Oncological whole-body staging in integrated 18F-FDG PET/MR: Value of different MR sequences for simultaneous PET and MR reading

Objective: To evaluate different magnetic resonance (MR) imaging sequences in integrated positron emission tomography (PET)/MR concerning their ability to detect tumors and allocate increased radionuclide uptake on F-18-fluorodeoxyglucose (F-18-FDG) PET in intraindividual comparison with computed tomography (CT) from PET/CT. Material and methods: Sixty-one patients (34 female, 27 male, mean age 57.6y) who were examined with contrast-enhanced PET/CT and subsequent PET/MR (mean delay for PET/MR after injection: 147 +/- 43 min) were included. A maximum of ten F-18-FDG-avid lesions per patient were analyzed on CT from PET/CT and with the following MR sequences from PET/MR: T2, turbo inversion recovery magnitude (TIRM), non-enhanced T1, contrast-enhanced T1, and diffusion-weighted imaging (DWI). All lesions were rated using a four-point ordinal scale (scored from 0 to 3) concerning visual detectability of the lesion against the surrounding background and anatomical allocation of the PET finding. In each category (detectability and allocation), Wilcoxon rank sum tests were performed. Bonferroni-Holm correction was performed to prevent ce.-error accumulation. Results: In 225 F-18-FDG-avid lesions (156 confirmed as malignant by radiological follow up, 69 by histopathology), visual detectability was comparably high on CT (mean: 2.5 +/- 0.9), TIRM (mean: 2.5 +/- 0.9), T2 (mean: 2.4 +/- 0.9), and DWI (mean: 2.5 +/- 1.0) and was significantly higher than on non-enhanced T1 (mean: 2.2 +/- 1.0). While anatomic allocation of the PET finding was comparable with CT (mean: 2.6 +/- 0.7), T2 (mean: 2.6 +/- 0.7), and TIRM (mean: 2.8 +/- 0.7), it was significantly higher compared to DWI (mean: 2.1 +/- 1.0) and non-enhanced T1 (mean: 2.4 +/- 0.8). Conclusion: In conclusion, T2, TIRM, and contrast-enhanced T1 provide a high quality of lesion detectability and anatomical allocation of FDG-avid foci. Their performance is at least comparable to contrast-enhanced PET/CT. Non-enhanced T1 may be omitted and the necessity of DWI should be further investigated for specific questions, such as assessment of the liver. (C) 2015 Elsevier Ireland Ltd. All rights reserved.