Abstract
397
Objectives: PET/MR imaging provides unique opportunity to acquire morphological, functional and metabolic data simultaneously, allowing different biological characteristics visualization of oncologic disorders with one examination. Diffusion weighted imaging (DWI) is a highly emerging advanced MR technique in cancer diagnostics. L-3,4-Dihydroxy-6-[18F]phenylalanine (F-DOPA) PET tracer was initially used for Parkinson imaging, however its neuro-oncologic application is also evolving in the recent years. The aim of our survey was to evaluate F-DOPA uptake and DWI MR parameters in different brain tumors using PET/MR. Methods: 39 F-DOPA PET/MR examinations, performed between September 2017 and November 2018 were retrospectively assessed. Imaging included UTE, T1W, T2W, FLAIR, DWI (b0, b1000) MR sequences and simultaneous PET acquisition, initiated 10 minutes after intravenous injection of 2 MBq/kg F-DOPA tracer. From the DWI MR images ADC maps were generated automatically. A total of 46 lesions were analysed /26 high grade glioma (HGG), 15 low-grade glioma (LGG) and 5 metastatic disorder/. Within the HGG group, a subgroup of glioblastoma multiforme (GBM) disorders was also generated, containing 21 lesions. PET SUVmax and SULpeak, MR DWImax(b1000) and ADCmin parameters were calculated for each lesion using ROI analysis and correlated with World Health Organization (WHO) tumor grade. Association of measured lesional PET and MR parameters were also assessed. Statistics included ANOVA, Tukey post-hoc test and correlation analysis.
Results: HGG disorders represented with higher SUVmax, SULpeak and DWImax and lower ADCmin values than low grade gliomas (mean SUVmax: 4.18 vs 2.36, p=0.0001; mean SULpeak: 2.61 vs 1.53, p=0.0005; mean DWImax: 481 vs 228, p<0.00001 and mean ADCmin: 228 vs 436, p<0.00001 respectively). GBM lesions also showed higher SUVmax, SULpeak, DWImax and lower ADCmin values than non-GBM disorders (mean SUVmax: 4 vs 2.8, p<0.00001; mean SULpeak: 2.48 vs 1.85, p=0.0104; mean DWImax: 492 vs 270, p<0.00001 and mean ADCmin: 263 vs 377, p<0.00001 respectively). Statistically significant (weak) correlation between the measured PET and MR parameters was only found in GBM subgroup, exclusively for SULpeak and ADCmin values (p=0.0405, r=0.1935). Conclusion: Based on our initial PET/MR data, significant divergence can be found in DOPA PET and DWI MR characteristics of different brain tumors. The simultaneously acquired DOPA PET related amino acid transport and the DWI MR derived molecular tissue water mobility information are suggested to represent different, autonomous biological aspects of tumor microenvironment. Our preliminary results need to be confirmed on a larger patient population, the survey is also planned to be extended with further parameter inclusion and cluster analysis.