A prospective technical validation study comparing Proton Density Fat Fraction by MRI and Liver Attenuation Index by CT for the quantitative assessment of hepatic steatosis in living liver donors with histopathological correlation as standard reference.

Abstract

Background:
Accurate quantification of hepatic steatosis is crucial in living liver donors to ensure safety and optimal graft function. Although liver biopsy is the reference standard, it is invasive and limited by sampling variability. Computed tomography (CT) derived liver attenuation index (LAI) is commonly used for non-invasive liver fat quantification; however, its accuracy is affected by confounding factors such as iron deposition, fibrosis, edema and medications, and involves ionizing radiation. Magnetic resonance imaging (MRI) derived proton density fat fraction (PDFF) has emerged as a more reliable non-invasive technique for liver fat quantification. The objective of the study is to compare the diagnostic accuracy of MRI- derived PDFF and CT - derived LAI in quantifying hepatic steatosis in living liver donors using histopathology as the standard reference.
Methods:
In this prospective observational study, 30 voluntary liver donors underwent non-contrast CT and multi-echo MRI. LAI was calculated as the difference between mean hepatic and splenic attenuation on CT, and MRI-PDFF was obtained using mDIXON Quant sequence. Histopathological steatosis grading served as the reference standard. A point biserial correlation was done to find out correlation between LAI, PDFF and steatosis. Receiver operating characteristic (ROC) curves were created to find out diagnostic characteristics of LAI and PDFF for predicting steatosis.
Results:
Histopathology demonstrated steatosis in 5 out of 30 donors (16.7%). CT-LAI demonstrated moderative negative correlation with steatosis (r= -0.545, p=0.002) and limited diagnostic performance (AUC 0.116; sensitivity 40%, specificity 24% at cut off <6.5HU). MRI-PDFF demonstrated strong correlation (r=0.825, p=0.001) and excellent diagnostic accuracy (AUC 0.992, sensitivity 100%, specificity 96% at cut off >6.75%).
Conclusion:
MRI-PDFF provides superior accuracy compared with CT-LAI for detecting hepatic steatosis, particularly mild disease, and represent a reliable, non-invasive, radiation-free alternative for pre-operative evaluation of living liver donors.