Methods: Data from five clinical studies (n=543) with participants suspected of NAFLD were pooled or used for individual participant data meta-analysis. The diagnostic accuracy of the MRI biomarkers to stratify NASH patients was determined using Area Under the Receiver Operating Characteristic curve (AUROC).
Results: A stepwise increase in cT1 and MRI liver fat with increased NAFLD severity was demonstrated, and cT1 was significantly higher in NASH participants with fibrosis grade ≥2 (high-risk NASH). The diagnostic accuracy (AUROC [95% CI]) of cT1 to identify those with NASH was 0.78 [CI: 0.74-0.82], for liver fat was 0.78 [CI: 0.73-0.82], and when combined with MRI liver fat was 0.82 [CI: 0.78-0.85]. The diagnostic accuracy of cT1 to identify those with high-risk NASH was good (AUROC: 0.78 [CI: 0.74-0.82]), was superior to MRI liver fat (AUROC: 0.69 [CI: 0.64-0.74]) and was not substantially improved by combining it with MRI liver fat (AUC: 0.79, [CI: 0.75-0.83]). The meta-analysis showed similar performance to the pooled analysis for these biomarkers.
Conclusions: This study demonstrates that quantitative MRI derived biomarkers cT1 and liver fat are suitable for identifying those with NASH, and cT1 is a better non-invasive technology than liver fat to identify NASH patients at greatest risk of disease progression. MRI cT1 and liver fat therefore have important clinical utility to help guide appropriate use of interventions in NAFLD and NASH clinical care pathways.