Int. J. Radiation Oncology Biol. Phys., Vol. 59, No. 1, pp.

197–207, 2004
Copyright © 2004 Elsevier Inc.
Printed in the USA. All rights reserved
0360-3016/04/$–see front matter

doi:10.1016/j.ijrobp.2003.10.018

CLINICAL INVESTIGATION Bladder

RADIOTHERAPY FOR MUSCLE-INVASIVE CARCINOMA OF THE BLADDER:

RESULTS OF A RANDOMIZED TRIAL COMPARING CONVENTIONAL
WHOLE BLADDER WITH DOSE-ESCALATED PARTIAL BLADDER
RADIOTHERAPY
RICHARD A. COWAN, M.D., M.R.C.P., F.R.C.R.,* CATHERINE A. MCBAIN, M.R.C.P., F.R.C.R.,*
W. DAVID J. RYDER, B.SC., GRAD.STAT.,† JAMES P. WYLIE, M.R.C.P., F.R.C.R.,*
JOHN P. LOGUE, M.R.C.P., F.R.C.R.,* SANDRA L. TURNER, M.B.B.S.,‡
JOHANNES VAN DER VOET, F.R.C.R.,§ CONOR D. COLLINS, F.R.C.R.,㛳
VINCENT S. KHOO, F.R.A.C.R.,* AND GRAHAM R. READ, F.R.C.P., F.R.C.R.¶
*Departments of Clinical Oncology and †Medical Statistics, Christie Hospital NHS Trust, Manchester, United Kingdom;
‡
Department of Radiation Oncology, Westmead Hospital, Sydney, Australia; §Department of Radiotherapy and Oncology, South
Cleveland Hospital, Middlesborough, United Kingdom; 㛳Department of Diagnostic Radiology, St. Vincent’s University Hospital,
Dublin, Ireland; ¶Department of Clinical Oncology, Royal Preston Hospital, Preston, United Kingdom

Purpose: To investigate whether delivering an increased radiation dose to the tumor-bearing region of the
bladder alone would improve local disease control without increasing treatment toxicity.
Methods and Materials: A total of 149 patients with unifocal T2-T3N0M0 bladder carcinoma were randomized
between whole bladder conformal radiotherapy (WBRT, 52.5 Gy in 20 fractions, n ⴝ 60) and partial bladder
conformal RT (PBRT) to tumor alone with 1.5-cm margins within either 4 weeks (PBRT4, 57.5 Gy in 20 fractions,
n ⴝ 44) or 3 weeks (PBRT3, 55 Gy in 16 fractions, n ⴝ 45). The response was assessed cystoscopically after 4
months.
Results: The 5-year overall and CFS rate was 58% and 47%, respectively, for the whole population. The CR rate
was 75% for WBRT, 80% for PBRT4, and 71% for PBRT3 (p ⴝ 0.6), with a 5-year local control rate of 58%,
59%, and 34%, respectively (p ⴝ 0.18). Solitary new tumors arose within the bladder, outside the irradiated
volume, in 6 (7%) of 89 patients who underwent PBRT. The 5-year overall survival and cystectomy-free survival
rate was 61% and 49% for WBRT, 60% and 50% for PBRT4, and 51% and 41% for PBRT3 (p ⴝ 0.81 and p
ⴝ 0.59). The treatment toxicity was mild and equivalent across the three trial arms.
Conclusion: The reduction in treatment volume allowed delivery of an increased radiation dose without a
reduction in local tumor control or the development of excess toxicity. However, this dose-escalated partial
bladder approach did not result in significantly improved overall survival. © 2004 Elsevier Inc.

Bladder cancer, Radiotherapy, Partial bladder, Dose.

INTRODUCTION
Bladder cancer is the fifth most commonly occurring ma-
lignancy in both Europe and North America, with around
12,500 new cases diagnosed in the United Kingdom every
year (1). Around 20% of patients present with muscle-
invasive disease, for whom the options for treatment lie
between radical cystectomy and organ conservation pro-
grams that incorporate radiotherapy (RT). The results of
series of patients treated with RT alone have been disap-
pointing, with 5-year survival rates of 30 –50% (2, 3).
There is compelling evidence to support the presence of
a radiation dose–response curve in transitional cell bladder
cancer (4, 5). However, the dose that is possible to deliver
to the bladder is limited by late radiation effects, including
telangiectasia, fibrosis, and bladder shrinkage. The severity
and incidence of late radiation toxicity is closely related to
the proportion of the organ irradiated (6), with the radiation
tolerance of part of the bladder greater than that of the organ
as a whole (7).
Up to 95% of tumors recur at their original site within the
bladder (8) and failure to control the primary tumor is
present in 90% of patients who die of bladder cancer (2). It
was, therefore, hypothesized that delivering an increased
dose of RT to the tumor-bearing region of the bladder alone

Reprint requests to: Richard A. Cowan, M.D., M.R.C.P.,
F.R.C.R., Department of Clinical Oncology, Christie Hospital
NHS Trust, Wilmslow Road, Withington M20 4BX Manchester,
United Kingdom. Tel: (⫹44) 161-446-3332; Fax: (⫹44) 161-446-
3265; E-mail: richard.cowan@christie-tr.nwest.nhs.uk
Presented as an oral presentation (Abstract 101) at the 44th
Annual Meeting of the American Society for Therapeutic Radiol-
ogy and Oncology, New Orleans, LA, 2002.
Received Apr 1, 2003, and in revised form Oct 1, 2003. Ac-
cepted for publication Oct 15, 2003.

197
198 I. J. Radiation Oncology ● Biology ● Physics
Table 1. Prescribed radiotherapy doses
Planning target
Trial arm volume (cm3)
PBRT3 (16 fractions) ⬍350
350–500
⬎500
PBRT4 (20 fractions) ⬍350
350–500
⬎500
Abbreviation: PBRT ⫽ partial bladder radiotherapy.
may result in improved local control without increasing
toxicity and that improved local control may translate into a
survival benefit.
To test this hypothesis, a prospective randomized con-
trolled trial was instituted to compare local disease control,
overall survival (OS), and toxicity in patients receiving
conventional RT to the whole bladder with those receiving
dose-escalated RT to the tumor-bearing region of the blad-
der alone.
METHODS AND MATERIALS
Patients
Between 1993 and 1999, 149 patients referred to the
Christie Hospital with primary muscle-invasive carcinoma
of the bladder were recruited. For trial eligibility, their
tumor had to be a unifocal, muscle-invasive transitional cell
carcinoma (TCC), Stage T2-T3N0M0, and measuring ⱕ7
cm. A central pathology review was undertaken on all
specimens. However, because it was not possible to com-
plete this before randomization, patients were entered into
the study on the basis of the histologic diagnosis made at
their referring hospital. Patients were required to have a
documented bladder capacity of ⱖ200 mL, to have pro-
vided written informed consent to trial entry, and to be
medically fit to undergo radical treatment and follow-up.
Patients with prior malignancy (other than skin cancer or
superficial bladder cancer), nodal or distant metastases, or
multifocal tumors within the bladder were excluded. Trans-
urethral resection of the bladder tumor was performed as
completely as the referring surgeon believed it was possible
at initial cystoscopy and biopsy. Patients were not referred
for repeated transurethral resection. Before trial entry, pa-
tients underwent baseline staging investigations comprising
full blood count, biochemical profile, liver function tests,
chest X-ray, and MRI of the pelvis and lower abdomen. In
cases in which the cystoscopic and radiologic disease stage
did not concur, the stage was recorded as the greater of the
two. The pretreatment bladder and bowel symptoms were
recorded using the World Health Organization (WHO) tox-
icity scoring system (9).
Radiotherapy
Patients were prospectively randomized to one of three
treatment arms: whole bladder RT (WBRT, control arm) or
Volume 59, Number 1, 2004
Prescribed Patients Trial arm
dose (Gy) (n) (%)
55 32 71
52.5 8 18
50 5 11
57.5 29 66
55 13 30
52.5 2 4
partial bladder RT (PBRT) using two different dose/frac-
tionation regimens (trial arms). The target volumes were
defined per the International Commission on Radiation
Units and Measurements, Report 50 (10), with the RT dose
prescribed to the treatment isocenter.
Patients randomized to the control arm (WBRT) received
RT to the whole bladder (clinical target volume) with a
1.5-cm expansion in all directions to form the planning
target volume (PTV). A total dose of 52.5 Gy in 20 fractions
of 2.63 Gy was prescribed.
In the two study arms, RT was delivered to the tumor-
bearing region of the bladder alone (clinical target volume),
identified using information from the cystoscopic assess-
ment, MRI and RT planning CT scan. This was expanded
by 1.5 cm in all dimensions to form the PTV. In the first trial
arm, RT dose was increased by 10% to 57.5 Gy in 20
fractions of 2.88 Gy (partial bladder RT within 4 weeks,
PBRT4). Assuming an ␣/␤ ratio of 10 for bladder cancer,
this is biologically equivalent to 61.7 Gy in 2 Gy/d fractions
(on the basis of the linear-quadratic model [11]). In the other
trial arm, a biologically equivalent dose was administered in
a 3-week accelerated hypofractionated regimen, resulting in
the prescription dose of 55 Gy in 16 fractions of 3.44 Gy
(PBRT3). This regimen was included because our institu-
tion had had previous experience of such an accelerated
hypofractionated treatment, and we wished to examine it
formally in a prospective randomized trail. Owing to the
increased risk of toxicity from delivering high doses to large
bladder volumes, the dose prescribed in the two trial arms
was reduced in increments if the PTV exceeded 350 cm3
and 500 cm3 (Table 1).
All RT was CT planned and conformally delivered using
8 –20 MV photons in a four-field box arrangement with
beam shaping (control arm) or multileaf collimation (trial
arms). Treatment was given in a single phase using a
once-daily fractionation regimen, treating 5 d/wk. Patients
were asked to void immediately before treatment; formal
immobilization devices were not used. Verification was
performed using weekly megavoltage portal imaging. The
dose–volume histograms were not plotted routinely.
Follow-up and assessment of response and toxicity
Response was assessed cystoscopically 4 months after
treatment completion, with subsequent cystoscopic fol-
low-up at 6-month intervals. A complete response (CR) was
 Whole vs. partial bladder RT
defined as the absence of any visualized abnormality. Where
suspicious areas were visualized and biopsied, the absence
of histologic evidence of malignancy was also accepted as a
CR. Urologists were asked to record the site of any tumor
visualized at each cystoscopy and document whether this
was at the original or a new tumor site. This information
was cross-checked with RT plans to assess whether local
recurrences arose inside or outside the irradiated bladder
volume. The response was also assessed radiologically with
MRI 3 and 12 months after treatment completion. Patients
were evaluated clinically at 6 weeks, 3 months, and then
every 6 months. Toxicity data were recorded using the
WHO toxicity scoring system on the final treatment day and
at every follow-up visit.
Statistical analysis
The sample size was calculated for the local control
outcome. The calculation was for a WBRT vs. PBRT com-
parison. For 80% power in a two-tailed log–rank test at the
5% level of significance, it was anticipated that 123 failures
would be required if the hazard ratio for PBRT to WBRT
was really 0.60 (determined by an improvement in the
5-year local control rate from 55% to 70%). It was calcu-
lated that an accrual of 125 patients annually for 4 years
with a minimal follow-up of 18 months would be expected
to yield the required number of events. No interim analyses
were planned.
Accrual to the trial was far slower than had originally
been anticipated with only 149 patients accrued within 6
years. An unplanned interim analysis was performed, and
because this showed no evidence to suggest a developing
benefit in the trial arms, with a trend shown for outcomes in
the PBRT3 arm to be inferior, the investigators elected to
close the trial to recruitment at the end of 1999. The anal-
yses presented here used the additional follow-up data that
accrued after closure to recruitment, and 72 patients had
failure of local control (compared with the target of 123).
No formal adjustment was made in this report for the early
closure of the trial, but we remind the reader to bear this fact
in mind.
Patients were randomized in a 4:3:3 ratio in favor of the
standard WBRT arm. No stratification for prognostic factors
was used. The randomization list was generated by the
Medical Statistics Department at the Christie Hospital using
a permuted block scheme with randomly selected block
sizes of 10, 20, or 30. The list was held in a locked cabinet
in the trials office at the hospital and was kept concealed
from the clinical investigators. Clinical investigators tele-
phoned the trials office on a dedicated “randomization line”
to enter patients and a member of the trials office staff
recorded the patient details and returned the patient’s trial
number and allocated treatment. No blinding was used.
The time-to-event curves were estimated by the Kaplan-
Meier method, and comparisons between trial arms were
made with the log–rank test. All periods were measured
from the date of randomization.
Local relapse was defined as the recurrence of disease
● R. A. COWAN et al. 199
within the bladder. Patients with tumor detected in the
bladder at the first follow-up cystoscopy (n ⫽ 27) or who
were unfit for cystoscopy (n ⫽ 10, see “Results”) were
assumed to have never achieved local control and to have
had failure at time 0. The development of pelvic or abdom-
inal nodal metastases was classified as metastatic disease.
OS was defined as the time to death from any cause, and
cystectomy-free survival (CFS) was defined as the time to
the first cystectomy or death from any cause.
Group proportions (e.g., response rates) were compared
using chi-square tests or Fisher’s exact test if the expected
numbers were too small. All quoted tests were two-tailed.
Statistical analysis was performed using Statistical Package
for Social Sciences software (SPSS), version 10, and S-Plus
2000 for Windows.
RESULTS
Patient and treatment characteristics
At the time of analysis, the median follow-up of the 77
patients who remained alive was 5.8 years. The patient
characteristics, disease stage, and tumor size and grade were
comparable across the three trial arms (Table 2). After
central pathology review, the initial histologic diagnosis of
TCC was changed to squamous cell carcinoma, mixed squa-
mous cell carcinoma/TCC or other histologic subtype in 12
patients. These patients were overrepresented in the PBRT3
arm, but subgroup analysis excluding patients with a histo-
logic subtype other than TCC showed that this did not affect
the trial outcome. Six protocol violations occurred: 2 pa-
tients were treated in trial arms different from their random-
ized allocation (one after chemotherapy for nodal metasta-
ses). One patient did not receive RT. Two patients were
included with Stage T1 disease and one with T4 disease.
None were excluded, and all data was analyzed on an
intention-to-treat basis (Fig. 1).
Treatment of the tumor-bearing region of the bladder
alone resulted in a 61% reduction in the median irradiated
volume in the two PBRT trial arms compared with the
standard WBRT arm. The median RT volumes in the three
trial arms were whole bladder, 815 cm3 (range, 490 –1362
cm3); PBRT4, 324 cm3 (range, 201–567 cm3); and PBRT3,
315 cm3 (range, 203– 626 cm3). Most patients in the two
trial arms (66% in PBRT4 and 71% in PBRT3) received the
full, intended RT dose; the prescribed dose was reduced in
the remainder in accordance with the trial protocol. The RT
doses received are detailed in Table 1.
Initial tumor response
The initial tumor response to RT was assessed cysto-
scopically in 139 of 149 patients; in 112 (75%) of 149
patients, the response was complete. The CR rates by trial
arm were WBRT, 75% (45 of 60); PBRT4, 80% (35 of 44);
PBRT3, 71% (32 of 45;) p ⫽ 0.6, chi-square test). Nine
patients were unfit for the first-check cystoscopy owing to
the development of early metastatic disease (n ⫽ 5), early
death from bladder cancer (n ⫽ 2), early intercurrent death
200 I. J. Radiation Oncology ● Biology ● Physics Volume 59, Number 1, 2004

Table 2. Patient and tumor characteristics*

Characteristic WBRT PBRT4 PBRT3 Whole group

Gender
Male 44 (73) 36 (82) 28 (62) 108 (72)
Female 16 (27) 8 (18) 17 (38) 41 (28)
Age (y)
Median 67.4 67.2 67.8 67.4
Range 40–82 41–80 46–76 40–82
WHO PS (0 or 1) 53 (88) 39 (89) 38 (84) 130 (87)
Hydronephrosis (Present) 18 (30) 12 (27) 9 (20) 39 (26)
Histologic type
TCC 55 (92) 44 (100) 38 (85) 137 (92)
Mixed TCC/SCC 0 0 5 (11) 5 (3)
SCC 1 (1) 0 1 (2) 2 (2)
Other 4 (7) 0 1 (2) 5 (3)
Grade
1 1 (1) 0 0 1 (1)
2 8 (13) 5 (12) 6 (13) 19 (9)
3 51 (85) 39 (88) 39 (87) 129 (87)
Median size (cm)
Cystoscopic assessment 3.0 3.0 3.0 3.0
MRI assessment 4.0 4.0 3.5 3.9
T stage (n)
1 0 1 (2) 1 (2) 2 (1)
2 17 (28) 10 (23) 16 (36) 43 (29)
3 42 (70) 33 (75) 28 (62) 103 (69)
4 1 (2) 0 0 1 (1)

Abbreviations: WBRT ⫽ whole bladder radiotherapy; PBRT ⫽ partial bladder RT; WHO PS ⫽ World Health Organization performance
status; TCC ⫽ transitional cell carcinoma; SCC ⫽ squamous cell carcinoma.
Data presented as number of patients, with percentages in parentheses, unless otherwise noted.
* Because of rounding, percentages may not total 100%.

(n ⫽ 1), or intercurrent illness (n ⫽ 1). Cystoscopy was
deferred in 1 patient treated with initial chemotherapy. The
proportion of patients who did not undergo follow-up cys-
toscopy was similar in all three trial arms (WBRT, 5 [8%]
of 60; PBRT4, 4 [9%] of 44; PBRT3, 1 [2%] of 45; and
overall, 10 [7%] of 149).
Residual disease was found at the original tumor site in
25 of the 27 patients who did not achieve a CR. Two of
these patients also had new, coexisting tumors lying outside
the treated area. Two patients had complete regression of
the primary tumor but had new, superficial disease outside
the irradiated region.
Local recurrence
Local recurrence arose in 35 of 112 patients after an
initial CR: 10 patients from the WBRT arm, 10 from the
PBRT4 arm, and 15 from the PBRT3 arm. The actuarial
5-year local control rates for the three trial arms were
WBRT, 58%; PBRT4, 59%; and PBRT3, 34% (p ⫽ 0.18,
log-rank test), corresponding to an overall local control rate
of 50% at 5 years. Figure 2 shows the Kaplan-Meier curves
of local control.
In the PBRT4 arm, 6 of 10 tumors recurred within the
irradiated bladder volume. The remaining four relapses
were due to the development of new tumors outside the
treated area. In the PBRT3 arm, all 15 recurrences arose at
the original tumor site within the RT field. One patient had
a simultaneous recurrence outside the irradiated volume. A
summary of the site and histologic type of the recurrences in
the three trial arms is shown in Table 3. Solitary new tumors
arose outside the treated area in 6 (7%) of 89 patients who
underwent PBRT. This was muscle invasive in only 1 case.

Management of residual and recurrent disease

The histologic type and management of residual and
Fig. 1. Flow diagram summarizing patient randomization and recurrent disease is summarized in Table 4. Twenty-eight
analysis. patients (19% overall) underwent salvage cystectomy, 21 as
 Whole vs. partial bladder RT ● R. A. COWAN et al. 201

Fig. 2. Local control by trial arm. Local control was defined as absence of intravesical disease on cystoscopy. WB ⫽
WBRT (control arm); PBRT4 ⫽ PBRT, 57.5 Gy in 20 fractions; PBRT3 ⫽ PBRT, 55 Gy in 16 fractions.

first-line treatment of residual or recurrent disease and 7 for
superficial or muscle-invasive disease arising after failure of
superficial therapy. No patient required salvage cystectomy
for toxicity. The 5-year OS rate after salvage cystectomy
was 54%.
Metastatic failure
Metastatic failure was documented in 17%, 25%, and
22% of patients in the WBRT, PBRT4, and PBRT3 trial
arms, respectively; this difference was not statistically sig-
nificant. Thirteen patients (9%) developed relapse in the
pelvic lymph nodes; this was the only site of failure in 4
patients (2.7%).
Survival and bladder preservation rates
No statistically significant differences were found in OS,
cause-specific survival (CSS), or CFS rates between the
control arm and two PBRT trial arms (Table 5). The OS and
CFS rates by trial arm are shown in Fig. 3. The overall
5-year CFS rate was 47%; 85% of patients who were alive
at 5 years had retained their native bladder.
Figure 4 shows the OS and CFS curves for the whole

Table 3. Site of residual and recurrent disease

Residual disease at first Local recurrence after

cystoscopy (n) initial CR (n)
Trial Patients Total
arm (n) Site Invasive Superficial Invasive Superficial (% of trial arm)

WBRT 60 Within radiation field 3 7 6 4 20 (33)

PBRT4 44 Within radiation field 1 3 1 5 10 (23)
Outside radiation field 0 1 1 3 5 (11)
Inside and outside radiation field 0 0 0 0 0
PBRT3 45 Within radiation field 5 4 7 7 23 (51)
Outside radiation field 0 1 0 0 1 (2)
Inside and outside radiation field 0 2 0 1 3 (7)
Total 149 9 18 15 20 62 (42)

Abbreviations: CR ⫽ complete response; WBRT ⫽ whole bladder radiotherapy (control arm); PBRT4 ⫽ partial bladder RT, 57.5 Gy in
20 fractions; PBRT3 ⫽ partial bladder RT, 55Gy in 16 fractions.
202 I. J. Radiation Oncology ● Biology ● Physics
Table 4. Management of residual and recurrent disease
Patients Cystectomy
(n) (n)
Residual at first cystoscopy
Muscle invasive 9 5
Superficial 18 3 0
Recurrence after initial CR
Muscle invasive 15 10
Superficial 20 3 0
Total 62 21
Abbreviation: CR ⫽ complete response.
group. The median OS and CSS for all 149 patients was 6.9
years and 7.9 years, respectively, with 5-year OS and CSS
rate of 58% and 65%, respectively. The outcomes were
highly dependent on the T stage, with a 5-year OS rate of
70% for T2 and 51% for T3 (p ⫽ 0.033, log–rank test). The
CSS for T2 was 81% and for T3 was 57% (p ⫽ 0.0022,
log–rank test). The 5-year CFS rate of patients with T2
disease was 60% compared with 40% for those with T3
disease (p ⫽ 0.025, log–rank test).
Toxicity
All three RT regimens were well tolerated, with all
treated patients completing their prescribed regimen without
interruption. No statistically significant difference was
noted in the incidence or severity of the acute radiation
reactions observed in the three trial arms, although a trend
was observed for more severe acute reactions, particularly
urinary frequency, in the PBRT3 arm. Figure 5 illustrates
the acute GU and GI toxicity profiles observed. One patient
in the WBRT arm developed acute toxicity requiring sur-
gical intervention after perforation of the small bowel 10
days after treatment completion. The parts of the WHO
symptom scoring system we used are summarized for ref-
erence in Table 6.
Full documentation of late toxicity was available in 63 of
74 patients who remained free of disease 2 years after
treatment completion. This comprised 25 patients from the
WBRT control arm, 22 from the PBRT4 arm, and 16 from
the PBRT3 arm. The percentage of patients describing any
bladder or bowel symptom rated WHO Grade 2 or worse
was equivalent among the three trial arms. Urinary symp-
toms WHO Grade 2 or worse were described by 1 (4%), 4
(18%), and 1 patient (6%) evaluated in the WBRT, PBRT4,
and PBRT3 arms, respectively, with late bowel symptoms
of Grade 2 or worse present in 2 patients in each arm,
Table 5. Five-year survival rates
WBRT (%)
Overall survival 61
Cause-specific survival 69
Cystectomy-free survival 49
Abbreviations as in Table 3.
Volume 59, Number 1, 2004
Medical/surgical contraindications Superficial resection/intravesical
to cystectomy (n) chemotherapy (n)
4 0
15
5 0
17
9 32
representing 8%, 9%, and 12% of those evaluated, respec-
tively. These differences were not statistically significant.
Late toxicity requiring surgical intervention occurred in 4
patients. Two patients (both in the PBRT3 arm) underwent
urinary diversion for severe urinary symptoms with poor
bladder function, and one (PBRT4 arm) required diathermy
of telangiectasia. One patient in the whole bladder arm
required surgery for GI complications (dilation of a sigmoid
colon stricture).
DISCUSSION
The goal of all combined-modality, organ-conservation
programs for muscle-invasive bladder cancer is to achieve
high rates of local disease control, translating into high rates
of OS, coupled with preservation of a well-functioning
natural bladder. A growing number of studies, including one
randomized trial, have reported the use of chemotherapy
concurrently with RT (12–14) such that chemoradiation is
rapidly becoming the standard of care for organ-conserva-
tion programs. The published outcomes of patients treated
with RT alone have been disappointing, with survival rates
inferior to cystectomy (15). However, many of these series
were published up to 20 years ago, often without the benefit
of accurate radiologic staging or computer-based treatment
planning, and reflect a negative selection bias. Although the
addition of concurrent chemotherapy to RT may well confer
benefit over RT alone, issues remain over how best to
optimize RT delivery.
This is the first prospective randomized trial to compare
RT of the tumor-bearing region of the bladder alone directly
with RT of the whole organ using external beam, megavolt-
age RT. We demonstrated that this approach results in a
reduction of the PTV by ⬎60%. Concern has been raised
that any reduction in the PTV may lead to inadequate
PBRT4 (%) PBRT3 (%) p Whole group (%)
60 51 0.81 58
67 58 0.76 65
50 41 0.59 47
 Whole vs. partial bladder RT ● R. A. COWAN et al. 203

Fig. 3. (a) Overall survival by trial arm from date of randomization. (b) Cystectomy-free survival (percentage of patients
alive with intact bladder) by trial arm from date of randomization. WB ⫽ WBRT (control arm); PBRT4 ⫽ PBRT, 57.5
Gy in 20 fractions; PBRT3 ⫽ PBRT, 55 Gy in 16 fractions.
204 I. J. Radiation Oncology ● Biology ● Physics
Fig. 4. Overall survival and cystectomy-free survival curves for all randomized patients (n ⫽ 149).
coverage of the tumor, reducing intravesical control. Our
results have shown this not to be the case. A partial bladder
approach permitted delivery of an increased radiation dose
without increased toxicity, but this dose escalation did not
lead to improved local control or OS. These observations
raise a number of interesting issues relating to the optimal
PTV, site of intravesical failure, radiation dose, and tumor
localization.
First, were the radiation doses used in this study inade-
quate to result in improved local control? The decision to
increase the delivered dose by 10% was made on the basis
of data suggesting a steep dose–response curve for bladder
cancer (4, 5). In addition, previous studies had shown that
57.5 Gy in 20 fractions delivered to the whole bladder
resulted in unacceptable late bladder and rectal toxicity (4,
16); therefore, escalation of the radiation dose to ⬎57.5 Gy
in this study was thought to be unjustified. When the bio-
logically effective doses calculated from the linear-qua-
dratic model are considered, the dose escalation used
equated to 12%.
Although it is apparent that this dose escalation was
insufficient to achieve the trial’s end points of improved
local control and OS, a nonstatistically significant trend for
improved control of muscle-invasive primary tumor in the
PBRT4, 57.5-Gy arm did appear to exist. Complete disease
regression at the primary tumor site 4 months after RT was
seen in 36 of 44 patients, giving an initial CR rate at the
primary site of 82% compared with 75% in the control arm.
After the initial CR, disease recurred at the original site in
6 of 10 patients with intravesical relapse in the PBRT4 arm
Volume 59, Number 1, 2004
compared with 10 of 10 patients in the WBRT control arm.
Fewer of these recurrences were muscle invasive in the
higher dose arm: 1 of 6 in the PBRT4 arm compared with 6
of 10 in the control arm. Although the number of patients
treated in this study was clearly too small to draw firm
conclusions, there is a suggestion that the delivery of a
higher RT dose may improve control of muscle-invasive
disease at the primary site. However, this was not adequate
to translate into reduced development of distant metastases,
resulting in equivalence of OS.
New bladder tumors developed outside the irradiated
volume in 5 patients in the PBRT4 arm (1 detected at the
first posttreatment cystoscopy, 4 during later follow-up). In
4 of 5 cases, this disease was superficial. Therefore, 33% of
the residual or recurrent disease seen in this trial arm de-
veloped in the unirradiated portion of the bladder. This may
have negated any trend for improved local control at the
primary site, resulting in the equivalent rates of local control
observed in the PBRT4 and control arms.
The nonstatistically significant, but clearly apparent,
trend for inferior local control in the PBRT3 arm is more
difficult to account for. Unlike the PBRT4 arm, inferior
local control was due to a failure to control the tumor at the
primary site, rather than to disease arising at other sites
within the bladder. The normalized tumor doses of the two
trial arms were equivalent, as were the RT and planning
parameters such as the median treatment volume and the
proportion of the bladder mucosa irradiated. The use of a
3-week hypofractionated regimen cannot be recommended
and no longer features in our clinical or research practice.
 Whole vs. partial bladder RT ● R. A. COWAN et al. 205

Fig. 5. Acute WHO toxicity profiles by trial arm. (a) Genitourinary (GU) (b) Gastrointestinal (GI). At 6 months, only
1 patient (WBRT arm) reported G3 bladder symptoms and 2 (1 WBRT, 1 PBRT3) reported G3 bowel symptoms. WB
⫽ WBRT (control arm); PBRT4 ⫽ PBRT, 57.5 Gy in 20 fractions; PBRT3 ⫽ PBRT, 55 Gy in 16 fractions.

The incidence of disease arising at new sites within the
bladder reported in this series is in keeping with the inci-
dences described in brachytherapy series. Van der Werf-
Messing et al. (8) has reported that second tumors arose at
new sites within the bladder in 7% of patients with T2
tumors. Mazeron et al. (17) also reported a second tumor
rate of 7% in patients with T1–T3 disease treated with
partial cystectomy and interstitial iridium. In the present
trial, second tumors developed at new sites within the blad-
der in 6 (7%) of 89 patients undergoing PBRT. That all
recurrences in the WBRT arm arose at the primary site, with
no second tumors at distant sites within the bladder, sug-
gests that a dose of 52.5 Gy in 20 fractions to the whole
urothelium is adequate to suppress the development of new
malignancy.
Exploitation of the higher radiation tolerances of partial
bladder volumes (6, 7) relies on reducing the proportion of
the organ irradiated. Concern was always present that in
reducing the PTV in this study, we did not encompass the
entire tumor with the planned dose. The high rates of local
control at the original tumor site in the PBRT4 arm tend to
refute this. However, issues, including accurate tumor lo-
calization and improved daily treatment verification, need to
be addressed before programs delivering dose-escalated
206 I. J. Radiation Oncology ● Biology ● Physics Volume 59, Number 1, 2004

Table 6. WHO symptom scoring system (abbreviated)

System/Grade 0 1 2 3

Bladder
Frequency ⱕ1⫻/2 h 1⫻/1–2 h 2–3⫻/h ⱖ4⫻/h
Nocturia 0–1 2–3 4–6 ⱖ7
Hematuria None visible Occasional Continuous Clots with obstruction
Dysuria/pain None Mild Moderate Severe
Bowel
Frequency Daily or less Twice daily 3–4 ⫻ daily ⱖ5 ⫻ daily
Diarrhea None Mild—medication unnecessary Moderate—needs medication Not controlled
Pain None Mild/occasional Moderate Severe
Bleeding/discharge None Spots, only with motion Persistent Hemorrhage

Abbreviation: WHO ⫽ World Health Organization.

PBRT can be adequately evaluated. Turner et al. (18) dem-
onstrated that at least one bladder wall moved by ⬎1.5 cm
in 60% patients imaged weekly during RT, and Miralbell et
al. (19) showed that a change in bladder volume of just 100
cm3 resulted in significant displacement of the target vol-
ume. Emerging techniques of image-guided RT and on-line
correction offer potential solutions to these problems.
Shirato et al. (20) implanted fiducial markers into 5 patients
with bladder tumors to facilitate real-time tumor tracking.
Another alternative is the use of on-line CT imaging using
devices such as the cone-beam CT imager developed by
Jaffray et al. (21) or the University of Wisconsin tomo-
therapy machine (22). We are currently evaluating these
approaches.
CONCLUSION
We present the results of a randomized trial involving
149 MRI-staged bladder cancer patients treated with CT-
planned conformal RT. Taken as a whole, the CR rate of
75% and the 5-year OS rates of 70% for T2 and 51% for T3
tumors, with a 5-year CFS rate of 47%, were as good as
recently published series using chemoradiotherapy (13, 14).
The stage-for-stage OS and CSS was also comparable to
that achieved with radical cystectomy (23). Treatment of the
tumor-bearing region of the bladder alone permitted deliv-
ery of an increased radiation dose without increased toxic-
ity, but no statistically significant differences were found in
the rates of local control, OS, or CCS. However, our results
demonstrating comparable outcomes with PBRT compared
with WBRT suggest that this concept merits further explo-
ration. PBRT may offer benefit in combination with che-
motherapy as a means of reducing the radiation volume. It
may also be of value as a means of boosting the primary
tumor site after conventional treatment to the whole bladder.
New techniques of on-line localization and verification will
improve our ability to deliver such higher dose, small-
volume treatments accurately, with the potential to improve
local control and survival further in organ-conservation
programs for this disease.

REFERENCES
1. www.cancerresearchuk.org/about cancer/statistics.
2. Duncan W, Quilty PM. The results of a series of 963 patients
with transitional cell carcinoma of the urinary bladder treated
by radical megavoltage X-ray therapy. Radiother Oncol 1986;
7:299–310.
3. Gospodarowicz MK, Hawkins NV, Rawlings GA, et al. Rad-
ical radiotherapy for muscle invasive transitional cell carci-
noma of the bladder: Failure analysis. J Urol 1989;142:1448–
1454.
4. Morrison R. The results of treatment of cancer of the blad-
der—A clinical contribution to radiobiology. Clin Radiol
1975;26:67–75.
5. Maciejewski B, Majewski S. Dose fractionation and tumour
repopulation in radiotherapy for bladder cancer. Radiother
Oncol 1991;21:163–170.
6. Marks LB, Carroll PR, Dugan TC, et al. The response of the
urinary bladder, urethra and ureter to radiation and chemo-
therapy. Int J Radiat Oncol Biol Phys 1995;31:1257–1280.
7. Emami B, Lyman J, Brown A, et al. Tolerance of normal
tissue to therapeutic irradiation. Int J Radiat Oncol Biol Phys
1991;21:109–122.
8. Van der Werf-Messing B, Menon RS, Hop WCJ. Cancer of
the urinary bladder category T2, T3 (NxM0) treated by inter-
stitial radium implant: Second report. Int J Radiat Oncol Biol
Phys 1983;9:481–485.
9. World Health Organisation. WHO handbook for reporting
results of cancer treatment, vol 48. Geneva: WHO; 1979.
10. International Commission on Radiation Units and Measure-
ments. Report 50: Prescribing, recording and reporting photon
beam therapy. Bethesda: ICRU, 1993.
11. Steel GG. Basic clinical radiobiology. 3rd ed. London: Arnold
Publishers; 2002.
12. Coppin CM, Gospodarowicz MK, James K, et al. Improved
local control of invasive bladder cancer by concurrent cispla-
tin and preoperative or definitive radiation. J Clin Oncol
1996;14:2901–2907.
13. Shipley WU, Kaufman DS, Zehr E, et al. Selective bladder
preservation by combined modality protocol treatment: Long-
term outcomes of 190 patients with invasive bladder cancer.
Urology 2002;60:62–67.
14. Rodel C, Grabenhauer, Kuhn, R et al. Combined modality
treatment and selective organ preservation in invasive bladder
cancer: Long term results. J Clin Oncol 2002;20:3061–3071.
15. Shelly MD, Barber J, Wilt T, et al. Surgery versus radiother-
 Whole vs. partial bladder RT
apy for muscle invasive bladder cancer (Cochrane review). In:
The Cochrane Library, Issue 1, 2002. Oxford: Update Soft-
ware; 2002.
16. Quilty PM, Duncan W, Kerr GR. Results of a randomised
study to evaluate influence of dose on morbidity in radiother-
apy for bladder cancer. Clin Radiol 1985;36:615–618.
17. Mazeron J, Crook J, Chopin D, et al. Conservative treatment
of bladder carcinoma by partial cystectomy and interstitial
iridium 192. Int J Radiat Oncol Biol Phys 1988;15:1323–
1330.
18. Turner SL, Swindell R, Bowl N, et al. Bladder movement
during radiation therapy for bladder cancer: Implications for
treatment planning. Int J Radiat Oncol Biol Phys 1997;39:
355–360.
19. Miralbell R, Nouet P, Rouzaud M, et al. Radiotherapy of
bladder cancer: Relevance of bladder volume changes in plan-
● R. A. COWAN et al. 207
ning boost treatment. Int J Radiat Oncol Biol Phys 1998;41:
741–746.
20. Shirato H, Shimizu S, Kitamura K, et al. Four-dimensional
treatment planning and fluoroscopic real-time tumour tracking
radiotherapy for moving tumor. Int J Radiat Oncol Biol Phys
2000;48:435–442.
21. Jaffray DA, Drake DG, Moreau M, et al. A radiographic and
tomographic imaging system integrated into a medical linear
accelerator for localization of bone and soft-tissue targets. Int
J Radiat Oncol Biol Phys 1999;45:773–789.
22. Welsh JS, Patel RR, Ritter MA, et al. Helical tomotherapy: An
innovative technology and approach to radiation therapy.
Technol Cancer Res Treat 2002;1:311–316.
23. 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 2001;19:666 – 675.