Clinical and molecular characterization of steatotic liver disease in the setting of immune-mediated inflammatory diseases

Abstract

Background & aims: Growing evidence suggests an increased prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) in the context of immune-mediated inflammatory diseases (IMIDs). We aimed to clinically and mechanistically characterize steatotic liver disease (SLD) in a prospective cohort of patients with IMID compared to controls.

Methods: Cross-sectional, case-control study including a subset of patients with IMID. Controls from the general population were age-, sex-, type 2 diabetes-, and BMI-matched at a 1:2 ratio. SLD was established using controlled attenuation parameter. Liver biopsies were obtained when significant liver fibrosis was suspected. Total RNA was extracted from freshly frozen cases and analyzed by RNA-seq. Differential gene expression was performed with 'limma-voom'. Gene set-enrichment analysis was performed using the fgsea R package with a preranked "limma t-statistic" gene list.

Results: A total of 1,456 patients with IMID and 2,945 controls were included. Advanced SLD (liver stiffness measurement ≥9.7 kPa) (13.46% vs. 3.79%; p <0.001) and advanced MASLD (12.8% vs. 2.8%; p <0.001) prevalence were significantly higher among patients with IMID than controls. In multivariate analysis, concomitant IMID was an independent, and the strongest, predictor of advanced SLD (adjusted odds ratio 3.318; 95% CI 2.225-4.947; p <0.001). Transcriptomic data was obtained in 109 patients and showed 87 significant genes differentially expressed between IMID-MASLD and control-MASLD. IMID-MASLD cases displayed an enriched expression of genes implicated in pro-tumoral activities or the control of the cell cycle concomitant with a negative expression of genes related to metabolism.

Conclusions: The prevalence of advanced SLD and MASLD is disproportionately elevated in IMID cohorts. Our findings suggest that IMIDs may catalyze a distinct MASLD pathway, divergent from classical metabolic routes, highlighting the need for tailored clinical management strategies.

Impact and implications: The prevalence of steatotic liver disease with advanced fibrosis is increased in patients with immune-mediated inflammatory diseases, independent of classic metabolic risk factors or high-risk alcohol consumption. Transcriptomic analysis revealed a unique gene expression signature associated with cellular activities that are compatible with a liver condition leading to an accelerated and aggressive form of steatotic liver disease. Our findings underscore the importance of heightened screening for advanced liver disease risk across various medical disciplines overseeing patients with immune-mediated inflammatory diseases.

Keywords: IMID; MASLD; SLD; advanced fibrosis; transcriptome.

Graphical abstract

Fig. 1

Study flowchart. A1AT, alpha-1 antitrypsin; AIH, autoimmune hepatitis; ALD, alcohol-associated liver disease; axial SpA, axial spondyloarthritis; IBD, inflammatory bowel disease; PBC, primary biliary cholangitis; PSC, primary sclerosing cholangitis.

Fig. 2

Logistic regression analysis of SLD and advanced SLD risk factors. (A) Global logistic regression of SLD risk factors, representing adjusted odds ratios with 95% confidence intervals. (B) Global logistic regression of advanced SLD risk factors, representing adjusted odds ratios with 95% confidence intervals. (C) Logistic regression of advanced SLD risk factors in the IMID population, representing adjusted odds ratios with 95% CIs. High-risk OH (alcohol consumption): >20 g/day in women and >30 g/day in men. Stepwise multivariate conditional logistic regression analyses were performed for SLD (A), and advanced SLD (B and C). The models were adjusted for all the variables shown in the table. Statistically significant variables are indicated in red. IMID, immune-mediated inflammatory diseases; OR, odds ratio; SLD, steatotic liver disease; T2D, type 2 diabetes; WC, waist circumference.

Fig. 3

Advanced SLD prevalence. (A) Supra-additive effect of IMID and hazardous WC on the prevalence of advanced SLD. (B) Advanced fibrosis prevalence attributable to SLD stratified according to the absence of T2D or obesity in IMID and controls. Fisher's exact tests were performed to obtain p values. IMID, immune-mediated inflammatory diseases; SLD, steatotic liver disease; T2D, type 2 diabetes; WC, waist circumference.

Fig. 4

Differential gene expression in IMID-MASLD vs. control-MASLD. (A) Heatmap showing mRNA expression values and clustering of the 87 differentially expressed genes (row-scaled logCPM values) in the TSC cohort of MASLD cases (n = 109). Pink and blue boxes correspond to IMID-MASLD and control-MASLD cases, respectively. (B) Representative selection of upregulated (logFC >0.40) and downregulated genes (logFC <-0.40), (padj <0.05) from those shown in (A). Activ, activated; Col, cholangiocyte; Fav, favorable; GLI, glioma; HCC, hepatocellular carcinoma; Hep, hepatocytes; IMID, immune-mediated inflammatory diseases; LogFC, log fold-change; MASLD, metabolic dysfunction-associated steatotic liver disease; MEL, melanoma; Prog, progression; RCC, renal cell carcinoma; TSC, translational study cohort; Unf, unfavorable.

Fig. 5

Differential gene set-enrichment analysis in IMID-MASLD vs. control-MASLD. A selection of positively (red) and negatively (blue) differential gene set-enrichment (stratified by biological activities) between IMID- and control-MASLD patients, (padj <0.05). Numbers in the scale show NESs. Numbers in brackets indicate pathway position in the full report (supp. Table 8). Pathway positions range 1-149 for positive NES and 150-169 for negative NES. IMID, immune-mediated inflammatory diseases; MASLD, metabolic dysfunction-associated steatotic liver disease; NESs, normalized enrichment scores.