Diffusion-weighted magnetic resonance imaging for the assessment of liver fibrosis in chronic viral hepatitis

Background Accurate noninvasive methods for the assessment of liver fibrosis are urgently needed. This prospective study evaluated the diagnostic accuracy of diffusion-weighted magnetic resonance imaging (DWI) for the staging of liver fibrosis and proposed a diagnostic algorithm using DWI to identify cirrhosis in patients with chronic viral hepatitis. Methods One hundred twenty-one treatment-naive patients with chronic hepatitis B or C were evaluated with DWI followed by liver biopsy on the same day. Breath-hold single-shot echo-planar DWI was performed to measure the apparent diffusion coefficient (ADC) of the liver and spleen. Normalized liver ADC was calculated as the ratio of liver ADC to spleen ADC. Results There was an inverse correlation between fibrosis stage and normalized liver ADC (p<0.05). For the prediction of fibrosis stage >= 2, stage >= 3, and cirrhosis, the area under the receiver-operating curve of normalized liver ADC was 0.603, 0.704, and 0.847, respectively. The normalized liver ADC value <= 1.02 x 10(-3) mm(2)/s had 88% sensitivity, 81% specificity, 25% positive predictive value (PPV), and 99% negative predictive value (NPV) for the diagnosis of cirrhosis. Using a sequential approach with the Fibrosis-4 index followed by DWI, normalized liver ADC <= 1.02 x 10(-3) mm(2)/s in patients with Fibrosis-4 > 3.25 yielded an 80% PPV for cirrhosis, and a 100% NPV to exclude cirrhosis in patients with Fibrosis-4 between 1.45 and 3.25. Only 15.7% of patients would require a liver biopsy. This sequential strategy can reduce DWI examinations by 53.7%. Conclusion Normalized liver ADC measurement on DWI is an accurate and noninvasive tool for the diagnosis of cirrhosis in patients with chronic viral hepatitis.