Detection of liver metastases from adenocarcinoma of the colon and pancreas: Comparison of mangafodipir trisodium-enhanced liver MRI and whole-body FDG PET

OBJECTIVE. The objective of our study was to assess the relative performance of mangafodipir trisodium-enhanced liver MRI and whole-body FDG PET for the detection of liver metastases from adenocarcinoma of the colon and pancreas. MATERIALS AND METHODS. Imaging data of 34 patients (23 men, I I women; age range, 44-78 years) with adenocarcinoma of the colon (n = 27) or adenocarcinoma of the pancreas (n = 7) who had undergone mangafodipir trisodium-enhanced liver MRI and whole-body FDG PET were retrospectively reviewed for the presence and number of liver metastases. Histopathology (n = 25) or follow-up imaging (n = 9) served as the standard of reference. Breath-hold T1-weighted gradient-recalled echo, respiratory-triggered T2-weighted fast spin-echo, and mangafodipir trisodium-enhanced axial fat-saturated high- spatial -resolution (256 x 512) T1 -weighted gradient-recalled echo images were obtained on a 1.5-T scanner. FDG PET was performed after the injection of 15-20 mCi (555-740 MBq) of FDG. The sensitivity, positive predictive value, and accuracy were calculated for each technique. The performances of the two techniques were compared using the Fisher's exact test. RESULTS. Thirty patients had hepatic metastases and four had no hepatic metastases according to the standard of reference. The total number of metastases was 79, including 33 that measured less than I cm. Based on a per-patient analysis, MRI and FDG PET showed sensitivities of 96.6% and 93.3%, positive predictive values of 100% and 90.3%, and accuracies of 97.1% and 85.3%, respectively. According to a per-lesion analysis, MRI and FDG PET showed sensitivities of 81.4% and 67.0%, positive predictive values of 89.8% and 81.3%, and accuracies of 75.5% and 64.1%, respectively. MRI detected more hepatic metastases than FDG PET (p = 0.016). Of the 33 subcentimeter lesions confirmed on the standard of reference, all were identified on MRI, whereas only 12 were detected on FDG PET (p = 0.0001). CONCLUSION. In patients with colon and pancreatic adenocarcinoma, high- spatial -resolution mangafodipir trisodium-enhanced liver MRI and whole-body FDG PET were comparable in the detection of patients with liver metastases. FDG PET provided additional information about extrahepatic disease and was useful in initial staging. However, significantly more and smaller liver metastases were detected on MRI than on FDG PET.