Physica Medica: European Journal of Medical Physics - Volume 32 - Supplement 1 - February 2016

33 / 146

Conclusions: The IVD results showed that our CBCT protocol (five consec-
utive CBCT scans in the first week of treatment and a CBCT once a week
in the following weeks) is not completely adequate. Small random varia-
tions due to residual errors in the set-up verification or organ motion may
occur involving R value larger than ±5%.
A.93
PATIENT’S BREATHING DURING PROSTATE RADIOTHERAPY SESSIONS: IS
IT REALLY A SECOND ORDER EFFECT?
T. Giandini C.M.V. Panaino M. Carrara S.C. Frasca B. Avuzzi
S. Morlino D. Bosett N. Bedini S. Villa T. Rancati D. Bettega
R. Valdagni E. Pignoli a Fondazione IRCCS Istituto Nazionale dei Tumori,
Milano, Italy; b Università degli Studi di Milano, Milano, Italy
Introduction: An electromagnetic (ELM) system (Calypso, Varian Medical
System) based on high frequency localization of transponders (TRN) per-
manently implanted in the prostate (PR) allows to track its motion with
sub-millimeter accuracy. The aim of this work was to use this ELM system
to quantify the amplitude of PR motion induced only by patient’s breath-
ing and to find possible correlations between breathing frequency (BF) and
PR excursions.
Materials and methods: A software (SW) was developed to analyse and
filter the signal produced by the ELM TRN during external beam radio-
therapy (EBRT) sessions of implanted PRs. The SW was first validated with
data recorded with QUASAR moving phantom, provided with a home-
made insert with TRN. BFs between 10 and 24 cycles/min were simulated.
After validation, the tracks of 8 patients who underwent PR EBRT were
analysed for a total of 248 treatment sessions. For each session, the cor-
responding PR maximum total excursion along the three main directions
was obtained. Possible correlations between BF and PR motion amplitude
were analysed.
Results: For the in-phantom validation, the developed SW automatically
computed the correct cycles/min within a 0.5% uncertainty. BFs ranged
between 13.6 and 26.2 cycles/min, with a very limited intra-patient stan-
dard deviation. The PR total excursions over all treatment sessions induced
only by breathing ranged between 0.06–0.50 mm, 0.10–0.62 mm and 0.22–
0.74 mm for the L-R, A-P and C-C directions, with very low intra-patient
variability. No significant correlations were found between BF and PR motion
amplitude.
Conclusions: Although apparently negligible if compared to other sources
of intra-session PR localization uncertainties, these preliminary results show
that the PR motion associated to patient’s breathing could be considered
to improve tracking-based delivery technologies. Very limited intra-
patient standard deviations of both BFs and PR excursions suggest possible
prospective modelling of patient-related PR motion.
Acknowledgement: This work was partially funded by AIRC (IG-14300).
A.94
DOSIMETRIC IMPACT OF CONTOURING VARIABILITY FOR PROSTATE
CANCER PATIENTS UNDERGOING COMBINED CARBON IONS/PHOTONS
TREATMENT (AIRC-IG 14300)
T. Giandini D. Ciardo M. Carrar S. Russ F. Pansini A. Bazani
N. Bedini S.P. Colangion S. Morlino B. Vischion D. Bosett D. Zerini
B. Avuzzi A. Hasegawa F. Cattani E. Pignoli F. Valvo
B.A. Jereczek-Fossa R. Valdagn R. Orecchia a Istituto Europeo di
Oncologia, Milano, Italy; b Fondazione IRCCS Istituto Nazionale dei Tumori,
Milano, Italy; c Fondazione CNAO, Pavia, Italy; d NIRS – National Institute for
Radiological Sciences, Chiba, Japan; e Università degli Studi di Milano, Milano,
Italy
Introduction: Within a research project dealing with carbon ion boosts for
high risk prostate cancers, involving two centres for radiotherapy and one
for hadron-therapy, a planning study was performed to evaluate the do-
simetric impact of contouring variability (CV) between centres.
Materials and methods: To evaluate CV, a dummy-run on 5 patients was
already performed by 7 radiation oncologists (RO) of the 3 centres in-
volved in the project. Out of these, one emblematic patient was chosen for
this dosimetric study. The single structures delineated by the 7 RO were
used to generate mean structures (MS): a voxel belonged to a MS if se-
lected by more than 3 RO. On the MS, each centre planned a VMAT treatment
for prostate (PTV-P), seminal vesicles (PTV-VS) and pelvic lymph nodes (PTV-
N), whereas a boost with ions was planned for PTV-P. Common planning
objectives were V95% > 95% for all PTVs, V95% < 2 cc for rectum and
V95% < 10 cc for bladder. The single dosimetric parameters, in particular
D98% as minimum dose index, were then analysed by superimposing on
single structures the plans obtained on MS.
Results: All plans satisfied the planning objectives for the MS. VMAT plans
showed comparable dose distributions. On single PTV-VS and PTV-N, V95%
and D98% dropped on average by 5% and 3% (max 17% and 14%) and by
5% and 7% (max 9% and 14%), respectively. On single PTV-P, V95% and D98%
dropped on average by 1.5% and 2% (max 4.5% and 5%) for VMAT and by
4.5% and 30% (max 8% and 75%) for ions, respectively. Small variations of
V95% both for VMAT and ions were observed for single bladders, whereas
rectal V95% increased more with VMAT than with ions (max 5 cc vs <1 cc).
Conclusions: On average, target CV showed a limited dosimetric impact
for VMAT. For ions, instead, the minimum dose to single PTV-P was dra-
matically reduced, whereas high doses to rectum resulted less sensitive to
CV.
These preliminary results suggest that further efforts should be done to
reduce the variability of PTV-P contouring.
Acknowledgement: This work was partially funded by AIRC IG-14300.
A.95
CLINICAL IMPLEMENTATION OF BREAST IMRT
L. Gianusso *, F. De Monte, U. Nastasi. SSCVD Fisica Sanitaria Citta’ della Salute
e della Scienza Presidio San Giovanni Antica Sede, Torino, Italy
Introduction: Breast irradiation in radiotherapy is of great interest for the
complex geometry of the target and the presence of adjacent critical struc-
ture. The standard irradiation technique consists of two tangential opposite
photon beams, but in the same cases it is not enough to respect the ob-
jectives of the treatment: deliver dose prescribed to tumour and spare OARs
maintaining an acceptable treatment time.
In this work IMRT step and shoot for breast was implemented with the fol-
lowing objectives:
1) identification of an IMRT that can be the right compromise between
dosimetric objectives and treatment time
2) comparison between IMRT inverse planning (IP) and forward plan-
ning (FP).
Materials and methods: In the breast IMRT two modality of planning are
used in our Centre, FP and IP depending on position and geometry of target,
OARs and patient contour.
In FP, also called field in field (FF) technique, the planner manually adds
one or two segments inside the standard tangent fields, while in IP, the
planner adds 2 or more fields with different orientations with respect to
standard and the TPS makes a segmentation of all beams.
In our hospital, FF is in use since 2006 while IP was not available yet.
In the present work, we study 4 irradiation techniques for breast with si-
multaneous boost: FF, IMRT Hybrid and IMRT with 4 and 5 fields.
Results: Seven patients were enrolled for this study. We calculated the con-
formation number: CN =(TVri)^2/(Vri × TV) where TV is the target volume
and ri is the reference isodose. The results obtained are reported in a cu-
mulative histogram in which it can be see that techniques IP have a higher
CN score than FF (5 fields is the best).
Conclusion: In this study IMRT for breast was implemented and four ir-
radiation techniques were compared.
Techniques IP show a better CN, while volumes of OARs that receive low
doses are lower in FF.
Patients enrolled are still too low for definitive conclusions but more robust
statistic will give us the reliability of these preliminary results.
e28 Abstracts/Physica Medica 32 (2016) e1–e70