*Purpose: Body mass index (BMI) is a poor marker of obesity and outcomes in liver transplant (LT) candidates. Data suggest that body composition profiling (BCP), such as adipose tissue and muscle volumes, may be more accurate markers of LT risk. However, implementation of BCP in LT has been hindered by arduous analysis methods. This study sought to assess the feasibility of a rapid magnetic resonance (MR)-protocol for BCP in LT candidates and to describe the distribution and change over time in BCP in relation to BMI.

*Methods: This was a prospective pilot study of adults listed for LT between 9/2017-3/2018 at an urban, tertiary care facility who underwent BCP using a 6-minute MR-protocol at the time of LT listing, with repeat BCP performed 6 months later. Spearman correlation coefficients assessed associations between baseline BCP and BMI. Wilcoxon rank scores assessed associations between change in BCP by LT status.

*Results: Among 28 patients enrolled, 18 (64.3%) completed baseline BCP (mean age 58 years, 72% male, 89% non-Hispanic white). Reasons for dropout included LT prior to baseline BCP (n=3), waitlist removal (n=2), critical illness (n=2) and patient time constraints (n=3). Indications for LT were Hepatitis C (39%), alcohol (33%) and NASH (22%). Mean BMI was 30.9±15.3 kg/m², median lab MELD score was 13 (range, 6-23) and 23% of patients had ascites. Among 18 patients with baseline BCP, 14 (78%) had repeat BCP at a median interval of 172 days (range 78-308 days), and six (33%) received a LT. Visceral adipose tissue (VAT) increased by a mean of 5.5% over time, while subcutaneous adipose tissue decreased by 3.4%. Liver proton density fat fraction increased by 8.0 percentage points (pp, 29% relative increase) in LT recipients vs. 6.4 pp (9% increase) in waitlist candidates though this difference was not statistically significant (p=0.26). Interestingly, BMI was not correlated with liver fat fraction (r= -0.35, p=0.16) or VAT (r=0.34, p=0.18). However, BMI correlated strongly with waist circumference (r=0.97, p<0.0001) and muscle fat infiltration (r=0.76, p=0.0002).

*Conclusions: Rapid MR-derived body composition profiling is feasible in LT candidates and demonstrates that body composition changes substantially over time in this population. As anticipated, BMI is a poor marker of underlying changes in fat and muscle volume in patients with liver disease. Future studies will focus on whether integration of body composition profiling can assist with prediction of clinical outcomes in LT candidates.

« Back to 2019 American Transplant Congress