MRI protocol

To measure hepatic PDFF, a state-of-the-art MR imaging technique was performed. The protocol utilized an MR-sequence product for the creation of PDFF maps for each liver (IDEAL-IQ; GE Healthcare). Imaging was performed at 1.5T on GE Optima MR450w. Axial T2 single-shot fast spin-echo sequences (SSFSE) were utilized for anatomic correlation with PDFF maps. The technique is composed of a 3D fast spoiled gradient-echo sequence with a low flip angle (FA) to minimize T1 bias, and it acquires multiple echoes to calculate triglyceride fat and water in each pixel based on their phase differences. Data obtained at each of the echo times are then passed to a nonlinear least-squares fitting algorithm that estimates and corrects T2* effects. This allows for more accurate modeling of the fat signal as multiple spectral peaks and estimates fat and water proton densities from which the fat content is calculated [5]. Using analysis software, a mathematical model was then applied pixel by pixel on the source images to generate parametric PDFF maps that depicted the quantity and distribution of fat throughout the entire liver. Imaging PDFF was measured in regions of interest (ROI) approximately 300–400 mm² in area placed on the PDFF parametric maps, avoiding vessels, bile ducts, lesions and artefacts. ROIs were placed in all 9 of the liver segments on the MR exams (counting segment 4A and 4B as separate segments). Two experienced radiologists (AD-15+ years’ and WA- 5+ years’ experience) who were blinded of the clinical and histological information interpreted the PDFF values independently, and the per-liver PDFF measurements were then averaged. A representative measurement on two different segments of the liver by ex-vivo MRI is presented in Fig 1.

This approach required a dedicated well-trained MRI staff, and a research coordinator who was available 24 hrs and 7 days (on phone) to coordinate the availability of the MRI room, and staffing of MRI technician for the MRI imaging. In general, as soon as an organ was available, the transplant coordinator notified the principal investigator (SKS). The PI and the research coordinator then coordinated the timing of the MRI based on an approximate organ arrival time with the MRI technician. The center has two dedicated state of the art MRI rooms, and one the room was kept open in anticipation of the organ arrival at least 30 minutes prior to organ arrival to avoid any delay in transportation-related time to the MRI imaging enroute to the operating room.

The decision to proceed with liver transplant was not based on the interpretation of the fat quantification as determined by MRI-PDFF methods, and the transplant surgeons were blinded to these results at the time of the liver transplant.

Outcome assessment:

To objectively evaluate the differences in the post-transplant outcome, we specifically evaluated the following outcomes:

EAD (Early allograft dysfunction).

Bilirubin level, Alanine Aminotransferase (ALT), aspartate aminotransferase (AST), Alkaline phosphatase (ALP), and INR in the immediate post-transplant period.

Need for operations post-transplant within 90 days.

The length of ICU stays in the immediate post-transplant period.

Length of hospitalization.

We used a previously validated definition of early allograft dysfunction (EAD) described by Olthoff in the present study. EAD was thus defined as the presence of 1 or more of the following variables: [1] bilirubin ≥ 10 mg/dL on post-operative day 7, [2] INR ≥ 1.6 on postoperative day 7, and [3] an aminotransferase level (ALT or AST) ≥ 2000 IU/mL within the first 7 postoperative days[6].

The Institutional Review Boards of the University of Tennessee Health Science Center approved the study and informed consent was obtained from all participants (organ recipients) before inclusion in the study. Institutional Review Boards of the University of Tennessee Health Science did not request additional consent from the donor as the organ after procurement was de-identified, and any clinical information related to the donor is linked with a code.