Detection of hepatic graft steatosis and fibrosis after liver transplantation using MRI-PDFF & MRE techniques

Abstract

Background: Although non-alcoholic fatty liver disease (NAFLD) is an uncommon indication for liver transplantation (LT) in Chinese, post-LT de novo NAFLD (PTNL) is increasingly observed, leading to graft dysfunction and fibrosis. We aimed to determine the prevalence and predictors of PTNL and graft advanced fibrosis/cirrhosis (F3/F4) by magnetic resonance imaging (MRI) and transient elastography (TE) in subjects who underwent LT for non-NAFLD related causes. Methods: Post-LT subjects were recruited for MRI and TE. MRI-proton density fat fraction (MRI-PDFF) and MR elastography (MRE) was performed using a 1.5 Tesla Optima 450W MR scanner with a 3D volumetric sequence. PTNL and F3/F4 was defined as MRI-PDFF ≥5% and MRE ≥3.64kPa, respectively. TE was performed by Fibroscan with controlled attenuation parameter (CAP) ≥248 dB/m and liver stiffness (LS) values obtained. Graft dysfunction was defined as persistent alanine aminotransferase (ALT) ≥40 U/L at ≥2 consecutive measurements at follow-up (FU). Metabolic dysfunction was assessed by anthropometric measurements and fasting bloods. Results: 131 subjects (71.8% male, median age at LT: 51.8) were recruited. The majority (77.1%) was transplanted for hepatitis B virus (HBV)-related complications. MRI was performed at a median time of 12.5 years post-LT. The median body mass index increased from 21.7 (38.2% overweight) to 23.8 kg/m2 (61.8% overweight), with 33.6%, 70.2%, 25.2% and 63.4% subjects having central obesity, prediabetes/diabetes (pre-LT: 11.5%), dyslipidemia (pre-LT: 0.8%) and hypertension (pre-LT: 9.2%), respectively. New-onset dyslipidaemia was more common among patients with PTNL (p=0.004) (Figure 1A). According to MRI criteria, PTNL (95% with metabolic dysfunction) and F3/F4 was present in 30.5% and 11.5%, respectively. HBV-LT was associated with lower MRE than non-HBV-LT (p=0.022) (Figure 1B). LS highly correlated with MRE (r=0.853, p<0.001; AUROC for F3/F4: 0.95; Figure 1C), while CAP moderately correlated with MRI-PDFF (r=0.58, p<0.001). Graft dysfunction was seen in 11.5% subjects, which was associated with MRI-PDFF (OR 1.13, 95%CI 1.11-1.24, p=0.02), time after LT (OR 1.23, 95%CI 1.003-1.50, p=0.047) and HBV-LT (OR 0.23, 95%CI 0.06-0.85, p=0.027). No significant inter-segmental variations of MRI-PDFF values were observed (Figure 1D). Conclusion: MRI-PDFF and MRE identified a high prevalence of PTNL and F3/F4, associated with metabolic dysfunction, among Chinese subjects transplanted for non-NAFLD.