Microvascular invasion in small hepatocellular carcinoma: Is it predictable with preoperative diffusion-weighted imaging?

Background and Aim: The presence of microvascular invasion (MVI) is an independent risk factor affecting recurrence-free survival following surgical treatment for small hepatocellular carcinoma (HCC). Our aim in this study was to investigate whether diffusion-weighted imaging (DWI) could be useful in predicting MVI for small HCC. Methods: Breath-hold DWI (b-value 0, 500 s/mm(2)) and gadopentate dimeglumine-enhanced dynamic imaging of preoperative magnetic resonance imaging of 109 surgically proven small HCCs from 92 patients were retrospectively analyzed. The signal intensity ratio on DWI and apparent diffusion coefficients (ADCs) for lesions were quantitatively measured. Signal intensity ratio and ADC of DWI, tumor size, tumor shape, tumor capsule, peritumoral enhancement on arterial phase images, and dynamic enhancement pattern were analyzed as radiological parameters reflecting MVI and were compared with histopathological references. The chi-square test, Fisher's exact test, Mann-Whitney U test, and the independent t-test were used for univariate analysis. To identify the independent predictors of MVI among these radiological parameters and to evaluate their diagnostic performance, multivariate logistic regression analysis and receiver operating characteristic curve analysis were performed, respectively. Results: A univariate analysis showed that a lower ADC value (P = 0.005) and irregular circumferential enhancement (P = 0.020) showed statistically significant associations with MVI. A multiple logistic regression analysis showed that the ADC value and irregular circumferential enhancement were independent predictors of MVI. With a cut-off of 1.227 x 10(-3) mm(2)/s, the ADC value provided a sensitivity of 66.7% and a specificity of 78.6% in the prediction of MVI with an odds ratio of 7.63 (P < 0.01). Conclusions: Lower ADC values (< 1.227 x 10(-3) mm(2)/s) on DWI with b-value of 0.500 s/mm(2) can be a useful preoperative predictor of MVI for small HCCs.