You Zhou1, Matej Orešič2, Marja Leivonen3, Peddinti Gopalacharyulu4, Jenni Hyysalo5, Johanna Arola6, An Verrijken7, Sven Francque8, Luc Van Gaal7, Tuulia Hyötyläinen2, Hannele Yki-Järvinen9. 1. Minerva Foundation Institute for Medical Research, Helsinki, Finland; Systems Immunity University Research Institute and Division of Infection and Immunity, School of Medicine, Cardiff University, Cardiff, United Kingdom. 2. Steno Diabetes Center, Gentofte, Denmark. 3. Department of Surgery, University of Helsinki and Helsinki University Hospital, Helsinki, Finland. 4. Institute for Molecular Medicine Finland, University of Helsinki, Helsinki, Finland. 5. Minerva Foundation Institute for Medical Research, Helsinki, Finland; Department of Medicine, University of Helsinki, Helsinki, Finland. 6. Department of Pathology, University of Helsinki and Helsinki University Hospital, Helsinki, Finland. 7. Department of Endocrinology, Diabetology and Metabolism, Antwerp University Hospital, University of Antwerp, Antwerp, Belgium. 8. Department of Gastroenterology and Hepatology, Antwerp University Hospital, University of Antwerp, Antwerp, Belgium. 9. Minerva Foundation Institute for Medical Research, Helsinki, Finland; Department of Medicine, University of Helsinki, Helsinki, Finland. Electronic address: hannele.yki-jarvinen@helsinki.fi.
Abstract
BACKGROUND & AIMS: Use of targeted mass spectrometry (MS)-based methods is increasing in clinical chemistry laboratories. We investigate whether MS-based profiling of plasma improves noninvasive risk estimates of nonalcoholic steatohepatitis (NASH) compared with routinely available clinical parameters and patatin-like phospholipase domain-containing protein 3 (PNPLA3) genotype at rs738409. METHODS: We used MS-based analytic platforms to measure levels of lipids and metabolites in blood samples from 318 subjects who underwent a liver biopsy because of suspected NASH. The subjects were divided randomly into estimation (n = 223) and validation (n = 95) groups to build and validate the model. Gibbs sampling and stepwise logistic regression, which fulfilled the Bayesian information criterion, were used for variable selection and modeling. RESULTS: Features of the metabolic syndrome and the variant in PNPLA3 encoding I148M were significantly more common among subjects with than without NASH. We developed a model to identify subjects with NASH based on clinical data and PNPLA3 genotype (NASH Clin Score), which included aspartate aminotransferase (AST), fasting insulin, and PNPLA3 genotype. This model identified subjects with NASH with an area under the receiver operating characteristic of 0.778 (95% confidence interval, 0.709-0.846). We then used backward stepwise logistic regression analyses of variables from the NASH Clin Score and MS-based factors associated with NASH to develop the NASH ClinLipMet Score. This included glutamate, isoleucine, glycine, lysophosphatidylcholine 16:0, phosphoethanolamine 40:6, AST, and fasting insulin, along with PNPLA3 genotype. It identified patients with NASH with an area under the receiver operating characteristic of 0.866 (95% confidence interval, 0.820-0.913). The NASH ClinLipMet score identified patients with NASH with significantly higher accuracy than the NASH Clin Score or MS-based profiling alone. CONCLUSIONS: A score based on MS (glutamate, isoleucine, glycine, lysophosphatidylcholine 16:0, phosphoethanolamine 40:6) and knowledge of AST, fasting insulin, and PNPLA3 genotype is significantly better than a score based on clinical or metabolic profiles alone in determining the risk of NASH.
BACKGROUND & AIMS: Use of targeted mass spectrometry (MS)-based methods is increasing in clinical chemistry laboratories. We investigate whether MS-based profiling of plasma improves noninvasive risk estimates of nonalcoholic steatohepatitis (NASH) compared with routinely available clinical parameters and patatin-like phospholipase domain-containing protein 3 (PNPLA3) genotype at rs738409. METHODS: We used MS-based analytic platforms to measure levels of lipids and metabolites in blood samples from 318 subjects who underwent a liver biopsy because of suspected NASH. The subjects were divided randomly into estimation (n = 223) and validation (n = 95) groups to build and validate the model. Gibbs sampling and stepwise logistic regression, which fulfilled the Bayesian information criterion, were used for variable selection and modeling. RESULTS: Features of the metabolic syndrome and the variant in PNPLA3 encoding I148M were significantly more common among subjects with than without NASH. We developed a model to identify subjects with NASH based on clinical data and PNPLA3 genotype (NASH Clin Score), which included aspartate aminotransferase (AST), fasting insulin, and PNPLA3 genotype. This model identified subjects with NASH with an area under the receiver operating characteristic of 0.778 (95% confidence interval, 0.709-0.846). We then used backward stepwise logistic regression analyses of variables from the NASH Clin Score and MS-based factors associated with NASH to develop the NASH ClinLipMet Score. This included glutamate, isoleucine, glycine, lysophosphatidylcholine 16:0, phosphoethanolamine 40:6, AST, and fasting insulin, along with PNPLA3 genotype. It identified patients with NASH with an area under the receiver operating characteristic of 0.866 (95% confidence interval, 0.820-0.913). The NASH ClinLipMet score identified patients with NASH with significantly higher accuracy than the NASH Clin Score or MS-based profiling alone. CONCLUSIONS: A score based on MS (glutamate, isoleucine, glycine, lysophosphatidylcholine 16:0, phosphoethanolamine 40:6) and knowledge of AST, fasting insulin, and PNPLA3 genotype is significantly better than a score based on clinical or metabolic profiles alone in determining the risk of NASH.
Authors: Cyrielle Caussy; Veeral H Ajmera; Puneet Puri; Cynthia Li-Shin Hsu; Shirin Bassirian; Mania Mgdsyan; Seema Singh; Claire Faulkner; Mark A Valasek; Emily Rizo; Lisa Richards; David A Brenner; Claude B Sirlin; Arun J Sanyal; Rohit Loomba Journal: Gut Date: 2018-12-19 Impact factor: 23.059
Authors: Robert A Egnatchik; Alexandra K Leamy; Sarah A Sacco; Yi Ern Cheah; Masakazu Shiota; Jamey D Young Journal: J Biol Chem Date: 2018-12-18 Impact factor: 5.157
Authors: Tagreed A Mazi; Kamil Borkowski; John W Newman; Oliver Fiehn; Christopher L Bowlus; Souvik Sarkar; Karen Matsukuma; Mohamed R Ali; Dorothy A Kieffer; Yu-Jui Y Wan; Kimber L Stanhope; Peter J Havel; Valentina Medici Journal: Free Radic Biol Med Date: 2021-06-25 Impact factor: 7.376