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

7 / 146

respectively. The sensitivity of IQM is suitable to detect all types of error
and a good correlation is observed with other metrics.
Conclusion: IQM provides optimal performances in terms of signal repro-
ducibility of complex IMRT plans. It also shows sensitivity sufficient for
monitoring delivery errors in clinical practice and the good correlation ob-
served with other metrics suggests that IQM signal could be used to quantify
delivery errors.
A.04
IN VIVO DOSIMETRY WITH DIODE AND SETUP VERIFICATION THROUGH
GAF CHROMIC RTQA IN IORT FOR BREAST IRRADIATION
B. Augelli E. Roberto a A.U.S.L. Umbria 2 U.O. Radioterapia Oncologica,
Spoleto, Italy; b A.S.O. Santa Croce E Carle S.S.D Fisica Sanitaria, Cuneo, Italy
Introduction: 105 patients with different diseases (breast, stomach, pan-
creas and rectum) have been treated by IORT. In the breast cancer IORT
treatment, the organ with higher risk is the lung so we have estimated the
dose to the lung by in vivo dosimetry (IVD) using diodes and the align-
ment of the disk protection with applicator by GAF chromic RTQA.
Materials and methods: The treatment of patients with breast cancer has
been performed by giving a single dose of 10 Gy (boost treatment) or 21 Gy
(exclusive treatment). We use a LIAC by Sordina with electron beams of
6, 8, 10 and 12 MeV, applicator with a diameter from 3 cm to 10 cm and
shielding protection disk made of steel–PTFE (3 + 3 mm). We have char-
acterized the diode checking, at reference condition, the repeatability,
reproducibility, linearity; we have determined the calibration factor using
RW3 phantom by intercomparison with Roos or Markus chamber for dif-
ferent energy of clinical use (6, 8 and 10 MeV). The verification of the
alignment of the applicator with protection disk has been performed with
GAF chromic RTQA. The surgeon places the film and then the diode under
protection disk. Finally, he reconstructs the mammary gland.
Results: The error in diode calibration is 3.7% for all energies. We have ana-
lyzed the data of 10 patients treated with 10 Gy and 2 patients treated with
21 Gy. The results obtained from IDV diode-based compared with exper-
imental measurements allow to evaluate an error of 13%. However the
average dose to the lung for treatment boost is less than 30 mGy and for
exclusive treatment it is less than 85 mGy. Measurements performed with
GAF chromic films showed an average setup error equal to 0.9 cm.
Conclusions: The dose delivered to the chest wall, measured by diode, allows
to estimate a dose to the lung less than 1% of the prescribed one. As for
the error setup, it is possible to suggest to enlarge the margin required for
entry into the cavity of a disk 2 cm in diameter greater than the diameter
of the applicator.
A.05
A DOSE-VOLUME PARAMETER ANALYSIS OF THE BLADDER IN THREE-
DIMENSIONAL CONFORMAL RADIOTHERAPY FOR LOCALIZED PROSTATE
CANCER
P. Bagalà G. Ingross S. Petrichell M. D’Andre M. Rago A. Lancia
C. Brun E. Ponti R. Santon M.D. Falc a Tor Vergata University General
Hospital, Roma, Italy; b Campus Bio-Medico University, Roma, Italy; c National
Cancer Institute Regina Elena, Roma, Italy
Introduction: In prostate cancer radiotherapy, the relationship between
genitourinary (GU) toxicity and clinical and dosimetric parameters is debated.
The aim of this study is to analyze the associations among GU toxicity and
dosimetric and clinical parameters in patients treated with 3D conformal
radiotherapy for localized prostate cancer.
Materials and methods: Our study was a retrospective analysis of late and
acute GU toxicity in 86 consecutive patients treated with conformal ra-
diotherapy for localized prostate cancer. We considered the bladder both
as “whole bladder” (defined in its entirety as a solid organ) and “inferior
bladder” (corresponding to the distal part of the bladder). We used the point
biserial correlation coefficient to correlate GU toxicity with dose-volume
parameters, i.e. the volume of bladder receiving at least a defined × Gy dose
level (Vx(cc)). The normal tissue complication probability (NTCP) cut-off
volume model was fitted to toxicity data. Finally, we performed a univari-
ate analysis between GU toxicity and clinical parameters (age, smoke, “whole
bladder” and “inferior bladder” volumes, pre-radiotherapy urinary symp-
toms, hormonal therapy).
Results: We observed an acute GU toxicity (grade ≥ 2) in 60 patients and
a late GU toxicity (grade ≥ 2) in 6 patients. A statistically significant cor-
relation between acute GU toxicity and Vx dose levels was found at 80 Gy
(P < 0.05) for both whole and inferior bladder. For late GU toxicity statis-
tically significant correlations (P < 0.05) were found from 77 Gy and 77.5 Gy
for the whole and inferior bladder, respectively. The fitting of the NTCP model
indicated a bladder volume of 6 cc as the cut-off volume corresponding to
a genitourinary toxicity of 50% at doses ≥ 77 Gy. A correlation between acute
GU toxicity and smoke (P < 0.001) was obtained from the univariate analysis.
Conclusions:The analysis showed a correlation between the bladder highest
doses, quantified as hotspots, and the GU toxicity.
A.06
LOCAL NTCP TO PREDICT LUNG TISSUE DENSITY CHANGES ON FOLLOW-
UP CT AFTER LUNG CANCER IMRT
S. Barbiero M. Avanzo M. Trovo J. Stancanello C. Furlan
C. Cappelletto E. Capra a Medical Physics Unit, Centro di Riferimento
Oncologico, Aviano, Italy; b Radiation Oncology Department, Centro di
Riferimento Oncologico, Aviano, Italy; cMRI Applications and Workflow, General
Electric, Buc, France
Purpose: The goal of this work was to develop a normal tissue complica-
tion probability (NTCP) model to predict voxel-by-voxel density changes
of lung tissue on follow-up CT of patients treated with intensity modu-
lated radiotherapy (IMRT).
Materials and methods: Follow-up free breathing CT scans acquired at 6
months after the end of RT of 20 patients treated with LINAC (60 Gy/25
fractions) or helical tomotherapy (40–52 Gy in 5–10 fractions) IMRT for lung
tumors were registered to free-breathing pre-treatment scans using rigid
(6 degrees of freedom) followed by a b-spline (>27 degrees of freedom)
deformable registration, performed with the Slicer 3D freeware software
suite. Accuracy of registration was assessed by comparing the displace-
ment of anatomical landmarks in target and registered images. The voxel
relative electron density (RED) was calculated on images using measured
CT-RED curves. The differences in voxel RED between the two images were
calculated in the healthy lung, defined as combined lungs after subtrac-
tion of PTV, among all patients. The function of the Lyman NTCP model
with two parameters: D0.5, the dose giving 0.5 increase in RED, and m, the
slope of the dose–response curve, was used to fit voxel RED changes versus
dose. No correction was used for fractionation.
Results: The fitting process yielded D0.5 of 99.5 Gy (95%CI = 84.0–114.9 Gy)
and m of 0.34 (0.30–0.38). To validate the robustness of the model, a sta-
tistical analysis is under development.
Conclusion: Differential diagnosis between benign changes on follow-up
CT from progression or recurrence is a difficult task in highly conformal
RT because areas of dense consolidation usually develop around the treated
tumor. The model describes well the RED change in follow-up CT scans of
patients treated with RT and can be used to generate maps of predicted
RED changes to be visualized on follow-up CT scans, as a support for dif-
ferential diagnosis between benign changes from progression or recurrence.
A.07
FLATTENING FILTER FREE (FFF) RAPID ARC FOR SINGLE FRACTION SBRT
OF THE LUNG REDUCES TREATMENT TIME AND IS DOSIMETRICALLY
EQUIVALENT TO FLATTENING FILTER VMAT
S. Barbiero F. Matteucc D. Fedele M. Avanzo a Centro di Riferimento
Oncologico, Aviano, Italy; b Ospedale Universitario Santa Chiara, Pisa, Italy;
c Casa di Cura S. Rossore, Pisa, Italy
Purpose: To study dosimetric results of flattening filter free (FFF) volu-
metric modulated arc therapy (VMAT) plans for stereotactic body
radiotherapy (SBRT) in patients with lung lesions, delivered in a single frac-
tion of 24 Gy, and compare them against flattening filtered (FF) VMAT.
e2 Abstracts/Physica Medica 32 (2016) e1–e70