Background:
Initial graft function is a major determinant of morbidity and mortality after LT, making early identification of vascular, biliary, and parenchymal complications crucial for improving transplant prognosis 1,2.
Early liver graft assessment primarily relies on ultrasound to evaluate graft perfusion and parenchymal integrity, supplemented by standard blood tests to monitor biochemical markers 3,4. However, these methods do not directly assess hepatocyte function, providing only indirect insights into graft health.
Gd-EOB-DTPA-enhanced MRI assesses hepatic functionality through hepatocyte contrast uptake and excretion, mediated by liver-specific transport proteins such as anion-transporting polypeptides (OATP1,3) and multidrug resistance-associated proteins (MRP2,3) 5-7. These methods have shown promise in evaluating various liver conditions but are not yet widely adopted for assessing early graft function after LT.
By detecting subtle parenchymal changes not captured by standard imaging metrics, Radiomics holds the potential to improve diagnostic precision and outcome prediction 8.
Methods:
Graft function was evaluated on postoperative days (POD) 3-7, 10-21, and 26-30 in consecutive transplant recipients using Gd-EOB-DTPA-enhanced MRI and standard blood tests.
Inclusion criteria required:
- primary LT at our institution between 2021 and 2023,
- age over 18 years, and
- MRI completion at the specified intervals
Exclusion criteria included:
- an estimated glomerular filtration rate (eGFR) < 30 mL/min/1.73m²,
- severely compromised health status, and
- contraindications to gadolinium-based contrast agents.
Relative liver enhancement (RLE):
RLE was calculated from pre- and post-contrast T1WI using the formula:
RLE (%)= ((SI liver 20 min. post-contrast - SIliver pre contrast) / SIliver pre contrast)*100.
Patients were classified as having either
a) good liver function: RLE ≥48%, or b) impaired liver function: RLE < 48% 9.
Functional liver imaging score (FLIS):
FLIS was obtained from T1WI acquired 20 minutes post-contrast, as described by Poetter-Lang et al.5, determining the following three critera:
1) liver enhancement quality, 2) excretion quality through the bile ducts, and 3) portal vein sign quality.
This scoring system served to categorize patients into:
a) good liver function: FLIS ≥ 4, and b) impaired liver function: FLIS 0-3.
Texture analysis
Texture analysis was conducted with LIFEx software version 7.6.1 (French Alternative Enervies and Atomic Energy Commission, http://www.lifexsoft.org), processing T1WI obtained 20 minutes post-contrast 10.
For each patient and interval, three 20 mm2 three-dimensional voxels of interest (VOI) were manually segmented from different liver lobes using a 3D drawing tool, excluding vessels and bile ducts > 5mm, focal lesions and artifacts.