849
A BLOOD-BASED BIOMARKER PANEL FOR NON-INVASIVE DIAGNOSIS OF METABOLIC DYSFUNCTION-ASSOCIATED STEATOHEPATITIS
Date
May 20, 2024
Explore related products in the following collection:
Background: Metabolic dysfunction-associated steatotic liver disease (MASLD) and its more severe form metabolic dysfunction-associated steatohepatitis (MASH) are highly prevalent worldwide. However, current non-invasive diagnostic approaches for MASLD and MASH are suboptimal. In this study, we aimed to develop a blood-based biomarker panel for the diagnosis of MASLD and MASH.
Methods: Serum samples were collected from 695 subjects including 590 patients with biopsy-proven MASLD and 105 healthy controls across four independent cohorts from Asia and America. The training cohort comprised of 112 MASLD patients and 72 healthy controls from Hong Kong, China. The performance of established biomarker panels was then validated in three independent cohorts from San Diego, USA (140 MASLD patients and 33 healthy controls), Wenzhou (235 MASLD patients) and Hong Kong, China (103 MASLD patients). Serum levels of protein biomarkers including C-X-C motif chemokine ligand 10 (CXCL10), squalene epoxidase (SQLE), carbonic anhydrase III (CA3), p62, cytokeratin 18 M30 (CK-18), and fibroblast growth factor 21 (FGF21), were measured by enzyme-linked immunosorbent assay. Random forest was used for biomarker selection and the model was trained using logistic regression and support vector machine. Discrimination performance of the biomarker panel was tested and confirmed by area under the receiver-operating characteristic curve (AUROC).
Results: A panel of 6 biomarkers (CXCL10, CK-18, HbA1c, ALT, AST, and BMI) was formulated for the diagnosis of MASLD (AUROC = 0.969) and MASH (AUROC = 0.865). We then optimized the panel to 3 biomarkers (CXCL10, CK-18, and BMI), termed as N3-MASH, with comparable performance to the 6-biomarker panel. N3-MASH discriminated MASLD patients from healthy controls with an AUROC of 0.949. Among MASLD patients, N3-MASH could identify MASH patients with an AUROC of 0.845, outperforming serum ALT (AUROC = 0.632, P < 0.001). N3-MASH demonstrated 90.0% specificity, 61.2% sensitivity, 74.1% accuracy and 88.4% positive predictive value (PPV) to discriminate MASH from metabolic dysfunction-associated steatotic liver (MASL) at an upper cutoff of 0.682; and 90.3% sensitivity, 52.0% specificity, 73.2% accuracy and 81.3% negative predictive value (NPV) at a lower cutoff of 0.339 to exclude MASH patients. The grey or indeterminate zone comprised 37 (33%) of 112 patients. The performance of N3-MASH in distinguishing MASH from MASL was confirmed in three validation cohorts, with AUROC of 0.795, 0.800 and 0.823, respectively. N3-MASH achieved PPV values of 88.1%, 95.4%, and 85.7% for diagnosing MASH in the validation cohorts.
Conclusion: We developed a robust blood-based panel consisting of 2 protein biomarkers and 1 clinical parameter for non-invasive diagnosis of MASH, which might help clinicians to identify MASH patients and reduce unnecessary liver biopsies.
Methods: Serum samples were collected from 695 subjects including 590 patients with biopsy-proven MASLD and 105 healthy controls across four independent cohorts from Asia and America. The training cohort comprised of 112 MASLD patients and 72 healthy controls from Hong Kong, China. The performance of established biomarker panels was then validated in three independent cohorts from San Diego, USA (140 MASLD patients and 33 healthy controls), Wenzhou (235 MASLD patients) and Hong Kong, China (103 MASLD patients). Serum levels of protein biomarkers including C-X-C motif chemokine ligand 10 (CXCL10), squalene epoxidase (SQLE), carbonic anhydrase III (CA3), p62, cytokeratin 18 M30 (CK-18), and fibroblast growth factor 21 (FGF21), were measured by enzyme-linked immunosorbent assay. Random forest was used for biomarker selection and the model was trained using logistic regression and support vector machine. Discrimination performance of the biomarker panel was tested and confirmed by area under the receiver-operating characteristic curve (AUROC).
Results: A panel of 6 biomarkers (CXCL10, CK-18, HbA1c, ALT, AST, and BMI) was formulated for the diagnosis of MASLD (AUROC = 0.969) and MASH (AUROC = 0.865). We then optimized the panel to 3 biomarkers (CXCL10, CK-18, and BMI), termed as N3-MASH, with comparable performance to the 6-biomarker panel. N3-MASH discriminated MASLD patients from healthy controls with an AUROC of 0.949. Among MASLD patients, N3-MASH could identify MASH patients with an AUROC of 0.845, outperforming serum ALT (AUROC = 0.632, P < 0.001). N3-MASH demonstrated 90.0% specificity, 61.2% sensitivity, 74.1% accuracy and 88.4% positive predictive value (PPV) to discriminate MASH from metabolic dysfunction-associated steatotic liver (MASL) at an upper cutoff of 0.682; and 90.3% sensitivity, 52.0% specificity, 73.2% accuracy and 81.3% negative predictive value (NPV) at a lower cutoff of 0.339 to exclude MASH patients. The grey or indeterminate zone comprised 37 (33%) of 112 patients. The performance of N3-MASH in distinguishing MASH from MASL was confirmed in three validation cohorts, with AUROC of 0.795, 0.800 and 0.823, respectively. N3-MASH achieved PPV values of 88.1%, 95.4%, and 85.7% for diagnosing MASH in the validation cohorts.
Conclusion: We developed a robust blood-based panel consisting of 2 protein biomarkers and 1 clinical parameter for non-invasive diagnosis of MASH, which might help clinicians to identify MASH patients and reduce unnecessary liver biopsies.
Presenter
Speakers
University of California San Diego
Tracks
Related Products
FAECALIBACTERIUM PRAUSNITZII INHIBITS GASTRIC CANCER AND ANTAGONIZES HELICOBACTER PYLORI COLONIZATION
BACKGROUND: The roles of probiotics in inhibiting gastric cancer (GC) and effecting _H. pylori_ infection are largely unclear. We identified _Faecalibacterium prausnitzii_ is depleted in GC. We aimed to evaluate the antitumorigenic function and molecular mechanism of _F…
