Table of Contents
A total of 879 patients underwent LT from January 2003 to December 2014. Of these, 100 patients were transplanted before the age of 18 years, and were excluded. Of the remaining 779 patients, 571 had transient elastography performed. The remaining 208 patients did not have liver stiffness or CAP measurement because of death, re-transplantation, suboptimal medical conditions, or were lost to follow up (Fig. 1). Twenty-two had invalid scans, resulting in 549 patients with valid measurements (Table 1). The median scan age was 59 (range, 23–78) years, with a median scan time from LT of 77 (range, 6–166) months.
Prevalence of metabolic factors
At the time of LT, 114 (14.6%), 71 (9.1%), and 17 (2.2%) had established DM, hypertension, and dyslipidemia respectively. After LT, 192 patients developed new onset DM, with cumulative rate of 17.9%, 24.0%, 32.3%, and 38.4% at 1, 5, 10, and 15 years respectively, 389 patients had new onset hypertension, with cumulative rate of 23.8%, 44.4%, 65.3%, and 72.5% at 1, 5, 10, and 15 years respectively, and 164 patients developed new onset dyslipidemia, with cumulative rate of 2.5%, 14.4%, 24.7%, and 36.6% at 1, 5, 10, and 15 years respectively.
Correlation of CAP score with implantation biopsy
Of the 549 patients with valid transient elastography performed, 422 (76.9%) had < 5% steatosis, 101 (18.4%) had 5–33% steatosis, 18 (3.3%) had 33–66% steatosis, and 8 (1.4%) had > 66% steatosis from the liver biopsy taken at the time of implantation. There was no significant difference in post-transplant CAP scores when correlating with the degree of graft steatosis at the time of transplant. For patients with < 5%, 5–33%, 33–66%, and > 66% steatosis at implant biopsy, the median CAP scores were 225 (range, 114–400), 229 (range, 121–400), 207 (range, 136–305), and 217 (range, 133–318) dB/m respectively (p = 0.935) (Fig. 2A).
Median CAP scores according to steatosis grade at (A) the time of liver implantation and (B) liver biopsy within 1 year of CAP measurement
Correlation of CAP with liver biopsy after liver transplantation
Liver biopsies performed within a year of valid CAP measurements were available for 42 patients. Of these, 22 (52.4%) had minimal (< 5%) steatosis, 14 (33.3%) had mild (5–33%) steatosis, 2 (4.8%) had moderate (34–66%) steatosis, and 4 (9.5%) had severe (≥ 67%) steatosis. A significantly higher CAP score was observed for higher grades of steatosis, from 200, 254, 318, and 324 dB/m for minimal, mild, moderate, and severe steatosis respectively (p = 0.003) (Fig. 2B). Given the small numbers of those with moderate and severe steatosis, these were grouped together for analysis. The AUROC for diagnosing at least mild, moderate, and severe steatosis using CAP measurement was 0.740, 0.954, and 0.951 respectively (Supplementary Table 1), with an optimal cut-off of 266, 293, and 301 dB/m respectively (Supplementary Fig. 1A, B and C).
Prevalence and risk factors for moderate-severe steatosis and MAFLD
Using the optimal CAP score cut-offs as derived, the prevalence of at least mild (≥ S1), moderate (≥ S2), and severe steatosis (S3) was 28.9%, 17.1%, and 14.0% respectively in the 549 patients with CAP measurements. Of the 159 patients with NAFLD, 152 (95.6%) fulfilled the criteria for MAFLD.
A total of 94 patients had moderate-severe (S2/3) steatosis. No difference was observed in the diagnosis of HBV, HCV, or alcoholic liver disease between those with and without S/3 steatosis, with a higher rate of cryptogenic cirrhosis (7.4% vs. 2.0% respectively, p = 0.011). Ninety-three (98.9%) patients had MAFLD, of which 88 (94.6%) had overweight/obesity, 46 (49.5%) had type 2 DM, and 33 (35.5%) had at least 2 metabolic risk abnormalities. Only 1 patient did not fulfil the MAFLD criteria with standalone hypertension, and interestingly, this patient had intermittent alcohol intake post transplantation for hepatitis B-related cirrhosis.
Patients with S2/3 steatosis, when compared to S0/1 steatosis, had higher cumulative rates of DM (54.3% vs. 35.2% respectively, p = 0.001), hypertension (85.1% vs. 60.2% respectively, p < 0.001), and dyslipidemia (39.4% vs. 23.7% respectively, p = 0.002). The incidence of new-onset hypertension was also higher among those with S2/3 steatosis compared to those without (70.2% vs. 48.6%, p < 0.001), while similar incidence rates for new-onset DM and dyslipidaemia were observed (Supplementary Fig. 2). A higher BMI at the time of transplant and at CAP measurement was observed for those with S2/3 steatosis (23.5 kg/m2 vs. 20.7 kg/m2, p < 0.001 and 26.6 kg/m2 vs. 23.1 kg/m2, p < 0.001 respectively). There was an increasing rate of moderate-severe steatosis observed for those with BMI < 18.5, 18.5–22.9, 23.0-24.9, 25-29.9, and ≥ 30 kg/m2 (0%, 3.1%, 10.0%, 36.0%, and 48.6% respectively, p < 0.001, Supplementary Fig. 3).
Patients with S2/3 steatosis, compared to those without, had higher fasting glucose (6.3 vs. 5.4 mmol/L respectively, p < 0.001), low density lipoprotein (LDL)-cholesterol (2.6 vs. 2.2 mmol/L respectively, p < 0.001), triglyceride (1.5 vs. 1.0 mmol/L respectively, p < 0.001), and lower HDL-cholesterol 1.1 vs. 1.3 mmol/L respectively, p < 0.001) at the time of transient elastography. The results are summarized in Supplementary Table 2.
After multivariate analysis, only BMI at the time of CAP measurement (and not at the time of transplant), cryptogenic cirrhosis as a primary liver disease, the presence of DM, hypertension, HDL-cholesterol and LDL-cholesterol levels remained significant factors associated with S2/3 steatosis (Table 2). The degree of donor graft steatosis at the time of implant (i.e. implant biopsy) was also not associated with S2/3 steatosis (Supplementary Table 4).
Steatosis and graft dysfunction
Patients with moderate-severe steatosis, compared to those without, had higher alanine aminotransferase (ALT) (28 vs. 23 U/L respectively, p < 0.001), and a higher rate of graft dysfunction, defined as serum ALT > 40 U/L (25.5% vs. 14.1%, p = 0.008; Supplementary Fig. 2). There was no significant difference between the 2 groups with respect to bilirubin, alkaline phosphatase (ALP), aminotransferase aspartate, and gamma glutamyl transpeptidase (GGT) levels (p = 0.964, p = 0.559, p = 0.461, and p = 0.066 respectively. The higher prevalence of biliary disease/complications and cholestasis observed in post-LT settings may have masked the difference in GGT levels. By excluding patients with elevated ALP, a significantly higher GGT level was able to be observed for those with S2/3 steatosis (39 vs. 28 IU/L, p = 0.003). Using the local laboratory cut offs of 58 and 45 U/L for males and females respectively, 16 patients (all with MAFLD) had elevated ALT at the time of CAP measurement, with a median level of 62 U/L (range, 57–143). Seven patients had liver biopsies within 1 year, all with steatosis and 3 with evidence of steatohepatitis.
Steatosis and graft fibrosis
There was no significant correlation between CAP and liver stiffness scores (p = 0.886), and no significant difference in the median liver stiffness between those with S2/3 and S0/1 steatosis (5.4 kPa vs. 4.9 kPa, p = 0.091). Of the 94 patients with moderate-severe steatosis, 4 (4.3%) had liver stiffness ≥ 12 kPa, compared to 3.5% among those with minimal or mild graft steatosis (p = 0.761; Supplementary Fig. 2). The characteristics of these 4 patients (all with MAFLD) are summarized in Supplementary Table 3. One patient had steatosis with bridging fibrosis with liver stiffness of 14.3 kPa, who was transplanted for acute flare of chronic hepatitis B, remained negative for serum hepatitis B surface antigen (HBsAg) and hepatitis B virus DNA after transplant without any episodes of rejection.
