A multi-component classifier for nonalcoholic fatty liver disease (NAFLD) based on genomic, proteomic, and phenomic data domains

Abstract

Non-alcoholic fatty liver disease (NAFLD) represents a spectrum of conditions that include steatohepatitis and fibrosis that are thought to emanate from hepatic steatosis. Few robust biomarkers or diagnostic tests have been developed for hepatic steatosis in the setting of obesity. We have developed a multi-component classifier for hepatic steatosis comprised of phenotypic, genomic, and proteomic variables using data from 576 adults with extreme obesity who underwent bariatric surgery and intra-operative liver biopsy. Using a 443 patient training set, protein biomarker discovery was performed using the highly multiplexed SOMAscan® proteomic assay, a set of 19 clinical variables, and the steatosis predisposing PNPLA3 rs738409 single nucleotide polymorphism genotype status. The most stable markers were selected using a stability selection algorithm with a L1-regularized logistic regression kernel and were then fitted with logistic regression models to classify steatosis, that were then tested against a 133 sample blinded verification set. The highest area under the ROC curve (AUC) for steatosis of PNPLA3 rs738409 genotype, 8 proteins, or 19 phenotypic variables was 0.913, whereas the final classifier that included variables from all three domains had an AUC of 0.935. These data indicate that multi-domain modeling has better predictive power than comprehensive analysis of variables from a single domain.

Plasma proteome profiling discovers novel proteins associated with non‐alcoholic fatty liver disease

“…AFM has been closely linked to metabolic syndrome, insulin resistance, NAFLD, and alcoholic liver disease (Bell et al , ; Liu et al , ; Neuman et al , ; Kollerits et al , ). LGALS3BP has already been used to build multi‐component classifiers for the prediction of NAFLD (Wood et al , ) and fibrosis in patients with hepatitis C infection (Cheung et al , ). The elevated levels of LGALS3BP and vitronectin (VTN), another extracellular matrix (ECM) protein, are likely a reflection of remodeling of the ECM in liver disease.…”

“…We identified a panel of six proteins in the two NAFLD subtypes: three in NAFLD without T2D (PIGR, ALDOB, and VTN) and four in NAFLD with T2D (PIGR, LGALS3BP, AFM, and APOM). Of these, AFM and LGALS3BP have been reported as potential markers for NAFLD (Bell et al, 2010;Wood et al, 2017). AFM has been closely linked to metabolic syndrome, insulin resistance, NAFLD, and alcoholic liver disease (Bell et al, 2010;Liu et al, 2011;Neuman et al, 2014;Kollerits et al, 2017).…”

“…LGALS3BP has already been used to build multi-component classifiers for the prediction of NAFLD (Wood et al, 2017) and fibrosis in patients with hepatitis C infection (Cheung et al, 2010). The elevated levels of LGALS3BP and vitronectin (VTN), another extracellular matrix (ECM) protein, are likely a reflection of remodeling of the ECM in liver disease.…”

Limitations of non-invasive tests for assessment of liver fibrosis

“…106 This single nucleotide polymorphism is strongly associated with more hepatic fat deposition and fibrosis, with recent data linking it with a higher risk of liverrelated events and death in patients with NAFLD. 107 Other approaches using plasma DNA methylation, 108 modified singlenucleotide aptamer-based assays, 109 circulating microRNA, 110 and gut microbiome metagenomic profiling, are examples of the wealth of promising data that is generated by technological advances, 111 but also highlight the need for further studies to validate these candidate NASH biomarkers. There will be issues regarding cost, reproducibility, and high-throughput capability.…”

Predicting and elucidating the etiology of fatty liver disease using a machine learning-based approach: an IMI DIRECT study

“…Omics technologies have been used in a small number of studies to identify molecular 443 biomarkers of NAFLD (39)(40)(41). This includes tests utilizing genetic data such as 444 FibroGENE for staging liver fibrosis (42), and tests using metabolomic data derived 445 from liver tissue to differentiate simple hepatitis from NASH (43), as well as a multi-446 we selected cut-offs that maximize balanced accuracy (considering both sensitivity and 472 specificity); these features are especially important in screening algorithms, where the 473 cost of false negatives can be high.…”