Non-invasive tests as a prediction tool to assess MASH resolution score.

Challenge

Non-invasive tests are individually imperfect predictors of MASH resolution, and the potential for sequential NIT algorithms—combining multiple tests in a stepwise fashion—to reduce diagnostic indeterminates while maintaining accuracy had not been explored in a real-world biopsy-confirmed MASH cohort.

Solution

Paired liver biopsy and NIT data from TARGET-NASH were used to evaluate twelve sequential two- and three-test NIT algorithms using FIB-4, NFS, APRI, and VCTE for predicting NAS-derived MASH resolution, comparing performance characteristics and indeterminate rates across combinations.

Impact

Demonstrating that sequential NIT algorithms—particularly FIB-4 followed by NFS and VCTE—can substantially reduce indeterminate classifications while maintaining accuracy provides a practical framework for non-invasive MASH monitoring in clinical practice and drug development, supporting biopsy-sparing trial designs.