Can negligible hepatic steatosis determined by magnetic resonance imaging–proton density fat fraction obviate the need for liver biopsy in potential liver donors?

Abstract

The purpose of this study is to determine whether magnetic resonance (MR)-proton density fat fraction (PDFF) estimate of negligible hepatic fat percentage (<5%) can exclude significant hepatic steatosis (≥10%) in living liver donor candidates obviating the need for liver biopsy and to perform intraindividual comparisons between MR-PDFF techniques for hepatic steatosis quantification. In an ethics-approved retrospective study, 144 liver donor candidates with magnetic resonance spectroscopy (MRS) and 6-echo Dixon magnetic resonance imaging (MRI) between 2013 and 2015 were included. A subset of 32 candidates underwent liver biopsy. Hepatic fat percentage was determined using MR-PDFF and histopathology-determined fat fraction as the reference standard. A receiver operating characteristic analysis with positive predictive value, negative predictive value (NPV), sensitivity, and specificity was performed to discriminate between clinically significant steatosis (≥10%) or not (<10%) at MRS-PDFF and MRI-PDFF thresholds of 5% and 10%. Pearson correlation and Bland-Altman analyses between MRS-PDFF and MRI-PDFF were performed for intraindividual comparison of hepatic steatosis estimation. There was significant association between MRS-PDFF and MRI-PDFF with HP-FP. High NPV of 95% (95% confidence interval [CI], 78%-99%) and 100% (95% CI, 76%-100%) as well as an area under the curve of 0.90 (95% CI, 0.79-1.0) and 0.93 (95% CI, 0.84-1.0) were obtained with a cutoff threshold of 5% MRI-PDFF and MRS-PDFF, respectively, to exclude clinically significant steatosis (≥10%). Intraindividual comparison between MRS-PDFF and MRI-PDFF showed a Pearson correlation coefficient of 0.83. Bland-Altman analysis showed a mean difference of 1% with 95% limits of agreement between -1% and 3%. MR-PDFF estimate of negligible hepatic fat percentage (<5%) has sufficient NPV for excluding clinically significant hepatic steatosis (≥10%) in living liver donor candidates obviating the need for liver biopsy. It may be sufficient to acquire only the multiecho Dixon MRI-PDFF for hepatic steatosis estimation. Liver Transplantation 24 470-477 2018 AASLD.

Accuracy of multi-echo Dixon sequence in quantification of hepatic steatosis in Chinese children and adolescents

“…Kang et al[18] also found that the correlation in the right lobe was higher than that in the left lobe. Satkunasingham et al[19] conducted a retrospective study on liver donor candidates with liver biopsy as a reference; the researchers demonstrated a significant correlation ( r = 0.83) between ME Dixon and MRS (HISTO), and they showed that these two methods could obviate the need for liver biopsy when proton density fat fraction (PDFF) was < 5%. Furthermore, Bannas et al[20] performed MRI, MRS, liver biopsy, and biochemical triglyceride extraction on ex vivo human livers that were unsuitable for transplantation.…”

The diminishing role of liver biopsy in living donor liver transplantation

“…In the current issue of Liver Transplantation , Satkunasingham et al compared MRI‐based estimations of hepatic steatosis to liver biopsy in 32 living donor candidates . Specifically, the MRS‐PDFF and 6‐echo Dixon MRI (MRI‐PDFF) were evaluated in their ability to discriminate between candidates with clinically significant histologic steatosis of ≥10%.…”

Novel Method for Ultrasound‐Derived Fat Fraction Using an Integrated Phantom

“…These thresholds have been used in several studies as inclusion criteria for liver transplantation and for therapeutic trials. [46][47][48] Last, both UDFF and PDFF were compared to the body mass index (BMI) by computing the slopes of regression and 95% CIs versus BMI.…”

“…An additional receiver operating characteristic analysis was performed to evaluate the performance of UDFF for identifying steatosis defined as MRI‐PDFF higher than 5% or 10%. These thresholds have been used in several studies as inclusion criteria for liver transplantation and for therapeutic trials 46–48 . Last, both UDFF and PDFF were compared to the body mass index (BMI) by computing the slopes of regression and 95% CIs versus BMI.…”