Background
Liver biopsy is the current gold standard to stage metabolic associated fatty liver disease (MAFLD). Various non-invasive tools have been used for the diagnosis and staging of liver disease including serum biomarkers and imaging techniques. This systematic review and meta-analysis aimed to identify a biomarker with the highest accuracy to diagnose liver disease in adults with obesity.
Methods
This systematic review and meta-analysis used PRISMA guidelines and was registered (CRD42023444990) on PROSPERO. Databases searched included PubMed, Embase, Medline and Web of Science up to 8th August 2023. The search strategy included all human studies with keywords “imaging” OR “biomarker” AND “obesity” AND “liver” AND “ROC” (receiver operator characteristic). For the meta-analysis, we have chosen the papers that had ROC curve values available. Number of participants, specificity and sensitivity data were collected from selected articles. Review Manager 5.4 software was used for data analysis.
Results
A total of 343 articles were retrieved using the search terms. After removal of duplicates articles (115), screening titles (24) and abstracts (82), 122 articles were eligible for full-text assessment. A further 36 articles were excluded with a final total of 86 papers were assessed in the systematic review. Circulating biomarkers were identified (n=10 studies), which included microRNAs, cytokines and liver related peptides/proteins. Imaging techniques (n=24 studies) included magnetic resonance, ultrasound and elastography. Liver steatosis and fibrosis ROC curve data available from 18 of those imaging-based studies were included in the meta-analysis. Overall, SROC analysis showed higher sensitivity/specificity for steatosis compared to fibrosis.
Conclusions
Several serum biomarkers and imaging techniques have been identified in this systematic review, and meta-analysis for imaging techniques will indicate the method with the highest accuracy to diagnose liver steatosis in people with obesity.