ABSTRACT
To retrospectively describe imaging characteristics of liver metastases on fused FDG-PET/MRI data sets and to compare the diagnostic accuracy of MRI and fused FDG-PET/MRI data sets for the detection of liver metastases in patients undergoing systemic anticancer treatment. 43 oncological patients [mean age: 56 +/- 11 years] were investigated by FDG-PET/CT and liver MRI. FDG-PET data from PET/CT scans were fused with MRI. 556 lesions were evaluated. 5 different evaluation algorithms were used for FDG-PET/MRI evaluation. The sensitivity, specifity, PPV, NPV and accuracy of MRI and FDG-PET/MRI data for the detection of liver metastases were calculated. A mean follow-up of 647 days served as reference standard. McNemar's test was used to test for statistically significant differences between MRI and FDG-PET/MRI [p<0.5]. The sensitivity, specificity, PPV, NPV and accuracy of MRI for the detection of liver metastases were 86%, 81%, 97%, 47%, and 85% and 50%, 100%, 100%, 22%, and 56%, for FDGPET/MRI. FDG-PET/MRI was significantly less accurate than MRI alone [p<.001]. In opposite to patients before systemic anticancer therapy the fusion of FDG-PET data with liver MRI cannot be recommended for the detection of liver metastases in patients undergoing systemic oncological therapy