Int. J. Radiation Oncology Biol. Phys., Vol. 66, No. 4, pp.

1072–1083, 2006
Copyright © 2006 Elsevier Inc.
Printed in the USA. All rights reserved
0360-3016/06/$–see front matter

doi:10.1016/j.ijrobp.2006.06.005

CLINICAL INVESTIGATION Prostate

HYPOFRACTIONATED VERSUS CONVENTIONALLY FRACTIONATED

RADIATION THERAPY FOR PROSTATE CARCINOMA: UPDATED RESULTS
OF A PHASE III RANDOMIZED TRIAL

ERIC E. YEOH, M.D., F.R.C.P. (EDIN), F.R.C.R.,* RICHARD H. HOLLOWAY, M.D., F.R.A.C.P.,†
ROBERT J. FRASER, PH.D., F.R.A.C.P.,‡ ROCHELLE J. BOTTEN, B.SC. (HONS),*
ADDOLORATA C. DI MATTEO, B.SC. (HONS),* JULIE BUTTERS,* SUJEEVA WEERASINGHE, M.D.,*
AND PRASAD ABEYSINGHE, M.D.*

Departments of *Radiation Oncology and †Gastroenterology, Royal Adelaide Hospital, Adelaide, Australia; ‡Gastrointestinal
Investigation Unit, Repatriation General Hospital, Daw Park, Australia

Purpose: The aim of this study was to compare the toxicity and efficacy of radiation therapy (RT) for localized
carcinoma of the prostate, using a hypofractionated (55 Gy/20 fractions/4 weeks) vs. a conventionally fractionated
(64 Gy/32 fractions/6.5 weeks) dose schedule.
Methods and Materials: A total of 217 patients were randomized to either the hypofractionated (108 patients) or
the conventional (109 patients) dose schedule, with planning with two-dimensional (2D) CT scan planning
methodology in the majority of cases. All patients were followed for a median of 48 (6 –108) months. Gastroin-
testinal (GI) and genitourinary (GU) toxicity was evaluated before RT and after its completion using modified
late effects of normal tissue—subjective, objective, management, analytic (LENT-SOMA) scales and the Euro-
pean Organization for Research and Treatment of Cancer sexual function questionnaire. Efficacy of RT based
on clinical, radiologic, and prostate-specific antigen data were also evaluated at baseline and after RT.
Results: Gastrointestinal and GU toxicity persisted 5 years after RT and did not differ between the two dose
schedules other than in regard to urgency of defecation. However, 1-month GI toxicity was not only worse in
patients with the hypofractionated RT schedule but also adversely affected daily activities. Nadir prostate-specific
antigen values occurred at a median of 18.0 (3.0 –54.0) months after RT. A total of 76 biochemical relapses, with
or without clinical relapses, have occurred since; of these, 37 were in the hypofractionated and 39 in the
conventional schedule. The 5-year biochemical ⴞ clinical relapse–free and overall survival was 55.9% and 85.3%
respectively for all patients, and did not differ between the two schedules.
Conclusions: Radiation therapy for prostate carcinoma causes persistent GI toxicity that is largely independent
of the two dose schedules. The hypofractionated schedule is equivalent in efficacy to the conventional schedule.
© 2006 Elsevier Inc.

Radiation therapy, Prostate carcinoma, Hypofractionation, LENT/SOMA toxicity scales, Quality of life.

INTRODUCTION dality confirm the higher fraction sensitivity of prostate

The premise that multiple small fractions results in better
sparing of late-responding normal tissues compared with
early-responding tissues, including those constituting most
malignant tumors, has been challenged in the curative radi-
ation therapy (RT) of prostate cancer (1, 2). Until now,
studies reporting a higher RT fraction size sensitivity of
prostate cancer relative to late responding normal tissues (1,
2) have assumed that external beam RT and brachytherapy
are biologically equivalent therapies (3). If prospective
studies involving different schedules of the same RT mo-
cancer, multiple small-fraction RT would not only be sub-
optimal but would also represent inefficient use of re-
sources. Retrospective series of hypofractionated RT have
reported efficacy and toxicity comparable to conventionally
fractionated RT (4, 5). However, prescriptions of hypofrac-
tionated RT are currently limited to a small number of
institutions worldwide because efficacy and toxicity data in
previous studies being physician based and retrospectively
analyzed are subject to bias (4, 5). Until now, prospective
data of phase III studies of hypofractionated vs. convention-

Reprint requests to: Eric E. Yeoh, M.D., F.R.C.P. (Edin),
F.R.C.R., Department of Radiation Oncology, Royal Adelaide
Hospital, Adelaide, Australia 5000. Tel: (⫹61) 8-8222-4815; Fax:
(⫹61) 8-8222-2016; E-mail: eyeoh@mail.rah.sa.gov.au
Supported by the Cancer Council of South Australia.
Acknowledgments—The authors are indebted to the Cancer Coun-
cil of South Australia for research grant support; to Professor Jack
Fowler for calculating the ␣/␤ ratios for prostate cancer and for
reviewing the manuscript; and to Ms. Kristyn Wilson for assis-
tance in statistical analysis.
Received March 14, 2006, and in revised form June 6, 2006.
Accepted for publication June 6, 2006.

1072
 Radiation therapy for prostate carcinoma
ally fractionated RT for localized prostate carcinoma in-
cluding our own have been limited by either the relatively
short follow-up of the patients (6, 7) or the use of physician-
based recording of gastrointestinal toxicity (7, 8); one such
system is the Radiation Therapy Oncology Group (RTOG)
system, which does not include evaluation of anorectal symp-
toms such as urgency of defecation and fecal incontinence
(9, 10). Although the late effects of normal tissue—subjec-
tive, objective, management, analytic (LENT-SOMA) tox-
icity scales addresses some deficiencies of the RTOG system
(9, 10), there are few data available regarding LENT-
SOMA GI toxicity after RT for prostate cancer (6, 9).
Our preliminary results of this phase III randomized
study comparing the toxicity and efficacy of hypofraction-
ated vs. conventional RT for localized prostate carcinoma
have been published (6). Data on the final number of 217
patients recruited into the study followed for a median of 48
months (range, 6 –108 months) is now presented. As in the
previous report, evaluation of GI and GU toxicity is based
on a modification of the LENT-SOMA scales (9) and treat-
ment efficacy assessed based on (1) serum prostate-specific
antigen (PSA) and the American Society for Therapeutic
Radiology and Oncology (ASTRO) definition of biochem-
ical failure (11) and on (2) clinical and radiologic signs of
relapse (6).
METHODS AND MATERIALS
The primary end point of the study was to compare late RT
morbidity between the treatment groups after a minimum fol-
low-up of 2 years. The trial, designed to detect a difference in the
frequency of mild late RT morbidity of 20% (40% vs. 20%) with
80% to 90% power and a 2-tailed ␣-value of 0.05, was required to
recruit 80 to 120 patients into each RT schedule. Accrual of
patients was terminated after 217 patients, after the full implemen-
tation of 3-D conformal RT at our institution (discussed below).
Subjects
A total of 217 men (median age, 69 years; range, 44 – 82 years),
with T1/T2, N0 M0 (International Union Against Cancer [UICC]
1992) prostate carcinoma, were recruited between July 1996 and
August 2003. Androgen deprivation therapy, which was not stan-
dard practice for T2b disease at the time, was one of the exclusion
criteria for the study (6). A total of 108 patients were randomized
Table 1. Modified LENT-SOMA gastrointestinal (GI) toxicity scales
GI symptom/Score 0 1
Stool frequency ⱕ1 ⫻ /day 2–3 ⫻ /day
Stool consistency None No medication
needed
Rectal pain None Mild (occasional &
minimal)
Rectal mucous discharge Never About once a month
Urgency of defecation When toilet Up to one hour
available
Rectal bleeding Never About once a month
● E. E. YEOH et al. 1073
to receive 55 Gy/20 fractions/4 weeks and 109 patients were
randomized to receive 64 Gy/32 fractions/6.5 weeks; the latter was
the standard of care for two-dimensional (2D) RT in Australia. The
choice of the hypofractionated schedule, based on a large, single-
institution study (4), was previously reported to be equi-toxic to
the conventional schedule (6). The dose was prescribed to the
isocenter of a computer-generated plan encompassing the prostate
gland only with a 1.5-cm 95% isodose margin, and RT delivered
using a predominately four-field (anterior/posterior and laterals),
external-beam, megavoltage (6 –23 MV) photon technique as pre-
viously reported (6); the standard practice then was not to include
pelvic nodes in the target. RT planning was based on 2D computed
tomography (CT) data in the majority (156) of the patients. In the
remaining (61) patients, three-dimensional (3D) planning CT data
were used after the department acquired 3D conformal RT capa-
bility but before an in-house treatment set-up study established that
margins around the prostate gland could be safely reduced and thus
allow higher radiation doses to be prescribed to the current Aus-
tralian minimum dose of 70 Gy / 35 fractions / 7 weeks. Anterior
field shielding was allowed in the patients with 2D planning as
previously described (6). In the patients with 3D planning, mul-
tileaf collimators were used to shape each beam around the
planned target volume, which was derived by applying a margin of
1.5 cm around the contoured prostate gland using the automatic
expansion tool of the planning (ADAC Pinnacle; Sun Corporation,
Santa Clara, CA) system.
Protocol
After the 1-month visit after completion of RT, patients were
followed at 3-month intervals for the next 2 years, then 6-monthly
for another 3 years. Five years after RT, the interval between fol-
low-up visits was extended to yearly until death. GI and GU
symptoms were evaluated before RT and at each visit after RT to
5 years. Venous blood was obtained before RT and at each
3-month, 6-month, or yearly visit after its completion for assay of
the serum PSA (Abbot AXSym; Abbott Laboratories, Abbott Park,
IL). Written, informed consent was obtained from all patients and
the study protocol was approved by the Human Ethics Committee
of the Royal Adelaide Hospital.
Methodology
The following GI symptoms were assessed by questionnaire:
stool frequency, stool consistency, rectal pain, rectal mucous dis-
charge, urgency of defecation, and rectal bleeding.
The scoring of each GI symptom was based on a 5-point
categorical scale (0 – 4) as previously reported (6) and summarized
2 3 4
4–8 ⫻ /day ⬎8 ⫻ /day Uncontrolled
Moderate (medication Severe (not well Uncontrolled
needed) controlled with
medication)
Moderate (intermittent Severe (persistent & Constant and
& tolerable) intense) excruciating
About once a week Almost every day Every time
Within minutes Immediately Occasional accidents
About once a week Almost every day Major hemorrhage
1074 I. J. Radiation Oncology ● Biology ● Physics Volume 66, Number 4, 2006

Table 2. Modified LENT-SOMA genitourinary (GU) toxicity scales

GU symptom/Score 0 1 2 3 4

Diurnal frequency of ⬎4 hours Every 3–4 hours Every 2–3 hours Every 1–2 hours More than every hour
micturition
Nocturia ⬎4 hours Every 3–4 hours Every 2–3 hours Every 1–2 hours More than every hour
Hematuria Never About once a month About once a week Almost every day Every time
Urgency of micturition When toilet Up to 1 hour Within minutes Immediately Occasional accidents
available
Dysuria None Mild (occasional Moderate (intermittent Severe (persistent Constant and excruciating
and minimal) and tolerable) and intense)

in Table 1. The effects of GI symptoms on the daily activities of
patients since completion of RT were recorded and were graded as
follows (9): 0 ⫽ no effect; 1 ⫽ little effect but noted change; 2 ⫽
moderate effect, needing changes to daily activities; 3 ⫽ severe,
practically housebound.
The following GU symptoms were evaluated: diurnal frequency
of micturition, nocturia, hematuria, urgency of micturition, and
dysuria. The scoring of each GU symptom was as previously
reported (6) and summarized in Table 2. The effect GU symptoms
had on the daily activities of patients after RT, graded as for GI
symptoms above (9), was also documented.
Sexual function was evaluated using The European Organiza-
tion for Research and Treatment of Cancer (EORTC) question-
naire (12), as in our earlier report (6).
Data analysis
The median (range) individual and total GI and GU symptom
scores were calculated for all the patients at baseline, at 1 month,
and annually after RT. The percentage of patients who had in-
creases in total GI and GU symptom scores since RT and the
effects of these symptoms on daily activities was also determined.
Mean (⫾SE) and median (range) serum PSA was calculated for
all patients and according to RT schedule at baseline and annually
after RT. The mean (⫾SE) and median (range) nadir serum PSA
was also derived for the total patient population and for each RT
schedule. Patient numbers in each of 3 pretreatment PSA and
histologic prognostic subgroups of ⬍10, 11 to 20, and ⬎20 ␮g/l
and Gleason scores of 2 to 6, 7, and 8 to 10 respectively were
determined for all patients and according to assigned RT schedule.
Statistical analysis
The Friedman repeated-measures analysis of variance (ANOVA)
on ranks was used to evaluate (1) the individual and total GI and
GU symptom scores at baseline and after RT, and (2) the effect GI
and GU symptoms had on the daily activities of patients since RT.
The serum PSA levels at baseline, the nadir PSA, and PSA levels
annually after RT were analyzed using 2-way repeated-measures
ANOVA with means comparison. The Mann-Whitney U test and
unpaired t test were respectively used to compare the individual
and total GI and GU symptom scores and the serum PSA values at
baseline and after RT between the treatment groups. The ␹ 2 test
was used to compare (1) the proportion of patients with increased
GI and GU symptoms after RT according to dose schedule and
treatment (2D vs. 3D) technique, and (2) the distribution of pa-
tients in each of the three pretreatment PSA and histologic prog-
nostic subgroups according to assigned RT schedule. The actuarial
5-year biochemical ⫾ clinical relapse-free and overall survival for
all patients was determined by the Kaplan-Meier method and the
log-rank test was used to compare the biochemical ⫾ clinical
relapse-free and overall survival between the two dose schedules.
The log-rank test was also used to compare the actuarial 2-year
biochemical ⫾ clinical relapse-free and overall survival between
the patients treated with 2D and 3D techniques.
The relationships among RT volumes in the patients and indi-
vidual and total GI and GU symptom scores since RT were
evaluated by linear regression analysis. Univariate and multiple
logistical regression analysis was also performed to examine the
prognostic variables that predict: (1) biochemical failure (with or
without clinical failure) after RT, and (2) increase in total GI and
GU symptom scores at 2, 3, 4, and 5 years using the Cox propor-
tional hazards and log binomial regression methodology respec-
tively. For biochemical fairure with or without clinical failure, the
predictor variables examined were: (a) T1 vs. T2 disease; (b) PSA
ⱕ10, 11–20, and ⬎ 20 subgroups; (c) Gleason scores 2 to 6, 7, and
8 to 10; (d ) hypofractionated vs. conventionally fractionated RT
schedule; (e) RT volume (as a continuous variable); and (f ) 2D vs.

Table 3. Individual and total gastrointestinal (GI) symptom scores in all patients at baseline and after radiation therapy

Baseline 1 Month 1 Year 2 Years 3 Years 4 Years 5 Years Overall

(n ⫽ 217) (n ⫽ 217) (n ⫽ 209) (n ⫽ 185) (n ⫽ 111) (n ⫽ 74) (n ⫽ 65) p value

Stool frequency 0 (0–3) 1 (0–3)* 1 (0–3)* 1 (0–2)* 1 (0–3)† 1 (0–2)‡ 1 (0–3) ⬍0.0001
Stool consistency 0 (0–2) 0 (0–3)† 0 (0–2)* 0 (0–2)* 0 (0–2)‡ 0 (0–4) 0 (0–3) ns
Rectal pain 0 (0–1) 0 (0–3)* 0 (0–3)† 0 (0–2)* 0 (0–2)‡ 0 (0–2)‡ 0 (0–2)‡ ⬍0.0001
Rectal mucous discharge 0 (0–3) 0 (0–4)* 0 (0–4)* 0 (0–3)* 0 (0–4)* 0 (0–3)† 0 (0–3)† ⬍0.0001
Urgency of defecation 0 (0–3) 1 (0–4)* 1 (0–4)* 1 (0–4)* 0 (0–4)* 0 (0–4)* 1 (0–4)* ⬍0.001
Rectal bleeding 0 (0–3) 0 (0–3)* 0 (0–3)* 0 (0–3)* 0 (0–3)* 0 (0–3)† 0 (0–2)† ⬍0.01
Total GI symptoms 1 (0–7) 3 (0–12)* 3 (0–12)* 3 (0–11)* 3 (0–10)* 2 (0–10)* 2.5 (0–10)* ⬍0.0001

* p ⬍ 0.001 vs. baseline.

†
p ⬍ 0.01 vs. baseline.
‡
p ⬍ 0.05 vs. baseline.
 Radiation therapy for prostate carcinoma ● E. E. YEOH et al. 1075

Table 4. Proportion of patients who had increased gastrointestinal (GI) symptoms and in whom daily
activities were affected after radiation therapy

1 Month 1 Year 2 Years 3 Years 4 Years 5 Years

Stool frequency 38% 31% 27% 26% 27% 15%

Stool consistency 23% 28% 26% 23% 20% 23%
Rectal pain 34% 13% 14% 9% 14% 17%
Rectal mucous discharge 26% 29% 24% 21% 19% 23%
Urgency of defecation 42% 36% 33% 38% 39% 48%
Rectal bleeding 17% 28% 36% 28% 26% 25%
Effect on daily activities 62% 50% 50% 44% 43% 51%

3D technique. For increase in total GI and GU symptom scores at (c) RT volume (as a continuous variable); (d) total GI and GU
2, 3, 4, and 5 years, the baseline predictor variables evaluated were symptoms of ⱖ3; and (e) 2D vs. 3D technique with or without the
(a) age; (b) hypofractionated vs. conventionally fractionated RT; post-treatment variable of patients with increases in total GI and

(a) (b)
1.2 1.8 ** *
1.1 * 1.6
1 1.4
.9
1.2
Stool .8 Urgency of 1
Frequency .7 Defecation
.8
.6
.5 .6
.4 .4
.3 .2
.2 0
Pre 1 Mth 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr Pre 1 Mth 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr

(c) (d)
1.8 ** 7
1.6 *
6
1.4
1.2 5
1 4
Mucous Total GI
Discharge .8 Symptom
.6 3
Score
.4 2
*
.2
1
0
-.2 0
Pre 1 Mth 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr Pre 1 Mth 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr

(e)
1.4
1.3
**
1.2
1.1
Effect of 1
GI Symptoms .9
on Daily .8
Activities .7
.6
.5
.4
.3
1 Mth 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr

Fig. 1. (a) Stool frequency, (b) urgency of defecation, (c) mucous discharge, (d) total gastrointestinal (GI) symptom
score, and (e) effect of GI symptoms on daily activities in the hypofractionated (●) and conventional (‘) schedule before
RT and at 1 month and annually up to 5 years after radiation therapy (RT). *p ⬍ 0.05, **p ⬍ 0.01.
1076 I. J. Radiation Oncology ● Biology ● Physics Volume 66, Number 4, 2006

GU symptoms at 1 month. In relation to the 2D vs. 3D predictor
variables, each univariate and multivariate analysis was based on
the 2-year efficacy and toxicity follow-up data only.
A value of p ⱕ 0.05 was considered significant in all analyses.
RESULTS
Compliance with the experimental protocol was excellent
for the first 2 years, with 185 of the 217 patients (85%)
completing their GI and GU symptom questionnaires, al-
though most patients failed to complete the EORTC sexual
functioning questionnaire as previously reported (6). The
2-year PSA data were also available in 208 patients (96%).
After 2 years, however, 76 patients had elected to attend
elsewhere, and 35 patients had died (31 of these patients
died of causes unrelated to prostate cancer). The GI and GU
toxicity data at 5 years are therefore based on 65 (approx-
imately one third) of the original 217 patients. However,
5-year PSA as well as clinical and radiologic data were
available in 96 (nearly one half) of the 217 patients, because
of electronic access to laboratory databases and information
provided by attending physicians elsewhere on nearly one
half of the patients who were no longer attending our
facility for follow-up after 2 years.
Gastrointestinal symptoms
The increase in individual and total GI symptom scores
after RT persisted at 5 years except for stool consistency

Fig. 2. (a) Rectal bleeding, (b) total gastrointestinal (GI) symptom scores, (c) effect of GI symptoms on daily activities,
(d) total genitourinaray (GU) symptom scores, and (e) effect of GU symptoms on daily activities in radiation therapy
with 2D (’) vs. 3D (⽧) treatment planning before radiation therapy (RT) and at 1 month, 1 year, and 2 years after RT.
*p ⬍ 0.05, **p ⬍ 0.01.
 Radiation therapy for prostate carcinoma ● E. E. YEOH et al. 1077

Table 5. Multivariate analysis of variables that independently predict for increase in gastrointestinal (GI) symptom scores at 2, 3, 4, and
5 years after radiation therapy

2 Years 3 Years 4 Years 5 Years

Variable/increase in GI
symptoms p RR CI p RR CI p RR CI p RR CI

Hypofractionated RT ⬍0.01 1.32 1.07–1.64 ns ns ns

Increase in GI
symptoms at 1 month ⬍0.01 1.71 1.22–2.39 ⬍0.01 2.20 1.21–3.99 ⬍0.05 2.03 1.07–3.83 0.06 2.46 0.95–6.37

Abbreviations: CI ⫽ 95% confidence interval; RR ⫽ relative risk.

(Table 3). At this time, 15% to 48% of patients reported
increased GI symptoms since completion of RT, and 51%
reported some effect on their daily activities (Table 4).
One month after RT, stool frequency, urgency of defeca-
tion, mucous discharge, total GI symptom scores, and the
effect of GI symptoms on daily living were significantly
worse in the patients treated with hypofractionated RT
compared with the conventional schedule (Figs. 1a–1e).
These differences, however, did not persist beyond 1 month
other than a worsening of scores for urgency of defecation
in patients treated with the hypofractionated schedule at 5
years (Fig. 1b).
Rectal bleeding, total GI symptom score, and the effect of
GI symptoms on daily activities were surprisingly worse in
the subset of patients treated by 3D vs. 2D RT (Figs. 2a–2c).
However, the length of the radiation fields for 3D patients of
10.6 cm (range, 9 –13 cm) was greater than the 9.0 cm
(range, 7.5–11.0 cm) for the 2D RT patients ( p ⬍ 0.0001),
although the median volume encompassed by the multileaf
collimators of the radiation fields in the 3D group (628 mLs,
range, 422–1039 mLs) was smaller than the median volume
encompassed by the 50% isodose of the radiation fields in
the 2D group of patients (855 mLs, range, 478 –1391 mLs)
(p ⬍ 0.0001).
Multivariate logistical analysis including baseline vari-
ables that were significant on univariate analysis, showed
that hypofractionated RT independently predicted increased
total GI symptoms at 2 years (Table 5). When increase in
total GI symptoms at 1 month was included in the multi-
variate analysis, this variable also predicted for increased
total GI symptoms at 2, 3, 4, and possibly 5 years (Table 5).
Genitourinary symptoms
Individual GU and total GU symptom scores other than
hematuria increased 1 month after RT (Table 6). The in-
creases were sustained for nocturia and total GU symptom
scores only, the remaining urinary symptoms tending to
decrease at 5 years (Table 6). At this time, 32% of patients
reported an increase in nocturia, and 46% of patients re-
ported some effect of urinary symptoms on their daily
activities after RT (Table 7).
Total GU symptom scores were significantly worse in the
hypofractionated compared with the conventional schedule
at 2 and 3 years (Fig. 3a), although the magnitude of the
change from baseline was no different between the two
schedules. There was no difference in the effect of GU
symptoms on daily activities between the two treatment
schedules (Fig. 3b).
Total GU symptoms, but not their effect on daily activi-
ties, were also significantly worse in the 3D vs. 2D RT
patients up to 2 years after RT (Figs. 2d, 2e).
Linear regression analysis showed a relationship between
urgency of micturition scores at 4 years and radiation treat-
ment volumes (Fig. 4).
Multivariate logistical regression analysis, including base-
line variables that were significant on univariate analysis,
showed that 3D RT was of independent prognostic signifi-
cance for increased total GU symptoms at 2 years (Table 8).
When increase in total GU symptoms at 1 month was
included in the multivariate analysis, this variable also in-
dependently predicted for increased total GU symptoms at
2, 3, and 5 years (Table 8).
Because compliance with the EORTC sexual function

Table 6. Individual and total genitourinary (GU) symptom scores in all patients at baseline and after radiation therapy

Baseline 1 Month 1 Year 2 Years 3 Years 4 Years 5 Years Overall

(n ⫽ 217) (n ⫽ 217) (n ⫽ 209) (n ⫽ 185) (n ⫽ 111) (n ⫽ 74) (n ⫽ 63) p value

Diurnal frequency of micturition 1 (0–4) 2 (0–4)* 1 (0–4) 1 (0–4)† 1 (0–4) 1 (0–3) 1 (0–4) ⬍0.0001
Nocturia 0 (0–4) 1 (0–4)* 1 (0–4) 0 (0–4) 1 (0–4)‡ 1 (0–3)* 1 (0–4) ⬍0.0001
Hematuria 0 (0–4) 0 (0–4) 0 (0–4) 0 (0–4)† 0 (0–3) 0 (0–0) 0 (0–3) ⬍0.05
Urgency of micturition 0 (0–4) 1 (0–4)‡ 1 (0–4) 0 (0–4) 0 (0–4)† 0 (0–4) 0 (0–4) ns
Dysuria 0 (0–4) 0 (0–3)* 0 (0–3) 0 (0–1)* 0 (0–2)† 0 (0–1)† 0 (0–1) ⬍0.0001
Total GU symptoms 3 (0–12) 4 (0–13)* 3 (0–11) 3 (0–10) 3 (0–15) 2 (0–10) 3 (0–15) ⬍0.0001

* p ⬍ 0.001 vs. baseline.

†
p ⬍ 0.01 vs. baseline.
‡
p ⬍ 0.05 vs. baseline.
1078 I. J. Radiation Oncology ● Biology ● Physics Volume 66, Number 4, 2006

Table 7. Proportion of patients who had increased genitourinary (GU) symptoms and in whom daily
activities were affected after radiation therapy

Characteristic 1 Month 1 Year 2 Years 3 Years 4 Years 5 Years

Diurnal frequency of micturition 44% 22% 24% 23% 26% 22%

Nocturia 39% 17% 23% 30% 36% 32%
Hematuria 3% 4% 0% 7% 0% 8%
Urgency of micturition 30% 21% 21% 28% 25% 24%
Dysuria 24% 8% 4% 3% 1% 5%
Effect on daily activities 59% 39% 39% 37% 37% 46%

questionnaire has remained poor, no additional meaningful histopathologic and radiologic (bone and CT abdominal
data on sexual functioning can be provided since the pre- scan) reassessment. The actuarial 5-year biochemical ⫾
vious report (6). clinical relapse-free survival, was 55.9% for all patients and
57.4% and 55.5% for patients in the hypofractionated and
Treatment efficacy conventional schedules respectively (log-rank, NS; Fig. 6a),
Prostate-specific antigen levels fell after RT (Table 9), the (HR and CI, Table 10). Because more patients with PSA
median duration for the nadir being 18 months (range, 3–54 ⬎20 and not with Gleason scores ⬎7 were assigned to
months). PSA at baseline, nadir, and yearly PSA after RT the hypofractionated vs. the conventional RT schedule
did not differ between the two RT schedules (Table 9, Fig. (Tables 11 and 12), subset analysis performed excluding (1)
5). A total of 76 patients (37 in the hypofractionated sched- patients with PSA ⬎20, and (2) patients with PSA ⬎20 and
ule, 39 in the conventional schedule, ␹ 2 ⫽ NS) relapsed Gleason scores ⬎7 showed that the actuarial 5-year bio-
based on PSA criteria, 15 of whom also had clinical evi- chemical ⫾ clinical relapse-free survivals for all patients
dence of relapse on investigation. The sites of clinical and those in the hypofractionated and conventional dose
relapse in the 15 patients were local (7 patients), pelvic schedules in subset were respectively 56.9% and 59.1% and
nodal (3 patients), and bony metastatic (5 patients) based on 56.3% for all patients and those in the hypofractionated and
conventional dose schedules in subset 2 were respectively
57.1% and 59.3% and 56.5%, log-rank NS for both, HR and
(a) CI (Table 10). Calculations involving correction for the
1.9% and 2.8% difference in actuarial freedom from bio-
5.5
chemical ⫾ clinical relapse between the treatment groups
5
* for all patients and excluding those at high risk made on the
4.5
basis of the slope of the prostate cancer dose–response
Total GU 4
curve (13) indicate an ␣/␤ ratio of 2.3 and 2.2 (CI, ⫺6.0 –
Symptom 3.5 *
Score 10.6) Gy respectively.
3
Actuarial 5-year overall survival of 85.3% for the whole
2.5 group did not differ between the values 86.4% and 84.1%
2 respectively for the hypofractionated and conventional
1.5 schedules (log-rank, p ⫽ NS) (Fig. 6b). Only 4 of the 35
1 patient deaths at the time of data analysis were related to
Pre 1 Mth 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr
prostate carcinoma (1 treated with the hypofractionated
(b) schedule and 3 with the conventional schedule).
1.2

Effect of GU .8
Symptoms
on Daily .6
Activities
.4

.2

0
1 Mth 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr

Fig. 3. (a) Total genitourinary (GU) symptom scores and (b) effect
of GU symptoms on daily activities in the hypofractionated (●) and Fig. 4. Relationship between urgency of micturition scores and
conventional (‘) schedule before radiation therapy (RT) and at 1 radiation treatment volumes in patients 4 years after radiation
month and annually up to 5 years after RT. *p ⬍ 0.05. therapy.
 Radiation therapy for prostate carcinoma ● E. E. YEOH et al. 1079

Table 8. Multivariate analysis of variables that independently predict for increase in genitourinary (GU) symptom scores at 2, 3, 4, and
5 years after radiation therapy

2 Years 3 Years 4 Years 5 Years

Variable/increase in GU
symptoms p RR CI p RR CI p RR CI p RR CI

3D conformal RT ⬍0.01 1.87 1.19–2.94 — — —

Increase in GU symptoms
at 1 month ⬍0.001 2.97 1.68–5.25 ⬍0.05 1.95 1.11–3.45 ns ⬍0.05 3.10 1.23–7.80

Abbreviations: CI ⫽ 95% confidence interval; RR ⫽ relative risk.

The actuarial 2-year biochemical ⫾ clinical relapse free
survival was better for the 3D (92.9%) vs. the 2D (78.6%)
technique (Fig. 6c).
Multivariate logistical analysis including variables signif-
icant on univariate analysis showed that 3D RT and baseline
PSA of ⬍ 10 each independently predicted for reduced risk
of biochemical ⫾ clinical relapse at 2 years (Table 13).
DISCUSSION
Updated results of this randomized trial of hypofraction-
ated vs. conventional RT schedule for localized prostate
carcinoma indicate the following: (1) there is no difference
in treatment efficacy or in GI and GU toxicity between the
dose schedules at 5 years, except for urgency of defecation
and significantly worse individual and total GI symptom
scores at 1 month in the hypofractionated vs. the conven-
tional schedule; (2) 3D vs. 2D RT was associated with a
reduced risk of biochemical ⫾ clinical relapse but at the
expense of worse GI symptoms at 2 years; (3) increased GI
and GU symptom scores at 1 month independently pre-
dicted for increased GI symptom scores at 2, 3, and 4 years
and increased GU symptom scores at 2, 3, and 5 years; and
(4) baseline PSA ⬍ 10 and 3D RT each independently
predicted for reduced risk of biochemical ⫾ clinical relapse
at 2 years.
The 5-year freedom from biochemical ⫾ clinical relapse
rate in relation to the hypofractionated schedule in this study
of 57.4% compares favorably with 40% in a multicenter
Canadian study (8). The shorter median duration of fol-
low-up here compared with the multicenter study is unlikely
to explain the superior results in this study as only one
further relapse has occurred in the hypofractionated sched-
ule 5 and 9 years after RT, a finding consistent with a report
that relapses beyond 5 years are uncommon (14). The higher
total and particularly fractional dose in the hypofractionated
arm in our study (55 vs. 52.55 Gy, each in 20 fractions
treating 5 ⫻ week) is a more likely explanation for the
better relapse free survival. In addition, not only were the
total and fractional doses in the hypofractionated arm lower
in the Canadian study but the total dose in the conventional
arm was also higher compared with this study (66 vs. 64 Gy,
each in 2-Gy fractions treating 5 ⫻ week). This explains the
low ␣/␤ ratio of 1.12 (CI ⫽ ⫺3.3–5.6) Gy estimated for
prostate cancer in the Canadian study (15). Although con-
fidence limits of the ␣/␤ ratio estimate in this study are also
large, the estimate of 2.2 for similar risk patients is never-
theless consistent with a higher fraction sensitivity of pros-
tate cancer relative to most malignant tumors and also to the
late effects in the rectum, the ␣/␤ ratio of which has been
estimated to be between 3 and 5.4 Gy (2, 16).
The lower risk of biochemical ⫾ clinical relapse ob-
served in the subset of patients treated with 3D RT vs. the
2D technique in this study may just reflect the shorter
follow-up of the 3D patients with consequent reduced ex-
posure to the risk of relapse. However, should the reduced
risk of relapse in 3D patients be sustained beyond 2 years, the
possibility arises that the 2D RT may have resulted in geo-
graphic misses of the planned target volume in the superior–
inferior axis. We have reported significantly longer field
dimensions in a previous study involving the replanning of
treatment in patients with 2D treatment planning using the
automatic expansion tool of our 3D planning system to
apply the same 1.5-cm margins around the prostate (17).
Although the longer fields in this and our previous study are
likely to better account for the interfraction displacement of
the planned target volume in the superior–inferior axis (18,
19), it probably also explains the worse late GI and GU
effects observed at 2 years in the patient with 3D vs. 2D
planning in this study, as well as the disparity in the findings
with a previous study that randomized patients according to
the two treatment techniques (20). Nonetheless, the need to
account for the increase in the superior–inferior field dimen-
sion in 3D planning by reducing the margins in the other

Table 9. Mean (⫾ SE) baseline, nadir, and yearly prostate-specific antigen levels (␮g/l) in all patients and according to assigned
radiation therapy schedule

Baseline Nadir 1 Year 2 Years 3 Years 4 Years 5 Years Overall p value

All patients (n ⫽ 217) 13 ⫾ 0.6 1.2 ⫾ 0.1 2.2 ⫾ 0.4 3.2 ⫾ 0.8 2.4 ⫾ 0.5 6.8 ⫾ 2.7 11.5 ⫾ 4.8 ⬍0.0001
Hypofractionated (n ⫽ 108) 13.8 ⫾ 0.9 1.4 ⫾ 0.2 2.6 ⫾ 0.7 3.3 ⫾ 1.2 2.5 ⫾ 1.0 6.1 ⫾ 4.4 11.5 ⫾ 6.5 ⬍0.05
Conventional (n ⫽ 109) 12.3 ⫾ 0.9 1.1 ⫾ 0.1 1.8 ⫾ 0.4 3.1 ⫾ 1.0 2.2 ⫾ 0.3 7.5 ⫾ 2.8 11.5 ⫾ 7.1 ⬍0.001
1080 I. J. Radiation Oncology ● Biology ● Physics
20
18
16
14
Serum
PSA 12
Levels 10
(ug/L) 8 multivariate analysis that revealed that increased total GI
6 scores at 1 month independently predicted for increased
4
2
0 The influence of acute GI toxicity on the development of
Pre Nadir 1 Yr 2 Yr 3 Yr 4 Yr 5 Yr
Fig. 5. Serum prostate-specific antigen (PSA) levels (mean ⫾ SE)
in all patients () and in the hypofractionated (●) and conventional
(‘) schedule at baseline, nadir and at 1, 2, 3, 4, and 5 years after
radiation therapy.
axes around the prostate particularly in the context of radi-
ation dose escalation is highlighted.
Not only are more mature data on GI and GU toxicity
provided here since our preliminary report (6), but new
information about the prognostic variables that determine
risks of late toxicity and their impact on the daily activities
of the patients are also provided. In addition, the individual
GI and GU symptom scores of the patients have been
pooled here and analyzed as ranked sum scores rather than
expressed as percentages of patients with scores of ⱖ1 at
baseline and after RT and then compared using the ␹ 2 test
as previously described (6). Repeated-measures ANOVA of
ranks, used here, is a more robust test of statistical signifi-
cance for the nonparametric GI and GU toxicity data (21).
Furthermore, a similar modified version of the LENT-
SOMA GI and GU toxicity scales to this study has been
found to correlate with the only validated quality-of-life
instrument for treatment-related sequelae of prostate carci-
noma (22).
Although the toxicity data beyond 2 years in this study is
limited by the reduced numbers of patients still attending for
follow-up, thus lowering the statistical power of detectable
differences between the RT schedules after this time, it is
well recognized that the great majority of patients who will
develop late GI sequelae will do so within 2 years of
completing RT (23, 24). Therefore, the observation of an
absence of significant differences between the radiation
schedules relating to GI toxicity at 2 years is still statisti-
cally valid. At this time, not only were there no differences
in individual and total GI symptom scores between the two
dose schedules but also the effects of these sequelae on the
daily activities of the patients were similar, although the
hypofractionated schedule independently predicted for in-
creased GI symptoms at 2 years.
That increased GI and GU symptom scores at 1 month
were of independent prognostic significance in relation to
increased GI and GU symptoms at 2 years after RT and
beyond has implications for the design of future hypofrac-
tionated RT trials for prostate carcinoma. We have shown
that the hypofractionated schedule in this study resulted in
an increase in some of the individual as well as the total GI
symptom scores, which had an adverse impact on the daily
Volume 66, Number 4, 2006
activities of the patients at 1 month, although these effects
were not sustained other than a worsening of urgency of
defecation at 5 years for the hypofractionated vs. the con-
ventional schedule. The possibility that late GI toxicity may
be a consequential effect of early GI toxicity is supported by
total GI symptom scores at 2, 3, 4, and also possibly 5 years.
chronic GI sequelae has previously been reported (25, 26)
and may be able to be minimized with more protracted
hypofractionated schedules, particularly as accelerated re-
population during RT for prostate carcinoma has not been
reported for treatment courses of 9 weeks or less (27). In
relation to the hypofractionated schedule in our study, cal-
culations based on those proposed to achieve tolerable lev-
els of acute upper GI tract mucosal reactions (28), indicate
that acute GI toxicity may be able to reduced if the treat-
ment fractions were to be delivered 4 ⫻ week over 5 weeks.
In addition to exploiting accelerated mucosal repopulation
(28), more protracted hypofractionated schedules should
also maximize repair of sublethal damage between each
fraction, particularly as the half time of repair of damage to
normal tissues may be longer than previously realized (29).
Of the five individual GI symptoms persistently increased
at 5 years, urgency of defecation reported by nearly half the
patients was the most prevalent. At this time, the proportion
of patients who reported an effect of GI symptoms on daily
activities was also approximately 50%. We and others have
reported that this GI symptom, rather than rectal bleeding, is
not only the most common but also most likely to affect
daily activities after RT for prostate carcinoma (9, 10, 25,
30). Yet reports on GI effects after RT for prostate carci-
noma continue to use the RTOG/EORTC toxicity criteria,
which ignore anorectal symptoms such as urgency of defe-
cation and fecal incontinence. Although rectal bleeding and
increased stool frequency has been reported to be reduced
by 3D RT (20), there is limited information on the effect of
this treatment technique on anorectal function. We have
found that anorectal sensory function and symptoms are not
reduced at 2 years for similar patients with prostate carci-
noma treated with 3D RT vs. the 2D technique (31). It is
therefore likely that anorectal dysfunction is an underesti-
mated cause of morbidity of RT for prostate carcinoma
regardless of treatment technique. A priori, there is no
reason that anorectal dysfunction would be reduced by any
technique of conformal RT except possibly for brachyther-
apy. With external beam techniques, a portion of the rectal
mucosa— however small—inevitably receives the full radi-
ation dose; and we have shown that heightened sensitivity
of the rectal mucosa is an integral component of anorectal
dysfunction after pelvic irradiation including the compara-
tive study of 2D vs. 3D RT for prostate carcinoma involving
physiologic measurements of anorectal sensory and motor
function discussed above (25, 31, 32).
Excessive mucous production is a feature of inflamma-
tory bowel disease and irritable bowel syndrome. The mu-
 Radiation therapy for prostate carcinoma ● E. E. YEOH et al. 1081

(a)
100
90
80
70
% Patients 60
Biochemical 50
+ Clinical
40
Relapse Free
30
20
10
0
0 12 24 36 48 60 72 84 96 108

Hypofractionated (108) (106) (90) (66) (42) (22) (13) (10) (7)
Conventional (109) (108) (93) (63) (32) (22) (9) (5) (2)

(b)
100
90
80
70
% Overall 60
Survival
50
40
30
20
10
0
0 12 24 36 48 60 72 84 96 108

Hypofractionated (108) (108) (105) (91) (79) (61) (46) (38) (22)
Conventional (109) (109) (107) (93) (80) (64) (53) (33) (14)

(c) 100
90 p<0.05
80
70
% Patients 60
Biochemical
50
+ Clinical
Relapse Free 40
30
20
10
0
0 3 6 9 12 15 18 21 24 27
Standard (156) (156) (156) (156) (154) (147) (146) (137) (132)
Conformal (61) (61) (61) (61) (60) (57) (56) (53) (51)

Fig. 6. (a) Kaplan-Meier estimate of biochemical ⫾ clinical relapse free survival in all patients () and those assigned
to the hypofractionated (●) and conventional (‘) schedule. (b) Kaplan-Meier Estimate of overall survival in all patients
() and those assigned to the hypofractionated (●) and conventional (‘) schedule. (c) Kaplan-Meier estimate for
biochemical ⫾ clinical relapse–free survival at 2 years in patients with 2D (’) vs. 3D (⽧) radiation therapy planning.
1082 I. J. Radiation Oncology ● Biology ● Physics Volume 66, Number 4, 2006

Table 10. Hazard ratio (HR) and 95% confidence interval (CI) Table 12. Distribution of pretreatment Gleason scores between
of biochemical ⫾ clinical relapse free comparisons among all patients assigned to each of the two radiation therapy schedules
patients and subsets of patients assigned to the hypofractionated
and conventional radiation therapy schedules Gleason score 2–6 7 8–10

Patient group HR CI All patients (n ⫽ 217) 173 31 13

Hypofractionated (n ⫽ 108) 85 15 8
All patients 0.92 0.58–1.45 Conventional (n ⫽ 109) 88 16 5
Excluding patients with PSA ⬎20 ␮g/l 0.86 0.51–1.42
Excluding patients with PSA ⬎20 ␮g/l and
Gleason score ⬎7 0.88 0.51–1.48 Table 13. Multivariate analysis of variables that independently
predict reduced risk of biochemical ⫾ clinical relapse at 2 years
PSA ⫽ prostate-specific antigen.
Reduced risk of biochemical ⫾
clinical relapse at 2 years
Table 11. Distribution of pretreatment prostate-specific antigen,
subgroups between patients assigned to each of the two Variable p HR CI
radiation therapy schedules
3D conformal RT ⬍0.01 0.27 0.11–0.65
Patient group ⬍10 ␮g/l 10–20 ␮g/l ⬎20 ␮g/l Baseline PSA ⬍10 ␮g/l ⬍0.05 0.47 0.24–0.92
All patients (n ⫽ 217) 101 86 30 Abbreviations: CI ⫽ 95% confidence interval; HR ⫽ hazard
Hypofractionated ratio, PSA ⫽ prostate-specific antigen; RT ⫽ radiation therapy.
(n ⫽ 108) 47 41 20*
Conventional (n ⫽ 109) 54 45 10

* p ⬍ 0.05 vs. conventional.

tients evaluated with the same anorectal manometric assem-
bly before and up to 2 years after 2D RT (32).

cous discharge reported by approximately one quarter of the

patients at 5 years in this study is likely to reflect increased
rectal irritability as evidenced by the mucosal hypersensi-
tivity discussed above (25, 31, 32). As this often occurred
independently of defecation such as while passing flatus, a
degree of incompetence of the anal sphincters is likely
among the patients. This is supported by another of our
studies involving anorectal manometry in which we found
persistent anal sphincteric dysfunction in a subset of pa-
CONCLUSION
In summary, updated results of this phase III randomized
trial of hypofractionated vs. conventionally fractionated RT
for prostate carcinoma confirm that the two schedules are
equivalent in efficacy at 5 years. Worse GI toxicity in the
hypofractionated schedule at 1 month, sustained only for
urgency of defecation, has implications for treatment ap-
proaches and studies involving higher radiation doses.

REFERENCES
1. Brenner DJ. Hypofractionation for prostate cancer radiother-
apy—what are the issues? (Editorial). Int J Radiat Oncol Biol
Phys 2003;57:912–914.
2. Fowler J, Chappell R, Ritter, M. Is the ␣/␤ for prostate
tumors really low? Int J Radiat Oncol Biol Phys 2001;50:
1021–1031.
3. D’Souza WD, Thames HD. Is the ␣/␤ for prostate tumors
low? Int J Radiat Oncol Biol Phys 2001;51:1–3.
4. Lloyd-Davies RW, Collins CD, Swan AV. Carcinoma of
prostate treated by radical external beam radiotherapy using
hypofractionation. Twenty-two years’ experience (1962–1984).
Urology 1990;36:107–111.
5. Duncan W, Warde P, Catton CN, et al. Carcinoma of the
prostate: Results of radical radiotherapy (1970 –1985). Int J
Radiat Oncol Biol Phys 1993;26:203–210.
6. Yeoh E, Fraser RJ, McGowan RE, et al. Evidence for
efficacy without increased toxicity of hypofractionated ra-
diotherapy for prostate carcinoma: Early results of a Phase
III randomized trial. Int J Radiat Oncol Biol Phys 2003;55:
943–955.
7. Kupelian PA, Reddy CA, Willoughby PR, et al. Preliminary
observations on biochemical relapse-free survival rates after
short course intensity-modulated radiotherapy (70 Gy at 2.5
Gy/fraction) for localized prostate cancer. Int J Radiat Oncol
Biol Phys 2002;53:904 –912.
8. Lukka H, Hayter C, Julian JA, et al. Randomized trial com-
paring two fractionation schedules for patients with localized
prostate cancer. J Clin Oncol 2005;23:6132– 6138.
9. Franklin CIV, Parker CA, Morton KM. Late effects of radia-
tion therapy for prostate carcinoma: The patient’s perspective
of bladder, bowel and sexual morbidity. Australis Radiol
1998;42:58 – 65.
10. Denham JW, O’Brien PC, Dunstan RH, et al. Is there more
than one late radiation proctitis syndrome? Radiother Oncol
1999;51:43–53.
11. American Society for Therapeutic Radiology and Oncology
Consensus Panel. Consensus statement: Guidelines for PSA
following radiation therapy. Int J Radiat Oncol Biol Phys
1997;37:1035–1041.
12. da Silva FC, Fossa SD, Aaronson HK, et al. The quality of life
of patients with newly diagnosed MI prostate cancer: Experi-
ence with EORTC clinical trial 30853. Eur J Cancer 1996;
32A:72–77.
13. Cheung R, Tucker SL, Lee A, et al. Dose response character-
istics of low and intermediate risk prostate cancer treated with
external beam radiotherapy. Int J Radiat Oncol Biol Phys
2005;61:993–1002.
14. Shipley WU, Thames HD, Sandler HM, et al. Prediction of
biochemical failure after radiation therapy for localized pros-
tate cancer. JAMA 1999;281:1598 –1604.
 Radiation therapy for prostate carcinoma
15. Bentzen SM, Ritter MA. The ␣/␤ ratio for prostate cancer:
What is it, really? Radiother Oncol 2005;76:1–3.
16. Brenner DJ. Fractionation and late rectal toxicity. Int J Radiat
Oncol Biol Phys 2004;60:1013–1015.
17. Sale C, Yeoh E, Scutter S, Bezak E. 2 D versus 3 D radiation
therapy for prostate carcinoma: A direct comparison of dose
volume histograms. Acta Oncol 2005;44:348 –354.
18. Dawson LA, Mah K, Franssen E, Morton G. Target position
variability throughout prostate radiotherapy. Int J Radiat On-
col Biol Phys 1998;42:1155–1161.
19. Zelefsky MJ, Crean D, Mageras GS, et al. Quantification and
predictors of prostate position variability in 50 patients eval-
uated with multiple CT scans during conformal radiotherapy.
Radiother Oncol 1999;50:225–234.
20. Dearnaley DP, Khoo VS, Norman AR, et al. Comparison of
radiation side effects of conformal and conventional radiother-
apy in prostate cancer: A randomised trial. Lancet 1999;353:
267–272.
21. Kirkwood BR, Sterne JAC. Relaxing model assumptions. In:
Essential medical statistics. 2nd ed. Oxford: Blackwell Sci-
ence; 2003. p. 341–354.
22. Livsey J, Routledge J, Burns M, et al. Scoring of treatment-
related late effects in prostate cancer. Radiother Oncol 2002;
65:109 –121.
23. Smit WGJM, Helle PA, van Putten WJJ, et al. Late radiation
damage in prostate cancer patients treated by high dose exter-
nal radiotherapy in relation to rectal dose. Int J Radiat Oncol
Biol Phys 1990;18:23–29.
24. Schultheiss TE, Hanks GE, Hunt MA, et al. Incidence of and
factors related to late complications in conformal and conven-
tional radiation treatment of cancer of the prostate. Int J
Radiat Oncol Biol Phys 1995;32:643– 649.
● E. E. YEOH et al. 1083
25. Yeoh E, Botten R, Russo A, et al. Chronic effects of
therapeutic irradiation for localised prostatic carcinoma on
anorectal function. Int J Radiat Oncol Biol Phys 2000;47:
915–924.
26. O’Brien PC, Franklin CIV, Poulsen MG, et al. Acute symp-
toms, not rectally administered sucralfate predict for late ra-
diation proctitis: Longer term follow-up of a Phase III trial—
Trans-Tasman Radiation Oncology Group. Int J Radiat Oncol
Biol Phys 2002;54:442– 449.
27. Fowler J, Chappell R, Ritter M. The prospects of new treat-
ments for prostate cancer. Int J Radiat Oncol Biol Phys
2002;52:3–5.
28. Fowler Harrari, Leborgne. Acute radiation reactions in oral
and pharyngeal mucosa: Tolerable levels in altered fraction-
ation schedules. Radiother Oncol 2003;69:161–168.
29. Orton CG. High dose rate brachytherapy may be radiobiolog-
ically superior to low dose rate brachytherapy due to slow
repair of late responding normal tissue cells. Int J Radiat
Oncol 2001;49:183–189.
30. Widmark A, Fransson P, Tavelin B. Self-assessment question-
naire for evaluating urinary and intestinal late side effects after
pelvic radiotherapy in patients with prostate cancer compared
with an age-matched control population. Cancer 1994;74:
2520 –2532.
31. Di Matteo A, Botten R, Holloway R, Fraser R, Yeoh E. 3 D
versus 2 D RT for carcinoma of the prostate reduces anorectal
motor dysfunction but not symptoms. Gastroenterology 2005;
128(Suppl):A265.
32. Yeoh E, Holloway R, Fraser R, et al. Anorectal dysfunction
increases with time following radiation therapy for carcinoma
of the prostate. Am J Gastroenterol 2004;99:361–369.