Advertisement
Clinical Science| Volume 101, 154005, December 2019

Non-invasive diagnosis of non-alcoholic steatohepatitis and fibrosis with the use of omics and supervised learning: A proof of concept study

  • Author Footnotes
    1 Indicates equal co-first authorship.
    Nikolaos Perakakis
    Correspondence
    Corresponding authors at: SL-418, 330 Brookline Avenue, East campus, Beth Israel Deaconess Medical Center, Boston, MA 02215, USA.
    Footnotes
    1 Indicates equal co-first authorship.
    Affiliations
    Division of Endocrinology, Diabetes and Metabolism, Department of Internal Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Author Footnotes
    1 Indicates equal co-first authorship.
    Stergios A. Polyzos
    Footnotes
    1 Indicates equal co-first authorship.
    Affiliations
    First Department of Pharmacology, Faculty of Medicine, Aristotle University of Thessaloniki, Thessaloniki, Greece
    Search for articles by this author
  • Author Footnotes
    1 Indicates equal co-first authorship.
    Alireza Yazdani
    Footnotes
    1 Indicates equal co-first authorship.
    Affiliations
    Division of Applied Mathematics, Brown University, Providence, RI 02906, USA
    Search for articles by this author
  • Aleix Sala-Vila
    Affiliations
    CIBER de Fisiopatología de la Obesidad y la Nutrición (CIBEROBN), Instituto de Salud Carlos III, Madrid, Spain

    Institut d'Investigacions Biomediques August Pi i Sunyer (IDIBAPS), Hospital Clinic of Barcelona, Villarroel 170, Barcelona 08036, Spain
    Search for articles by this author
  • Jannis Kountouras
    Affiliations
    Second Medical Clinic, Faculty of Medicine, Aristotle University of Thessaloniki, Ippokration Hospital, Thessaloniki, Greece
    Search for articles by this author
  • Athanasios D. Anastasilakis
    Affiliations
    Department of Endocrinology, 424 General Military Hospital, Thessaloniki, Greece
    Search for articles by this author
  • Christos S. Mantzoros
    Correspondence
    Corresponding authors at: SL-418, 330 Brookline Avenue, East campus, Beth Israel Deaconess Medical Center, Boston, MA 02215, USA.
    Affiliations
    Division of Endocrinology, Diabetes and Metabolism, Department of Internal Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
    Search for articles by this author
  • Author Footnotes
    1 Indicates equal co-first authorship.
Published:November 08, 2019DOI:https://doi.org/10.1016/j.metabol.2019.154005

      Highlights

      • A lipidomic, glycomic and hormonal analysis was performed in healthy, NAFL and NASH subjects
      • Results were analyzed with 5 different machine learning techniques in two different platforms
      • Diagnostic models with excellent accuracy for diagnosing healthy vs. NAFL vs. NASH sumulatneously were developed
      • A predictive model consisting of 10 lipids showed perfect accuracy for detecting the presence of liver fibrosis

      Abstract

      Background

      Non-alcoholic fatty liver disease (NAFLD) affects 25–30% of the general population and is characterized by the presence of non-alcoholic fatty liver (NAFL) that can progress to non-alcoholic steatohepatitis (NASH), liver fibrosis and cirrhosis leading to hepatocellular carcinoma. To date, liver biopsy is the gold standard for the diagnosis of NASH and for staging liver fibrosis. This study aimed to train models for the non-invasive diagnosis of NASH and liver fibrosis based on measurements of lipids, glycans and biochemical parameters in peripheral blood and with the use of different machine learning methods.

      Methods

      We performed a lipidomic, glycomic and free fatty acid analysis in serum samples of 49 healthy subjects and 31 patients with biopsy-proven NAFLD (15 with NAFL and 16 with NASH). The data from the above measurements combined with measurements of 4 hormonal parameters were analyzed with two different platforms and five different machine learning tools.

      Results

      365 lipids, 61 glycans and 23 fatty acids were identified with mass-spectrometry and liquid chromatography. Robust differences in the concentrations of specific lipid species were observed between healthy, NAFL and NASH subjects. One-vs-Rest (OvR) support vector machine (SVM) models with recursive feature elimination (RFE) including 29 lipids or combining lipids with glycans and/or hormones (20 or 10 variables total) could differentiate with very high accuracy (up to 90%) between the three conditions. In an exploratory analysis, a model consisting of 10 lipid species could robustly discriminate between the presence of liver fibrosis or not (98% accuracy).

      Conclusion

      We propose novel models utilizing lipids, hormones and glycans that can diagnose with high accuracy the presence of NASH, NAFL or healthy status. Additionally, we report a combination of lipids that can diagnose the presence of liver fibrosis. Both models should be further trained prospectively and validated in large independent cohorts.

      Abbreviations:

      NAFLD (non-alcoholic fatty liver disease), NAFL (non-alcoholic fatty liver), NASH (non-alcoholic steatohepatitis), HCC (hepatocellular carcinoma), IR (insulin resistance), TG (triglycerides), HDL (high-density lipoprotein), LDL (low-density lipoprotein), BMI (body mass index), SAM (significance analysis of microarrays), sPLS-DA (sparse Partial Least Squares – Discriminant Analysis), PCA (principal component analysis), t-SNE (t-distributed stochastic neighbor embedding), ROC (receiver operating characteristic), REF (recursive feature elimination), SVM (support vector machine), AUC (area under curve), OvR (One-vs-Rest), RBF (radial basis function), MCCV (Monte-Carlo cross validation), DG (diglycerides), PG (phosphatidylglycerols), PA (phosphatidic acids), AcCa (acylcarnitines), Che (cholesterol esters), Co (coenzyme Q10), LPC (lysophosphatidylcholines), SM (sphingomyelines), PE (Phosphatidylethanolamines), PC (Phosphatidylcholines), C16:0 (palmitic acid), C16:1n7cis (cis-palmitoleic acid), C18:2n6 (polyunsaturated linoleic acid), C20:4n6 (arachidonic acid), kNN (k-nearest neighbor), Ck-18f (cytokeratin-18 fragment), C18:1n9 (oleic acid), C18:3n6 (gamma-linoleic acid)

      Keywords

      To read this article in full you will need to make a payment

      Purchase one-time access:

      Academic & Personal: 24 hour online accessCorporate R&D Professionals: 24 hour online access
      One-time access price info
      • For academic or personal research use, select 'Academic and Personal'
      • For corporate R&D use, select 'Corporate R&D Professionals'

      Subscribe:

      Subscribe to Metabolism - Clinical and Experimental
      Already a print subscriber? Claim online access
      Already an online subscriber? Sign in
      Institutional Access: Sign in to ScienceDirect

      References

        • Younossi Z.M.
        • Koenig A.B.
        • Abdelatif D.
        • Fazel Y.
        • Henry L.
        • Wymer M.
        Global epidemiology of nonalcoholic fatty liver disease-meta-analytic assessment of prevalence, incidence, and outcomes.
        Hepatology. 2016; 64 ([Epub 2015/12/29]): 73-84
        • Polyzos S.A.
        • Kountouras J.
        • Mantzoros C.S.
        Obesity and nonalcoholic fatty liver disease: from pathophysiology to therapeutics.
        Metabolism: clinical and experimental. 2018; 92 ([Epub 2018/12/07]): 82-97
        • Argo C.K.
        • Northup P.G.
        • Al-Osaimi A.M.
        • Caldwell S.H.
        Systematic review of risk factors for fibrosis progression in non-alcoholic steatohepatitis.
        J Hepatol. 2009; 51 ([Epub 2009/06/09]): 371-379
      1. Chalasani N, Younossi Z, Lavine JE, Charlton M, Cusi K, Rinella M, et al. The diagnosis and management of nonalcoholic fatty liver disease: Practice guidance from the American Association for the Study of Liver Diseases. Hepatology. 2018;67(1):328–57. Epub 2017/07/18.

        • Sumida Y.
        • Nakajima A.
        • Itoh Y.
        Limitations of liver biopsy and non-invasive diagnostic tests for the diagnosis of nonalcoholic fatty liver disease/nonalcoholic steatohepatitis.
        World J Gastroenterol. 2014; 20 ([Epub 2014/02/28]): 475-485
        • Castera L.
        • Friedrich-Rust M.
        • Loomba R.
        Noninvasive assessment of liver disease in patients with nonalcoholic fatty liver disease.
        Gastroenterology. 2019; 156 (e4. [Epub 2019/01/21]): 1264-1281
        • Perakakis N.
        • Yazdani A.
        • Karniadakis G.E.
        • Mantzoros C.
        Omics, big data and machine learning as tools to propel understanding of biological mechanisms and to discover novel diagnostics and therapeutics.
        Metabolism: clinical and experimental. 2018; 87 ([Epub 2018/08/12]): A1-A9
        • Musso G.
        • Cassader M.
        • Paschetta E.
        • Gambino R.
        Bioactive lipid species and metabolic pathways in progression and resolution of nonalcoholic Steatohepatitis.
        Gastroenterology. 2018; 155 (e8. [Epub 2018/06/16]): 282-302
      2. Puri P, Baillie RA, Wiest MM, Mirshahi F, Choudhury J, Cheung O, et al. A lipidomic analysis of nonalcoholic fatty liver disease. Hepatology. 2007;46(4):1081–90. Epub 2007/07/27.

      3. Puri P, Wiest MM, Cheung O, Mirshahi F, Sargeant C, Min HK, et al. The plasma lipidomic signature of nonalcoholic steatohepatitis. Hepatology. 2009;50(6):1827–38. Epub 2009/11/26.

        • Vilar-Gomez E.
        • Chalasani N.
        Non-invasive assessment of non-alcoholic fatty liver disease: clinical prediction rules and blood-based biomarkers.
        J Hepatol. 2018; 68 ([Epub 2017/11/21]): 305-315
        • Blomme B.
        • Van Steenkiste C.
        • Callewaert N.
        • Van Vlierberghe H.
        Alteration of protein glycosylation in liver diseases.
        J Hepatol. 2009; 50 ([Epub 2009/01/23]): 592-603
        • Callewaert N.
        • Van Vlierberghe H.
        • Van Hecke A.
        • Laroy W.
        • Delanghe J.
        • Contreras R.
        Noninvasive diagnosis of liver cirrhosis using DNA sequencer-based total serum protein glycomics.
        Nat Med. 2004; 10 ([Epub 2004/05/22]): 429-434
      4. Vanderschaeghe D, Laroy W, Sablon E, Halfon P, Van Hecke A, Delanghe J, et al. GlycoFibroTest is a highly performant liver fibrosis biomarker derived from DNA sequencer-based serum protein glycomics. Molecular & cellular proteomics : MCP. 2009;8(5):986–94. Epub 2009/02/03.

      5. Liu XE, Desmyter L, Gao CF, Laroy W, Dewaele S, Vanhooren V, et al. N-glycomic changes in hepatocellular carcinoma patients with liver cirrhosis induced by hepatitis B virus. Hepatology. 2007;46(5):1426–35. Epub 2007/08/09.

      6. Liu X, Perakakis N, Gong H, Chamberland JP, Brinkoetter MT, Hamnvik OR, et al. Adiponectin administration prevents weight gain and glycemic profile changes in diet-induced obese immune deficient Rag1−/− mice lacking mature lymphocytes. Metabolism: clinical and experimental. 2016;65(12):1720–30. Epub 2016/11/12.

      7. Perakakis N, Kokkinos A, Peradze N, Tentolouris N, Ghaly W, Tsilingiris D, et al. Follistatins in glucose regulation in healthy and obese individuals. Diabetes, obesity & metabolism. 2019;21(3):683–90. Epub 2018/11/06.

        • Boutari C.
        • Perakakis N.
        • Mantzoros C.S.
        Association of Adipokines with development and progression of nonalcoholic fatty liver disease.
        Endocrinol Metab (Seoul). 2018; 33 ([Epub 2018/03/29]): 33-43
      8. Perakakis N, Upadhyay J, Ghaly W, Chen J, Chrysafi P, Anastasilakis AD, et al. Regulation of the activins-follistatins-inhibins axis by energy status: Impact on reproductive function. Metabolism: clinical and experimental. 2018;85:240–9. Epub 2018/05/13.

        • Polyzos S.A.
        • Kountouras J.
        • Anastasilakis A.D.
        • Triantafyllou G.
        • Mantzoros C.S.
        Activin A and follistatin in patients with nonalcoholic fatty liver disease.
        Metabolism: clinical and experimental. 2016; 65 ([Epub 2016/09/14]): 1550-1558
        • Polyzos S.A.
        • Perakakis N.
        • Mantzoros C.S.
        Fatty liver in lipodystrophy: a review with a focus on therapeutic perspectives of adiponectin and/or leptin replacement.
        Metabolism: clinical and experimental. 2019; 96 ([Epub 2019/05/10]): 66-82
        • Polyzos S.A.
        • Toulis K.A.
        • Goulis D.G.
        • Zavos C.
        • Kountouras J.
        Serum total adiponectin in nonalcoholic fatty liver disease: a systematic review and meta-analysis.
        Metabolism: clinical and experimental. 2011; 60 ([Epub 2010/11/03]): 313-326
        • Polyzos S.A.
        • Kountouras J.
        • Anastasilakis A.D.
        • Geladari E.V.
        • Mantzoros C.S.
        Irisin in patients with nonalcoholic fatty liver disease.
        Metabolism: clinical and experimental. 2014; 63 ([Epub 2013/10/22]): 207-217
        • Bondia-Pons I.
        • Castellote A.I.
        • Lopez-Sabater M.C.
        Comparison of conventional and fast gas chromatography in human plasma fatty acid determination.
        J Chromatogr B Analyt Technol Biomed Life Sci. 2004; 809 ([Epub 2004/08/19]): 339-344
        • Kalivodová A.
        • Hron K.
        • Filzmoser P.
        • Najdekr L.
        • Janečková H.
        • Adam T.
        PLS-DA for compositional data with application to metabolomics.
        J Chemometr. 2015; 1: 21-28
        • van den Berg R.A.
        • Hoefsloot H.C.
        • Westerhuis J.A.
        • Smilde A.K.
        • van der Werf M.J.
        Centering, scaling, and transformations: improving the biological information content of metabolomics data.
        BMC Genomics. 2006; 1: 142
        • Lê Cao K.A.
        • Boitard S.
        • Besse P.
        Sparse PLS discriminant analysis: biologically relevant feature selection and graphical displays for multiclass problems.
        BMC bioinformatics. 2011; 1: 253
      9. Pedregosa FaV, G. and Gramfort, A. and Michel, V. and Thirion, B. and Grisel, O. and Blondel, M. and Prettenhofer, P. and Weiss, R. and Dubourg, V. and Vanderplas, J. and Passos, A. and Cournapeau, D. and Brucher, M. and Perrot, M. and Duche. Scikit-learn: machine learning in python. Journal of Machine Learning Research 2011;12:2825–30.

        • Guyon I.
        • Weston J.
        • Barnhill S.
        • Vapnik V.
        Gene selection for cancer classification using support vector machines.
        Machine learning. 2002; 1-3: 389-422
        • Bishop C.M.
        Pattern Recognition and Machine Learning.
        Springer, 2006
        • Chong J.
        • Xia J.
        MetaboAnalystR: an R package for flexible and reproducible analysis of metabolomics data.
        Bioinformatics. 2018; 34 ([Epub 2018/06/30]): 4313-4314
      10. Perakakis N. Data sets - NAFLD. figshare. Dataset 2019;https://doi.org/10.6084/m9.figshare.9774809.

      11. Perakakis N. R command history in MetaboanalystR - NAFLD. figshare. Dataset 2019;https://doi.org/10.6084/m9.figshare.9773843.

      12. Yazdani A. Python code - NAFLD. 2019;https://github.com/alirezayazdani1/NAFLD.

      13. Caussy C, Ajmera VH, Puri P, Hsu CL, Bassirian S, Mgdsyan M, et al. Serum metabolites detect the presence of advanced fibrosis in derivation and validation cohorts of patients with non-alcoholic fatty liver disease. Gut. 2018. [Epub 2018/12/21].

      14. Kwok R, Tse YK, Wong GL, Ha Y, Lee AU, Ngu MC, et al. Systematic review with meta-analysis: non-invasive assessment of non-alcoholic fatty liver disease--the role of transient elastography and plasma cytokeratin-18 fragments. Alimentary pharmacology & therapeutics. 2014;39(3):254–69. Epub 2013/12/07.

        • Musso G.
        • Gambino R.
        • Cassader M.
        • Pagano G.
        Meta-analysis: natural history of non-alcoholic fatty liver disease (NAFLD) and diagnostic accuracy of non-invasive tests for liver disease severity.
        Ann Med. 2011; 43 ([Epub 2010/11/03]): 617-649
      15. Cusi K, Chang Z, Harrison S, Lomonaco R, Bril F, Orsak B, et al. Limited value of plasma cytokeratin-18 as a biomarker for NASH and fibrosis in patients with non-alcoholic fatty liver disease. Journal of hepatology. 2014;60(1):167–74. Epub 2013/08/27.

        • Buzzetti E.
        • Pinzani M.
        • Tsochatzis E.A.
        The multiple-hit pathogenesis of non-alcoholic fatty liver disease (NAFLD).
        Metabolism: clinical and experimental. 2016; 65 ([Epub 2016/01/30]): 1038-1048
      16. Reccia I, Kumar J, Akladios C, Virdis F, Pai M, Habib N, et al. Non-alcoholic fatty liver disease: a sign of systemic disease. Metabolism: clinical and experimental. 2017;72:94–108. Epub 2017/06/24.

        • Clarke J.D.
        • Novak P.
        • Lake A.D.
        • Hardwick R.N.
        • Cherrington N.J.
        Impaired N-linked glycosylation of uptake and efflux transporters in human non-alcoholic fatty liver disease.
        Liver international : official journal of the International Association for the Study of the Liver. 2017; 37 ([Epub 2017/01/18]): 1074-1081
      17. Poynard T, Ratziu V, Charlotte F, Messous D, Munteanu M, Imbert-Bismut F, et al. Diagnostic value of biochemical markers (NashTest) for the prediction of non alcoholo steato hepatitis in patients with non-alcoholic fatty liver disease. BMC gastroenterology. 2006;6:34. Epub 2006/11/14.

      18. Younossi ZM, Jarrar M, Nugent C, Randhawa M, Afendy M, Stepanova M, et al. A novel diagnostic biomarker panel for obesity-related nonalcoholic steatohepatitis (NASH). Obesity surgery. 2008;18(11):1430–7. Epub 2008/05/27.

      19. Crossan C, Tsochatzis EA, Longworth L, Gurusamy K, Davidson B, Rodriguez-Peralvarez M, et al. Cost-effectiveness of non-invasive methods for assessment and monitoring of liver fibrosis and cirrhosis in patients with chronic liver disease: systematic review and economic evaluation. Health Technol Assess. 2015;19(9):1–409, v-vi. Epub 2015/01/31.

        • Mato J.M.
        • Alonso C.
        • Noureddin M.
        • Lu S.C.
        Biomarkers and subtypes of deranged lipid metabolism in non-alcoholic fatty liver disease.
        World J Gastroenterol. 2019; 25 ([Epub 2019/07/12]): 3009-3020
      20. Mayo R, Crespo J, Martinez-Arranz I, Banales JM, Arias M, Minchole I, et al. Metabolomic-based noninvasive serum test to diagnose nonalcoholic steatohepatitis: results from discovery and validation cohorts. Hepatology communications. 2018;2(7):807–20. Epub 2018/07/22.

        • Pabst B.
        • Futatsugi K.
        • Li Q.
        • Ahn K.
        Mechanistic characterization of long residence time inhibitors of Diacylglycerol Acyltransferase 2 (DGAT2).
        Biochemistry. 2018; 57 ([Epub 2018/11/14]): 6997-7010
        • Vitali C.
        • Khetarpal S.A.
        • Rader D.J.
        HDL cholesterol metabolism and the risk of CHD: new insights from human genetics.
        Curr Cardiol Rep. 2017; 19 ([Epub 2017/11/06]): 132
        • Machado M.V.
        • Coutinho J.
        • Carepa F.
        • Costa A.
        • Proenca H.
        • Cortez-Pinto H.
        How adiponectin, leptin, and ghrelin orchestrate together and correlate with the severity of nonalcoholic fatty liver disease.
        Eur J Gastroenterol Hepatol. 2012; 24 ([Epub 2012/06/28]): 1166-1172
      21. Blomme B, Francque S, Trepo E, Libbrecht L, Vanderschaeghe D, Verrijken A, et al. N-glycan based biomarker distinguishing non-alcoholic steatohepatitis from steatosis independently of fibrosis. Digestive and liver disease : official journal of the Italian Society of Gastroenterology and the Italian Association for the Study of the Liver. 2012;44(4):315–22. Epub 2011/11/29.

      22. Kobayashi T, Ogawa K, Furukawa JI, Hanamatsu H, Hato M, Yoshinaga T, et al. Quantifying Protein-Specific N-Glycome Profiles by Focused Protein and Immunoprecipitation Glycomics. Journal of proteome research. 2019;18(8):3133–41. Epub 2019/07/04.

      23. Yamasaki Y, Nouso K, Miyahara K, Wada N, Dohi C, Morimoto Y, et al. Use of non-invasive serum glycan markers to distinguish non-alcoholic steatohepatitis from simple steatosis. Journal of gastroenterology and hepatology. 2015;30(3):528–34. Epub 2014/08/30.

      24. Sriharan M, Reichelt AJ, Opperman ML, Duncan BB, Mengue SS, Crook MA, et al. Total sialic acid and associated elements of the metabolic syndrome in women with and without previous gestational diabetes. Diabetes care. 2002;25(8):1331–5. Epub 2002/07/30.

        • He J.
        • Mao W.
        • Zhang J.
        • Jin X.
        Association between serum Sialic acid levels and nonalcoholic fatty liver disease: a cross-sectional study.
        Ann Nutr Metab. 2015; 67 ([Epub 2015/09/01]): 69-75
        • Lu Z.
        • Ma H.
        • Xu C.
        • Shao Z.
        • Cen C.
        • Li Y.
        Serum Sialic acid level is significantly associated with nonalcoholic fatty liver disease in a nonobese Chinese population: a cross-sectional study.
        Biomed Res Int. 2016; 2016 ([Epub 2016/04/05]): 5921589
        • Ma J.
        • Sanda M.
        • Wei R.
        • Zhang L.
        • Goldman R.
        Quantitative analysis of core fucosylation of serum proteins in liver diseases by LC-MS-MRM.
        J Proteomics. 2018; 189 ([Epub 2018/02/11]): 67-74
      25. Shimizu K, Katoh H, Yamashita F, Tanaka M, Tanikawa K, Taketa K, et al. Comparison of carbohydrate structures of serum alpha-fetoprotein by sequential glycosidase digestion and lectin affinity electrophoresis. Clinica chimica acta; international journal of clinical chemistry. 1996;254(1):23–40. Epub 1996/10/15.

      26. Tawara S, Tatsumi T, Iio S, Kobayashi I, Shigekawa M, Hikita H, et al. Evaluation of Fucosylated Haptoglobin and Mac-2 Binding Protein as Serum Biomarkers to Estimate Liver Fibrosis in Patients with Chronic Hepatitis C. PloS one. 2016;11(3):e0151828. Epub 2016/03/24.

      27. Adinolfi LE, Rinaldi L, Guerrera B, Restivo L, Marrone A, Giordano M, et al. NAFLD and NASH in HCV Infection: Prevalence and Significance in Hepatic and Extrahepatic Manifestations. International journal of molecular sciences. 2016;17(6). Epub 2016/05/28.