Characterization of Hepatic Focal Lesions by Diffusion Tensor Imaging and How Far it Can Predict Post-Treatment Response?

Abstract Background: Liver diseases have been known to be a major health problem principally because of their world-wide distribution. Focal liver disease is a common diagnostic problem referred to radiologists for evaluation owing to its nonspecific clinical presentation and marked interobserver variation on clinical examination. Aim of Study: The purpose of this study is to evaluate the role of diffusion tensor imaging in characterization of hepatic focal lesions and its value in post-treatment response. Patients and Methods: This prospective study included 30 patients previously diagnosed to have hepatic focal lesions (HFLs) by their characteristic triphasic computed tomography (CT), dynamic magnetic resonance imaging (MRI) features and/or biopsy. The study included 43 lesions in 30 patients (16 benign lesions, 23 malignant lesions, and 4 treated malig-nant lesions). MRI with diffusion tensor imaging (DTI) was performed for all patients. Apparent diffusion coefficient (ADC) and fractional anisotropy (FA) values were evaluated for all lesions. Results: There was a statistically significant difference in ADC values between the benign and malignant lesions (p < 0.001) and between the treated malignant and untreated malignant lesions (p=0.002). There was a statistically signif-icant difference in FA values between the benign and malignant lesions (p < 0.001) and between the treated malignant and untreated malignant lesions (p=0.004). The best cut-off ADC value to differentiate between benign and malignant lesions respectively was >1.42x 10-3 mm2/s with 95.7% sensitivity and 82.8% specificity. The best cut-off ADC value to differ-entiate between treated malignant and malignant lesions respectively was >1.65x 10-3 mm2/s with 97.8% sensitivity and 95.7% specificity. The best cut-off FA value to differentiate between malignant and benign lesions respectively was >0.29 with 95% sensitivity and 70% specificity. The best cut-off FA value to differentiate between treated malignant and untreated malignant lesions respectively was <0.297 with 100% sensitivity and 69.2% specificity. Conclusion: Diffusion tensor imaging is an evolving technique that can be used to characterize different hepatic focal lesions either benign or malignant with significant additive value to dynamic contrast enhanced MRI examination. It can also be used to monitor treatment response.