ESTRO 36 Abstract Book

940 / 1096

S924 ESTRO 36
_______________________________________________________________________________________________
correlate the extra-vascular volume fraction (ve) and
apparent diffusion coefficient (ADC) from DCE-MRI/CT and
DWI MRI done at the same time points in patients with
brain metastases.
Material and Methods
A total of 26 tumours in 19 patients were treated under
ethics-approved trials with stereotactic radiosurgery (SRS)
alone (n=14) or SRS on day 7 of sunitinib (n=12) and
underwent multi-parametric imaging at baseline, post
drug (if applicable), then 7 and 21 days post-
radiosurgery. Each patient received a comparative DCE-
MRI scan on the same day as the CT imaging on a Verio 3T
System (IMRIS) with Variable Flip Angle (VFA) T1
quantification and 3D-FLASH and Gadolinium injection
(Magnevist 20cc); T2-weighted imaging; DWI (echo-planar
imaging with TR/TE 7700/110; 3D-diffusion gradient
encoding). Volumetric DCE-CT was acquired following a
60cc Visipaque320® injection in an intermittent time
sequence up to 3 mins (Toshiba, Aquilion ONE). A temporal
dynamic analysis (TDA) method for voxel-based CT
perfusion [1] was remodeled to enable using GPU-based
optimization on a high throughput cluster to include
various complimentary transport processes into a common
framework. As DCE-CT is considered a gold standard for
tracer-kinetic validation given its signal linearity, we
compared extravascular extracellular volume maps from
DCE-CT to those from DCE-MRI and ADC values by Pearson
correlation on a voxel-by-voxel basis as well as other
kinetic parameters using the Modified Tofts model (AUC,
Ktrans, Kep).
Results
Voxel-wise Pearson’s analysis showed statistically
significant correlations in Ktrans (P<0.001) between DCE-CT
and DCE-MRI over all imaging time points as well as
excellent agreement with very little bias (see Figure 1).
The correlation between ADC and ve values were strong in
the Sunitinib cohort (R=0.6, p<0.01, all days) and peaked
at day 3 post SRS (R=0.75, p<0.008). No such statistically
significant correlation was seen between ADC and ve in the
SRS alone group. Correlation of ADC histogram parameters
between imaging days was highly correlated however,
again peaking at Day 7 (R=0.85, p<0.001).
Conclusion
Using a common analysis platform has improved the
correlations in pharmaco-kinetic parameters, Ktrans and
ve, obtained from CT and MR than previously reported.
Consistent with our hypothesis that ADC and ve values
would describe a similar physiological effect, the observed
correlation between extravascular volume fraction and
ADC values was high for the cohort treated with Sunitinib
and SRS, but this correlation was not seen for the SRS
alone group.
[1] Coolens C et al. IJROBP. 2015;91(1):48-57.
EP-1692 Multi-device textural analysis on 18F-FDG PET
images for predicting cervical cancer recurrence
S. Reuzé1,2,3, F. Orlhac3,4, C. Chargari 1, C. Nioche4, F.
Riet 1, A. Escande1, C. Haie-Meder 1, L. Dercle5, I. Buvat4,
E. Deutsch1,2,3, C. Robert1,2,3
1 Gustave Roussy, Radiotherapy, Villejuif, France
2 Univ. Paris-Sud, Université Paris-Saclay, Le Kremlin-
Bicêtre, France
3 INSERM, U1030, Villejuif, France
4 IMIV, CEA- Inserm- CNRS- Univ. Paris-Sud- Université
Paris-Saclay- CEA-SHFJ, Orsay, France
5 Gustave Roussy, Nuclear Medicine and Endocrine
Oncology, Villejuif, France
Purpose or Objective
The aim of this study was to evaluate the possibility of
gathering images from 2 different PET devices in a
radiomic study, and to propose a signature of local
recurrence for locally advanced cervical cancer (LACC).
Material and Methods
118 patients with LACC were retrospectively included. All
patients underwent a 18F-FDG PET-CT scan before
treatment. They were classified in 2 groups depending on
the PET device used for acquisition (G1: Siemens Biograph
installed in 2003, N=79; G2: GE Discovery installed 2011,
N=39). Treatment consisted in a concomitant
chemoradiation delivering 45 Gy in 25 fractions of 1.8 Gy
to the pelvis +/- the para-aortic area followed by a pulse-
dose rate image-guided uterovaginal brachytherapy of 15
Gy.
The primary tumor was delineated on the PET images using
a fixed threshold (40% of SUVmax) within a manually drawn
volume of interest (VOI), called VOI-T. A 73 mL sphere was
drawn in the healthy liver considered as homogeneous
(VOI-L). For each VOI, 5 conventional indices (SUVmean,
SUVmax , SUVpeak in a 1 mL sphere, metabolic volume, tumor
lesion glycolysis) and 6 3D textural indices were calculated
after resampling the VOI SUV between 0 and 40 using 128
gray levels: Homogeneity and Entropy from the Gray-Level
Co-occurrence Matrix, Short-Run Emphasis (SRE), Long-
Run Emphasis (LRE), Low Gray-level Zone Emphasis (LGZE)
and High Gray-level Zone Emphasis (HGZE).
Wilcoxon’s tests were performed between G1 and G2 in
VOI-L to determine to what extent technological
differences and image properties influence radiomic
feature values. A stepwise model selection using the
Akaike Information Criterion was applied to determine the
best 4-feature signature for local recurrence prediction in
both groups, used successively for training and validation.
Delong’s tests between AUC were performed to evaluate
if the signature was more powerful than SUVmax only.