Diagnostic Performance of Abbreviated MRI for HCC Detection in Patients with Non-alcoholic Fatty Liver Disease

Background/aim: Hepatocellular carcinoma (HCC) surveillance is recommended in nonalcoholic fatty liver disease (NAFLD)-related cirrhosis. The performance of ultrasound (US) is impaired in NAFLD. This study aimed to evaluate the diagnostic performance of non-contrast abbreviated magnetic resonance imaging (AMRI) for HCC detection in NAFLD.

Methods: Consecutive contrast-enhanced magnetic resonance imaging (CE-MRI) scans of NAFLD patients between June 2017 and December 2021 were retrieved. A radiologist extracted and anonymized a noncontrast AMRI dataset comprising T2-weighted, T1-weighted, and diffusion-weighted imaging (DWI) sequences. Two radiologists blinded to CE-MRI reports and treatment details independently reviewed the AMRI for liver lesion and portal vein (PV) characteristics. HCC and malignant PV thrombosis were diagnosed based on the original dynamic CE-MRI diagnostic reports. The diagnostic performance of AMRI and the interobserver agreement for detecting HCC and malignant PV thrombosis were calculated.

Results: Seventy-five patients (52 males; mean age (±SD), 56 ± 17.6 years; 61 cirrhotic) were included. Nine patients had HCC (14 HCCs). The sensitivity, specificity, positive predictive value, and negative predictive value of AMRI for detecting HCC were 100%, 93.9%, 69.2%, and 100%, respectively, and malignant PV thrombosis was 100%, 98.5%, 80%, and 100%, respectively. There was substantial interobserver agreement for detecting HCC (kappa = 0.721) and malignant PV thrombosis (kappa = 0.645) on AMRI.

Conclusion: AMRI has high diagnostic performance in HCC detection in patients with NAFLD. However, prospective studies must compare the diagnostic performance of AMRI with that of US.