PPAR-γ-induced changes in visceral fat and adiponectin levels are associated with improvement of steatohepatitis in patients with NASH - PubMed (original) (raw)

Randomized Controlled Trial

. 2021 Nov;41(11):2659-2670.

doi: 10.1111/liv.15005. Epub 2021 Jul 21.

Affiliations

Randomized Controlled Trial

PPAR-γ-induced changes in visceral fat and adiponectin levels are associated with improvement of steatohepatitis in patients with NASH

Amalia Gastaldelli et al. Liver Int. 2021 Nov.

Abstract

Background and aims: Peroxisome proliferator-activated receptor (PPAR)-γ agonists decrease hepatic/visceral fat (VF) and improve necroinflammation despite subcutaneous (SC) fat weight-gain. Understanding the impact of changes in VF, VF-to-SC fat distribution (VF/SC) and adiponectin (ADPN) levels in relation to histological improvement after weight-loss or pioglitazone is relevant as novel PPAR-γ agonists are being developed for treating non-alcoholic steatohepatitis (NASH).

Methods: Fifty-five patients with NASH received a -500 kcal/d hypocaloric diet and were randomized (double-blind) to pioglitazone (45 mg/d) or placebo for 6-months. Before and after treatment patients underwent a liver biopsy and measurement of hepatic/peripheral glucose fluxes, hepatic/adipose tissue-IR and, in 35 patients, hepatic and VF/SC-fat was measured by magnetic resonance spectroscopy/imaging. Data were examined by multivariable statistical analyses combined with machine-learning techniques (partial least square discriminant analysis [PLS-DA]).

Results: Both pioglitazone (despite weight-gain) and placebo (if weight-loss) reduced steatosis but only pioglitazone ameliorated necroinflammation. Using machine-learning PLS-DA showed that the treatment differences induced by a PPAR-γ agonist vs placebo on metabolic variables and liver histology could be best explained by the increase in ADPN and a decrease in VF/SC, and to a lesser degree, improvement in oral glucose tolerance test-glucose concentrations and ALT. Decrease in steatosis and disease activity score (ballooning plus lobular inflammation) kept a close relationship with an increase in ADPN (r = -.71 and r = -.44, P < .007, respectively) and reduction in VF/SC fat (r = .41 and r = .37, P < .03 respectively).

Conclusions: Reduction in VF and improved VF/SC-distribution, combined with an increase in ADPN, mediate the histological benefits of PPAR-γ action, highlighting the central role of fat metabolism and its distribution on steatohepatitis disease activity in patients with NASH.

Keywords: NASH; PPAR-y; adiponectin; fatty liver; insulin resistance; pioglitazone; type 2 diabetes mellitus; visceral fat.

© 2021 The Authors. Liver International published by John Wiley & Sons Ltd.

PubMed Disclaimer

Conflict of interest statement

S.S, CF, GM, FB, RA, PV, DB, SK have nothing to disclose.

AG has received honorarium from Novo Nordisk and is consultant for Boehringer Ingelheim, Eli Lilly, Gilead, Inventiva, and Pfizer.

KC has received research support as principal investigator for the University of Florida from the National Institute of Health (NIH), Cirius, Echosens, Inventiva, Novartis, Novo Nordisk, Poxel and Zydus. KC is also a consultant for Allergan, Altimmune, Arrowhead, AstraZeneca, BMS, Boehringer Ingelheim, Coherus, Eli Lilly, Fractyl, Hanmi, Genentech, Gilead, Intercept, Janssen, Pfizer, Prosciento, Madrigal and Novo Nordisk.

SAH: Scientific advisor or consultant for Akero, Alentis, Altimmune, Arrowhead, Axcella, Canfite, Cirius, CiVi Biopharma, Cymabay, Echosens, Fibronostics, Forest Labs, Galectin, Genfit, Gilead, Hepion, HistoIndex, Intercept, Madrigal, Medpace, Metacrine, NGM Bio, Northsea, Novartis, Novo Nordisk, PathAI, Poxel, Liminal, Ridgeline, Sagiment, Terns, Viking, 89 Bio. Stock options: Akero, Cirius, Galectin, Genfit, Hepion, HistoIndex, PathAI, Metacrine, NGM Bio, Northsea. Grant/Research support: Akero, Axcella, BMS, Cirius, CiVi Biopharma, Conatus, Cymabay, Enyo, Galectin, Genentech, Genfit, Gilead, Hepion, Hightide, Intercept, Madrigal, Metacrine, NGM Bio, Novartis, Novo Nordisk, Northsea, Pfizer, Sagimet, Viking.

Figures

FIGURE 1

FIGURE 1

Changes in clinical parameters in patients with NASH after pioglitazone (“PIO”; blue bars), weight loss following dietary counselling (“BW‐loss”; red bars), or dietary failure to nutritional counselling (“Diet‐fail”; green bars). Panel A: body weight (BW); Panel B: Subcutaneous adipose tissue; Panel C: Visceral fat; Panel D: Liver fat *P < .05 vs pretreatment. MRI scan of a study patient before (Panel E) and after (Panel F) treatment with pioglitazone. Clinical characteristics: 39 year old male with type 2 diabetes: Fasting plasma glucose decreased from 8.8 to 7.0 mM and HbA1c from 7.9% to 6.2%; BMI increased from 35.6 to 37.3 kg/m2. The ratio of visceral‐to‐subcutaneous fat distribution (VF/SC) decreased from 0.54 to 0.36, while liver fat decreased from 8.4% to 2.5%

FIGURE 2

FIGURE 2

Machine learning approach with partial least square discriminant analysis (PLS‐DA) to discriminate the effect of pioglitazone (PIO) vs placebo (either “BW‐loss” or “Diet fail”), using all metabolic (log2 post/pre) and histological (post‐pre) variables. PLS‐DA is able to determine the discriminative power of each variable, improving model prediction. Panel A: Scores plot of PLS‐DA in PIO (red points) vs placebo (grey points) subjects’ classification. Panel B: Variables contribution to the PLS‐DA model, measured through VIP index. The red box features variable with VIP > 1 that are considered relevant in the discrimination. Adipo‐IR, adipose tissue IR index; ADPN, adiponectin; ALT, alanine transaminase; ballooning, Ballooning score in liver biopsy; OGTT‐glu, mean glucose concentration during OGTT; Periph ISI, peripheral insulin sensitivity index calculated as OGIS

FIGURE 3

FIGURE 3

Changes (post‐pre values) in liver histological parameters in patients with NASH after pioglitazone (“PIO”; red bars), weight loss following dietary counselling (“BW‐loss”; blue bars), or dietary failure to nutritional counselling (“Diet‐fail”; green bars). Panel A: steatosis; Panel B: disease activity score (as hepatocyte ballooning and lobular inflammation); Panel C: fibrosis; and Panel D: NAS. *P < .05 vs pretreatment. Panel E: correlation matrix between changes in metabolic variables and individual liver histological parameters (steatosis, ballooning, lobular inflammation, their combined activity score or fibrosis). The size and intensity of the colour indicate the value of the Spearman correlation coefficient (according to the colour bar on the right, that is, red circles indicate a positive correlation while blue circles a negative correlation). * indicates a correlation coefficient with P < .05. Key variables (changes in adiponectin, VF/SC, glucose during the OGTT, ALT and liver fat by MRS; as well as weight and other metabolic parameters) have been organized from left to right based on their relative importance from the machine learning approach with PLS‐DA. Of note, PLS‐DA discriminated well the treatment effect of pioglitazone on necroinflammation, while changes in steatosis (and NAS) were less discriminatory for the effect between pioglitazone and placebo. NAS, NAFLD Activity Score; NASH, non‐alcoholic steatohepatitis; OGTT, oral glucose tolerance test; PLS‐DA, partial least square discriminant analysis; VF/SC, visceral‐to‐subcutaneous fat ratio

FIGURE 4

FIGURE 4

Univariate regression analysis in the entire cohort of NASH patients showed strong association between changes in liver fat with respect to VF. Patients treated with pioglitazone (“PIO” red squares) were those with the strongest decrease in both VF and hepatic fat. In placebo group those that did not lose weight (“Diet‐fail”; green triangles) and those that lost weight (“BW‐loss”; blue circles) showed the same association between changes in VF and liver fat. NASH, non‐alcoholic steatohepatitis; VF, visceral fat

FIGURE 5

FIGURE 5

Panel A: Correlation matrix showing the univariate correlation between changes (post‐ pre) in intrahepatic TG (liver fat) and VF/SC ratio and changes in metabolic parameters in the entire cohort. The size and intensity of the colour indicate the value of the Spearman correlation coefficient (according to the colour bar on the right, that is, red circles indicate a positive correlation while blue circles a negative correlation). *P < .05. Key variables (changes in adiponectin, VF/SC fat, glucose during the OGTT, ALT and liver fat by MRS; as well as weight and other metabolic parameters) have been organized from left to right based on their relative importance from the machine learning approach with PLS‐DA showed in Figure 2. Multiple regression analysis for changes in plasma adiponectin and adipose tissue insulin resistance (Adipo‐IR) (Panel B); HOMA and mean glucose excursions during an OGTT (Panel C); plasma ALT and plasma AST (Panel D), with respect to change in liver fat and VF/SC fat ratio, in patients with NASH that lost weight following dietary counselling (“BW‐loss”; blue circles), fail the diet (“Diet‐fail”; green triangles) or were treated with pioglitazone (“PIO” red squares). OGTT, oral glucose tolerance test; PLS‐DA, partial least square discriminant analysis; VF/SC, visceral‐to‐subcutaneous fat ratio

References

    1. Younossi ZM, Koenig AB, 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(1):73‐84. - PubMed
    1. Morelli M, Gaggini M, Daniele G, Marraccini P, Sicari R, Gastaldelli A. Ectopic fat: the true culprit linking obesity and cardiovascular disease? Thromb Haemost. 2013;110(4):651‐660. - PubMed
    1. Neeland IJ, Ross R, Després J‐P, et al. Visceral and ectopic fat, atherosclerosis, and cardiometabolic disease: a position statement. Lancet Diabetes Endocrinol. 2019;7(9):715‐725. - PubMed
    1. Stefan N, Fritsche A, Schick F, Haring HU. Phenotypes of prediabetes and stratification of cardiometabolic risk. Lancet Diabetes Endocrinol. 2016;4(9):789‐798. - PubMed
    1. Azzu V, Vacca M, Virtue S, Allison M, Vidal‐Puig A. Adipose tissue‐liver cross talk in the control of whole‐body metabolism: implications in nonalcoholic fatty liver disease. Gastroenterology. 2020;158(7):1899‐1912. - PubMed

Publication types

MeSH terms

Substances

LinkOut - more resources