Purpose: The superparamagnetic iron oxide (SPIO) Resovist is a contrast media with shortening of both T(1) and T(2) relaxation time. This study evaluates the impact of SPIO-enhanced T(1)- and T(2)-weighted images for the differentiation of liver lesions.
Materials and methods: SPIO-enhanced MRI examinations (1.5 T, Symphony Quantum) of 61 patients were analyzed. Thirty-seven patients had malignant liver lesions (metastases n = 32, HCC n = 5) proven by biopsy or laparotomy, 11 patients had benign liver lesions (FNH n = 2, hemangiomas n = 4, benign cysts n = 5, normal liver on laparoscopy n = 13). After unenhanced T(1)- and T (2)-weighted imaging, a bolus injection of 1.4 ml SPIO (Resovist) was given, followed by T(1)-weighted imaging at 20 s, 60 s, and 5 min and T(2)-weighted imaging at 10 min post injection. A score from 1 (benign) to 5 (malignant) was used by three blinded radiologist for the ROC analysis of the unenhanced T(1)-/T(2)-weighted images (set 1) and of the combinations of unenhanced T(1)/T(2)-weighted and SPIO T(1)-weighted images (set 2), unenhanced T(1)/T(2) and SPIO T(2)-w images (set 3) and all images (set 4).
Results: The accuracy of plain MRI (set 1: 56 %) was increased by SPIO-enhanced T(1)-weighted images (set 2: 81 %) and SPIO-enhanced T(2)-weighted images (set 3: 90 %). Best results were obtained using unenhanced T(1)-weighted, unenhanced T(2)-weighted and both SPIO T(1)-weighted and T(2)-weighted images (set 4: 93%). The accuracy of predicting histopathologic diagnosis was 91%.
Conclusion: For the differentiation of liver lesions, SPIO-enhanced T(2)-weighted images had a greater impact on the accuracy of MRI than T (1)-weighted images, but SPIO-enhanced T(1)-weighted images provided additional information in some patients and should not be deleted.