Fat fraction quantification by MRI predicts diagnosis and prognosis of HBV-related steatohepatitic hepatocellular carcinoma.

Abstract

This study explored the clinical prognosis and lipidomics of hepatitis B virus steatohepatitic hepatocellular carcinoma (HBV-SHHCC) and aimed to identify a noninvasive and convenient method to diagnose this phenotype and guide treatment using MRI. A total of 433 HBV-infected HCC patients were enrolled in this retrospective study. Survival data were analyzed using Cox regression analyses, and lipidomics was used to study HCC tissue composition. Logistic regression identified an independent predictor for HBV-SHHCC, and receiver-operating characteristic (ROC) analysis verified its discrimination. HBV-SHHCC patients had longer disease-free survival (DFS, p < 0.0001) and overall survival (OS) time (p = 0.00097). Compared with common HCC (cHCC), SHHCC was associated with significantly higher mean triacylglyceride (p = 0.010) and diacylglyceride contents (p = 0.002) in tumor tissues. Fat fraction (FF) was linearly correlated with lipid composition and fatty acid degradation (FAD) subtype, which could help in treatment options for HCC. The univariate and multivariate logistic regression indicated FF (p < 0.001) as an independent predictor for diagnosing this phenotype. ROC analysis confirmed excellent discrimination (area under the curve (AUC), 0.914; sensitivity, 92.3%; specificity, 78.7.0%). After using the optimal cutoff point, the DFS time of patients with SHHCC stratified by FF was significantly higher than that of patients with cHCC. The biological behavior and prognosis of HBV-SHHCC were better than those of other types. FF is a valuable tool for the clinical diagnosis of SHHCC, prognosis prediction, and treatment guidance in patients with HCC. Question Can the diagnosis of steatohepatitic hepatocellular carcinoma (SHHCC) be made noninvasively? Findings Fat fraction (FF) correlated with lipid composition and could be used to diagnose SHHCC with an AUC of 0.914, sensitivity of 92.3%, and specificity of 78.7%. Clinical relevance MRI-based FF could be used to diagnose HBV-related SHHCC, indicate prognosis, and guide the clinical treatment of patients with HCC.