Differential fuel utilization in liver transplant recipients and its relationship with non‐alcoholic fatty liver disease

Abstract Metabolic flexibility is the ability to match biofuel availability to utilization. Reduced metabolic flexibility, or lower fatty acid (FA) oxidation in the fasted state, is associated with obesity. The present study evaluated metabolic flexibility after liver transplantation (LT). Methods Patients receiving LT for non‐alcoholic steatohepatitis (NASH) (n = 35) and non‐NASH (n = 10) were enrolled. NASH was chosen as these patients are at the highest risk of metabolic complications. Metabolic flexibility was measured using whole‐body calorimetry and expressed as respiratory quotient (RQ), which ranges from 0.7 (pure FA oxidation) to 1.0 is (carbohydrate oxidation). Results The two cohorts were similar except for a higher prevalence of obesity and diabetes in the NASH cohort. Post‐prandially, RQ increased in both cohorts (i.e. greater carbohydrate utilization) but peak RQ and time at peak RQ was higher in the NASH cohort. Fasting RQ in NASH was significantly higher (0.845 vs. 0.772, p < .001), indicative of impaired FA utilization. In subgroup analysis of the NASH cohort, body mass index but not liver fat content (MRI‐PDFF) was an independent predictor of fasting RQ. In NASH, fasting RQ inversely correlated with fat‐free muscle volume and directly with visceral adipose tissue. Conclusion Reduced metabolic flexibility in patients transplanted for NASH cirrhosis may precede the development of non‐alcoholic fatty liver disease after LT.


| INTRODUC TI ON
Weight gain and obesity are common after liver transplantation (LT) and are associated with increased risk of cardiovascular disease, dyslipidemia, diabetes and reduced survival after LT. [1][2][3][4][5][6] Obesity is also a known as risk factor for the development of post-LT nonalcoholic fatty liver disease (NAFLD), which occurs in nearly a third of patients receiving LT for non-NAFLD cirrhosis and universally among patients receiving LT for non-alcoholic steatohepatitis (NASH) cirrhosis. 4 Efficient energy homeostasis is central to weight maintenance and perturbations in energy homeostasis are linked to obesity and the development of NAFLD in the non-LT population. 9,10 In normal physiology, energy metabolism is characterized by periodic shifts between glucose and fatty acid (FA) oxidation by the skeletal muscle depending on fuel availability. 11 In the fed state, when carbohydrate supply is ample, meal-induced insulin secretion facilitates its use as the preferred fuel source while inhibiting lipolysis and promoting the storage of excess carbohydrates as fat in adipose tissue. Conversely, in the fasted state, when dietary carbohydrate intake declines, the decrease in serum insulin levels promotes lipolysis. This results in a steady supply of FA to be used as the major fuel source during fasting. 11,12 The ability to preferentially use available biofuel for the energy-demanding biological process is referred to as metabolic flexibility and is associated with weight maintenance. 11 Metabolic inflexibility is the inability of skeletal muscle to preferentially utilize FA for oxidation during the fasted state and is associated with weight gain. 10,13,14 Blunted FA oxidation results in the recycling of unused free FA back to adipose tissue, which is then esterified as triacylglycerol for storage, thereby, promoting adiposity. 15 The aim of the current study was to better understand energy homeostasis and metabolic flexibility in LT recipients. As patients receiving LT for NASH cirrhosis have the greatest propensity for weight gain and development of NAFLD, we hypothesized that patients transplanted for NASH cirrhosis will have reduced metabolic flexibility compared with patients transplanted for non-NASH indications. Furthermore, since skeletal muscle is the central organ in energy homeostasis, an inverse relationship between metabolic flexibility and skeletal muscle function will be observed.

| Study design and participants
The study was approved by the institutional review board (IRB) and all authors have approved the manuscript prior to submission. Adult (age ≥ 18 years) LT recipients who were at least 6 months post-LT were invited to participate in the study. The study enrolled subjects transplanted for NASH and non-NASH cirrhosis. The diagnosis of NASH as aetiology of cirrhosis requiring LT was established if patients had (1) a prior biopsy showing NASH with progression to cirrhosis, (2) evidence of steatohepatitis or steatosis on explant or (3) prior history of metabolic co-morbid conditions even if explant did not show NASH/NAFLD (i.e. burnt out NASH) after a negative serological evaluation and less than moderate alcohol consumption. 16 Patients transplanted for non-NASH cirrhosis who developed post-LT NAFLD as evidenced by controlled attenuation parameter on vibration-controlled transient elastography greater than 270 dB/m were excluded in this proof of concept study to avoid the potential confounders from post-LT de novo NAFLD. 17 Additional exclusion criteria included multi-organ transplants, renal failure requiring haemodialysis, prednisone use, gastroparesis, non-dermatological malignancy and poorly controlled diabetes defined by HbA1c > 8.5%.
Patients with acute/chronic rejection, vascular and biliary complications within 6 months of screening were also excluded.  19 Following the administration of the standardized meal, the patients continued to fast for an additional 12 h. Total energy expenditure (EE), CO 2 production and O 2 consumption were recorded every 1 min for a total of 18 h. While the patient was inside the Conclusion: Reduced metabolic flexibility in patients transplanted for NASH cirrhosis may precede the development of non-alcoholic fatty liver disease after LT.

K E Y W O R D S
carbohydrates, energy expenditure, fatty acids, liver transplantation, metabolic flexibility, nonalcoholic steatohepatitis chamber, only normal physical activity was allowed (i.e. no exercise).
Mitochondrial fuel use was quantified by measuring whole-body CO 2 production relative to O 2 consumption or respiratory quotient (RQ). The RQ oscillates between 0.7 and 1.0, which is indicative of either predominantly FA or glucose oxidation respectively.

Anthropometric measurements including height and weight
were recorded at the time of admission to the CRSU. Body composition was quantified via magnetic resonance imaging (MRI). After an overnight fast, patients were scanned in a research-dedicated Phillips Ingenia 3.0T MRI scanner using a 6-min dual-echo Dixon protocol, providing water-and fat-separated volumetric data set covering neck to knees. Body composition profiling was performed using AMRA® Researcher. 20

| Statistical analysis
Data are presented as means with standard deviation or frequency and percentage as appropriate. The RQ (CO 2 production to O 2 consumption ratio) was plotted against time at 1-min intervals for 18 h.
The RQ curves were smoothed using local linear regression and the smoothing parameter was selected via cross-validation. This was subsequently modelled with RQ on the y-axis and time on the x-axis using a cubic B-spline to better fit the data. The graph was interrogated to identify biologically relevant times points that included RQ at the time of standardized meal administration (180 min). To determine efficient and maximal carbohydrate utilization, the time to peak RQ after administration of standard meal was determined.
Next, to determine if both groups had the equal capacity to utilize carbohydrates, the peak RQ between the two groups was compared.
To determine the whole-body FA oxidation during the fasting state, RQ 360 min after a standardized meal was compared between the two groups as well as the lowest RQ after meal administration. To better understand the relationship between biofuel utilization (i.e. RQ) and bioclinical parameters, linear model of residuals of the cubic B-spline models against gender, diabetes, immunosuppression (tarcrolimus use), age, body mass index (BMI), FFMVi, VATi, and MFI adj were generated.
To better understand the relationship between obesity, NAFLD and metabolic flexibility (i.e. fasting RQ), a staged analysis was completed. In the first step, the relationship between NASH diagnosis and BMI was evaluated by using both as co-variates in predicting fasting RQ in the entire cohort. Next, to better understand the interplay between NASH diagnosis and BMI, subgroup analysis was performed in patients transplanted for NASH cirrhosis in which BMI as a predictor of fasting RQ was performed using multiple linear regression. In the final step, in subgroup analysis in a cohort of patients receiving LT for NASH cirrhosis, liver fat as measured by MRI-PDFF and BMI were used as co-variates in predicting fasting RQ using multiple linear regression.
For all patients, EE was measured directly at 1-min intervals in the whole room calorimeter. EE was graphed versus time and smoothed using local linear regression and the smoothing parameter was selected via cross-validation. Resting EE (REE) was measured in the fasting state, while the subjects were resting in a quiet surrounding at 24°C. The REE curves of the two groups were compared. The relationship between REE and gender, diabetes, immunosuppression (tarcrolimus use), age, BMI, FFMVi, VATi and MFI adj was evaluated using linear models of the cubic B-spline models. A nominal p-value of <.5 was considered statistically significant.

| Patient characteristics
The study cohort consisted of 45 subjects that underwent LT for NASH (n = 35) and non-NASH (n = 10) indications. The mean age of patients transplanted for NASH vs. non-NASH cirrhosis was similar (61 ± 9 vs. 60 ± 12 years, p = .75). The cohorts were similar with regard to gender and ethnicity (Table 1). While the prevalence of hypertension was similar across the two cohorts, patients transplanted for NASH cirrhosis were more likely to have diabetes and dyslipidemia (Table 1). Serum aminotransferase, alkaline phosphatase and bilirubin levels were similar between the two cohorts. Expectedly, patients transplanted for NASH cirrhosis had lower serum high density lipoprotein cholesterol (HDL-C) (43 ± 12 vs. 53 ± 10 mg/dl, p = .02) and higher triglyceride (170 ± 112 vs. 102 ± 77, p = .043) levels. Finally, the two cohorts were similar with regard to transplantrelated metrics including time from LT and immunosuppressant use.

| Body composition
Patients transplanted for NASH cirrhosis had a higher BMI compared to those transplanted for non-NASH cirrhosis (37.1 ± 5.5 vs. 26.2 ± 5.0 kg/m 2 ; p < .0001). Similarly, patients transplanted for NASH cirrhosis compared to non-NASH cirrhosis had higher VATi, ASATi and lower FFMVi ( Figure 1A). Patients receiving LT for NASH cirrhosis had significantly higher MFI than patients transplanted for non-NASH cirrhosis ( Figure 1B). In multivariate analysis, FFMVi was positively associated with BMI and tacrolimus (vs. cyclosporine use) and inversely with female gender and presence of diabetes (Table 2).
MFI was positively associated with BMI and diabetes. The ASATi was positively associated with male gender, BMI and presence of diabetes but none of the other parameters including immunosuppression and time from LT. Finally, VATi was directly associated with BMI and cyclosporine use.

| Metabolic flexibility
The whole-body energy utilization in patients transplanted for NASH vs. non-NASH cirrhosis is depicted in Figure 2. Immediately prior to standardized meal administration, patients transplanted for NASH cirrhosis had higher baseline RQ (p < .05). Post-prandially, the RQ increased for both cohorts, indicative of increasing carbohydrate utilization following a standardized meal. In the post-prandial period, the time to reach RQ peak or maximal carbohydrate metabolism was similar in patients transplanted for NASH vs. non-NASH cirrhosis (95% CI: 416 ± 238 vs. 385 ± 144 min, p = .39). However, patients in the NASH cohort had a higher peak RQ (95% CI: 0.849 ± 0.002 vs. 0.829 ± 0.003 min, p < .001), indicative of greater carbohydrate utilization. Furthermore, patients in the NASH cohort had higher remained at peak RQ significantly longer than the non-NASH cohort (p < .001; Figure 2A

| Risk factors for reduced metabolic flexibility
To gain a deeper understanding of clinical predictors of metabolic flexibility, we generated a linear model in which RQ overtime was correlated with clinical and body composition parameters (

| Relationship between BMI, liver fat and metabolic flexibility
In this analysis, using the whole cohort, BMI and NASH diagnosis  Figure 3A,B).

| Energy expenditure
Energy expenditure was measured directly for each participant within the whole room calorimeter (Figure 4)

| DISCUSS ION
In the present study, we demonstrate that patients transplanted for NASH cirrhosis have reduced metabolic flexibility, which correlates with the quality of skeletal muscle. Physiologically, the body readily transitions between carbohydrate metabolism during the fed state to FAs during the fasted state and reflects the relative abundance of these biofuels during these states. 11 Rapid transitions between available substrates are possible because of the ability of mitochondria to readily utilize the most abundantly

| Relationship between weight, NALFD and metabolic flexibility
The clinical impact of reduced metabolic flexibility is higher recycling

| Strengths and limitations
The findings from our observations must be evaluated in the context of their limitations. By design, this is a proof of concept study to evaluate biofuel utilization after LT. Thus, these findings should not be extrapolated to non-LT patients with NASH. Furthermore, patients were dichotomized into NASH vs. non-NASH, however, additional prospective studies are required to provide granularity on how other causes of chronic liver disease (i.e. alcohol, hepatitis C, In conclusion, we demonstrated that patients transplanted for NASH cirrhosis have impaired biofuel utilization which is associated with skeletal muscle structure. These findings underscore the importance of better defining the pathophysiology of post-LT weight gain and obesity so as to provide targeted therapy with the longterm goal of improving outcomes in LT recipients.