Performance of Noninvasive Tests for Metabolic Dysfunction-Associated Steatohepatitis and Liver Fibrosis Resolution after Bariatric Surgery

Background Noninvasive tests (NITs) to monitor metabolic dysfunction-associated steatohepatitis (MASH) progression and response to interventions are needed because of the risks of liver biopsy. A monocytes-based diagnostic test using perilipin-2 (PLIN2) and Ras-related protein-14 (RAB14) predict the severity of MASH and fibrosis. Here we compared the performances of PLIN2 and RAB14 with cytokeratin-18 (CK18) assessed by Ella™ or M65 ELISA in predicting MASH and fibrosis resolution following bariatric surgery in a longitudinal and histologically characterized cohort of individuals with obesity. Methods Participants in the BRAVES randomized controlled trial underwent ultrasound-guided needle liver biopsy at baseline and 1 year after surgery. We evaluated NITs’ performance using area under the receiver operating characteristic and calculated accuracy, sensitivity, and specificity based on the Youden threshold. Univariable and multivariable logistic models were used to assess the role of recorded covariates in predicting MASH and fibrosis severity, as well as resolution or improvement. Results After surgery, patients who experienced MASH improvement or resolution showed a significant decrease in PLIN2 expression as compared to those who did not, while patients with fibrosis improvement displayed an increase in RAB14. No differences were found for CK18. The diagnostic accuracy of PLIN2 and RAB14 for the prediction of MASH resolution or fibrosis improvement was superior to CK18 assessed by either Ella or M65 ELISA. Conclusions PLIN2 and RAB14, but not CK18, are markers for monitoring improvements in MASH and fibrosis after interventions such as bariatric surgery. This may reduce or eliminate the need for frequent liver biopsies. ClinicalTrials.gov Registration Number: NCT03524365