Non-invasive quantitative magnetic resonance imaging and spectroscopic biomarkers in nonalcoholic fatty liver disease and other cardiometabolic diseases associated with ectopic fat deposition

Abstract

The increasing prevalence of obesity worldwide has prompted the development of non-invasive quantitative imaging biomarkers (QIBs) of hepatic steatosis, percentage and type of fat accumulation in skeletal muscle, and absolute and relative amounts of subcutaneous and visceral body fat. These biomarkers have utility in various contexts of use (COUs) in nonalcoholic fatty liver disease (NAFLD) drug development clinical trials, including as inclusion criteria, and in the assessment of drug safety and efficacy. Magnetic resonance imaging (MRI)-estimated proton density fat fraction (MRIPDFF) is well suited for drug development trials where assessment of hepatic steatosis is required, including those in which a baseline level is needed for study inclusion, and for studies in which drug efficacy, safety, and mechanism are being assessed. MRI-PDFF is accurate compared to both MRS-PDFF as reference and to histologic steatosis grade as reference and is reproducible across scanner manufacturer and field strength which facilitates implementation in multi-center clinical trials. Moreover, metaanalysis has shown it to have good linearity and low bias. Capability to acquire MRI-PDFF scans is widely available, and central analysis by groups experienced in the analysis of these scans for clinical trials is also available. Levels and changes of PDFF that are clinically and biologically meaningful have been suggested but have not yet been established. MRS has a limited role in assessing hepatic steatosis in drug development clinical trials, mainly due to challenges of implementation in multi-center studies. However, it may be helpful in single-center early exploratory or mechanistic studies. Other non-invasive imaging techniques, such as computed tomography (CT), ultrasound (US) and MR techniques that measure uncorrected signal fat fraction rather than PDFF have thus far been less useful because they are confounded by biological, physical, and technical factors. MRS is the only non-invasive technique currently able to differentiate fat accumulating within skeletal muscle cells (intramyocellular lipid [IMCL]) from skeletal muscle fat accumulating outside muscle cells (extramyocellular lipid [EMCL]). MRS can also be used to assess the overall percentage of skeletal muscle fat. Thus, MRS non-invasive assessment of IMCL may have a limited role in drug development, and although there are limitations, MRS is feasible for single-center trials where expertise in acquisition and analysis is available. Conclusions In the context of developing new drugs for NAFLD and related cardiometabolic conditions (obesity, type 2 diabetes, cardiovascular disease, metabolic syndrome), advanced MRI currently stands out as the best method to assess hepatic steatosis across the entire liver. The advantages of accuracy, precision, availability for multi-as well as single-center trials, and suitability for cross-sectional and longitudinal trials outweigh the cost disadvantage. Harmonization of acquisition and analysis techniques across MRI scanner vendors will help ensure that results across sites, vendors, and time will be comparable. Advanced ultrasound techniques are being developed to assess hepatic steatosis, and if validated may also become useful in clinical trials. CT is less useful than MRI but may be of value when advanced MRI is not available or contraindicated. MRS is currently the only way to non-invasively assess IMCL, and so if that is required in single-center or small multi-center trials, it may be of use. Finally, semi-automated MRI is emerging as the leading method to assess fat and muscle body composition.