In/ J Radmion Oncology Bm/. Phm, Vol. 19, pp. 693-699 Printed in the U.S.A. All rights reserved.

Copyright

0360-3016/90 $3.00 + Ml 0 1990 Pergamon Press plc

??Phase I/II Clinical Trial

COMBINED

5-FLUOROURACIL AND IRRADIATION FOR TRANSITIONAL CELL CARCINOMA OF THE URINARY BLADDER

K. J. RUSSELL, M.D.,’ M. A. BOILEAU, M.D.,3 C. HIGANO, M.D.,* C. COLLINS, M.D.,* A. H. RUSSELL, M.D.,’ W. KOH, M.D.,’ S. B. COLE, M.D.,4 W. H. CHAPMAN, M.D.3 AND T. W. GRIFFIN, M.D.’ ‘Dept. of Radiation Oncology, ‘Division of Medical Oncology, 3Department of Urology, University of Washington Medical Center; 4Department of Radiation Oncology, Providence Medical Center, Seattle, WA and 5Radiological Associates of Sacramento, Sacramento, CA

Thirty-four patientshavecompleted treatmentona bladder-preservation protocolusing primary irradiation combined with infusion 5-fluorouracil (5-FU). 4,000 cGy pelvic irradiation was delivered in 5 weeks, with 1,000 mg/m2/day of 5-FU administered as a 96 hr infusion on days l-4 of week 1 and 4. After a 3-week rest period, patients eligible for cystectomy underwent cystoscopy and biopsy. Those with residual tumor underwent cystectomy, and those without tumor received an additional cycle of chemotherapy and irradiation. Patients ineligible for cystectomy for reasons medical, surgical, or refusal received a third cycle without the 4-week delay or re-evaluation. With a median follow-up of 18 months (range 2-45 months), and with 25/34 patients having T3 (16) or T4 (9) tumors, 17 patients are NED, 4 have died of intercurrent deaths, 7 have died with bladder cancer, and 6 are alive with tumor (2 confined to the bladder). The actuarial cancer-specific survival for the entire group of patients is 64% (t-12%) at 45 months, with a freedom from relapse of invasive cancer of 54% (*lo%). Twenty-four of the 34 patients retained intact bladders, with 20/24 reporting entirely normal voiding. Of 18 potential surgical candidates, 13/16 (81%) who underwent pathologic re-staging after 2 cycles of chemoradiotherapy had no histologic evidence of residual cancer. Of these 13 patients, 8 remain NED and 2/13 have locally recurrent non-invasive tumors only. Treatment was well-tolerated, with 28/34 patients having received 100% of the planned 5-FU and 34/34 having received >80%. This regimen appears more successful than radiotherapy alone in achieving complete tumor responses, and is an attractive alternative for patients who are unable to receive more aggressive chemotherapy/radiation combinations. Bladder cancer, 5-fluorouracil.

INTRODUCTION

The 5-year survival rate for patients without such pathological “downstaging” of the tumor is only 27-48% (25, 8, 14, 19, 24). For patients cured of their bladder cancers by primary radiation, normal bladder function is retained by the vast majority, with only 2-l 1% experiencing late radiation complications (7, 9, 10, 25). With these observations as historical precedence, this clinical trial was designed to address two questions: (a) can the percentage of patients with complete clinical responses to radiation be increased by the addition of the radiation potentiating drug Sfluorouracil, and does this increased response rate result in a better prognosis, and (b) can selected patients with radioresponsive tumors be treated successfully by chemoradiotherapy alone, provided that there is no histologic evidence of residual tumor at

Extensive clinical experience has demonstrated the efficacy of external beam radiotherapy for selected patients with radioresponsive bladder cancers. Patients who achieve a complete clinical response within 6 months of completion of radical radiotherapy enjoy a 45-69% probability of surviving 5 years, as opposed to patients with unresponsive tumors who experience a 6- 14% 5-year sur-

vival rate. The relationship between tumor radioresponsiveness and prognosis is also observed for patients undergoing combined pre-operative radiation followed by planned cystectomy. The 14-43% of patients whose cystectomy specimens reveal no histologic evidence of residual cancer as a result of the radiation enjoy a favorable prognosis in the range of a 64-80% 5-year survival rate. Presented at the 3 I st annual ASTRO meeting, San Francisco,

Dr. K. Russell is the recipient of an American Cancer Society

CA, l-6 October 1989. Reprint requests to: K. J. Russell, M.D., University

Career Oncology Development

Cancer Center, NN- 111, University of Washington Medical Center, RC08, 1959 N.E. Pacific St., Seattle, WA 98195.

Award.

Accepted for publication 22 February 1990.

693

694

I. J. Radiation Oncology 0 Biology 0 Physics

the time that planned cystectomy would normally be performed.

September 1990, Volume 19, Number 3 SURGICAL CANDIDATES 4,000 cGy

CYSTECTOMY bweeks

5-FU :.I: x 2d

METHODS

CYSTOSCOPY

2,000 cGy

_ \

AND MATERIALS

This trial was initiated in September 1985 as a single institution study at the University of Washington. One year later, the study expanded as a multi-institutional pilot under the auspices of the Puget Sound Oncology Consortium, a regional collaborative group affiliated with the Southwest Oncology Group (SWOG). Accordingly, patients have been treated at seven participating regional facilities in Washington and Oregon. All patients had biopsy proven transitional cell carcinoma. Patients with tumors containing squamous elements were eligible, but patients with pure squamous cell carcinoma or adenocarcinoma were excluded. Patients with both muscle-invading and non-invasive tumors were enrolled. Eligible patients with non-invasive tumors were those who had experienced recurrence of their tumors following intravesical chemotherapy and for whom cystectomy was deemed the appropriate conventional therapy of their tumor recurrence. The pre-treatment evaluation in all patients included a history and physical examination, routine serum chemistries and hematologic indices, cystoscopic evaluation, computed tomography of the pelvis (CT), and chest radiographs. At the time of cystoscopy, information was collected as to the location and focality (solitary vs multifocal) of the tumors, the tumor(s) surface morphology, (papillary versus solid or ulcerative), and any attempts at transurethral tumor resection (TURBT). The majority of the patients underwent radionuclide bone scanning, and all patients gave written informed consent prior to the initiation of therapy. At the conclusion of staging, patients were placed into one of two treatment groups. Those patients who were candidates for cystectomy were assigned to a treatment group which involved a cystoscopic re-evaluation for tumor response after 2 cycles of chemoradiotherapy. Those patients who were not candidates for cystectomy were treated without a mid-course cystoscopic re-evaluation. Patients were considered candidates for cystectomy if their tumors were confined to the bladder or had extravesical spread limited to the vagina or prostate. Patients were considered non-surgical candidates for one of three conditions: surgical inoperability (tumor extension to the abdominal wall or rectum, or pelvic lymph node metastases on CT), medical inoperability (physical frailty), or inoperability due to patient refusal. The treatment schema for these two groups of patients is summarized in Figure 1. All patients received an initial course of 4,000 cGy pelvic radiation therapy given with two 96-hr infusions of 5-fluorouracil ( WU) over a period of 5 weeks. Pelvic irradiation was administered for an initial 2 weeks, with 5-FU administered on days l-4 of

/

RE-BtOPSY

+

5-FU &.I. x 1

NON-SURGICAL CANDIDATES 4,400 cGy PELVIS

+

5-FU C.I. x 3

6,000 cGy BLADDER

Fig. 1. Treatment fusion.

schema.

C.I. = continuous

intravenous

in-

the radiation therapy (1,000 mg/m2/day by continuous intravenous infusion, with a maximum of 1,800 mg/day). This first 2-week cycle of treatment was followed by 1 week without treatment, and then a second 2-week cycle of combined treatment, again delivering 2,000 cGy to the pelvis with concurrent 5-FU on days l-4. Patients ineligible for cystectomy received a third cycle of treatment 1 week after completion of the second cycle. This cycle differed from the first two by the shrinking of the radiation treatment portals after the first 400 cGy of this last cycle to include only the bladder for the remaining 1,600 cGy. Therefore, the complete treatment consisted of 4,400 cGy to the pelvic lymph nodes, 6,000 cGy to the entire bladder, and 3 concurrent cycles of 5-FU, delivered over 8 week’s time. Patients eligible for cystectomy were allowed a 3-week period of rest after the completion of the first 2 treatment cycles in order to allow both tumor regression as well as patient recovery from any acute side-effects of therapy. These patients then underwent a repeat cystoscopy under anesthesia, with deep-muscle biopsies performed of the original tumor site(s). Those patients with histologic evidence of residual tumor underwent radical cystectomy l-2 weeks later. Those without evidence of residual tumor received the third cycle of treatment, identical to that described for the non-surgical patients. Treatment for these patients was therefore completed in 11 weeks. All radiotherapy was delivered by megavoltage linear accelerators with an energy of 4 MeV or greater. Pelvic fields extended from the L5-Sl interspace to the bottom of the obturator foramina. At the time of both treatment simulations for pelvic and boost fields, contrast was instilled into both bladder and rectum. Lateral field borders on the anterior and posterior pelvic fields incorporated the bony pelvic brim with 1.5-2.0 cm margin. The posterior field edge of the lateral pelvic ports bisected the posterior sweep of the sigmoid colon and provided at least 2 cm posterior margin on the contrast visualized in the bladder or any posterior extravesical extension of tumor. The anterior margin of the lateral pelvic fields remained 2 cm in front of introduced air in the dome of the bladder, which had also been distended with 90- 120 cc of contrast

695

5-FU/XRT for bladder cancer 0 K. J. RUSSELLeta/. medium

via Foley catheter.

In no case did the anterior

field edge lie behind the symphysis pubis. The boost fields covered the entire bladder, with a 2 cm margin on the visualized contrast or known extravesical extension of tumor, as measured from pre-treatment computerized tomography. Corner blocking was inserted on both pelvic and boost fields to minimize treatment of uninvolved normal tissues, and all 4 fields were treated each day. After completion of therapy, all patients received routine followup at 3-month intervals. Patients with retained bladders underwent follow-up cystoscopy and directed biopsies, as clinically warranted, at the same intervals. Follow-up ranges from 2-45 months, with a median of 18 months. Survival and relapse analysis was performed by the actuarial method of Kaplan and Meier (+ 1 standard error) ( 12). RESULTS A total of 40 patients have been enrolled. Thirty-four patients have completed the prescribed treatments and form the study population. Six patients were removed from study prior to completing the protocol, (see “Toxicity of Therapy”, below), and are excluded from survival and tumor response analysis. The age range of the patients was 39-86 years, with a median of 69 years; 27 patients were male and 7 were female. Of the 34 patients, 18 were considered surgical candidates and 16 were not. For the non-surgical candidates, the criteria for inoperability were: surgical inoperability (5) medical inoperability (5), and patient refusal (6). The distribution of tumor stages for the 34 patients, subgrouping by treatment cohort and treatment outcome by stage is shown in Table 1. For the entire study population, the actuarial cancer-specific survival is 64% (a 12%) at 45 months, with a freedom from relapse of invasive cancer of 54% (+-10%). Results-surgical

candidates

Of the 18 patients who were surgical candidates, 16 underwent the full urologic re-evaluation and biopsy after the initial two cycles of treatment, and 2 underwent cystoscopy without biopsy. The two patients who were not biopsied had no evidence of tumor by visual inspection and received the third cycle of chemoradiotherapy as if they had achieved a histologically confirmed tumor clearance. Of the 16 patients who were biopsied, 11/ 16 (69%) had no tumor in the biopsy specimen and 5116 had histologic evidence of residual tumor. Of the five patients who underwent planned cystectomy l-2 weeks later, 2/ 5 had no evidence of tumor in the cystectomy specimen. These two patients had extensive circumferential tumors at initial presentation which were not resectable with a simple biopsy. If one allows that the additional 2 weeks prior to cystectomy allowed for the completion of tumor

Table 1. Distribution Surg. cand.

of tumor stages and treatment outcomes

Non-surg. cand. 1 3

Fraction alive

Fraction NED

3 6

313 416

213 316

517 519

317 319

517 J/2

517 J/2

23134

17/34*

Total

Tl T2

2 3

T3A

4

3

I

T3B T4A T4B

6 3 0

3 4 2

9 7 2

18

16

34

* Includes 2 patients with non-invasive successfully treated with TURBT.

tumor recurrences

regression in these two patients, these patients may be included as complete responders, yielding a percentage of tumor clearance after the initial 2 cycles of chemoradiotherapy of 8 1% ( 13/16).

Tumor stage, focality, morphology, and pre-treatment resectability via transurethral resection (TURBT) were tabulated for the 16 patients who were evaluated for complete tumor clearance after the initial two cycles of therapy. This analysis was performed in order to see how these potential predictive indicators (22) might associate with a negative biopsy after 4,000 cGy. These data are summarized in Table 2. Of the 18 patients in this treatment arm, 9/ 18 remain without evidence of bladder cancer (NED), 2/ 18 have locally recurrent, non-invasive bladder cancer at different locations within the bladder successfully managed by TURBT (both patients having had invasive tumors at presentation), l/l 8 is alive with metastatic disease, 2/18 have died of intercurrent illness without bladder cancer, and 4 are dead of bladder cancer. Actuarial survival and determinental survival (censoring intercurrent, non-cancer deaths) for the 18 patients is 52% (+ 17%) and 61%

Table 2. Tumor characteristics and pre-treatment TURBT in relation to histologic clearance of tumor (PO) at 4,000 cGY (13/l 6 pts attained PO) Fraction PO Tumor morphology Papillary Solid Unspecified Pre-treatment TURBT TURBT Performed No TURBT Tumor multifocality Multifocal Solitary Tumor Stage Tl T2 T3A T3B T4A

616 214

516 9112 414 719 617 112 212 313 516 213

696

1. J. Radiation Oncology 0 Biology 0 Physics

(a 19%) at 30 months, and freedom from relapse (FFR) and FFR of invasive cancer is 32% (2 14%) and 6 1% (213%). In the subgroup of the 13 patients who were complete responders following two cycles of treatment, 8/l 3 are NED (62%) l/l 3 has recurrent, non-invasive bladder cancer, l/ 13 died of intercurrent illness without bladder cancer, and 3/ 13 died of bladder cancer ( 1 patient died 6 weeks after treatment with widespread liver metastases). Results-non-surgical candidates Of the 16 patients who were not surgical candidates, 6/ 16 remained relapse-free, l/ 16 is NED after a salvage cystectomy for locally recurrent disease (7/ 16 patients or 44% overall NED), 4/ 16 are alive with bladder cancer, 2/ 16 have died of intercurrent illness, and 3/ 16 are dead of bladder cancer. The actuarial survival and determinental survival for these 16 patients is 55% (+15%) and 66% (+ 16%) at 45 months, and the FFR and FFR of invasive cancer are 4 1% (k 14%) and 49% (t 14%). Patterns offailure The sites of first tumor recurrence have been scored for the 18 “surgically eligible” patients (S), the 13 patients in the subgroup of these 18 who were without histologic evidence of tumor in the repeat biopsies (PO), and the 16 non-surgical (NS) candidates. These data are summarized in Table 3. To date, the overall distant metastatic rate is only 18% (6/34), but as the median follow-up is short, it is premature to conclude that this favorable trend will continue. Bladder preservation Of the 34 patients, 24 (7 1%) have retained their bladders. Ten patients underwent cystectomy, of which 6/ 10 were planned and 4/ 10 were salvage cystectomies for local tumor recurrence (l/ 11 surgical candidates with a negative biopsy after the first two cycles of treatment underwent a planned cystectomy out of concern that the large initial tumor had been inadequately sampled at repeat biopsy. Complete tumor clearance was confirmed by the cystectomy specimen). The information regarding bladder preservation is summarized in Table 4. Of the 24 patients with retained bladders, 20124 (83%) have normal voiding and 4124 have symptoms of frequency or urgency which developed after the conclusion of the treatment.

Table 3. Sites of first tumor LF PO (13) Surg. cand. (18) Non-surg. cand. ( 16) PO = Patients

with completely

recurrence LF/DM

2 4 5

1 1 0 downstaged

tumors.

DM 1 2 3

September 1990, Volume 19, Number 3

Table 4. Bladder status (34 pts) Stage

Fraction

retaining

Tl T2 T3 T4

tumor-free

bladders

J/3 516 5116 719 Total = 18/34

Note: 1O/34 cystectomy (6 primary-4 salvage); 24/34 retained bladder (7 1%): 1S/24 NED, 3/24 non-invasive tumor recurrence, 3/24 invasive tumor recurrence; 20/24 retained bladders (7 1%).

Toxicity of therapy Of the 40 patients originally enrolled on study, six were unable to complete the treatment. Three of these patients developed 5-FU associated coronary vasospasm and angina while receiving the first cycle of 5-FU and were subsequently treated with cis-platinum off-study. An additional two patients went off-study at their insistence after the first cycle of treatment, one due to vulvovaginal mucositis and the other because of nausea. One patient experienced a fatal complication of treatment while completing the protocol. This patient developed severe diarrhea during the first cycle of treatment, ultimately culminating in sepsis and a fatal outcome. For the 34 patients completing the entire treatment, 28134 received 100% doses of 5-FU, and 34134 received 80% or greater of the planned chemotherapy. Specific on-treatment side-effects of treatment for these 34 patients included one patient with excess bleeding following biopsy at 4,000 cGy, which resulted in a 2-week delay in resuming the last cycle of treatment. Two patients developed dysuria, and most patients developed diarrhea, limited to the weeks under treatment, and controlled with oral anti-diarrhea1 medications. Late complications of treatment occurred in one planned cystectomy patient, who developed small bowel obstruction requiring surgical intervention at the edge of the radiation fields, 6 months after his initial surgery.

DISCUSSION Tumor radioresponsiveness is a well established favorable prognostic factor in bladder cancer. Complete tumor clearance with radiation portends a favorable outcome of treatment, whether the treatment is combined radiation and cystectomy or primary radiation alone (2-5, 8, 14, 19, 24). This trial has addressed prospectively whether the treatment of the individual bladder cancer patient can be directed by using tumor radioresponsiveness as the criterion with which to decide the need for cystectomy. This hypothesis is not original to this study, but has been proposed in the radiotherapy literature of bladder cancer for some time (3). Although the patient follow-up in this study

5-FU/XRT for bladder cancer 0 K. J. RUSSELLetal.

is still too short to allow one to draw firm conclusions, there is the suggestion that this treatment approach is a valid one. The survival and recurrence data for the 13 patients with rapid histologic clearance of their cancer is encourgaging, however further follow-up will be required to address both the long-term incidence of locally recurrent tumors as well as the metachronous, non-invasive cancers which have developed despite eradication of the original invasive primary. Additionally, this pilot study has investigated whether the overall percentage of patients with radioresponsive tumors can be increased by the use of a radiation potentiator such as 5-FU, and whether those increases yield parallel improvements in prognosis in addition to conservation of bladder function. The 69-8 1% of downstaging among the 18 patients evaluated after initial 2 cycles of treatment is higher than the expected frequency of 1443% of patients found historically to be completely downstaged by pre-operative radiation doses of 4-5,000 cGy (2, 4, 5, 14, 19, 24, 25). To the extent that a deep-muscle biopsy detects residual cancer with the same certainty as pathological examination of the entire cystectomy specimen, these preliminary results may be expected to remain favorable. The requirement in this study of histologic confirmation of tumor clearance within 3 weeks of completing relatively low-dose radiation would seem to be a more conservative measure of radioresponsiveness than a cystoscopic assessment, without biopsy, at 6 months following completion of high dose radiotherapy. This latter approach has been the accepted endpoint used in prior investigations (3, 4). Nonetheless, recent studies correlating biopsy and cystectomy data to assess combination chemotherapy-induced complete bladder tumor responses suggest that biopsies may underestimate the presence of residual tumor relative to information obtained by examination of the full cystectomy specimen by as much as 34% (20). To date, one can only conclude from this current series of patients that the incidence of local tumor recurrence in those patients achieving biopsy-proven complete responses has been low. Much debate has been engendered among investigators as to which pre-treatment clinicopathological features of invasive bladder tumors may predict for radioresponsiveness. Tumor stage, size, grade, focality, morphology, resectability by TURBT, and obstruction of the distal ureters, as well as patient age and hemoglobin, have all been proposed as pre-treatment parameters useful in selecting patients suitable for primary radiotherapy (1, 4, 15, 22). Given the small subgroups in the 16 biopsied patients, it is fair only to conclude that there are no striking trends in radioresponsiveness predictable by these patients’ tumor stage, morphology, multifocality, or resectability. For the whole group of 34 patients, 26 patients had some degree of tumor resection prior to treatment and 8 patients had none. Of the 26 patients who underwent TURBT, 14 remain NED (54%) compared with 3/S (38%) without a TURBT. Given the small numbers of patients

697

and the range of tumor stages treated, it is hard to conclude that pre-treatment TURBT contributed to a successful treatment outcome independent of other prognostic variables. The current investigation is one of a number of ongoing clinical trials involving bladder preservation which address the same two general issues: (a) improving absolute tumor response rates to radiation by the addition of concurrent chemotherapy, and (b) developing a flexible treatment approach based upon individual tumor responsiveness. At the present time, cis-platinum. alone or in combination with other cytotoxic drugs, has been the drug most frequently used with radiation, owing to its high singleagent activity in metastatic transitional cell cancers. Complete tumor response rates of 77% have been reported by Shipley et al. when platinum has been given concurrently with radiation as the primary treatment of 70 patients with invasive bladder cancer (2 1). Of the responders, 73% maintained this complete response, resulting in a 4year survival of 57%. Similar results have been reported by Sauer et al., who treated 4 1 patients with continuous infusion cis-platinum and radiation with an 85% (35/4 1 pts.) histologically confirmed complete response rate, 83% retention of normal bladder function, and only a 17% (6/35 pts.) salvage cystectomy rate for local tumor recurrence. Two-year actuarial survivals of 7 l/6 l/50% were reported for patients with stage T 1, T2-T3, and T4 tumors, respectively ( 18). Jakse et a/. have also combined concurrent cis-platinum and irradiation in a prospective study carried out with 22 patients. With a mean follow-up of 14 months, the incidence of tumor-free bladders was 77% (17/22), with three patients experiencing sequelae of bladder contracture, and two patients recurring with non-invasive metachronous tumors elsewhere in the bladder (11). The Radiation Therapy Oncology Group (RTOG) has carried out a trial, (RTOG #85-12), using virtually the same response-based treatment algorithm as used in this trial but employing cis-platinum rather than 5-FU. Results from this trial remain forthcoming (16). The results of a trial employing neoadjuvant CMV (cisplatinum, vinblastine, and methotrexate) followed by concurrent platinum and radiation has been reported by Marks et ~11.This treatment yielded an incidence of complete tumor downstaging ranging from 2 l-50%, depending on the subgroup of patients analyzed, and is presently active as an RTOG study (13). Combined 5-fluorouracil and radiation for the treatment of bladder cancer in patients unsuitable for surgery has been previously reported by Rotman et al. ( 17). Nineteen patients with locally advanced tumors were treated with irradiation and comcomitant 5-FU, with or without bolus mitomycin. Of the 19 patients, 6 1% (11 pts.) obtained a complete local response within 3-6 months of treatment and an additional 28% (5 patients) showed tumor regression to be subsequently controlled by TURBT and intravesical chemotherapy. Life-table analysis yielded

698

I.

J. Radiation Oncology 0 Biology0 Physics

an adjusted survival of 62.5% at 5 years, with only 2 patients developing treatment-related cystitis. With respect to the toxicity experienced in the current trial, the combined 5-FU and radiation regimen has been well-tolerated by most patients, with 80% of patients receiving full doses of 5-FU. Three patients developed a significant side-effect of 5-FU infusion, coronary artery vasospasm, (clinically manifested as angina), which has been previously well described as a complication of continuous infusion 5-FU (6). Patients who retained their bladders have experienced little in the way of treatmentrelated bladder injury, with 83% reporting completely normal urinary function. As platinum containing combination chemotherapy has been reported to exhibit significant morbidity, with an incidence of neutropenia-associated sepsis of 20%, renal toxicity of 3 l%, and a treatment-related fatality rate of 4% (23) 5-FU combined with irradiation is an attractive alternative for those medically frail bladder cancer patients for whom the systemically more active drug combinations combined with irradiation might prove too toxic.

September 1990, Volume 19, Number 3

CONCLUSIONS The combination of 5-FU with radiotherapy in the treatment of bladder cancer appears to yield a higher percentage of complete tumor responses than historical experiences using radiation alone, and with little additional morbidity from the addition of the chemotherapy. This combined modality approach has permitted bladder preservation in the majority of the patients treated, may prove to be an improvement over radiation alone for patients ineligible for cystectomy, and is an attractive alternative regimen for patients unsuitable for more aggressive chemoradiotherapy approaches. A treatment approach to the management of invasive bladder cancer based upon individual tumor responsiveness deserves continued clinical investigation. It permits bladder preservation in selected patients with a high probability of success, and it defines a group of patients with a favorable prognosis for whom the toxicity of additional adjuvant combination chemotherapy may not be warranted.

REFERENCES 1. Batata, M. A.; Chu, F. C. H.; Hilaris, B. S.; Kim, Y. S.; Lee, M. Z.; Chung, S.; Whitmore; W. F. Factors of prognostic

and therapeutic significancein patients with bladder cancer. Int. J. Radiat. Oncol. Biol. Phys. 7575; 1981. 2. Batata, M. A.; Whitmore, W. F., Jr.; Chu, F. C.; Hilaris, B. S.; Unal, A.; Chung, S. Patterns of recurrence in bladder cancer treated by irradiation and/or cystectomy. Int. J. Radiat. Oncol. Biol. Phys. 6:155-159; 1980. 3. Blandy, J. P.; England, H. R.; Evans, S. J. W.; Hope-Stone, H. F.; Mair, G. M. M.; Mantell, B. S.; Oliver, R. T. D.; Paris, A. M. I.; Ridson, R. A. T3 bladder cancer-the case for salvage cystectomy. Brit. J. Urol. 52:506-510; 1980. 4. Bloom, H. J. G.; Hendry, W. F.; Wallace, D. M.; Skeet, R. G. Treatment of T3 bladder cancer: controlled trial of preoperative radiotherapy and radical cystectomy versus radical radiotherapy: second report and review. Br. J. Urol. 54:136-151; 1982. 5. Boileau, M. A.; Johnson, D. E.; Chan, R. C.; Gonzalez, M. 0. Bladder carcinoma. Results with pre-operative radiation therapy and radical cystectomy. Urology 16:569576; 1980. 6. Collins, C.; Weiden, P. L. Cardiotoxicity of 5-fluorouracil. Cancer Treat. Rep. 7 1:733-738. 7. Duncan, W.; Quilty, P. M. The results of a series of 963 patients with transitional cell carcinoma of the urinary bladder primarily treated by radical megavoltage x-ray therapy. Radiother. Oncol. 7:299-310; 1986. 8. Goffinet, D. R.; Schneider, M. J.; Glatstein, E. J.; Ludwig, H.; Ray, G. R.; Dunnick, N. R.; Bagshaw, M. A. Bladder cancer: results of radiation therapy in 384 patients. Radiology 117:149; 1975. 9. Goodman, G. B.; Hislop, T. G.; Elwood, J. M.; Balfour, J. Conservation of bladder function in patients with invasive bladder cancer treated by definitive irradiation and selective cystectomy. Int. J. Radiat. Oncol. Biol. Phys. 7:569-573; 1981. 10. Hope-Stone, H. F.; Oliver, R. T. D.; England, H. R.; Blandy,

11.

12.

13.

14. 15.

16.

17.

18.

19.

J. P. T3 bladder cancer: salvage rather than elective cystectomy after radiotherapy. Urology 24:3 15-3 19; 1984. Jakse, G.; Frommhold, H.; zur Nedden, D. Combined radiation and chemotherapy for locally advanced transitional cell carcinoma of the urinary bladder. Cancer 55: 1659- 1664; 1985. Kaplan, E. S.; Meier, P. Non parametric estimation from incomplete observation. Am. Stat. Assoc. J. 53:457-480; 1958. Marks, L. B.; Kaufman, S. D.; Prout, G. R., Jr.; Heney, N. M.; Griffin, P. P.; Shipley, W. U. Invasive bladder carcinoma: preliminary report of selective bladder conservation by transurethral surgery, upfront MCV (methotrexate, cisplatin, and vinblastine)’ chemotherapy and pelvic irradiation plus cisplatin. In. J. Radiat. Oncol. Biol. Phys. 15:877-883; 1988. Miller, L. S. TJ bladder cancer. The case for higher radiation dosage. Cancer 45:1875-1878; 1980. Quilty, P. M.; Kerr, G. R.; Duncan, W. Prognostic indices for bladder cancer: an analysis of patients with transitional cell carcinoma of the bladder primarily treated by radical megavoltage x-ray therapy. Radiother. Oncol. 7:3 1 l-32 1; 1986. Radiation Therapy Oncology Group. No. 85-12: phase II study of combined modality program with organ preservation in carcinoma of the bladder. Rotman, M.; Macchia, R.; Silverstein, M.; Aziz, H.; Choi, K.; Rosenthal, J.; Braverman, A.; Laungani, G. B. Treatment of advanced bladder carcinoma with irradiation and concomitant 5-fluorouracil infusion. Cancer 59:7 1O-7 14; 1988. Sauer, R.; Schrott, K. M.; Dunst, J.; Thiel, H.-J.; Hermanek, P.; Bomhof, C. Preliminary results of treatment of invasive bladder carcinoma with radiotherapy and cisplatin. Int. J. Radiat. Oncol. Biol. Phys. 15:871-875; 1988. Scanlon, P. W.; Scott, M.; Se.gura, J. W. A comparison of

5-F’U/XRT for bladder cancer 0 K. J. RUSSELLetal. short-course, low-dose and long-course, high-dose preoperative radiation for carcinoma of the bladder. Cancer 52: 1153-l 159; 1983. 20. Scher, H. I.; Yogoda, A.; Herr, H. W.; Sternberg, C. N.; Bosl, G.; Morse, M. J.; Sogani, P. C.; Watson, R. C.; Dershaw, D. D.; Reuter, V.; Geller, N.; Hollander, P. S.; Vaugh, E. D., Jr.; Whitmore, W. F.; Fair, W. R. Neoadjuvant MVAC (methotrexate, vinblastine, doxorubicin and cisplatin) effect on the primary bladder lesion. J. Ural. 139:470-474; 1988. 21. Shipley, W. U.; Prout, G. R., Jr.; Einstein, A. B.; Coombs, L. J.; Wajsman, Z.; Soloway, M. S.; Englander, L.; Barton, B. A.; Hafermann, M. D. Treatment of invasive bladder cancer by cisplatin and radiation in patients unsuited for surgery. J.A.M.A. 258:931-935; 1987. 22. Shipley, W. U.; Rose, M. A.; Perrone, T. L.; Mannix, C. M.; Heney, N. H.; Prout, G. R. Full dose irradiation for patients with invasive bladder carcinoma: clinical and his-

699

tologic factors prognostic of improved survival. J. Urol. 134: 679-683; 1985. 23. Stemberg, C. N.; Yagoda, A.; Scher, H. I.; Watson, R. C.; Herr, H. W.; Morse, M. J.; Sogani, P. C.; Vaughan, E. D., Jr.; Bander, N.; Weiselberg, L. R.; Geller, N.; Hollander, P. S.; Lipperman, R.; Fair, W. R.; Whitmore, W. F., Jr. MVAC (methotrexate, vinblastine, doxorubicin and cisplatin) for advanced transitional cell carcinoma of the urothelium. J. Urol. 139:461; 1988. 24. Whitmore, W. F.; Batata, M. A.; Hilaris, B. S.; Reddy, G. N.; Vnal, A.; Ghoneim, M. A.; Grabstald, H.; Chu, F. A comparative study of two preoperative radiation regimens with cystectomy for bladder cancer. Cancer 40:1077-1086; 1977. 25. Yu, W. S.; Sagerman, R. H.; Chung, C. T.; Dalal, P. S.; King, G. A. Bladder carcinoma. Experience with radical and preoperative radiotherapy in 42 1 patients. Cancer 56: 1293-1299; 1985.