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

72 / 146

A.225
CHARACTERISTICS AND CRITICALITIES OF OCTAVIUS-4D DETECTOR
SYSTEM FOR IMRT AND RAPIDARC PLAN VERIFICATION
P. Urso *, R. Lorusso, L. Marzoli, D. Corletto, A. Pepe, P. Imperiale, M. Mira,
L. Bianchi. Ospedale di Circolo di Busto Arsizio, Busto Arsizio, Italy
Aim: Octavius-4D is a treatment plan quality assurance tool for IMRT/
VMAT verification. It is composed of a polystyrene cylindrical phantom,
rotating together with the gantry, coupled with a 729 ionization chamber
array. Beginning from the device commissioning, main response depen-
dences were studied.
Method and materials: Commissioning consisted of the assessment of static
and rotational field delivery, arc interruption recovery and intercalibration
with PTW-Semiflex ionization chamber. Clinical dose distributions of 30
treatment plans were evaluated. The sample included head/neck (13) and
pelvis (17), treated with IMRT (22) and RapidArc (RA) (8) techniques at the
Ospedale di Circolo di Busto Arsizio. 2D, 3D and volumetric γ-index (both
local and global) were compared, as well as their dependency on the mea-
surement plane orientation in the phantom.
Results: Commissioning indicated that high dose gradient areas are crit-
ically managed by the assembly, as confirmed by clinical plan results. The
effect was smoother in RA technique, where also time-resolution was co-
herent with planned gantry speed. 2D γ-index showed worse agreements
versus 3D and volumetric ones (p < 0.001), the last ones statistically com-
parable (p > 0.100). Volumetric plan evaluation was coherent with the passing
rates average on the three phantom planes (r > 0.9), but transversal section
provided best agreements versus sagittal and coronal ones (p < 0.050). No
differences emerged when comparing the two studied districts (p > 0.300).
Conclusion: Better agreements obtained with RA vs IMRT technique, due
to the larger number of time-related data in the algorithm interpolation,
confirmed Octavius-4D as specifically conceived for rotational delivery. If
a 2D evaluation is needed, results suggest that the transversal section better
fits the planned dose distribution grid. In addition to set-up and analysis
simplicity, pointed out strengths are time-resolved assessment and dosi-
metric evaluation on the whole volume.
A.226
COMPARISON OF DOSIMETRIC SYSTEMS IN RADIOTHERAPY PLAN
VERIFICATION
P. Urso *, D. Corletto, L. Marzoli, R. Lorusso, P. Imperiale, A. Pepe, M. Mira,
L. Bianchi. Ospedale di Circolo di Busto Arsizio, Busto Arsizio, Italy
Aim: Because of extremely modulated dose distributions in radiation
therapy, dosimetric plan verification with a simple, fast and reliable
system is a crucial step. Three different techniques were studied, i.e.
Gafchromic (EBT3) films in anthropomorphic phantom (CIRS ATOM®-
701), 729 ionization chamber matrix in cylindrical phantom rotating
together with the gantry (Octavius®-4D), and ionization chamber in
water phantom (PTW-MP1/MP3) with independent absolute point dose
(APD) measurements. The aim was to characterize the three approaches
and their relative agreement.
Methods and materials: The studied sample involved 30 treatment plans
for head/neck (13) and pelvis regions (17), delivered with intensity modu-
lated radiation therapy (IMRT) (22) and RapidArc (RA) (8) techniques on
two Varian linear accelerators (DHX and Trilogy) at the Ospedale di Circolo
di Busto Arsizio. Agreement between planned and delivered dose distri-
butions was studied by both local (Lγ) and global (Gγ) γ-index (3%/3 mm),
while APD in water was evaluated by the average of absolute errors (<3%).
Results: Lγ approach provided coherent plan evaluation results among the
three dosimetric measurement techniques, as passing rates were statisti-
cally comparable (p > 0.300) and rejected plans were also identified by APD.
Conversely, Gγ furnished not coherent results (p < 0.001), neither con-
firmed by the point dose evaluation. Results were generally comparable
for the two treated districts, while Octavius-4D showed different agree-
ments depending on delivery techniques.
Conclusion: Octavius-4D is characterized by practical advantages and a very
good response for RA technique. Distribution differences related to the con-
sidered anatomical regions do not affect the results of the studied
dosimetrical approaches, as comparable agreement was observed. The com-
parison of the selected methods suggests a more reliable plan verification
through the Lγ approach, as confirmed by APD measurements.
A.227
NEW EXPERIENCE WITH BREAST VOLUMETRIC MODULATION ARC
THERAPY (VMAT) REALIZED WITH TRILOGY LINAC: COMPARISON WITH
TOMOTHERAPY, PROS AND CONS
E. Vaccara * S. Agostinelli M. Bevegni F. Cavagnetto S. Garell
M. Gusinu M. Guenzi R. Corvo G. Taccini a Medical Physics, IRCCS
University Hospital San Martino – IST, Genoa, Italy; b Radiotherapy, IRCCS
University Hospital San Martino – IST, Genoa, Italy
Introduction: In the past years our Institute treated an average of 581 breasts
and most critical patients (5%) are voted to helical TomoTherapy. Now –
after Trilogy Varian Commissioning – we have two arc-based approaches
to deliver breast IMRT: helical and VMAT.
By summer 2015, we start to elaborate breast treatment plans with VMAT.
The aim of the present study is to compare VMAT Trilogy breast patients
– actually delivered – and TomoTherapy simulation on the same contour-
ing from physical and quality point of view, just to set up a criterion to
optimize the machines scheduling.
Materials and methods: TomoTherapy is usually a full 360° 6 MV photon
delivery; VMAT is a mix of two or more arcs of approximately 240 degrees
plus eventual directional blocks for ipsilateral lung or heart sparing. The
treatment qualities (6 or 6FFF MV) are selected according to the size of the
PTV.
Every breast plan is calculated with Eclipse Anisotropic Analytical Algo-
rithm and Progressive Resolution Optimizer (vs. 13.5) for Trilogy. For each
patient, a rival simulated TomoTherapy® Plan (vs. 2.0, Hi-Art 5.0) has being
elaborated. The work of comparison is in progress to focalize on beam-on
time, coverage of planning target volume and surrounding healthy tissue
sparing.
Results: Early results for breast VMAT indicate very short beam-on time
and optimal coverage and conformity index. The dose-volume histo-
grams (DVH) of the treated breasts always comply with the standard limits
of the literature. Moreover independent 3D dose verification with Octavius
4D (PTW) gives a mean γ (4 mm; 4%) index over 95%.
The first comparison with TomoTherapy Plan indicates a strong reduction
of beam-on time and of low dose ‘bath’.
Conclusions: Both VMAT and helical TomoTherapy are capable of provid-
ing highly uniform target dose and healthy tissue sparing; VMAT technique
is able to provide a consistent reduction in beam-on time as compared with
TomoTherapy and this is an undisputed advantage to reduce motion during
dose delivery.
A.228
MONTE CARLO-CORRECTED GTV DOSE PRESCRIPTION ON LUNG TUMORS
TREATED WITH CYBERKNIFE: A POSSIBLE RECIPE
A. Vai P. Bonfanti A.S. Martinott I. Redaelli F. Ri A. Berganti
a Centro Diagnostico Italiano, Progetto Diventerò Grantee Fond. Bracco, Milan,
Italy; b Centro Diagnostico Italiano, Cyberknife Center, Milan, Italy; c Centro
Diagnostico Italiano, Progetto Diventerò Alumnus Fond. Bracco, Milan, Italy
Introduction: Comparison of Ray-Tracing algorithm (effective path length
method, EPL) and Monte Carlo (MC) dose calculations in Cyberknife SBRT
treatments for lung lesions provides a possible recipe for a MC-corrected
dose prescription.
Material and Methods: 40 consecutive patients from July to October 2015,
treated with Cyberknife SBRT using an advanced target tracking system (lung
optimized treatment, LOT), were considered. Primary and metastatic pul-
monary lesions, different sizes (volume cutoff: V = 14 cc) and locations,
prescription dose and fractionation schemes were included. Treatment plans
were optimized using RT algorithm (RT), with prescription isodose line of
80% providing 95% PTV coverage (PTV = GTV + 5 mm), and re-calculated with
MC algorithm (1 × 1 × 1 mm3 dose grid, uncertainty=1%), using the same
beam angles and number of monitor units (MCrecalc).
e67 Abstracts/Physica Medica 32 (2016) e1–e70