To compare the diagnostic accuracy of gadoxetate disodium-enhanced magnetic resonance imaging (EOB-MRI) on a 3-T system and integrated contrast-enhanced 18F-fluorodeoxyglucose positron emission tomography/computed tomography (CE-PET/CT) for the detection of hepatic metastases from colorectal cancers.
Materials and Methods:
The approval from the institutional review board was obtained, and the requirement for informed consent was waived. We retrospectively evaluated 135 metastases in 68 patients (37 men, 31 women; mean age: 68 years; age range: 37–82 years) who underwent both EOB-MRI and CE-PET/CT. A total of 103 metastases were confirmed during surgery and 32 were confirmed by imaging findings during follow-up. The images were independently reviewed by 2 observers. The diagnostic accuracies of EOB-MRI and CE-PET/CT were determined by calculating the areas under each reader-specific receiver operating characteristic curve (Az). Patient-based lesion sensitivity and specificity were compared using the McNemar test.
The mean area under the Az on EOB-MRI versus CE-PET/CT was 0.94 versus 0.81 for all lesions (P < 0.001), 0.92 versus 0.60 for lesions ≤1 cm in size (P < 0.001), and 0.88 versus 0.96 for lesions >1 cm (P = 0.098), respectively. The sensitivity, specificity, positive predictive values, and negative predictive value on a patient basis were 100%, 71%, 97%, and 100% for EOB-MRI and 93%, 71%, 97%, and 57% for CE-PET/CT, respectively.
EOB-MRI using a 3-T system is more accurate than CE-PET/CT, especially for the detection of small (≤1.0 cm) lesions. Patient-based analysis revealed that EOB-MRI has a higher sensitivity and negative predictive value than CE-PET/CT.