Author information
1Arizona Liver Health, Chandler, AZ, USA.
2Perspectum Ltd, Oxford, UK.
3Akero Therapeutics Inc., South San Francisco, California, USA.
4Liver Center, Massachusetts General Hospital and Harvard Medical School, Boston, MA, USA.
5Division of Gastroenterology and Hepatology, Indiana University School of Medicine, Indianapolis, IN, USA.
6Division of Endocrinology, Diabetes and Metabolism, University of Florida, Gainesville, FL, USA.
7MASLD Research Center, University of California at San Diego, La Jolla, CA, USA.
8Clinica Di Radiologia EOC, Istituto Di Imaging Della Svizzera Italiana (IIMSI), Ente Ospedaliero Cantonale, Via Tesserete 46, 6900, Lugano, Switzerland; John Radcliffe Hospital, Oxford University Hospitals NHS Foundation Trust, OX3 0AG, Oxford, UK.
9Perspectum Ltd, Oxford, UK. Electronic address: andrea.dennis@perspectum.com.
Abstract
Background & aims: MRI biomarkers of liver disease are robust and reproducible alternatives to liver biopsy. Emerging data suggest that absolute reduction in iron corrected T1 (cT1) of ≥ 80 ms and relative reduction in liver fat content of 30% reflect histological improvement. We aimed to validate the associations of changes to these noninvasive biomarkers with histological improvement, specifically the resolution of steatohepatitis.
Methods: A retrospective analysis of participants from three interventional clinical trials who underwent multiparametric MRI to measure liver cT1 and liver fat content (LFC) (LiverMultiScan) alongside biopsies at baseline and end of study. Responders were defined as those achieving resolution of steatohepatitis with no worsening in fibrosis. Differences in the magnitude of change in cT1 and LFC between responders and non-responders was assessed.
Results: Individual patient data from 150 participants were included. There was a significant decrease in liver cT1 (-119 ms vs. -49 ms) and liver fat content (-65% vs. -29%) in responders compared to non-responders (P < .001) respectively. The diagnostic accuracy to identify responders was 0.72 (AUC) for both. The Youden's index for cT1 to separate responders from non-responders was -82 ms and for liver fat was a 58% relative reduction. Those achieving a ≥ 80 ms reduction in cT1 were 5-times more likely to achieve histological response (sens 0.68; spec 0.70). Those achieving a 30% relative reduction in liver fat were ∼4 times more likely to achieve a histological response (sens 0.77; spec 0.53).
Conclusions: These results, from three combined drug trials, demonstrate that changes in multiparametric MRI markers of liver health (cT1 and PDFF) can predict histological response for steatohepatitis following therapeutic intervention.
Impact and implications: There is great interest in identifying suitable biomarkers that can be used to replace liver biopsy, or to identify those patients who would benefit from one, in both the clinical management of MASH and in drug development. We investigated the utility of two MRI-derived non-invasive tests, iron corrected T1 mapping (cT1) and liver fat content from proton density fat fraction (PDFF), to predict histological improvement in patients who had undergone experimental treatment for MASH. Using data from 150 people who participated in one of three clinical trials, we observed that a reduction in cT1 by over 80 ms and a relative reduction in PDFF of over 58% were the optimal thresholds for change that predicted resolution of steatohepatitis. PDFF as a marker of liver fat, and cT1 as a specific measure of liver disease activity, are both effective at identifying those who are likely responding to drug interventions and experiencing improvements in overall liver health.
Clinical trial number(s): NCT02443116, NCT03976401, NCT03551522.