A complete response (CR) required the absence of any endoscopically visible tumor, the absence of any microscopic tumor in the biopsy specimen, as well as negative urine cytology. In case of CR, patients were observed at 3-month intervals for the first 2 years and every 6 months thereafter. Evaluations consisted of pertinent medical history, physical examination, complete blood counts and blood chemistry, urine cytology, and cystoscopy with biopsies of all suspected areas. In case of persistent or recurrent tumor, additional treatment, such as TUR followed by intravesical therapy for superficial tumors or salvage cystectomy for muscle-invasive tumors, was recommended and initiated at the earliest opportunity. Evaluation of late treatment-related toxicity was performed according to the grading system of late effects of normal tissue (LENT). 17 Statistics All patients were followed up until June At the time of analysis, the median follow-up for the entire group was 36 months and 60 months (range, 6 to 199 months) for all surviving patients; 142 patients have been followed up 5 years and more. Actuarial survival rates were calculated from the time of initial TUR to the time of the last follow-up visit or death using the method of Kaplan and Meier. Differences were analyzed using the log-rank test. The 2 test (twotailed) was used to determine statistical significance between proportions for the end point of initial response to RT/RCT (CR v non-cr). Level of significance was 0.05 (two-sided) in all of the statistical testing. Multivariate analyses and determination of odds ratios were performed using logistic regression analysis (initial response) and the Cox proportional hazards model (censored data). The following factors were tested for predictive and prognostic impact on initial response, local control after CR, distant metastases, and survival rates: sex, age, resection-status after initial TUR, T-category, grading, computed tomography evidence of pelvic lymph node metastases, presence of lymphatic vascular invasion, presence of Tis, presence of multiple foci, and treatment modality (RT, RCT with carboplatin, RCT with cisplatin, or RCT with 5-FU/cisplatin). The statistical analysis was performed using SPSS for Windows version 10.0 (SPSS, Inc, Chicago, IL). RESULTS Initial Response Initial TUR provided a visibly and microscopically complete resection (R0) in 28% (118 of 415 patients; Table 1). As shown in Fig 1, a complete remission after RT/RCT at restaging-tur was achieved in 288 patients (72%), 20 patients (5%) showed only superficial residual tumor (Noninvasive papillary tumor Ta/Tis/T1), and 90 patients (23%) showed muscle-invasive residual tumor. Predictive clinicopathologic and treatment-related factors influencing CR at restaging-tur 6 weeks after completion of RT/RCT are Fig 1. Initial response, local control, and salvage treatment after combined-modality therapy in 415 patients with invasive bladder cancer. listed in Table 2. Completeness of initial TUR, early tumor stage, pelvic lymph node status, and treatment modality revealed the strongest impact on initial response in univariate analysis. If patients with superficial high-risk T1 tumors were analyzed separately, only completeness of initial TUR was significantly (P.003) associated with CR at restaging-tur. In multivariate analysis, resection-status after initial TUR and treatment modality were independently associated with CR. The respective odds ratios for incomplete response are listed in Table 2. Local Control Among 288 patients who had no evidence of disease at restaging-tur, 186 (65%) have been continuously free of tumor in their bladder, 26 patients (9%) experienced a noninvasive (Ta/Tis) relapse, 15 patients (5%) showed a T1-recurrent tumor, 32 patients (11%) had a muscle-invasive relapse, and 10 patients (3%) had a pelvic recurrence (Fig 1). In 19 patients (7%), tumor stage of the recurrent tumor was unknown, and these patients were regarded to have muscle-invasive disease in further analyses. For the whole group of patients, local control without any relapse and freedom from muscle-invasive disease at 5 and 10 years

4 3064 RÖDEL ET AL Table 2. Predictive Factors for Complete Response at Restaging-TUR After Combined-Modality Treatment (TUR RT/RCT) in 398 Patients CR at Restaging- TUR (%) Univariate P Odds Ratio for Incomplete Response 95% Confidence Interval Multivariate P All patients 72 Age median 76.1 Age median 69 Male 73.4 Female 69 R R R T T T T cn cn G1/ G3/4 70 L L1 66 Tis 77.3 No Tis 71 Unifocal 75.3 Multifocal 69 RT alone RCT carboplatin RCT cisplatin RCT 5-FU/cisplatin are shown in Fig 2A. Interestingly, only multifocality of the primary tumor was associated with a higher risk for local relapse (Fig 2B; P.08), whereas none of the other established histopathologic markers, including resectionstatus and T-category or the treatment strategy (RT v RCT), did predict for local failure once CR had been achieved. Local control for T1 tumors and muscle-invasive carcinoma (T2 to T4) are shown in Fig 3A and 3B. The overall relapse-rate at 5 years was not different for T1 and muscleinvasive tumors (Fig 3A); however, progression of T1 tumors to muscle-invasive disease was restricted to 15% at 5 years (Fig 3B). Salvage Treatment for Nonresponding and Recurrent Carcinoma Overall, 83 of 415 patients (20%) underwent salvage cystectomy for invasive residual or recurrent tumor: 41 of 110 noncomplete responders (37%) and 42 of 288 patients (15%) with initial CR who experienced a local relapse (Fig 1). Salvage cystectomy was radical (R0) in all but five patients (in four patients with nonresponding tumors and in one patient with an invasive relapse, only R1/R2 resections Fig 2. (A) Local control without any relapse (ie, without superficial and muscle-invasive relapse) and freedom from muscle-invasive relapse in patients with CR after RT/RCT (n 288). (B) Freedom from any relapse after CR, stratified by unifocal (n 83) and multifocal (n 180) tumors. were achieved). It should be noted that in 35 patients with invasive persistent disease after RT/RCT and in 16 patients with a local relapse after initial CR, salvage cystectomy could not be performed because of poor general health, advanced age, or patients refusal, which were not regarded as exclusion criteria for RT or RCT. The median time interval between initial TUR and cystectomy for nonresponders was 6.2 months (range, 3.5 to 28 months) and 26 months (range, 9.6 to 114 months) for patients with an invasive relapse. Disease-specific survival rates at 5 and 10 years (as calculated from the time of salvage cystectomy) for patients treated by salvage cystectomy for an invasive relapse were 50% and 45%, respectively. These figures dropped to

5 ORGAN PRESERVATION IN BLADDER CANCER 3065 Fig 4. Disease-specific survival for patients after salvage cystectomy (as measured from time of cystectomy), stratified by cystectomy for nonresponders (n 41) and cystectomy for patients with muscle-invasive relapse after initial CR (n 42). 21% and 18%, respectively, for nonresponding patients treated by immediate salvage cystectomy (Fig 4; P.01). In patients with muscle-invasive recurrent or residual tumor who were not suitable for or declined salvage cystectomy, the median survival time was 9.7 months, with no patient alive at 5 years. In the subgroup of patients with superficial recurrent tumor after initial CR salvage treatment by TUR and instillation therapy resulted in 5- and 10-year disease-specific survival rates of 76% and 52%, respectively. Conversely, in patients with superficial residual tumor after RT/RCT who were treated by TUR and instillation therapy, the 5- and 10-year disease-specific survival rates were 40% and 28%, respectively. Fig 3. (A) Local control without any relapse (ie, without superficial and muscle-invasive relapse) after CR, stratified by T1- and muscle-invasive primary tumor (T2 to T4). (B) Freedom from muscle-invasive relapse after CR stratified by T1- and muscle-invasive primary tumor (T2 to T4). Distant Metastases Distant metastases were diagnosed in 98 patients with an actuarial rate of 29% and 35% at 5 and 10 years, respectively. Five-year metastases-free survival rate was 79% for patients with CR tumors but dropped to 52% for patients with residual tumor after combined-modality treatment (Fig 5). Prognostic factors for distant disease are listed in Table 3; only T-category, lymph node metastases, and vascular involvement were independently related to this end point. Interestingly, concurrent systemic chemotherapy had no impact on the development of distant disease, neither for the whole group of evaluated patients (n 415), nor for patients with CR after RT/RCT (n 288). Extension of the RT-target volume to the para-aortic lymph nodes in 79 patients was associated with a reduced metastases-free survival rate at 5 years (62% v 75% in patients without irradiation of the para-aortic region; P.04). However, note that only the subgroup of patients with a high risk of tumor dissemination to the para-aortic lymph nodes (cn, G3 tumors, and macroscopically residual tumor after initial TUR) received para-aortic irradiation. Survival and Bladder Preservation Overall survival and cause-specific survival rates for all patients were 51% and 56% at 5 years and 31% and 42% at 10 years, respectively. Of all surviving patients, more than 80% preserved their bladder (42% at 5 years and 27% at 10

6 3066 RÖDEL ET AL Fig 5. Freedom from distant metastases, stratified by initial tumor response (CR, n 288; non-cr, n 110). Fig 6. Overall survival and overall survival with preserved bladder for all patients (n 415). In the lower plot, salvage cystectomy was taken as an additional event. years, Fig 6). Overall survival at 5 and 10 years was 75% and 51% for T1 tumors and 45% and 29% for muscleinvasive disease, respectively. Prognostic clinicopathologic Table 3. Prognostic Factors for the Development of Distant Disease After Combined-Modality Treatment (TUR RT/RCT) in 415 Patients (see text) Metastases-Free Survival at Univariate 5 Years (%) P Odds Ratio for Metastases 95% Confidence Multivariate Interval P All patients 71 Age median 68.4 Age median 73 Male 72.7 Female 69 R R R T T T T cn cn G1/ G3/4 69 L L Tis 71.9 No Tis 70 Multifocal 71.9 Unifocal 71 RT alone 74.2 RCT carboplatin 64 RCT cisplatin-based 73 RCT 5-FU/cisplatin 65 and treatment-related factors influencing overall survival are summarized in Table 4. The strongest impact on overall survival was noted for T-category and resection-status after initial TUR. An independent value in multivariate analysis was also confirmed for treatment mode (RT v RCT with carboplatin v RCT with cisplatin 5-FU), age, and lymph vessel involvement. Acute Toxicity and Chronic Sequelae Typical acute radiation-induced side effects, such as transient urocystitis and enteritis, were easily managed by symptomatic treatment. The percentages of patients experiencing grade 3 or greater hematologic or nonhematologic acute toxicity and the impact of the different chemotherapy regimens, as well as the extension of the RT-target volume to the para-aortic region, are listed in Table 5. Grade 4 toxicity was mainly restricted to hematologic side effects; however, neither bleeding nor life-threatening infections occurred. One patient died from acute enteritis and electrolyte disorder during the course of RCT. The use of carboplatin was associated with a slightly increased risk of grade 3 to 4 thrombocytopenia. Gastrointestinal and bladder toxicity were increased with the use of a cisplatin-containing regimen as compared with RT alone or RCT with carboplatin. Extension of the RT-treatment volume to the para-aortic region markedly increased hematologic and nonhematologic acute toxicity. The percentages of patients sustaining chronic sequelae reported during the follow-up period of this study is listed in Table 6. Three patients underwent cystectomy because of a shrinking bladder (all three were treated with multiple TURs

7 ORGAN PRESERVATION IN BLADDER CANCER 3067 Table 4. Prognostic Factors for Overall Survival After Combined-Modality Treatment (TUR RT/RCT) in 415 Patients Overall Survival (%) At 5 Years At 10 years Univariate P Odds Ratio for Death 95% Confidence Interval Multivariate P All patients Age median Age median Male Female R R R T T T T cn cn G1/ G3/ L L Tis No Tis Multifocal Unifocal RT alone RCT carboplatin RCT cisplatin RCT 5-FU/cisplatin before RCT), and six patients experienced late gastrointestinal toxicity grade 4 requiring surgical intervention. One of these patients perioperatively died from cardiac failure. Two of the six patients with small bowel obstruction requiring surgery had received 5-FU/cisplatin-chemotherapy together with para-aortic irradiation. Six patients suffered from a reduced bladder capacity with less than 2-hour intervals of micturition. Mild dysuria, diarrhea, and urgency with nocturia occurred in 10% to 20% of assessable patients. DISCUSSION We demonstrate in this large series of patients with long-term follow-up that a bladder-sparing approach by combined modality treatment can be performed safely. CR of the tumor was achieved in more than 70% of patients. Durable local control without muscle-invasive relapse was maintained in the majority of these patients (72% after 5 years and 64% after 10 years). For patients with a muscleinvasive relapse, salvage cystectomy could still be per- Table 5. Acute Toxicity According to RCT-Regimen and RT-Treatment Volume % of Patients With Grade 3/4 Toxicity All Patients (N 415) RT Alone (n 126) RCT With Carboplatin (n 95) RCT With Cisplatin (n 145) RCT With 5-FU/Cisplatin (n 49) RT/RCT Para-aortic RT (n 79) RT/RCT Para-aortic RT (n 336) Leucopenia 13/3 21/4 20/3 13/3 19/9 10/1 Thrombocytopenia 7/3 12/6 6/2 10/4 9/6 6/2 Anemia 2/0 3/0 4/0 2/0 3/0 2/0 Creatinine elevation 3/0 2/0 4/0 4/0 3/0 3/0 Diarrhea 5/1* 3/1 4/2 8/1* 10/0 9/2 5/0 Cystitis 5/0 4/0 4/0 7/0 4/0 5/0 4/0 Nausea/vomiting 3/0 1/0 2/0 8/0 7/0 4/0 3/0 *Including one treatment-related death (grade 5).

8 3068 RÖDEL ET AL Table 6. Late Toxicity According to LENT/SOMA No. of Patients/Total % Grade 4 Salvage cystectomy due to contracted 3/186 2 bladder Bowel obstruction requiring surgical 6/ intervention Grade 3 Reduced bladder capacity (volume /186 3 ml) with 2-hour intervals of micturition Grade 2 Frequency with urgency and nocturia 18/ Dysuria, intermittent and tolerable 15/186 8 Moderate diarrhea 20/415 5 Proctitis 8/415 2 NOTE. For gastrointestinal toxicity, the entire cohort of 415 patients was considered, for bladder toxicity, 186 patients with preserved bladder and without local relapse were evaluated. formed and still had a curative potential (50% and 45% disease-specific survival rates after 5 and 10 years, respectively). Disease-specific survival was 56% at 5 years and 42% at 10 years, with more than 80% of long-term survivors maintaining their own bladder. The irradiated preserved bladder functioned well with only three cystectomies (2%) required because of a shrinking bladder within our observation period of 18 years. In this regard, it is important to stress that cystectomy because of a shrinking bladder, as well as the markedly reduced bladder capacity in further five patients (3%), exclusively occurred after multiple TURs preceding definitive RT/RCT. The primary goal of the bladder-sparing approach remains optimal patients survival. Thus, results of this organsparing approach need to be compared with the surgical standard. Unfortunately, primary cystectomy has not yet been tested against combined-modality bladder-sparing treatment in randomized trials. In a recently published series of 1,054 patients with bladder cancer treated between 1971 and 1997 with radical cystectomy, Stein et al 18 reported excellent overall survival rates that, according to the authors, should be considered as the surgical standard to which other treatment modalities should be compared. They reported a 5- and 10-year overall survival in the range of 60% and 43% for the whole group of patients, which compared favorably to our results in 415 patients with a 51% overall survival rate at 5 years and 31% at 10 years. 18 However, this cystectomy series included 213 patients (20%) with noninvasive (T0, Ta, Tis) bladder cancer and excluded 112 patients with inoperable tumor, unradical surgery, or intraoperative detection of distant metastases. Conversely, our series did not include superficial noninvasive tumors and did not exclude patients who were, for different reasons, unsuitable for surgery. If comparison is restricted to the respective groups, the 5-year overall survival rates of this radical cystectomy series and our bladder-sparing approach become equal: 74% and 75%, respectively, for T1-tumors, and 47% and 45%, respectively, for muscle-invasive disease. Notwithstanding the limitations of nonrandomized comparison, this analysis indicates that selective bladder preservation by trimodality treatment may result in longterm cure and survival rates comparable to the best cystectomy series, allowing this treatment to be considered a reasonable treatment option for patients who are deemed medically unfit for cystectomy and for those seeking an alternative to radical cystectomy. Based on these results, a prospective, randomized trial directly comparing cystectomy with trimodality selective bladder-sparing treatment would be warranted. Unfortunately, others 19 and we have found it impossible to have patients accept a randomization to bladder preservation versus radical cystectomy in a prospective trial. As more experience is acquired with organ-sparing treatment, it is clear that future directions of clinical and basic research will focus on two main topics: (1) the optimization of the treatment modalities, including incorporation of new cytotoxic agents, and (2) the proper selection of patients who will most probably benefit from the respective treatment alternatives. As demonstrated in our study, clinical criteria helpful in determining patients for bladder preservation include such variables as early tumor stage and a visibly and microscopically complete TUR. Because of its overwhelming predictive and prognostic impact, a TUR as thorough as safely possible should always be attempted. Ta and superficially infiltrating cancers with high or moderate differentiation (T1G1 to G2) can often be managed successfully by TUR alone. Patients in whom T1- tumors are less differentiated (G3), large, multiple, or associated with Tis are at greater risk of recurrent disease and tumor progression. The treatment recommendations for these high-risk T1-tumors differ widely. Some authors go on to favor radical cystectomy, 20,21 whereas most of the recent series propose an organ-preserving approach, including TUR with fulguration followed by intravesical bacille Calmette-Guérin or intravesical chemotherapy with thiotepa, mitomycin, or doxorubicin Intravesical therapy after TUR has been reported to significantly reduce recurrences in these high-risk T1-cancers; however, it also became evident that it mainly reduces favorable recurrences, whereas it had only a limited effect on progression. The American Urological Association has recently published treatment recommendations for superficial bladder cancer on the basis of a meta-analysis of studies with intravesical therapy. 25 According to this analysis, all types

9 ORGAN PRESERVATION IN BLADDER CANCER of drugs may reduce the rate of recurrences, and intravesical therapy is therefore recommended for T1- and high-risk Ta-tumors after TUR of the tumor. However, the recommendations include the remark that there is no effect of intravesical therapy on long-term progression. Given the possibility that, at the time of TUR, the tumor may be clinically understaged as superficial cancer and may have already invaded into the muscle wall, RT should exert certain advantages over intravesical instillation therapy because deeper cell deposits can be more effectively treated by RT. In our series of TUR plus RT/RCT in high-risk T1-tumors, only 15% of tumors had progressed to muscleinvasive disease after 5 years, which compares favorably to most studies of TUR plus immuno- or chemo-ablative therapy. Nevertheless, the high rate of overall relapses (44% at 5 years) clearly indicates a lifelong risk to develop recurrent disease and requires regular cystoscopies in these patients with superficial disease. In selected patients with muscle-invasive bladder cancer, Herr 26 recently published long-term results of TUR alone. Patients with tumors judged clinically to be confined to the bladder with no palpable mass or hydronephrosis, who presented with T0 or T1-tumors on restaging-tur (n 151), were given the option of follow-up with regular cystoscopy and salvage treatment as needed (n 99) or immediate cystectomy (n 52). The 10-year diseasespecific survival in patients who received TUR as definitive treatment was 76% (57% with bladder preserved). This outcome was comparable or even better compared with those who chose radical cystectomy as the treatment alternative (71% at 10 years). This excellent subgroup compares, in part, with our patients who had a radical TUR (R0) before RT/RCT. The 5- and 10-year disease-specific survival in this favorable group was 82% and 69%, respectively, with a bladder preservation rate of 85% at 5 years. In our opinion, these patients with a good prognosis with regard to survival are optimal candidates for an organ-sparing approach because the low amount of residual cells after TUR can excellently be controlled by adjuvant RT/RCT, a situation comparable to the organ-sparing approach in early-stage breast cancer. The optimal regimen and combination of RT and chemotherapy remains to be established. In our series, RT with concurrent chemotherapy not only increased the rate of CR, but also was associated with a significantly improved overall survival. This is in line with the only prospective, randomized comparison of RT alone versus RCT in bladder cancer; an improved local control rate was reported when cisplatin was given in conjunction with RT. 3 It is noteworthy that the improved survival rates in our series with combined RCT compared with RT alone were primarily an 3069 effect of the higher initial response rates. The addition of chemotherapy did not show any impact on the development of distant metastases, which is also reflected in the contradictory, albeit mostly negative, results of adjuvant and neoadjuvant chemotherapy in cystectomy-based series The intensification of our concurrent chemotherapy protocols over the past 18 years clearly yielded an increasing rate of CR with 61% for patients treated with RT alone, 66% after RCT with carboplatin, 82% after RCT with cisplatin, and 87% after RCT with 5-FU/cisplatin. This remains also true if adjusted to the most important prognostic factor, the completeness of the initial TUR, with the rate of CR after incomplete TUR (R1/R2-resection) being 46% after RT alone, 57% after RCT with carboplatin, 78% after RCT with cisplatin, and 82% after RCT with 5-FU/cisplatin. Interestingly, a recently published phase III trial of combined radiation and chemotherapy in invasive bladder cancer did not show any advantage of upfront methotrexate, cisplatin, and vinblastine chemotherapy before concurrent RCT with cisplatin in terms of local control, bladder preservation, distant metastases, and survival. 8 Evidently, prolongation of the overall treatment time by sequential chemotherapy may be biologically less favorable than a more intensive course of concurrent RCT in bladder cancer. Newer chemotherapeutic agents, particularly gemcitabine 30 and the taxanes, 31,32 are now being tested in combination with RT and may further improve organ preservation in bladder cancer. To further optimize patient selection, it should be of pivotal interest to recognize the subgroup of tumors that do not respond to RT/RCT. In this study, patients with nonresponding tumors showed a 5-year disease-specific survival rate of only 21%, even when salvage cystectomy could be performed, and more than 40% developed distant metastases within the first 2 years. Evidently, these tumors have a biologically less favorable profile, and prompt cystectomy, possibly combined with more aggressive adjuvant chemotherapy, might be more effective in these patients. However, tumor heterogeneity is so great in bladder cancer that conventional histopathologic classification is inadequate for predicting the response to RCT for individual lesions. In our analysis, the presence of histopathologic markers that indicate more aggressive tumors, such as associated Tis, G3/G4, multifocality, and lymph vessel involvement, revealed no impact on initial response to RT/RCT. Translational research to identify molecular markers that may better identify a tumor s true malignant potential as well as its response to specific cytotoxic therapies are sorely needed. We have recently published an immunohistochemical study in 70 patients with invasive bladder cancer uniformly treated by RCT within this bladder-sparing protocol. 33 Higher rates of spontaneous apoptosis and proliferation

10 3070 RÖDEL ET AL were significantly related to better initial response and better local control with bladder preservation. This relationship has now been confirmed by other groups using the bladderpreserving approach. 34 Furthermore, in this present study, we identified multifocality of the tumor as a risk factor for local relapse in patients who have achieved CR by primary treatment. This may indicate that multifocality of the tumor is associated with transitional cell epithelium having a high propensity to develop recurrent or de-novo secondary carcinoma. Further studies are required to select tumors less likely to respond to or recur after RCT. In conclusion, we have learned over the 18-year experience with our bladder-sparing approach that any component of the trimodality therapy contributes considerably to the overall success. In the beginning of this series, the importance of a visibly or even microscopically complete TUR was not clearly recognized. Once this has been realized, a TUR as thorough as safely possible has always been attempted. If a patient is now conferred to our department with a TUR performed externally, re-resection biopsies of the periphery and the tumor bed and, if required, reresection of residual tumor are mandatory. The intensification of our RCT-regimen (RT alone, RCT with carboplatin, RCT with cisplatin, and RCT with 5-FU/cisplatin) has been associated with higher response rates and increased overall survival figures. However, we also noted an increase in late-onset sequelae, especially when RCT with 5-FU/cisplatin was combined with para-aortic irradiation, a strategy that was subsequently abandoned. Intensification of RT by hyperfractionated accelerated regimens might also be superior to conventional fractionation because of the proliferation rate of urothelial cancers 35 and because of the prognostic impact of the overall treatment time. 36 These regimens are currently being tested 9 but should not yet be considered standard. Moreover, patient surveillance after combinedmodality treatment with salvage cystectomy performed immediately for nonresponding or recurrent tumors remains an integral and important part of the overall success rate of this treatment option. All patients should be observed closely with surveillance cystoscopies by an urologist for some years, if possible throughout life, because they remain at risk for developing recurrences even beyond 5 to 10 years. Combined-modality treatment as an alternative to cystectomy is still investigative in bladder cancer, and it is recommended that such treatment be administered only by dedicated multimodality teams. However, our long-term results demonstrate that this organ-sparing approach is not only a curative treatment option for patients who are unsuitable for radical surgery. Ideal candidates for organ preservation are also those with early-stage and unifocal tumors, in whom a microscopically or at least visibly complete TUR is accomplished. 1. Housset M, Moulard C, Chretien Y, et al: Combined radiation and chemotherapy for invasive transitional-cell carcinoma of the bladder: A prospective study. J Clin Oncol 11: , Dunst J, Sauer R, Schrott KM, et al: Organ-sparing treatment of advanced bladder cancer: A 10-year experience. Int J Radiat Oncol Biol Phys 30: , Coppin CM, Gospodarowicz MK, James K, et al: Improved local control of invasive bladder cancer by concurrent cisplatin and preoperative or definitive radiation. J Clin Oncol 14: , Tester W, Caplan R, Heaney J, et al: Neoadjuvant combined modality program with selective organ preservation for invasive bladder cancer: Results of Radiation Therapy Oncology Group phase II trial J Clin Oncol 14: , Kachnic LA, Kaufmann DS, Heney NM, et al: Bladder preservation by combined modality therapy for invasive bladder cancer. J Clin Oncol 15: , Cervek J, Cufer T, Zakotnik B, et al: Invasive bladder cancer: Our experience with bladder sparing approach. Int J Radiat Oncol Biol Phys 41: , Sauer R, Birkenhake S, Kühn R, et al: Efficacy of radiochemotherapy with platin derivates compared to radiotherapy alone in organ-sparing treatment of bladder cancer. Int J Radiat Oncol Biol Phys 40: , Shipley WU, Winter KA, Kaufman DS, et al: Phase III trial of neoadjuvant chemotherapy in patients with invasive bladder cancer treated with selective bladder preservation by combined radiation REFERENCES therapy and chemotherapy: Initial results of Radiation Therapy Oncology Group J Clin Oncol 16: , Zietman AL, Shipley WU, Kaufman DS, et al: A phase I/II trial of transurethral surgery combined with concurrent cisplatin, 5-fluorouracil, and twice-a-day radiation followed by selective bladder preservation in operable patients with muscle-invading bladder cancer. J Urol 160: , Birkenhake S, Leykamm S, Martus P, et al: Concomitant radiochemotherapy with 5-FU and cisplatin for invasive bladder cancer. Acute toxicity and first results. Strahlenther Onkol 175:97-101, Kaufman DS, Winter KA, Shipley WU, et al: The initial results in muscle-invading bladder cancer of RTOG 95-06: Phase I/II trial of transurethral surgery plus radiation therapy with concurrent cisplatin and 5-fluorouracil followed by selective bladder preservation or cystectomy depending on the initial response. Oncologist 5: , Rödel C, Dunst J, Grabenbauer GG, et al: Radiotherapy is an effective treatment for high-risk T1-bladder cancer. Strahlenther Onkol 177:82-88, Dunst J, Rödel C, Zietman A, et al: Bladder preservation in muscle-invasive bladder cancer by conservative surgery and radiochemotherapy. Semin Surg Oncol 20:24-32, Hermanek PS: UICC-International Union Against Cancer: TNM Classification of Malignant Tumors (ed 4). Heidelberg, Germany, Springer-Verlag, 1987, pp

11 ORGAN PRESERVATION IN BLADDER CANCER 15. AJCC Cancer Staging Manual (ed 5). Philadelphia, PA, Lippincott-Raven, 1997, p Trotti A, Byhardt R, Stetz J, et al: Common toxicity criteriaversion 2.0: An improved reference for grading the acute effects of cancer treatment Impact on Radiotherapy. Int J Radiat Oncol Biol Phys 47:13-47, Rubin P, Constine L, Fajardo L, et al: Overview: Late effects of normal tissue (LENT) scoring system. Int J Radiat Oncol Biol Phys 31: , Stein JP, Lieskovsky G, Cote R, et al: Radical cystectomy in the treatment of invasive bladder cancer: Long-term results in 1,054 patients. J Clin Oncol 19: , Shipley WU (in reply to Shahab N): Bladder preservation trial: Radiation Therapy Oncology Group J Clin Oncol 17: , Amling CL, Thrasher JB, Frazier HA, et al: Radical cystectomy for stages Ta, Tis and T1 transitional cell carcinoma of the bladder. J Urol 151:31-36, Freemann JA, Esrig D, Stein JP, et al: Radical cystectomy for high risk superficial bladder cancer in the era of orthotopic urinary reconstruction. Cancer 76: , Herr HW, Schwalb DM, Zhang ZF, et al: Intravesical bacillus Calmette-Guerin therapy prevents tumor progression and death from superficial bladder cancer: Ten-year follow-up of a prospective randomized trial. J Clin Oncol 13: , Igawa M, Urakami S, Shirakawa H, et al: Intravesical instillation of epirubicin: Effect on tumor recurrence in patients with dysplastic epithelium after transurethral resection of superficial bladder tumor. Br J Urol 77: , Krege S, Giani G, Meyer R, et al: A randomized multicenter trial of adjuvant therapy in superficial bladder cancer: Transurethral resection only versus transurethral resection plus mitomycin C versus transurethral resection plus bacillus Calmette-Guerin. J Urol 156: , Smith JA, Labasky RF, Cockett AT, et al: Bladder cancer clinical guidelines panel summary report on the management of nonmuscle invasive bladder cancer (stages Ta, T1 and Tis): The American Urological Association. J Urol 162: , Herr HW: Transurethral resection of muscle-invasive bladder cancer: 10-year outcome. J Clin Oncol 19:89-93, Ghersi D, Stewart LA, Parmer MKB, et al: Does neoadjuvant cisplatin-based chemotherapy improve the survival of patients with locally advanced bladder cancer: A meta-analysis of individual patient data from randomized clinical trials. Br J Urol 75: , International Collaboration of Trialists on Behalf of the Medical Research Council Advanced Bladder Cancer Working Party: Neoadjuvant cisplatin, methotrexate, and vinblastine chemotherapy for muscle-invasive bladder cancer: A randomised controlled trial. Lancet 354: , Skinner DG, Daniel JR, Russel CA, et al: The role of adjuvant chemotherapy following cystectomy for invasive bladder cancer: A prospective comparative trial. J Urol 145: , Sternberg CN: Gemcitabine in bladder cancer. Semin Oncol 27:31-39, Dunst J, Weigel C, Heynemann H, et al: Preliminary results of simultaneous radiochemotherapy with paclitaxel for urinary bladder cancer. Strahlenther Onkol 175:7-10, Nichols RC, Sweetser MG, Mahmood SK, et al: Radiation therapy and concomitant paclitaxel/carboplatin chemotherapy for muscle invasive transitional cell carcinoma of the bladder: A well tolerated combination. Int J Cancer 90: , Rödel C, Grabenbauer GG, Rödel F, et al: Apoptosis, p53, bcl-2, and Ki-67 in invasive bladder carcinoma: Possible predictors for response to radiochemotherapy and successful bladder preservation. Int J Radiat Oncol Biol Phys 46: , Moonen L, Ong F, Gallee M, et al: Apoptosis, proliferation and p53, cyclin D1, and retinoblastoma gene expression in relation to radiation response in transitional cell carcinoma of the bladder. Int J Radiat Oncol Biol Phys 49: , Maciejewski B, Majewski S: Dose fractionation and tumor repopulation in radiotherapy for bladder cancer. Radiother Oncol 21: , De Neve W, Lybeert ML, Goor C, et al: Radiotherapy for T2 and T3 carcinoma of the bladder: The influence of overall treatment time. Radiother Oncol 36: , 1995