196 / 1082
S183
ESTRO 36 2017
_______________________________________________________________________________________________
of 41.4Gy (preoperative RT) or 50.4Gy (definitive RT). For
IMRT/VMAT, a DO strategy (i.e., assigning mass density to
gas pockets in the pCT) with three settings was used: no
DO (denoted as DO=0), DO=0.5, and DO=1 (equivalent to
an adipose-muscle mixture), resulting in 6 plans per
patient. Next, by copying the gas pockets derived from the
available CBCT to the pCT, a fractional CT was simulated
to calculate the fractional doses using all 6 plans. DVH
parameters of the CTV and organs at risk (OARs) were
compared between 1) the three DO settings, 2) IMRT and
VMAT, and 3) fractional and planned dose. Dose
distribution difference in the CTV between fractional and
planned dose was also compared.
Results
The range of initial gas volume measured in the pCT was
56–732ml. The gas volume fluctuated over the treatment
course with no time trend (range of mean: 33–519ml,
range of standard deviation: 20–162ml). For the fractional
dose, V
95%
of the CTV was always >98% for VMAT but not
for IMRT with DO=0 (Fig.1). For both IMRT and VMAT, DVH
parameters of the CTV were significantly larger for DO=1
than for DO=0 and 0.5 (
p
<0.05, Wilcoxon signed-rank test).
For an increasing gas volume, an overdose (>3.5% higher
than the planned dose) in the CTV was found in 72–88%/64–
77% cases for IMRT/VMAT with all three DO settings. The
amount of overdose increased as the gas volume increased
relative to the initial volume and was >5% when the
increase was >100ml (Fig.2). For a decreasing gas volume,
an underdose (>3.5% lower than the planned dose) in the
CTV was found for IMRT/VMAT in 34%/23% cases with
DO=0, 7%/0% cases with DO=0.5, and 0%/0% cases with
DO=1. The underdose became more severe as the gas
volume decreased for DO=0 and 0.5. An overdose (>3.5%)
still existed in up to 28% cases for DO=1 when the gas
volume decreased. DVH parameters of OARs in the
fractional dose were almost the same as in the planned
dose and below the clinical constraints for all scenarios.
Conclusion
For esophageal cancer RT, the use of VMAT with DO=0.5 in
treatment planning is preferable to avoid an
overdose/underdose in the CTV when the abdominal gas
volume decreases during treatment. However, when the
gas volume increases with >100ml, a DO strategy would
result in an overdose >5%. Therefore, in that case re-
planning may be a better solution.
OC-0349 Prediction of GTV median dose differences
benefit Monte Carlo re-prescription in lung SBRT
D. Dechambre
1
, Z.L. Janvary
1
, N. Jansen
1
, C. Mievis
1
, P.
Berkovic
1
, S. Cucchiaro
1
, V. Baart
1
, C. Ernst
1
, P. Coucke
1
,
A. Gulyban
1
1
C.H.U. - Sart Tilman, Radiotherapy department, Liège,
Belgium
Purpose or Objective
The use of Monte Carlo (MC) dose calculation algorithm for
lung patients treated with stereotactic body radiotherapy
(SBRT) can be challenging. Prescription in low density
media and time-consuming optimization conducted
CyberKnife centers to propose an equivalent path length
(EPL)-to-Monte Carlo re-prescription method, for example
on GTV median dose (Lacornerie T, et al. Radiat Oncol
2014;9:223). The aim of this study was to evaluate the
differences between the two calculation algorithms and
their impact on organs at risk (OAR) and to create a
predictive model for the re-prescription.
Material and Methods
One hundred and twenty seven patients (with 149 lesions)
were treated with CyberKnife (CK; Accuray, Sunnyvale,
US) between 2010 and 2012. A high-resolution grid (512³)
was used for the EPL and MC calculations (2% variance).
All re-calculation from EPL to MC maintained the number
of beams and their monitor units. Relative differences in
GTV D50 between the two algorithms were assessed and
uni/multivariate linear regression was performed using
prescription dose (Gy), tracking (ITV concept if not
available), location (peripheral or central) and volume (in
cc) of the lesion as input parameters. Statistical
significance was determined using F-test at p-value<0.05.
OARs volumetric dose constraints were applied from
Timmerman RD et al. (Semin Radiat Oncol 2008;18:215-
22). As tolerance limits were defined based on simple
heterogeneity correction algorithm (e.g. EPL), correlation
between EPL and MC OARs dose values was assessed
following the work from the Rotterdam team (van der
Voort van Zyp NC , et al. Radiother Oncol 2010;96:55–60).
Results
The observed difference (MC compared to EPL) varied
from 0 % to 48% (median = 10%, standard deviation = 9%).