NON INVASIVE METHODS FOR DIAGNOSING LIVER FIBROSIS

Abstract

The invention relates to a method comprising a) providing a blood sample from a subject b) determining the level of CPS-.sub.1 expression in said sample c) comparing the level of CPS-.sub.1 expression of (b) to the level of CPS-.sub.1 expression determined from a blood sample from a subject with mild to moderate fibrosis of the liver d) determining the level of glutamate in said sample e) comparing the level of glutamate of (d) to level of glutamate determined from a blood sample from a subject with mild to moderate fibrosis of the liver f) wherein if the level of glutamate of (d) and CPS-.sub.1 expression of (b) is higher than the level of glutamate and CPS-.sub.1 expression from a blood sample from a subject with mild to moderate fibrosis of the liver, it is inferred the subject has increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

Claims

1) A method comprising a) providing a blood sample from a subject b) determining the level of CPS-.sub.1 expression in said sample c) comparing the level of CPS-.sub.1 expression of (b) to the level of CPS-.sub.1 expression determined from a blood sample from a subject with mild to moderate fibrosis of the liver d) determining the level of glutamate in said sample e) comparing the level of glutamate of (d) to the level of glutamate determined from ablood samplefrom a subject with mild to moderate fibrosis of the liver f) wherein if the level of CPS-.sub.1 expression of (b) is higher than the level of CPS-.sub.1 expression determined from a blood sample from a subject with mild to moderate fibrosis of the liver, and the level of glutamate of (d) is higher than the level of glutamate determined from ablood sample from asubject with mild to moderate fibrosis of the liver, then it is inferred that the subject has an increased likelihood of having advanced or severe (F3/ F4) fibrosis of the liver.

2) A method comprising a) providing a blood sample from a subject b) determining the level of arginine in said sample c) comparing the level of arginine of (b) to the level of arginine determined from a blood sample from a subject with mild to moderate fibrosis of the liver d) determining the citrulline/ornithineratio in said sample e) comparing the citrulline/ornithine ratio of (d) to the citrulline/ornithine ratio determined from a blood sample from a subject with mild to moderate fibrosis of the liver f) determining the reserve capacity in said sample g) comparing the reserve capacity of (f) to the reserve capacity determined from a blood samplefrom a subject with mild to moderate fibrosis of the liver h) wherein if the level of arginine of (b) islower than thelevel of arginine determined from a blood samplefrom asubject with mild to moderate fibrosis of the liver, and the citrulline/ ornithine ratio of (d) is lower than the citrulline/ornithine ratio determined from a blood sample from asubject with mild to moderate fibrosis of the liver, and the reserve capacity of (f) is lower than the reserve capacity determined from a blood sample from a subject with mild to moderate fibrosis of the liver, then it isinferred that the subject has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

3) A method according to claim 1 wherein theCPS-1 level determined is the plasma CPS-1 level.

4) A method according to claim 1 or claim 3 wherein theCPS-1 level is determined by quantitative sandwich immunoassay.

5) A method according to claim 4 wherein said quantitative sandwich immunoassay comprises an ELISA assay.

6) A method according to any of claims 1, 3, 4 or 5 wherein the level of glutamate is determined using mass spectrometry.

7) A method according to claim 2 wherein thelevel of arginine is determined using mass spectrometry.

8) A method according to claim 2 or claim 7 wherein thelevelsof citrulline/ornithine are determined using mass spectrometry.

9) A method according to claim 2, 7 or 8 wherein the reserve capacity is determined as maximal OCR minus basal respiration.

10) A method according to claim 2, 7, 8 or 9 wherein the reserve capacity is determined using the ‘XF cell mito stress test kit’ from Agilent Technologies.

11) A method according to any preceding claim wherein the subject is suspected of having liver fibrosis.

12) A method according to claim 11 wherein the subject has been previously identified as having F0-F2 liver fibrosis, preferably F1-F2 liver fibrosis.

13) A method according to any preceding claim wherein the subject is suspected of having, or has, metabolic syndrome.

14) A method according to any preceding claim wherein the subject is suspected of having, or has, diabetes, preferably type 2 diabetes.

15) A method according to any preceding claim wherein the subject is suspected of having, or has, Non-Alcoholic Fatty Liver Disease (NAFLD).

16) A method according to claim 15 wherein the subject issuspected of having, or has, Non-Alcoholic Steatohepatitis (NASH).

17) A method comprising a) providing a first blood sample from a subject taken at afirst timepoint; b) either i. determining the level of CPS-.sub.1 expression in said sample and determining the level of glutamate in said sample, or ii. deter mining the level of arginine in said sample, determining the citrulline/ ornithine ratio in said sample and determining the reserve capacity in said sample; c) providing a second blood sample from a subject taken at a second timepoint; d) determining the same characteristics as were determined in step (b) for said second blood sample of step (c) e) comparing the values from step (b) to the values from step (d); f) inferring from the comparison of step (e) whether fibrosis has changed wherein if the values from step (b) and step (d) are different, then it is inferred that fibrosis has changed in the subject.

18) A method accordingto claim 17 wherein if thelevel of CPS-.sub.1 expression in said second sample and the level of glutamatein said second sample are higher than the levels for said first sample, then it isinferred that fibrosis has advanced or increased in said patient, and wherein if thelevel of CPS-.sub.1 expression in said second sample and thelevel of glutamate in said second sample are lower than the levels for said first sample, then it is inferred that fibrosis has receded or decreased in said patient.

19) A method according to claim 17 wherein if the level of argininein said second sample, the citrulline/ ornithine ratio in said second sample and the reserve capacity in said second sample are lower than the levels for said first sample, then it is inferred that fibrosis has advanced or increased in said patient, and wherein if thelevel of arginine in said second sample, the citrulline/ ornithine ratio in said second sample and the reserve capacity in said second sample are higher than the levels for said first sample, then it is inferred that fibrosis has receded or decreased in said patient.

20) A method of treating a subject with fibrosis of the liver, the method comprising performing a method according to any of claims 1 to 16 and if it is inferred that the subject has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver, then oneor more treatments selected from the group consisting of: reformed diet, exerciseregime, low calorie diet, and administration of GLPanalogue, such as a weekly injection of GLP analogue, is administered to said subject.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0247] The present invention will now be described by way of example, with reference to the accompanying drawings, in which:

[0248] FIG. 1 shows graphs; the fundamental parameters of mitochondrial function measured by XF cell mito stress test kit and a representative of a run of the experiment. Cellular mitochondrial profile in human PBMCs of NAFLD patient with mild /moderate fibrosis (blue) and severe fibrosis (red). Well-defined inhibitors, oligomycin (Oligo), FCCP and Rot/antimycin A (AntiA) are used in mito stress kit. PBMCs were counted and seeded at density of 300,000 cells/well. Using Seahorse XFp analyser, three measurement of basal respiration were taken, followed by an injection of oligomycin (0.75 .Math.M final) which indicated the ATP linked respiration. To assess maximal respiration, FCCP was injection at the 0.75 .Math.M working concentration. Reserve capacity was measured as a difference between maximal and basal respiration. Representative data shown as a Mean ± SEM, n=3 replicates per sample.

[0249] FIG. 2 shows bar charts - Mitochondrial dysfunction in live peripheral blood mononuclear cells (PBMCs) from NAFLD patients with severe fibrosis. PBMCs from HC (N= 5), mild/moderate liver fibrosis (N= 10) and patients with severe fibrosis (N=10) were isolated, seeded at 3×10.sup.5 cells/well, and the Seahorse XFp extracellular flux analyzer was used to measure Basal respiration (A) Maximal respiration (B) ATP linked respiration (C) Reserve capacity (D) Non mitochondria production (E) and Proton leak (F). Data presented as mean ± SEM and analysed by one-way ANOVA with Tukey’s test where *p<0.05, **p < 0.01

[0250] FIG. 3 shows bar charts – Glycolysis as represented by ECAR

[0251] FIG. 4 shows bar charts - Bioenergetic Health Index (Bill)

[0252] FIG. 5 shows bar charts - Phenotype Stress test showing metabolic potential

[0253] FIG. 6 shows bar charts - Citrulline/Ornithine Reversal in advanced Fibrosis in NAFLD

[0254] FIG. 7 shows bar charts - Reduced Arginine in advanced Fibrosis in NAFLD

[0255] FIGS. 8 shows plots; FIG. 8a - Principle Component Analysis to Detect Outliers. Heat map with Hierarchical Clustering discovers KCH-050319-18 as an Outlier; FIG. 8b - Partial least squares discriminant analysis Scores plot comparing healthy control (HC) plasma v NAFLD. R.sub.2 = 0.6, Q.sub.2 = 0.45, AUROC 0.96, CV ANOVA p<0.01.

[0256] FIG. 9 shows a bar chart - CPS-1 Levels by ELISA

[0257] FIG. 10 shows bar charts - Circulating Markers of Inflammation and Oxidative Stress in healthy controls, NAFLD patients with mild/moderate fibrosis and severe fibrosis. Plasma concentrations of (A) Interleukin-6 (IL-6) (B) lnterleukin-8 (IL-8) (C) TNF-alpha (D) Interleukin-13 (IL-13). Data are represented as means ± SEM (HC n=9, F1-F2 n=12, F3-F4 =8), *p < 0.05. **p < 0.01 and ***p <0.001

[0258] FIGS. 11 shows graphs - FIG. 11a: ROC curve analysis using glutamate + CPS-1 in F1-F2 versus F3-F4 NAFLD groups showed sensitivity of 0.95 and p=0.0007; FIG. 11b: ROC curve analysis using Citrulline/Ornithine + Arginine + Reserve capacity in F1-F2 versus F3-F4 NAFLD groups showed sensitivity of 0.94 and p=0.0006

[0259] FIG. 12 shows graphs.

EXAMPLES

Example 1

Methods

[0260] In this example, we investigated 30 subjects divided into 3 groups: subjects with NAFLD having mild fibrosis (n=10), subjects with NAFLD having severe fibrosis (n=10) and healthy controls (n=10). All NAFLD patients had biopsy proven fibrosis. Oxygen consumption rate (OCR) was measured in PBMCs using XFp Seahorse analyser. A mass spectrometric based untargeted metabolomic approach was used to analyze a broad range of human plasma metabolites in these samples. ELISA for CPS-1 and MSD for inflammatory markers was also done on corresponding plasma samples.

Findings

[0261] The mitochondrial bioenergetics showed reduced basal respiration, ATP linked respiration, maximal respiration and reserve capacity in PBMCs of NAFLD patients with severe fibrosis compared to NAFLD patients with mild fibrosis. The untargeted global metabolomic approach showed 13 metabolites which were significantly different between the mild and severe fibrosis of NAFLD patients. Most of these metabolites were involved in the pathways occurring in mitochondria.

Interpretation

[0262] Metabolites and bioenergetics measurements in peripheral blood samples can be incorporated together as a unique panel for diagnosing and staging of NAFLD.

Study Design and Participants

[0263] Patients were recruited between 2018-2019 with written informed consent from Kings College Hospital clinics under ethical approval from the regional Research Ethics Committee (REC; ref number 18/LO/1355). The cross-sectional study adhered to the Ethical Principles for Medical Research Involving Human Subjects, World Medical Association Declaration of Helsinki. The patients were studied in 3 groups: Group 1. Healthy control (N=10) who had no history of any chronic disease were recruited with informed consent, and were age and sex matched with the patient study group. Group 2 included patient with steatosis with mild/moderate fibrosis stage 1 or 2 (N=10) and group 3 were patient with advanced fibrosis stage 3 or 4 (N=10). All subjects in group 2 and 3 had liver biopsy consistent with NASH (defined as the presence of at least grade 1 steatosis, hepatocellular ballooning, and lobular inflammation according to the NAFLD Activity Score [NAS]) and fibrosis (1-4) according to the NASH CRN classification). There was no prior history of decompensated liver disease, chronic HBV and HCV infection, HIV positivity, other causes of liver disease or history of a malignancy or any other serious co-morbidities.

[0264] The 20 participants with NAFLD were aged 20-74 (Mean=52) and the female: male ratio was 8:12. The average BMI (kg/m2) in the group was 35 ± 6.5. Out of the 20 subjects, 14 (70%) had type 2 diabetes and 10 (50%) were diagnosed with hypertension. The mean systolic blood pressure was 136 ± 14 (mmHg) and diastolic blood pressure was 81 ± 9(mmHg). Fibroscan value for liver stiffness measured as LSM value was 11± 5.2 kPa and the controlled attenuation parameter (CAP) applied to assess the fat content value was 331±74. Table 1 shows the baseline characteristics of subjects with NAFLD used in this study.

TABLE-US-00006 Baseline characteristics of the NAFLD study cohort Variables F1- F2 Fibrosis F3-F4 Fibrosis Age (years) 49±15 54±14 (p=0.41) Gender (female:male) 3:7 5:5 BMI (kg/m.sup.2) 34±6 37±6(p=0.41) HbA1c (%) 6.8±1.2 6.8±1.3 (p=0.93) HTN 60% 40% DM 70% 70% Systolic BP (mmHg) 144±14 129±9* (p = 0.01) Diastolic BP (mmHg) 81±11 80±7 (p = 0.87) Cholesterol (mmol/l) 5.2±1.4 4.4±1.6 (p = 0.28) Triglycerides 2.2±0.9 2.6±1.6 (p = 0.49) HDL 1.1±0.3 1.1±0.3 (p = 0.67) LDL 3.1±1.2 1.8±1.0* (p = 0.03) FBS 139+62 125±41*(p = 0.60) Bilirubin-total 13±4 9+4* (p = 0.05) ALP 90±23 80±15 (p = 0.3) ALT 61±41 46±39 (p = 0.4) AST 37+1.4 32±19 (p = 0.55) GGT 80±89 66±37 (p = 0.7) Albumin 46±2 45±2 (p = 0.80) INR 1.1±0.1 1.0±0.1 (p = 0.21) Platelets 287±103 247±76 (p = 0.4) Uric Acid 399±59 329±54 (p = 0.08) FIB-4 Score 1.05±0.63 1.2±0.8 (p = 0.6) Fibroscan LSM value (kPa) 11.7+5.9 10.3±4.7 (p = 0.6) Fibroscan CAP value (dB/m) 357±45 310±89 (p = 0.15) Footnote: Data are means ± SD. BMI: Body Mass Index, HbA1c: Glycated haemoglobin, HTN: Hypertension, DM: Diabetes Mellitus, BP: Blood Pressure, HDL: High density lipoprotein, LDL: Low density lipoprotein, FBS: Fasting blood sugar, ALP: alkaline phosphatase, ALT: alanine aminotransferase, AST: aspartate amino-transferase, GGT: gamma glutamyl transferase. Key: *p<0.05

[0265] Currently no data is available to perform a formal sample size calculation for oxygen consumption rates in PBMCs of patients with NASH. The current study must be regarded as a pilot study. The aim of the study was to get an insight about the possible role of mitochondria dysfunction in the progression of simple steatosis to fibrosis in NAFLD. We estimated that 7-10 patients per group in NAFLD will be sufficient to get a reliable estimation of effect size. We proposed to include 5-10 healthy controls to establish reference values.

Procedures

Sample Preparation

[0266] Blood was collected in 8 ml Cell Preparation Tube (CPT) tubes (Becton Dickinson, Franklin Lakes, NJ, USA, ref. 362753) for separation of peripheral blood mononuclear cells (PBMCs) and plasma. The anticoagulant used in the tube is sodium heparin and cell separation is based on the principle using Ficoll method. The anticoagulated blood was collected by routine phlebotomy and tube inverted 8-10 times for mixing of anticoagulant and blood. The tubes were then centrifuged at room temperature (18-25° C.) in a horizontal rotor (swing-out head) for a minimum of 15 minutes at 1500 to 1800 x g. After centrifugation, PBMCs were visible as a whitish layer just under the plasma layer. The plasma was removed without disturbing the cell layer and separation of cells was done immediately following centrifugation for best results. After separation of PBMCs from plasma they were immediately counted, and viability checked with Countess Automated Cell Counter. PBMCs were plated on XFp 8-well of polystyrene plates designed for the Seahorse XFp analyser (Agilent Technologies Santa Clara, CA United States) within 2 hours. The plasma samples were immediately stored at -80° C. for metabolomic assays.

Measurement of Bioenergetics in Peripheral Blood Mononuclear Cells (PBMCs)

[0267] Cellular bioenergetics was performed using XF cell mito stress test kit in a Seahorse XFp analyzer (Agilent Technologies). PBMCs were suspended in XF media and 300,000 cells/well were seeded to Cell-Tak (Beckton Dickinson Ltd) coated XFp plates (Agilent Technologies). All experiments were performed with 3 replicate wells in the Seahorse XFp analyzer.

[0268] OCR, a measurement of mitochondrial respiration, and ECAR which correlates to number of protons released from the cell with potential contribution from glycolysis and the Krebs cycle, were measured in the presence of specific mitochondrial activators and inhibitors. Oligomycin (ATP synthase blocker) was used to measure ATP turnover and to determine proton leak, the mitochondrial un-coupler FCCP (carbonyl cyanide 4-[trifluoromethoxy] phenylhydrazone) was used to measure maximum respiratory function (maximal OCR). Reserve capacity was calculated as maximal OCR minus the basal respiration. At the end of the experiments, Rotenone (inhibitor of complex I) and Antimycin A (a blocker of complex III), were injected to completely shut the mitochondrial respiration down, to confirm that any changes observed in respiration were mitochondrial (Brand and Nicholls, 2011; Dranka et al., 2011) (FIG. 1)

[0269] For a measurement of basal respiration 3 measurements were taken before injecting ATP synthase inhibitor, Oligomycin at 0.75 .Math.M (final concentration). FCCP was then injected at 0.75 .Math.M. Finally, a mixture of rotenone and antimycin A (1 .Math.M) was injected. Mitochondrial basal respiration, proton leak, spare capacity and maximal respiration were measured after correcting for non-mitochondrial respiration. OCR and ECAR rates were normalised to cell count for PBMCs. Cell Energy Phenotype Test Kit which measures metabolic potential of cells was also used in a subset of experiments. The XF Mito stress test report generator automatically calculate the XF cell mito stress test parameters from Wave data that was exported to Excel.

Global Metabolomic Profiling:

[0270] A mass spectrometric based metabolomic approach based on the MxP® Global Profiling QUANT platform (Biocrates Life Sciences, Innsbruck, Austria) was used to analyze a broad range of human plasma metabolites covering various biochemical classes. A statistical method was applied to identify differences between the individual groups. The analyses included data normalization and transformation followed by univariate statistics with significance testing.

[0271] Plasma samples were extracted by a proprietary method and separated into lipid and polar fractions after precipitation of proteins (supplementary material; FIG. S1). Two types of mass spectrometry analysis were used for MxP® Global Profiling: Gas chromatography-mass spectrometry (GC-MS; Agilent 6890 GC coupled to an Agilent 5973 MS System, Agilent, Waldbronn, Germany) and Liquid chromatography-MS/MS (LCMS/MS; Agilent 1100 HPLC-System, Agilent, Waldbronn, Germany, coupled to an Applied Biosystems API4000 MS/MS-System, Applied Biosystems, Darmstadt, Germany) (van Ravenzwaay et al., 2007). For GC-MS analysis samples were sequentially derivatized before measurement. In LC-MS/MS analysis a Metanomics Health proprietary technology was applied which allows targeted and high sensitivity MRM (Multiple Reaction Monitoring) profiling in parallel to full screen analyses.

Normalization

[0272] Pooled reference samples derived from aliquots of all plasma samples were run in parallel throughout the entire analytical process and used to assess analytical process variability as part of the rigorous quality control performed at the metabolite, sample and experiment (study) level. Subsequently, data were normalized against the median in the pool reference samples to give pool-normalized ratios (performed for each sample per metabolite). This compensated for inter- and intra-instrumental variation.

[0273] To enable the MxP® Boost quantification and to allow for an experiment-to-experiment alignment of data, the MxPool™ concept was devised. The MxPool™ is a defined stock of the sample material to be analyzed and is stored in-house at Metanomics Health. Aliquots of the MxPool™ are analyzed within each experimental batch and within each project and all data from that batch are normalized to the MxPool™ in addition to the pool material generated for said batch, i.e. an additional normalization step of already experiment pool-normalized metabolite ratios is performed.

MxP® Boost Quantification

[0274] The metabolome of a large human plasma pool (MxPool™) was quantified by various methods, and results were cross-validated where possible. Up to now, more than 2.000 metabolites with absolute concentrations have been quantified in this material. Multiple aliquots of the MxPool TM material were measured within this project and used as one-point calibrator to yield metabolite concentrations. Metabolites that could not be quantified by this method were analyzed semi-quantitatively. Quality control is samples is given in supplementary section S1.

Carbamoyl Phosphate Synthase 1, Mitochondrial (CPS1) in Plasma by ELISA

[0275] Plasma CPS-1 concentration were measured using a commercial sandwich enzyme-linked immunosorbent assay (Human Carbamoyl Phosphate Synthase 1, Mitochondrial (CPS1) ELISA Kit Catalog No: RD-CPS1-Hu, RedDot Biotech limited, Kelowna, Canada) as per the manufacturer’s instructions (supplementary section S2).

Measuring Expression of Proinflammatory Cytokines

[0276] V-PLEX Proinflammatory Panel 1 Human Kit (Meso Scale Diagnostics, Rockville, USA) was used to investigate a range of cytokines namely IFN-γ, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70), IL-13, TNF-α in patients with NASH and HC as per manufacturer’s instructions (supplementary section S3).

Statistical Analysis

[0277] Statistical analysis was performed using GraphPad (GraphPad Software, Inc). The distribution of the data was tested using the Kolmogorov-Smirnov test (Graph pad) For parametric analysis, groups were compared using t-test (2 groups) or one-way ANOVA with post-hoc Tukey’s multiple comparison test (>2 groups). For non-parametric analysis, groups were compared using Mann-Whitney (2 groups) or Kruskal Wallis with Dunn’s post hoc test with Bonferroni correction (>2groups). Data was presented as mean ± Standard error of the mean.

[0278] Metabolomic data was analysed using univariate based on t-test and multi variate analysis comprising Principle Component Analysis (PCA) and Partial Least Squares-Discriminant Analysis (PLS-DA). Univariate statistics for the entire metabolomic dataset and each individual group to be analyzed statistically, the minimum, maximum, mean and median values were determined. Mean and median values were calculated on a logarithmic scale and then back-transformed to non-logarithmic scale. Variabilities within sample groups were evaluated by calculating the standard deviation of log10-transformed data. Subsequently, the standard deviation was transformed into a relative standard deviation (RSD) which was asymmetric due to the back-transformation from the logarithmic scale (i.e. RSD up and RSD down differ). The RSD down was calculated according to: RSD down =1-10^(-SDlog). The univariate analysis comprised one-way Analysis of Variance (ANOVA) and Tukey Honest Significant Difference (HSD) as post-hoc analysis. In contrast to multivariate analyses, single metabolite univariate models consider each metabolite independently. Accordingly, the results are not influenced by how many and which other metabolites are measured. In a first step, one-way ANOVA was used to assess whether the estimated sample means between the experimental groups differ from each other. This analysis revealed that the potential confounders gender and age had a significant effect on the data. Therefore, they were included in the univariate statistical model. Accordingly, the ANOVA model used GROUP, GENDER and AGE as factors, with GROUP being the factor of interest. In order to determine which of the multiple group comparisons produce significant differences, Tukey’s HSD test was applied as post-hoc analysis. The Tukey HSD test simultaneously compares all possible pairs of group means and additionally corrects for the type I error rate (Tukey, 1949).

Results

Mitochondrial Dysfunction in Patients With Advanced Fibrosis in NASH

[0279] We assessed bioenergetics in 3 groups: Healthy controls (N=5), patients of NAFLD with mild to moderate fibrosis (F1-F2) and patients of NAFLD with severe fibrosis (F3-F4).

[0280] The basal respiration (FIG. 2.1), ATP linked respiration (FIG. 2.2), maximal respiration (FIG. 2.3) and reserve capacity (FIG. 2.4) were all reduced in PBMCs of patients with severe fibrosis. Proton leak, non-mitochondrial respiration (FIG. 2.5 and FIG. 2.6) and ECAR (FIG. 3) were similar in the 3 groups. These data suggest that in patients with severe liver fibrosis there is mitochondrial dysfunction as manifested by a significant suppression of all mitochondrial parameters.

[0281] Basal respiration is the energetic demand of the cell under baseline conditions and reduction in this indicate that oxygen consumption used to meet cellular ATP demand is compromised in severe fibrosis (45 ± 6, n= 9) as compared to patients having mild fibrosis (86 ± 19, n= 7, p = 0.04, FIG. 2.1). Our results show that ATP-linked respiration by the mitochondria that contributes to meeting the energetic needs of the cell is also significantly reduced in severe fibrosis (40 ± 5, n= 9) versus mild/moderate fibrosis (74 ± 16, n= 7, p = 0.04) (FIG. 2). There was significant reduction in maximal respiration between mild/moderate fibrosis (242 ± 62, n= 7) and patients with F3 and F4 fibrosis (106 ± 25, n= 9, p ≥ 0.05) (FIG. 2). The reserve capacity was also significantly reduced in severe fibrosis (56 ± 16, n= 9) as compared to mild/moderate fibrosis (184 ± 42, n= 7, p = 0.0064) (FIG. 2).

[0282] The reduced reserve capacity and reduced maximal respiration are suggestive of a compromised response to stress in patients with advanced fibrosis.

Reduced Bioenergetic Health Index in PBMCs of NASH Patients With Advanced Fibrosis

[0283] The Bioenergetic Health Index (BHI) is a dynamic measure of the response of the body to stress. It is indicative of the dysfunctional metabolic response and any defects in the electron transport chain (ETC) will result in a lower BHI because of lower reserve capacity, ATP-linked respiration or increased uncoupling. BHI was calculated by using the bioenergetics data using the formula below as described by the Darley-Usmar group (Chacko et al., 2014): BHI= log (ATP-linked x reserve capacity)/ (proton leak x non-mitochondrial) The mean BHI value for patients with severe fibrosis (2.6 ± 0.2, n = 9, p = 0.0092) was significantly lower than in patients with mild/moderate fibrosis (3.7 ± 0.3, n == 6) (FIG. 4).

Defects in Metabolic Switching

[0284] In a sub-set experiment (n=9) we used Cell Energy Phenotype Test Kit (Agilent Technologies) which measures mitochondrial respiration and glycolysis under baseline and stressed conditions, to reveal the three key parameters of cell energy metabolism: Baseline Phenotype, Stressed Phenotype, and Metabolic Potential. By simultaneously measuring the two major energy producing pathways in live cells - mitochondrial respiration and glycolysis, we can determine the energy phenotypes of cells in patients with mild/moderate and advanced fibrosis in NASH.

[0285] Our results show that in patients with F3/F4 fibrosis there is reduced basal OCR, reduced stressed OCR and ECAR and reduced metabolic potential for mitochondrial respiration (FIG. 5). The basal ECAR and the metabolic potential for glycolysis is not significantly different between NAFLD patients with the mild and severe fibrosis.

Global Metabolomic Profiling Showing Altered Amino Acid Involved in Urea Cycle

[0286] Global untargeted metabolomics was performed on HC (n=9), patients with NAFLD with F1-F2 fibrosis (n=10) and F3-F4 fibrosis (n=10) using MxP® Global Profiling. Metabolite concentrations of each sample were determined in a single analysis. In the present study, we acquired data for a total of 493 metabolites, of which 401 were known metabolites and 92 unknown analytes. Furthermore, 25 metabolite to-metabolite ratios and sums were calculated and included in the statistical analyses. Univariate analysis was performed between the three groups.

[0287] Comparing mild/moderate with severe liver fibrosis in patients with NAFLD, 13 metabolites were significantly changed (p-value (F-Stats) < 0.05 and p-value (Tukey) < 0.05) out of the wide range of untargeted metabolomics data, (table 2). Out of these 13 metabolites, 5 are amino acids, 3 are choline ether lipids, 2 complex fatty acid lipids, 1 carbohydrate and 3 unknown metabolites (Table 2).

TABLE-US-00007 Metabolites significantly altered while comparing mild/moderate fibrosis NAFLD patients with severe fibrosis No Metabolites (p-value (F-Stats) < 0.05 and p-value (Tukey) < 0.05) 1 Glutamate 2 Unknown polar 3 Choline ether lipid (C40:5) 4 Choline ether lipid (C42:4) 5 Maltose 6 Unknown lipid 7 Citrulline / Ornithine 8 Histidine 9 Choline ether lipid (C44:6) 10 Sphingomyelin (C41:3) 11 Unknown lipid 12 Sum of Branched Amino Acids / Sum of Aromatic Amino Acids 13 Arginine 14 Phosphatidylcholine (C40:1)

[0288] Interestingly all the 5 amino acids which were significantly increased or decreased between mild/moderate fibrosis and severe fibrosis were part of urea cycle related to mitochondria. The citrulline/ ornithine ratio was reduced in severe fibrosis (0.26 ± 0.08, n = 10) compared to mild fibrosis (0.41 ± 0.15, n = 10, p = 0.02). Citrulline was low in severe fibrosis and ornithine was high. This was reversed in mild/moderate fibrosis with low ornithine and high citrulline (FIG. 6). Citrulline and ornithine are alpha amino acids and are part of Urea cycle in the liver. The conversion of ornithine to citrulline is the rate limiting step of urea cycle and occurs exclusively in mitochondria as compare to rest of the steps occurring in cytosol. Mitochondrial dysfunction can result in the disruption of the urea cycle and hence changes in the metabolites involved such as ornithine and citrulline. Arginine another component of urea cycle is also significantly reduced in severe fibrosis (40.4 ± 11.32, n = 10, p = 0.002) as compared to mild/moderate fibrosis (61.68 ± 15.22, n = 10) (FIG. 7). Histidine and Glutamate were upregulated in severe fibrosis (p < 0.001). Glutamate is a key compound in cellular metabolism and involved in both urea cycle and TCA cycle. Complex ether lipids which have antioxidant properties are also reduced in severe fibrosis. Maltose, a disachhride and precursor of glucose is high in F3-F4 fibrosis (p=0.01).

Changes in Metabolites in HC and NASH Patients

[0289] Metabolites were also analyzed between HC and patients having NASH. Our collaborator REVIVEMED performed Hierarchical Clustering to remove any outlier. Only 1 sample was an outlier (Supplementary material). Using their platform, molecular networks associated with significantly dysregulated metabolites between Healthy and NASH patients are shown in table.

[0290] Using Partial Least Squares-Discriminant Analysis (PLS-DA), healthy controls and patients with NAFLD could be distinguished (FIGS. 8)

3.6: Increased Carbamoyl Phosphate synthetase-1(CPS-.SUB.1.) Levels in Plasma Of Patients with Advanced Fibrosis in NASH

[0291] To further evaluate the changes in the metabolites of urea cycle especially the first-rate limiting step occurring in mitochondrial matrix we investigated the expression of CPS-1 levels in plasma of the corresponding patients. CPS-1 occurs in mitochondria and results in formation of carbamyl phosphatase which is utilized in conversion of ornithine to citrulline. As our results from metabolomics showed decreased citrulline and high ornithine in patients with high fibrosis we measured the plasma levels of CPS-1 to analyse any changes in this key enzyme of urea cycle. Plasma CPS.sub.1 levels were measured in patients with NAFLD having different degrees of fibrosis and healthy controls using the ELISA kit for quantitative sandwich immunoassay technique. In comparison to the HC (n=9), the NASH patients with mild/moderate (n=10) and severe fibrosis (n=10) had significantly high CPS-.sub.1 levels in plasma (HC versus F1-F2: p =0.02 and HC versus F.sub.3-F.sub.4: p= 0.0003) (FIG. 9). The CPS-.sub.1 values were also significantly high in F.sub.3-F.sub.4 versus F.sub.1-F.sub.2 (p =0.02).

3.7: Increased Expression of Pro Inflammatory Cytokines in Severe Fibrosis In Patient with NASH

[0292] Proinflammatory human cytokines kit (Meso Scale Diagnostics, Rockville, USA) was used to investigate a range of cytokines namely IFN-γ, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p70, IL-13 and TNF-α to understand the systemic inflammatory response due to oxidative stress. Interestingly proinflammatory cytokines IL-6, IL-8 and TNF-α were significantly increased in NASH patients with fibrosis compared to HC (FIG. 10). IL-13 which is considered as anti-inflammatory cytokine was significantly higher in HC as compared to NASH patients. These results showed that the systemic response resulting from mitochondrial dysfunction cause a cascade of inflammatory reactions which also is evident in the cytokines.

3.8: Biomarker Panel for Liver Fibrosis in NASH Patients

[0293] Combining the bioenergetic and metabolomic data we analyzed two separate panel of biomarkers: panel with increase expression of metabolites and panel with decrease expression of metabolites.

[0294] The panel for increased expression of metabolite included measuring glutamate and plasma levels of CPS-.sub.1. The ROC curve analysis showed the sensitivity of 95% and p=0.0007 (F1-F2 versus F.sub.3-F.sub.4) FIG. 11-a.

[0295] The other panel with decreased metabolites was based on levels of citrulline/ornithine ratio, Arginine and bioenergetic value of reserve capacity. The ROC curve analysis showed the sensitivity of showed sensitivity of 0.94 and p=0.0006 (F.sub.1-F.sub.2 versus F.sub.3-F.sub.4) FIG. 11-b.

Discussion

[0296] The main function of mitochondria is the generation of by oxidative phosphorylation, but apart from energy production mitochondria also have many other functions in cellular metabolism (Brand and Nicholas, 2011). As mitochondria regulate metabolic and energy homeostasis, its dysfunction is implicated in the pathophysiology of numerous chronic conditions such as obesity, insulin resistance, Type 2 Diabetes Mellitus (T2DM), diabetes-related complications (Holmstrom et al and Cjaka et al) and non-alcoholic fatty liver disease (NAFLD) (Perez-Carreras et al) etc. Maintaining normal mitochondrial function is critical for survival and maintenance of stable biological functions and cellular repair. We propose that mitochondrial dysfunction in hepatocytes can result in a systemic inflammatory response due to liver injury and oxidative damage. In the liver, this is accompanied with disturbance in the metabolic pathways which can be evaluated by measuring level of different metabolites in the peripheral blood. Mitochondrial functional changes cause impairment of fatty acid b-oxidation resulting in a vicious cycle of increased lipid intermediates, insulin resistance and reactive oxygen species leading to more inflammation and hepatocyte necrosis. These changes can be measured in peripheral cells due to a global immune response related to mitochondrial dysfunction in hepatocytes. Peripheral blood cells such as leucocytes and platelets can act as surrogate markers for different chronic diseases (A Perl et al, Cjaka et al, Rudkowska, I et al, Zharikov and Shiva).

[0297] In this study, we have shown for the first time an integrated approach for investigating mitochondrial function and untargeted metabolomics in patients with different stages of fibrosis in NAFLD. Our aim was to establish novel platform for non-invasive biomarkers in progression of NAFLD by using bioenergetics and metabolite changes.

[0298] The role of mitochondrial dysfunction in progression liver disease is not a new concept (Christie and Judah, 1954). Mitochondrial dysfunction in NASH has been hinted at histologically using electron microscopy where swollen and rounded hepatocellular mitochondria (Sanyal et al) were observed. However, with techniques to measure bioenergetics in live PBMCs it is possible to more fully study mitochondrial function. In our study, all parameters of mitochondrial function namely basal respiration, ATP linked respiration, maximal respiration and reserve capacity were significantly reduced in patients with severe (but not mild) fibrosis.

[0299] Reserve capacity serves to meet increase energy demands especially during periods of oxidative stress. Defects in the ETC can result in a lower BHI because of the lower reserve capacity, ATP-linked respiration or increased uncoupling. It is an indicator of bioenergetic health in real-time and can serve as prognostic value for identifying progressive deterioration in mitochondrial function. Metabolic potential for mitochondrial respiration was also reduced in advanced fibrosis but interestingly the potential for glycolysis was not significantly different between mild/moderate fibrosis and advanced fibrosis.

[0300] Mitochondria is an important cellular organelle as most significant metabolic pathways such as TCA cycle, beta-oxidation of fatty acids, rate-limiting steps of urea cycle, heme biosynthesis, cardiolipin synthesis, quinone and steroid biosynthesis occur completely or partly in the mitochondria. Progression of steatohepatitis likely involves both direct injury from excess fatty acids oxidation and increased oxidative stress within hepatocytes which forms the concept of double hits or insults to the hepatocyte. This creates a vicious cycle where fat accumulation causes defects in ETC leading to failure to generate reduced NAD and FAD, which further effects fatty acid oxidation and other metabolic pathways (Tarek Hussein).

[0301] Our untargeted metabolomic approach showed about 14 metabolites highly significant between the mild/moderate and severe fibrosis in NASH patients (Table 2). Out of these, five significant metabolites were amino acids which are all linked with the urea cycle or TCA. The citrulline/ornithine ratio was significantly reduced (P ≥ 0.05) in NASH patients with severe fibrosis (FIG. 7). Citrulline and ornithine are the most important metabolites in the Urea cycle pathway which is exclusively located in liver. This is the only rate limiting step in urea cycle which occurs inside mitochondria whereas rest of the steps of urea cycle are cytosolic. Recently De Chiara, Francesco et al. has shown that urea cycle enzymes are affected in rats models of NASH and humans resulting in hyperammonemia and impairment of urea synthesis. They suggested strategy of targeting ammonia as a potential treatment for NASH. In 2019, Canbay and Sowa also published a study suggesting the role ofl-ornithine 1-aspartate (LOLA) a known effective ammonia-lowering agent for treatment of NAFLD due to its actions of enhanced ammonia removal, increased anti-oxidative capacity, and attenuated lipid peroxidation by glutamine and glutathione and improved hepatic microcirculation due to 1-arginine-derived NO (Canbay and Sowa). K.L. Thomsen et al. also suggested hyperammonemia in NASH results in progression of fibrosis and hypothesized that treating ammonia can be a potential target for prevention of fibrosis progression of patients with NASH (K.L Thomsen et al).

[0302] The other metabolite significantly reduced in NASH patients with severe fibrosis was Arginine which also forms part of urea cycle. Arginine is hydrolyzed to form urea and ornithine. Reduced formation of Arginine can result in less urea production and more ammonia in the body. Thus, defects in urea cycle also effects the TCA which is a central driver of cellular respiration and vice versa. Glutamate another amino acid was significantly higher in NASH patients with severe fibrosis. Glutamate dehydrogenase (GDB) is an enzyme, present again only in the mitochondria and required for urea synthesis, that converts glutamate to α-ketoglutarate, and vice versa.

[0303] To further explore the role of metabolites involved in urea cycle as potential biomarkers for progression of fibrosis in NASH we measured CPS-.sub.1 levels in plasma of the same set of NASH patients and HC by ELISA. CPS1 is the most abundant protein in liver mitochondria. Our results showed highly significant levels of CPS-.sub.1 in plasma of NASH patients with severe fibrosis as compared to mild/moderate fibrosis and HC. CPS-.sub.1 was also significantly higher in NASH patients with mild/moderate fibrosis as compared with the HC. Previously CPS1 has been found in serum or plasma of sepsis animal models and plasma of human septic patients, suggesting that it might serve as a serum marker for detecting mitochondrial injury of the liver under septic conditions (Crouser et al; Struck et al). Another study has also shown that the hepatocyte-selective and most abundant mitochondrial matrix protein CPS1 is a marker for apoptotic and necrotic forms of hepatocyte death and injury and is released in the circulation after acute liver injury (Sujith et al). El-Sheikh et al. has recently shown that the tissue and serum CPS-.sub.1 correlated significantly in moderate and severe fibrosis in HCV patients and in these patients, there was significantly higher levels of serum CPS1 and lower mitochondrial counts than those with moderate fibrosis.

[0304] Proinflammatory cytokines IL-6, 1L-8 and TNF-a were significantly increased in NASH patients with fibrosis compared to HC indicating that mitochondrial dysfunction is accompanied with systemic immune response and utilizing combined approach of measuring mitochondrial bioenergetics and metabolomics we can elaborate novel biomarkers for progression of fibrosis in NASH. We propose platform of non-invasive biomarkers based on bioenergetics, metabolites involved in urea cycle notably citrulline/ornithine ratio, arginine, glutamate and CPS-.sub.1 levels in plasma for identifying degree of fibrosis in NASH patients.

[0305] Confirmatory measurement of bioenergetics in corresponding liver tissue is taught to confirm that the changes in hepatocytes are accompanied by systemic changes in immune cells which can be utilized as non-invasive biomarkers for liver fibrosis.

[0306] In conclusion, this is the first study to show mitochondrial functional changes in peripheral cells with accompanied with changes in urea cycle metabolites which can clearly differentiate mild and severe fibrosis in NAFLD patients.

Example 2 - Demonstrations of Sensitivity and Specificity

[0307] Combining the bioenergetic and metabolomic data we analysed two separate panels of biomarkers: panel with increased expression of metabolites and panel with decreased expression of metabolites.

[0308] In this example the panel for increased metabolites included measuring glutamate and plasma levels of CPS-.sub.1. The ROC curve analysis showed the sensitivity of 95% and p=0.0007 (F1-F2 versus F3-F4) See FIG. 12 (left graph).

[0309] In this example the panel for decreased metabolites was based on levels of citrulline/ornithine ratio, Arginine and bioenergetic value of reserve capacity. The ROC curve analysis showed the sensitivity of showed sensitivity of 0.94 and p=0.0006 (F1-F2 versus F3-F4) FIG. 12 (right graph).

[0310] In more detail in the graphs of FIG. 12 there is only one solid line on each and that shows Area Under the Curve (AUC) which shows relationship between sensitivity and specificity. So from FIG. 12 it can be seen that the ROC curves quantify the overall ability of the methods of the invention to discriminate between those individuals with the disease (advanced/severe fibrosis i.e. F3/F4) and those without the disease (mild or moderate i.e. F1/F2). The closer the curve follows the left-hand border and then the top border of the ROC space, the more accurate the test.

[0311] FIG. 12 (left) and 12 (right): ROC curve analysis using glutamate + CPS-.sub.1 in F1-F2 versus F3-F4 NAFLD groups showed sensitivity of 0.95 and p=0.0007 and for Citrulline/Ornithine + Arginine + Reserve capacity in F1-F2 versus F3-F4 NAFLD groups showed sensitivity of 0.94 and p=0.0006.

Example 3a - Clinical Application (Increasing Panel)(Increasing Values with Advanced Fibrosis)

[0312] In this example we compare values of a subject of interest with a comparator/reference subject (i.e. matched subject) with mild to moderate fibrosis (F1/F2 fibrosis).

[0313] In this example the subject with mild to moderate fibrosis has mild fibrosis.

[0314] In this example the subject of interest has F4 fibrosis.

[0315] 1) A method comprising [0316] a) providing a blood sample from a subject [0317] b) determining the level of CPS-.sub.1 expression in said sample [0318] in this example CPS-value=5.8 [0319] c) comparing the level of CPS-.sub.1 expression of (b) to the level of CPS-.sub.1 expression determined from a blood sample from a subject with mild to moderate fibrosis of the liver [0320] in this example comparing CPS-value=5.8 to CPS-value=1.2 from a subject with mild to moderate fibrosis [0321] d) determining the level of glutamate in said sample [0322] in this example Glutamate value= 309 [0323] e) comparing the level of glutamate of (d) to the level of glutamate determined from a blood sample from a subject with mild to moderate fibrosis of the liver [0324] in this example comparing Glutamate value= 309 to Glutamate value= 78 from a subject with mild to moderate fibrosis [0325] f) wherein if the level of CPS-.sub.1 expression of (b) is higher than the level of CPS-.sub.1 expression determined from a blood sample from a subject with mild to moderate fibrosis of the liver, and [0326] the level of glutamate of (d) is higher than the level of glutamate determined from a blood sample from a subject with mild to moderate fibrosis of the liver, [0327] then it is inferred that the subject has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

[0328] In this example the comparisons show that the subject of interest has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

[0329] In fact the subject of interest has F4 fibrosis. This shows the method of the invention in operation.

Optional Additional Step

[0330] Patient with advanced fibrosis: Average of glutamate and CPS scores = 157

[0331] Patient with mild fibrosis: Average of glutamate and CPS scores = 39

[0332] The comparison of the averages shows a large increase in the subject of interest.

[0333] This provides an enhanced confidence that the subject of interest has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

Example 3b - Clinical Application (Decreasing Panel) (Decreasing Values with Advanced Fibrosis)

[0334] In this example we compare values of a subject of interest with a comparator/reference subject (i.e. matched subject) with mild to moderate fibrosis (F1/F2 fibrosis).

[0335] In this example the subject with mild to moderate fibrosis has mild fibrosis.

[0336] In this example the subject of interest has F4 fibrosis.

[0337] 2) A method comprising [0338] a) providing a blood sample from a subject. [0339] b) determining the level of arginine in said sample [0340] in this example Arginine value =39 [0341] c) comparing the level of arginine of (b) to the level of arginine determined from a blood sample from a subject with mild to moderate fibrosis of the liver [0342] in this example comparing Arginine value= 39 to Arginine value= 55 from a subject with mild to moderate fibrosis [0343] d) determining the citrulline/ornithine ratio in said sample [0344] in this example citrulline/ornithine ratio value = 0.3 [0345] e) comparing the citrulline/ornithine ratio of (d) to the citrulline/ornithine ratio determined from a blood sample from a subject with mild to moderate fibrosis of the liver [0346] in this example comparing citrulline/ornithine ratio value = 0.3 to citrulline/ornithine ratio value = 0.5 from a subject with mild to moderate fibrosis [0347] f) determining the reserve capacity in said sample [0348] in this example Reserve capacity =30 [0349] g) comparing the reserve capacity of (f) to the reserve capacity determined from a blood sample from a subject with mild to moderate fibrosis of the liver [0350] in this example comparing Reserve capacity =30 to Reserve capacity =74 from a subject with mild to moderate fibrosis [0351] h) wherein if the level of arginine of (b) is lower than the level of arginine determined from a blood sample from a subject with mild to moderate fibrosis of the liver, and [0352] the citrulline/ornithine ratio of (d) is lower than the citrulline/ornithine ratio determined from a blood sample from a subject with mild to moderate fibrosis of the liver, and [0353] the reserve capacity of (f) is lower than the reserve capacity determined from a blood sample from a subject with mild to moderate fibrosis of the liver, [0354] then it is inferred that the subject has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

[0355] In this example the comparisons show that the subject of interest has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

[0356] In fact the subject of interest has F4 fibrosis. This shows the method of the invention in operation.

Optional Additional Step

[0357] Patient with advanced fibrosis: Average of arginine and citrulline/ornithine ratio and reserve capacity scores = 43

[0358] Patient with mild fibrosis: Average of arginine and citrulline/ornithine ratio and reserve capacity scores = 23

[0359] The comparison of the averages shows a large decrease in the subject of interest.

[0360] This provides an enhanced confidence that the subject of interest has an increased likelihood of having advanced or severe (F3/F4) fibrosis of the liver.

Example 4 - Clinical Application

[0361] The values for the biomarkers taught herein were studied in a substantial cohort of subjects.

[0362] The average values across the cohorts are shown in the tables below

TABLE-US-00008 Biomarker F1-F2 cohort F3-F4 cohort Glutamate 64 .Math.mol/L 137 .Math.mol/L CPS1 2.7 ng/ml 4.3 ng/ml Biomarker F1-F2 cohort F3-F4 cohort Arginine 62 .Math.mol/L 40 .Math.mol/L Citrulline/Ornithine Ratio 0.41 0.2 Reserve Capacity 185 pmol/min 56 pmol/min

[0363] The average values of the F1-F2 cohort serve useful purpose as comparative values (i.e. for comparison to the values from the subject of interest, thereby advantageously avoiding having to process additional samples/additional measurements of an F1/F2 subject each time the method is carried out).

[0364] Although illustrative embodiments of the invention have been disclosed in detail herein, with reference to the accompanying drawings, it is understood that the invention is not limited to the precise embodiments shown and that various changes or modifications can be effected by one skilled in the art without departing from the scope of the invention as defined by the appended claims and their equivalents.