Combination of celecoxib with percutaneous radiotherapy in patients with localised prostate cancer - a phase I study.

Abstract

Current approaches for the improvement of bNED for prostate cancer patients treated with radiotherapy mainly focus on dose escalation. However molecularly targeted approaches may also turn out to be of value. In this regard cyclooxygenase (COX)-2 inhibitors have been shown to exert some anti-tumour activities in human prostate cancer in vivo and in vitro. Although in vitro data indicated that the combination of COX-2 inhibition and radiation was not associated with an increased toxicity, we performed a phase I trial using high dose celecoxib together with percutaneous radiation therapy.
In order to rule out any increases of more than 20% incidence for a given side effect level 22 patients were included in the trial. Celecoxib was given 400 mg twice daily with onset of the radiation treatment. Risk adapted radiation doses were between 70 and 74 Gy standard fractionation. RTOG based gastrointestinal (GI) and genitourinary (GU) acute toxicity scoring was performed weekly during radiation therapy, at six weeks after therapy and three month after completing radiation treatment.
Generally no major increase in the level and incidence of side effects potentially caused by the combined treatment was observed. In two cases a generalised skin rash occurred which immediately resolved upon discontinuation of the drug. No grade 3 and 4 toxicity was seen. Maximal GI toxicity grade 1 and 2 was observed in 85% and 10%, respectively. In terms of GU toxicity 80% of the patients experienced a grade 1 toxicity and 10 % had grade 2 symptoms.
The combination of irradiation to the prostate with concurrent high dose celecoxib was not associated with an increased level of side effects.

Full text

Page 1

BioMed Central
Page 1 of 10
(page number not for citation purposes)
Radiation Oncology
Open Access
Research
Combination of celecoxib with percutaneous radiotherapy in
patients with localised prostate cancer – a phase I study
U Ganswindt1, W Budach2, V Jendrossek1, G Becker3, M Bamberg1 and
C Belka*1
Address: 1CCC Tübingen, Centre for Genitourinary Oncology, Department of Radiation Oncology, University of Tübingen, Tübingen, Germany,
2Department of Radiation Oncology, University of Düsseldorf, Düsseldorf, Germany and 3Department of Radiation Oncology, Klinik am Eichert,
Göppingen, Germany
Email: U Ganswindt - ute.ganswindt@med.uni-tuebingen.de; W Budach - wilfried.budach@uni-duesseldorf.de;
V Jendrossek - verena.jendrossek@uni-tuebingen.de; G Becker - radioonkologie@KaE.de; M Bamberg - michael.bamberg@med.uni-
tuebingen.de; C Belka* - claus.belka@uni-tuebingen.de
* Corresponding author
Abstract
Background: Current approaches for the improvement of bNED for prostate cancer patients
treated with radiotherapy mainly focus on dose escalation. However molecularly targeted
approaches may also turn out to be of value. In this regard cyclooxygenase (COX)-2 inhibitors have
been shown to exert some anti-tumour activities in human prostate cancer in vivo and in vitro.
Although in vitro data indicated that the combination of COX-2 inhibition and radiation was not
associated with an increased toxicity, we performed a phase I trial using high dose celecoxib
together with percutaneous radiation therapy.
Methods: In order to rule out any increases of more than 20% incidence for a given side effect
level 22 patients were included in the trial. Celecoxib was given 400 mg twice daily with onset of
the radiation treatment. Risk adapted radiation doses were between 70 and 74 Gy standard
fractionation. RTOG based gastrointestinal (GI) and genitourinary (GU) acute toxicity scoring was
performed weekly during radiation therapy, at six weeks after therapy and three month after
completing radiation treatment.
Results: Generally no major increase in the level and incidence of side effects potentially caused
by the combined treatment was observed. In two cases a generalised skin rash occurred which
immediately resolved upon discontinuation of the drug. No grade 3 and 4 toxicity was seen.
Maximal GI toxicity grade 1 and 2 was observed in 85% and 10%, respectively. In terms of GU
toxicity 80 % of the patients experienced a grade 1 toxicity and 10 % had grade 2 symptoms.
Conclusion: The combination of irradiation to the prostate with concurrent high dose celecoxib
was not associated with an increased level of side effects.
Published: 10 April 2006
Radiation Oncology2006, 1:9doi:10.1186/1748-717X-1-9
Received: 25 November 2005
Accepted: 10 April 2006
This article is available from: http://www.ro-journal.com/content/1/1/9
© 2006Ganswindt et al; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 2

Radiation Oncology 2006, 1:9 http://www.ro-journal.com/content/1/1/9
Page 2 of 10
(page number not for citation purposes)
Background
Prostate cancer is the most common malignant tumour in
men. At present, approximately 200.000 new diseases are
diagnosed per year in the USA leading to the death of
more than 30.000 patients. Due to the increased use of
PSA screening the number of patients diagnosed in local-
ised disease is rising strongly. Radical prostatectomy, per-
cutaneous radiotherapy and interstitial radiation methods
are available for curative treatment of localised stages.
Due to a lack of randomised studies, the optimal treat-
ment is still unclear. Based on the available data, however
it seems likely that all given methods are more or less
equivalent in terms of tumour control. Side effects in the
rectum predominate with percutaneous radiotherapy,
while mainly impotence and incontinence are seen after
prostatectomy [1].
Nevertheless, a crucial problem is still unsolved. The long
natural history of prostate cancer makes it difficult to
determine which type of local therapy is best in men with
life expectancies longer than 8–10 years at diagnosis. In
this regard, long-term follow-up data with overall survival
as endpoint and meticulous determination of side effects
will finally answer the question whether there is an opti-
mal therapy for localised prostate cancer.
Local control rates (defined as biochemically relapse-free
five-year survival) between ~ 50 and ~ 90% can be
achieved with percutaneous irradiation for localised
stages. All available data indicate the existence of a clear
dose-effect relationships for pathological control as well
as bNED [2-9]. Hence, strategies for increasing the radia-
tion dose are an important goal when trying to optimise
the outcomes after radiotherapy. In order to increase the
dose, intensity-modulated radiotherapy or particle based
therapy approaches are currently under investigation [10-
16].
In addition to an increased radiation dose, the blockade
of testosterone action was found to be an effective meas-
ure for improved radiation treatment results [17-20].
To further optimise the efficacy of radiation treatments,
molecular targeted approaches are currently under inves-
tigation [21].
Of special importance are drugs targeting tyrosine recep-
tor associated kinase pathways (EGF-R, VEGF-R, IGF-R)
downstream kinase molecules, and cell death signalling
pathways [22-26]. Beside this, numerous reports under-
line the importance of prostaglandin signalling during
cancer development and growth [27-30]. In addition it
has been suggested that the modulation of prostaglandin
generation may alter treatment responses towards chemo-
therapy and radiation [31-34].
A key enzyme involved in prostaglandin synthesis is the
inducible cyclooxygenase-2 molecule which is frequently
found to be overexpressed in human cancer cells, whereas
in non-malignant tissues COX-2 is predominantly found
in association with inflammatory processes [35-37]. The
development of selective COX-2-inhibitors thus theoreti-
cally allows a tumour specific response modulation.
Based on these findings, COX-2 inhibitors were shown to
be effective in patients with FAP, where the number of
polyps is strongly reduced when patients received 2 × 400
mg celecoxib per day. Importantly lower doses had less
effects on the development of adenomas [38].
Although the inhibition of the COX-2 enzyme by
celecoxib is important for the understanding of its effi-
cacy, several data suggest that celecoxib may exert non-
COX-related effects in cancer cells [39-43]. In this regard,
Waskewich [44] showed that celecoxib induces clono-
genic cell kill with similar IC50 values irrespectively of the
COX-2 expression status. Although the mechanisms of the
non-COX-dependent action of celecoxib are not com-
pletely understood, several data suggest that they are
related to the fact that celecoxib triggers a new apoptosis
mitochondrial apoptosis pathway or interferes with PKB
AKT signalling. Especially the pro-apoptotic effect was
found to require doses higher than needed for an inhibi-
tion of the regular target enzyme. In this regard the data
on FAP suppression are important, since there was a clear
dose response relationship above the anti-inflammatory
dose level.
Although celecoxib seems to be active alone, several
groups provided evidence that the drug is considerably
more effective when combined with a second anti-tumour
treatment option. A comparative study in animals showed
that the combination of radiotherapy with COX-2 inhibi-
tors produces a clearly improved response rate when com-
pared to radiotherapy alone. The TCD 50 values (FSA
sarcoma xenograft) were found to be halved in case of a
combined treatment [45,46].
Antitumour activities of COX-2 inhibitors have been
shown for various human malignant tissues including
colorectal [38,47,48], breast [29,49], non small cell lung
[50,51] and other epithelial cancers [42,52-54].
Therefore the role of the combination of an COX-2 inhib-
itor with other treatment modalities has mainly been
tested in lung cancer, cervical cancer, head and neck can-
cer and colorectal cancers.
Several lines of evidence point to a role of COX-2 inhibi-
tion as treatment approach for prostate cancer [39,43,55-
61] (table 1). Histological analysis of prostate carcinoma

Page 3

Radiation Oncology 2006, 1:9 http://www.ro-journal.com/content/1/1/9
Page 3 of 10
(page number not for citation purposes)
cells revealed an overexpression of COX-2 in tumour tis-
sue when compared to normal prostate stroma or benign
prostatic hyperplasia [59].
COX-2 contributes to the proliferation of prostate cancer
cells, while COX-2 inhibitors were clearly shown to
inhibit proliferation and to induce apoptosis [60].
In the setting of hormone refractory prostate cancer the
application of celecoxib in patients was associated with
some partial PSA responses [62]. Likewise in patients with
biochemical relapse after definitive therapy a significant
inhibition of serum PSA levels 3 months after treatment
with celecoxib was observed [63].
Furthermore, it could be shown in vitro that irradiation of
PC-3 cells triggers an increase in COX-2 expression [64].
In own studies, the combination of celecoxib with ionis-
ing radiation revealed an additive effect on cell kill in PC-
3 and DU-145 cells [65].
Based on murine data the combination of celecoxib with
irradiation seems not critical regarding toxicity [45]. How-
ever, recent clinical data suggest that at least in an multi-
modality setting the addition of celecoxib to a chemoradi-
otherapy protocol may be associated with increased toxic-
ity rates [66].
In order to rule out any safety concerns of a combination
of celecoxib with irradiation we prospectively determined
the toxicity of such an combination in prostate cancer
using the highest Food and Drug Administration-
approved dose of 800 mg celecoxib daily.
Methods
Aim of the study
Aim of the study was to determine the acute toxicity of a
celecoxib administration during percutaneous radiother-
apy of localised prostate cancer. The primary endpoint of
the study was the incidence of acute toxicity (up to three
months after therapy).
Inclusion and exclusion criteria
Patients with histologically proven prostate cancer, stages
cT1-cT3 cN0 cM0, G1-3, PSA ≤ 20 ng/ml, age up to 75
years and Karnofsky Index ≥ 80 %, were included after
providing informed consent. Further inclusion criteria
were normal levels of hemoglobin, leukocytes, platelets,
creatinine, urea, GGT, AP, AST, ALT, bilirubine, creatinine
clearance > 50 ml/min and no other clinically leading sec-
ondary disease. Any other NSAIDs were not allowed with
the exception of acetylsalicylic acid at a cardioprotective
dose. Patients after transurethral resection or prostatec-
tomy and patients with a known contraindication (e.g.
gastric ulcer) or allergy to COX-2 inhibitors were
excluded. Further exclusion criteria were severe heart, car-
diovascular, liver, renal, inflammatory intestinal or blood
coagulation disorders, collagenoses, former irradiation of
the prostate, secondary malignancies (exception non-
melanotic skin cancer) and regular intake of lithium or
fluconazole.
Staging examinations
The pre-therapeutic staging examinations included the
initial PSA value, biopsy with histological confirmation
and statement of the grading or Gleason score, rectal dig-
ital examination, transrectal endosonography and at least
Table 1: Overview on the available mechanistic data regarding the activity of coxibes in prostate cancer
Cancer TypeTreatmentInvestigationResultsReference
LNCaP PC 3CelecoxibIn vitroIncreased cell death/
apoptosis
G1 block/reduced DNA
synthesis/growth inhibition
by COX-2 independent
mechanism
Growth inhibition
Induction of apoptosis by
blocking Akt activation
independently of Bcl-2
Up-regulation of COX-2,
elevated PGE2 levels after
irradiation
Bax-independent pro-
apoptotic effect of
Celecoxib
Augmentation of
chemotherapeutic drug-
induced apoptosis by
activation of caspase 3 and
9
Kamijo 2001
PC 3CelecoxibIn vitro/Xenograft
Patel 1999
LNCaP PC 3
LNCaP PC 3
Celecoxib
Celecoxib
In vitro
In vitro
Srinath 2003
Hsu 2000
PC 3Celecoxib+ radiationIn vitro
Steinauer 2000
LNCaP PC3 DU-145Celecoxib+ radiation In vitro Handrick 2004
LNCaP DU-145 PC-3ML Celecoxib+ COL-3/
Docetaxel
In vitro/Xenograft
Dandekar 2005

Page 4

Radiation Oncology 2006, 1:9 http://www.ro-journal.com/content/1/1/9
Page 4 of 10
(page number not for citation purposes)
pelvic sonography, alternatively computed tomography
(CT) or magnetic resonance imaging (MRI), to evaluate
the lymph nodes. At PSA levels > 10 ng/ml a bone scintig-
raphy was mandatory.
Treatment course
All patients were treated with celecoxib 400 mg twice daily
in an open-label, unblinded trial during the entire series
of radiation. The intake of celecoxib was started on the
first day of radiotherapy, continued also on radiation-free
days (e.g. weekends) and stopped on the last day of radi-
otherapy. Celecoxib medication was discontinued, if a
patient developed ≥ grade 3 toxicity. The percutaneous
radiotherapy was planned with a three-dimensional (3D)
radiation planning system based on computed tomogra-
phies in supine position. A rectal balloon filled with 40 ml
of air was used in order to spare the posterior wall of the
rectum and for fixation of the prostate [67]. An additional
3D radiation planning without the rectal balloon was per-
formed simultaneously for use in case of non-tolerance of
the balloon. We used a conformal, isocentric 4-field tech-
nique with 15 MV photons. Target volume and dose con-
cept depended on a risk classification based on the
prognostic factors stage, grading and initial PSA level. The
patients received 5 × 2.0 Gy per week up to 70.0 Gy and
74.0 Gy cumulative dose, respectively. The planning target
volume (PTV) included the risk dependent clinical target
volumes (table 2) with a safety margin of 10 mm (with
rectal balloon) and 12 mm (without balloon), respec-
tively. The patients with a high risk of relapse treated with
74.0 Gy cumulative dose received a boost of 8 Gy with a
dorsal safety margin of 5 mm followed by 66 Gy as
described above. As organ at risk the whole rectum from
anal sphincter to the location where the rectum turned
horizontally into the sigmoid colon was defined. The
given radiotherapy doses were prescribed in line with
ICRU Report No 50 and the given volumes complied with
the definitions of ICRU Report No 62. Additional hor-
mone therapy could be freely used as part of the study.
Laboratory measurements
The creatinine clearance was examined prior to inclusion
into the study. Prior to treatment start, at week 2, 4, 6 of
the combined therapy and 3 months after the end of treat-
ment blood samples were taken. The measurements
included a blood count, coagulation parameters and
serum levels of electrolytes, creatinine, urea, GGT, AP,
AST, ALT and total bilirubine. PSA levels were measured
prior to treatment start and after three months.
Measurement of acute toxicity
Acute toxicity according to RTOG criteria (gastrointesti-
nal, genitourinary) was acquired at least once weekly dur-
ing the 7–8 week series of radiation treatment, 6 weeks
and 3 months after treatment. Beside the clinical examina-
tion documented on case report forms we used a stand-
ardised questionnaire that had to be filled by the patients
at the same time. Beside acute gastrointestinal and geni-
tourinary toxicity according to RTOG criteria any other
acute toxicity was described on the case report form. Late
toxicity is further ascertained as part of the radiotherapeu-
tic follow-up examination outside the study once a year.
Criteria for discontinuation/statistics
The acute toxicity data published by Storey et al. [68] with
cumulative doses from 70 to 78 Gy were the reference
basis for the toxicity to be anticipated in our study. The
study was powered to exclude an > 20% increase in the
incidence of grade 3 and 4 acute GI and GU toxicity.
Derived from these conditions the following criteria to
close the study prematurely were defined: If no grade 4
acute toxicity would occur in 20 patients, the 95% confi-
dence interval is 0 to 16.8%. The study would then be dis-
continued, because at 95% safety acute toxicity of 20% or
more could be ruled out. If exactly one grade 4 acute tox-
icity would occur, the 95% confidence interval is 0.1 to
24.9%. The sample size has then to be increased by further
15 patients. If there would remain just one case of grade 4
acute toxicity, the 95% confidence interval is 0.1 to
14.9%, with one further case 0.7 to 19,2%, i.e. it would
not include the critical value of 20%. If two cases of grade
4 toxicity would occur in the first 20 patients, further 15
patients would be recruited. In case of no further grade 3
or 4 toxicity, the 95% confidence interval is 0.7 to 19.2%.
If at least three cases of grade 4 toxicity would occur in just
the first 20 patients, the study would be discontinued.
Table 2: Target volume and dose concept depending upon stage, grading and PSA
Low risk: white Medium risk: light grey High risk: dark grey
Stage
PSA
G 1 Gleason 2–3
≤ cT2a
≤ 10
Prostate
≤ cT2c
< 20
cT3
< 20
Prostate & base of seminal vesicles
70 Gy
Prostate & base of seminal vesicles
70 Gy
Prostate & base of seminal vesicles
70 Gy
Prostate & base of seminal vesicles & visible
tumour 74 Gy
Prostate & base of seminal vesicles & visible
tumour 74 Gy
Prostate & base of seminal vesicles & visible
tumour 74 Gy
G 2 Gleason 4–6
Prostate
G 3 Gleason > 6
Prostate & base of seminal vesicles
70 Gy

Page 5

Radiation Oncology 2006, 1:9 http://www.ro-journal.com/content/1/1/9
Page 5 of 10
(page number not for citation purposes)
Even when treating additional 15 patients the predefined
acceptable toxicity level would have been exceeded.
Results
Patient characteristics
From 06/2003 to 07/2004 22 patients were included into
the study. All 22 patients completed the radiotherapy
without treatment break. In all cases the intake of
celecoxib started at the first day of radiotherapy in the
morning. Within 2 weeks after commencing treatment 2
of the 22 included patients displayed a general exanthema
with pruritus. Medication was stopped immediately and
the skin rash resolved completely afterwards. Therefore we
assumed that this reaction was a drug allergy. Both
patients were excluded from the trial. The other 20
patients completed the treatment according to the study
protocol with 400 mg celecoxib twice daily. 5 patients
received 74 Gy cumulative dose, 14 patients received 70
Gy cumulative dose and 1 patient was treated with 72 Gy.
Median age was 67 years (range 49 – 74 years); median
initial PSA-level was 8 ng/ml (range 2,4 – 18,3 ng/ml). 14
patients received hormone ablative therapy (table 3),
mostly started before and continued concurrently to radi-
otherapy. The rectal balloon was tolerated well, 2 patients'
radiotherapy treatment was continued without rectal bal-
loon after 40 and 46 Gy, respectively. The resulting dose-
volume-histograms of the rectum are shown for all
patients in figure 4.
Acute gastrointestinal and genitourinary toxicity
No gastrointestinal or genitourinary acute toxicity grade 3
or 4 (RTOG) occurred. Thus we finished patient recruit-
ment after complete treatment of 20 patients. 17 of 20
patients showed a gastrointestinal acute toxicity grade 1. 2
of 20 patients showed a gastrointestinal acute toxicity
grade 2. Most frequent grade 1 symptom was mild rectal
discomfort. Among he 2 patients with grade 2 gastrointes-
tinal toxicity 1 patient had diarrhoea and the other patient
required mild analgetics for his rectal symptoms (figure
1).
In 16 of 20 patients we observed a genitourinary acute tox-
icity grade 1, in 2 of 20 patients a genitourinary acute tox-
icity grade 2. Most frequent grade 1 symptom was slight
dysuria. Among the 2 patients with grade 2 genitourinary
toxicity 1 patient had bladder spasms, the other patient
presented with a bacterial cystitis 3 weeks after radiother-
apy, which completely resolved after treatment with ade-
quate antibiotics (figure 2).
Other toxicity
Considering the acute skin toxicity we observed 2 patients
with a grade 2 toxicity (circumscribed moist desquama-
tion measuring 1–2 cm per patient), 8 patients with a
grade 1 toxicity and 10 patients with no toxicity at all (fig-
ure 3). Based on the clinical examinations, the taken
blood samples and the questionnaires filled by the
patients we observed no other acute toxicity. With excep-
tion of the 2 patients described above who developed a
drug allergic reaction no cardiovascular, gastric, renal,
hepatic or bone marrow side effect of celecoxib occurred.
Discussion
Several approaches for the improvement of bNED in the
radiotherapeutic treatment of localised prostate cancer
were tested. Current strategies mainly focus on dose esca-
lation. In this regard, new radiation technologies for
example IMRT allow the application of high radiation
doses without increasing the toxicity. In addition, the
combination with hormonal treatment has been proven
to be suitable to increase local control and biochemical
relapse-free interval rates. The results of four major trials
[18-20], [69-72] revealed that a combined treatment is
advantageous for intermediate and high risk patients.
Patients with an intermediate risk profile benefit both
from radiation dose escalation and additional hormonal
treatment, even if there is no clear cut recommendation
regarding starting time and duration of hormonal treat-
ment for intermediate risk patients. However molecularly
targeted approaches may also turn out to be of value. In
this regard, preclinical studies suggest that COX-2 inhibi-
tors have an certain anti-tumour activity when given alone
and are even more active when combined with classical
anti-tumour treatment.
In case of prostate cancer, a clear dose response relation-
ship exists for the endpoint local control and bNED espe-
cially in patients with a low or intermediate risk profile.
Although in vitro data indicated that there is no increased
toxicity when COX-2 inhibitors are combined [45] with
radiation, there are few clinical data concerning the toxic-
Table 3: Patients Characteristics
CharacteristicsNo. of patients
Age
< 67
> 67
10
10
T-Stage
1c – 2a
2b – 2c
3
Initial PSA
≤ 10 ng/ml
> 10 ng/ml
Gleason Score
≤ 6
≥ 7
Hormonal ablation
Yes
No
12
6
2
13
7
13
7
14
6