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

8 / 146

Materials and methods: 25 patients were treated with FFF SBRT for lung
tumors with a Varian TrueBeam STx LINAC using VMAT. The lesions were
treated with single dose of 24 Gy. Two plans were created for each patient,
using Varian Eclipse treatment planning system, with and without the pres-
ence of flattening filter. Plans were compared and differences were analyzed
in terms of dose volume histograms (DVH), number of monitor units (MUs)
and beam on time.
Results: With FFF VMAT, the PTV conformity index was 1.03 ± 0.10.
Maximum doses to spinal cord, heart, esophagus and trachea were 2.9 ± 1.9,
0.8 ± 1.2, 3.3 ± 4.4 and 1.5 ± 1.7 Gy respectively. Average lungs V5, V20 and
mean doses were 14.6 ± 7.5%, 6.1 ± 3.7% and 1.1 ± 0.6 Gy. The average number
of MU was 7159 ± 609. No statistically significant differences were found
between the FFF and FF techniques with regard to dose indexes and MUs.
Beam delivery times were 6.7 minutes with FFF beams and 15.4 minutes
with flattening filter. The difference in beam on time was statistically sig-
nificant (t-Student test p < 0.05).
Conclusion: FFF and FF plans were equivalent in terms of coverage of the
PTV and doses to the main organ at risk (OAR). The use of FFF VMAT for
single fraction SBRT of lung cancer patients greatly reduces treatment de-
livery time compared to VMAT with flattening filter.
A.08
A COMPARISON BETWEEN DIFFERENT PATIENT QA DEVICES FOR IMRT
TREATMENTS ON VERO SYSTEM
A. Bazani F. Pansini C. Garibald S. Com E. Rondi G. Piperno
A. Ferrar B.A. Jereczek-Fossa F. Cattani a Department of Medical Physics,
European Institute of Oncology, Milan, Italy; b Department of Radiation
Oncology, European Institute of Oncology, Milan, Italy; c University of Milan,
Milan, Italy
Introduction: To compare the ability of OCTAVIUS 4D with 1000 SRS array
and ArcCHECK in detecting geometric and dosimetric errors intentionally
introduced into IMRT treatments delivered with VERO. The impact on the
DVH of PTVs and OARs was also investigated.
Methods and materials: From each of the original plans of 3 clinical cases
(prostate, partial breast irradiation-PBI and splenic lesion), 4 QA plans were
created with one intentional error each: gantry rotation (+3°), ring rota-
tion (+5°), MU (+2%) and isocenter translation (3 mm in caudal direction).
All plans were calculated with iPlan 4.5.3 (calculation grid: 2mm) on a math-
ematical phantom, for OCTAVIUS, and on the CT images (plug inserted), for
ArcCHECK.
The 3D γ evaluation method (3% local dose-3 mm and 2% local dose-
2 mm, 10% dose threshold) was applied, comparing the original calculated
distributions with the measured ones (with errors) using the related soft-
ware (VeriSoft, coronal projection, and SNC Patient, respectively).
An error was considered detected for γ failure rates > 5%.
The impact of the errors on the DVH of PTVs (D98%, D2% and Dmean),
rectum (D50% and D5%), ipsilateral lung (D40% and D10%) and spinal cord
PRV (Dmax) was evaluated. The Pearson’s correlation coefficient between
the γ passing rate variation and DVH variation was calculated.
Results: The results of the 3D γ evaluation show that the total detection
rate revealed by ArcCHECK was higher compared to OCTAVIUS (p = 0.045),
with 3%-3 mm criteria, while it was comparable with the 2%-2 mm crite-
ria (p = 0.480) (McNemar’s test).
In most cases the absolute value of the Pearson’s correlation coefficient was
<0.7.
Conclusion: A different sensitivity to errors was found, especially in the
case of ring and gantry rotations: this is due to the different dose recon-
struction methods used.
No significant correlation was found between the 3D γ analysis and the DVH
variations due to the intentional errors.
A.09
FDG-PET POSITIVE LYMPH NODE VARIATIONS DURING IMAGE-GUIDED
IMRT FOR HEAD AND NECK CANCER IDENTIFY NON-RESPONDING
PATIENTS
M. Belli I. Dell’Oca R. Raso F. Zerbetto A. Chiara G.M. Cattaneo
M. Picchio N. Di Muzio C. Fiorino R. Calandrino a Medical Physics, S.
Raffaele Scientific Institute, Milano, Italy; b Scuola di Specializzazione in Fisica
Medica, Università degli Studi di Milano, Milano, Italy; c Radiotherapy, S. Raffaele
Scientific Institute, Milano, Italy; d Scuola di Specializzazione in Radioterapia,
Università degli Studi di Milano Bicocca, Milano, Italy; e Nuclear Medicine, S.
Raffaele Scientific Institute, Milano, Italy
Introduction: We investigated the impact of FDG-PET positive lymph node
(PN) volume variation during radiotherapy on local control in head-and-
neck cancer patients.
Materials and methods: All patients (37 patients, T3–T4) were treated with
helical tomotherapy delivering (54 Gy, 66 Gy, 69 Gy in 30fr on precaution-
ary lymph nodal, primary and PET positive GTV, respectively, SIB approach).
Data on T, N and M relapses (rT, rN, rM) were collected for all patients. 42
PNs were manually contoured on weekly VCTs (fr10/20/30) by 3 experts.
Shrinkage and center-of-mass (CM) shifts were measured of PNs and the
residual error was estimated. The difference in PN volume changes, during
treatment, between patients with versus without relapses was tested (Mann–
Whitney test). The impact of shrinkage on the corresponding survival curves
(Cox proportional-hazard regression), divided between no/moderate versus
large shrinkage (based on ROC curve best cut-off value), was also
investigated.
Results: Volume reduction at the end of therapy and time trend CM shift
were significant for 27/37 and 14/37 patients respectively (vol, median:
71%, range: 27–94%, p = 0.93; CM, median: 5.1 mm at the end of treat-
ment, range: 1.0–8.9). 5 mm PTV margin results to be sufficient to cover
95% of all PN positions during treatment. Median follow-up was 27.4 m (3.7–
108.9) with 5, 4, 6 rT, rN, rM respectively. Differences in PN shrinkage were
found between patients with and without rT/rN at all considered timing
[fr20, rT: 0.56 vs 1.07 (median), p = 0.06; rN: 0.57 vs 1.25, p = 0.07]. Dif-
ferences were lower for rM. Survival curves provide high hazard ratios (HR)
between PN changes and rT/rN at all considered timing [fr20, rT: best cut-
off = 0.58, HR = 5.1 (95%CI 0.5–49.4), p= 0.12; rN: best cut-off= 0.98, HR = 14.9
(1.6–142.9), p = 0.01].
Conclusions: The residual error relative to PNs is relatively small. Poorer
outcome is associated with a limited shrinkage of PNs during the early treat-
ment phase. This information may help in developing rational strategies
for an early adaptation of the treatment.
A.10
DEVELOPMENT OF AN ADAPTIVE RADIOTHERAPY METHOD USING
PORTAL IMAGING AND CBCT
G. Benecchi L. Bonom F. Palleri E. Calabri R. Rossi G. Ceccon
C. Ghetti a Medical Physics Department, AOU Parma, Parma, Italy; b University
of Bologna, Bologna, Italy; c Radiotherapy Department, AOU Parma, Parma, Italy
Introduction: The aim of this study is to develop an innovative method
of adaptive radiation therapy that can be applied to patients treated with
VMAT technique.
Materials and methods: This study consists of two phases: in the first phase,
two types of analysis are performed on portal images, in particular recon-
struction of the dose at the isocenter and 2D gamma analysis. If the outcome
of this analysis is within the tolerance criteria, the control procedure will
be repeated again after some time. If at least one of the analyses fails, it is
necessary to proceed with the second phase of the method to investigate
the error highlighted in the first phase: a patient cone-beam CT is ac-
quired and a calibration in electron density is performed in order to have
a correct calculation of dose from TPS on CBCT. Meanwhile, the physician
performs the contouring operation of the structures of interest.
The treatment plan is calculated by TPS on CBCT and the 3D gamma anal-
ysis is performed on dose matrices calculated from the planning CT and
CBCT. Through a previous structure extraction and its registration on dose
matrices, it is possible to provide 3D gamma analysis indexes on PTV, CTV
and OAR of clinical interest.
e3 Abstracts/Physica Medica 32 (2016) e1–e70