Objective: The purpose of this study was to compare whole-body MRI including diffusion-weighted imaging (DWI) with (18)F-FDG PET/CT in the staging of newly diagnosed lymphoma.
Subjects and methods: Twenty-two consecutively registered patients with newly diagnosed lymphoma prospectively underwent whole-body MRI (22 with T1-weighted, STIR, and DWI sequences and 21 with T1-weighted and STIR sequences but not DWI) and FDG PET/CT. Whole-body MRI-DWI was independently evaluated by two blinded observers. Interobserver agreement was assessed, and whole-body MRI-DWI was compared with FDG PET/CT.
Results: The kappa values for interobserver agreement on whole-body MRI-DWI for all nodal regions together and for all extranodal regions together were 0.676 and 0.452. The kappa values for agreement between whole-body MRI-DWI and FDG PET/CT for all nodal regions together and for all extranodal regions together were 0.597 and 0.507. Ann Arbor stage according to whole-body MRI-DWI findings was concordant with that of FDG PET/CT findings in 77% (17/22) of patients. Understaging and overstaging relative to the findings with FDG PET/CT occurred in 0% (0/22) and 23% (5/22) of cases. In the care of 9% (2/22) of patients, overstaging with whole-body MRI-DWI relative to staging with FDG PET/CT would have had therapeutic consequences.
Conclusion: Our early results indicate that overall interobserver agreement on whole-body MRI-DWI findings is moderate to good. Overall agreement between whole-body MRI-DWI and FDG PET/CT is moderate. In the care of patients with newly diagnosed lymphoma, staging with whole-body MRI-DWI does not result in underestimation of stage relative to the results with FDG PET/CT. In a minority of patients, reliance on whole-body MRI-DWI leads to clinically important overstaging relative to the results with FDG PET/CT. FDG PET/CT remains the reference standard for lymphoma staging until larger-scale studies show that use of whole-body MRI-DWI results in correct staging in this minority of cases.