Purpose: This study aimed to investigate whether ADC similarity can predict microvascular invasion (MVI) in patients with bifocal hepatocellular carcinoma (HCC).
Materials and methods: Between January 2015 and September 2015, 51 patients with two HCC lesions were included. All patients underwent conventional magnetic resonance imaging including diffusion-weighted imaging (DWI) before the HCC lesions were surgically resected; the tumor specimens were examined histopathologically. Similarity between two HCC lesions regarding DWI signal intensity (SI) and ADC value was calculated as the difference between the two lesions: Value Similarity = [1-(|valuelarge lesion-valuesmall lesion|)/(valuelarge lesion + valuesmall lesion)] × 100%. Univariate and multivariate logistic regression analyses were performed to assess the presence of MVI.
Results: Risk factors significantly related to MVI of bifocal HCC in univariate analysis were cirrhosis (P = 0.010), histological grade (P = 0.040), DWI SI similarity (P = 0.027) and ADC similarity (P = 0.003). In multivariate analysis, cirrhosis (odds ratio 0.068, P = 0.022) and ADC similarity (odds ratio 1.204, P = 0.008) were independent risk factors for MVI of bifocal HCC.
Conclusion: In patients with two HCC lesions, highly similar ADC values for the two HCC lesions may be a preoperative predictor of MVI.
Keywords: Apparent diffusion coefficient; Diffusion-weighted MRI; Hepatocellular carcinoma; Magnetic resonance imaging; Microvascular invasion.