Keywords:
Oncology, Hybrid Imaging, Paediatric, PET-MR, MR, CT, Staging, Neoplasia, Metastases
Authors:
G. Orsatti1, A. Varotto1, F. Crimì1, P. Zucchetta1, M. Weber2, D. Cecchin1, R. Stramare1, C. Giraudo1; 1Padova/IT, 2Vienna/AT
DOI:
10.1594/ecr2018/C-1129
Methods and materials
Pediatric patients with histologically proven sarcoma who underwent 18F-FDG PET/MRI (i.e.,
fully-integrated) for initial staging or re-staging were included in this retrospective study.
The MRI protocol of the PET/MR scan had to include at least the following sequences: axial Turbo Inversion Recovery Magnitude (TIRM),
axial Half-Fourier Acquisition Single-shot Turbo spin Echo imaging (HASTE),
coronal T1w Turbo Spin Echo (TSE),
pre-contrast and contrast enhanced axial Volumetric Interpolated Breath-hold Examination (VIBE) and axial whole body Diffusion Weighted Imaging (DWI).
Each examination was evaluated independently and then in consensus by two teams,
each composed by one radiologist and one nuclear medicine physician; all raters were blinded to the clinical information and to conventional imaging results.
Each team assessed the presence of primary tumor,
local and distant nodal involvement,
skeletal and pulmonary metastases as well as metastatic lesions in any other organ/system.
A maximum of six lesions per investigated region was recorded.
PET/MRI was rated positive if at least two of the following criteria were satisfied: 1) focal pathologic uptake at PET; 2) morphological correlate at any MR sequence (excluding DWI); 3) restricted diffusion at DWI.
Cohen’s Kappa coefficient (κ) was used to investigate the inter-observer agreement between the two teams.
PET/MRI sensitivity and specificity were calculated considering histological examination and conventional imaging as reference standard (i.e.,
MRI of the area affected by the primary tumor for local staging and whole body contrast enhanced CT for the detection of distant metastases).