Department of Hepatology
Hospital Pitié Salpêtrière
Vlad Ratziu, MD, PhD, has disclosed that he has received consulting fees from Galmed, Genfit, and Intercept.
Several NASH therapeutic trials were completed in 2018 with results presented at the AASLD conference, and advances in noninvasive diagnostic techniques played a big part in making these trials possible. With easily quantifiable methods based on serum biomarkers or imaging procedures, we can now explore a wide variety of endpoints in assessing treatments for NASH and we can test endpoints that are aligned with the agent’s mechanism of action.
Here’s my take on how these new diagnostic techniques for NASH/NAFLD will affect studies and practice in the future.
Measuring Liver Fat
Since large phase III studies must rely on histology for confirmation of NASH diagnosis and for scoring, noninvasive methods are instead mostly useful in early-phase trials. Such trials are intended to provide proof of concept that a pharmacologic agent engages its molecular target and that this translates into a measurable biological effect.
MRI-based quantification of liver fat fraction by proton density fat fraction measurement (PDFF) is one noninvasive success story. Easy and highly quantitative, standardized MRI-PDFF can be used at the beginning of study, helping select patients with steatosis above a certain threshold if required by the inclusion criteria.
MRI-PDFF can also be helpful in defining study endpoints by measuring changes in liver fat induced by therapy (reported as mean absolute reduction, mean reduction relative to baseline, proportion of patients with a relative reduction of 30% or more from baseline, or proportion of patients with absolute reduction of 5% or more from baseline).
In a small study, the endpoint of a relative reduction of liver fat fraction by 30% was associated with a 2-point reduction in NAFLD activity score and improvement in histologic composite score of steatosis. The endpoint of absolute reduction of liver fat fraction by 5% has an even more robust validation—in a subset of the FLINT trial, it was shown to be associated with a 1-grade improvement in histologic steatosis.
MRI-PDFF also minimizes sampling variability by allowing for measurements before and after therapy in specific anatomical areas of the liver.
What’s more, changes in liver fat content by MRI-PDFF can be seen as early as 2 weeks following therapy initiation, with full effects seen after only 3-4 months. Therefore, MRI-PDFF measurement is particularly useful for detecting whether a drug that improves the metabolic condition or inhibits de novo lipogenesis will also induce a measurable reduction in steatosis.
Depending on a drug’s mechanism of action, in some cases a reduction in liver fat fraction could also coexist with resolution of steatohepatitis. However, whichever metric is used, we must remember that steatosis reduction does not necessarily imply improvement in inflammation; additional information on resolution of steatohepatitis necessitates histologic documentation.
Whether future MR-based imaging methods could detect an improvement in inflammation remains to be seen. For the moment, data on multiparametric MRI (LiverMultiScan) are not sufficiently conclusive for most expert radiologists.
What about the antifibrotic effects of agents? These are certainly more difficult to document in early-phase trials with noninvasive methods. Early-phase trials use MR-based elastometry (MRE) or serum fibrosis markers but, to a certain extent, both are confounded by inflammation and may not entirely reflect changes in fibrosis: Short-term changes could be influenced more by inflammatory changes than by fibrosis reduction.
In addition, attempting to detect short-term changes in fibrosis can be misleading because increases or decreases in fibrosis are slow to occur. At best, this will only underestimate the antifibrotic potency of a drug.
Although MRE is now widely used in short-term (3-6 months) trials, its sensitivity to early changes in fibrosis is unknown and probably suboptimal. Therefore, data are rarely conclusive when MRE is used at early time points.
To confidently use both noninvasive imaging and serum-based biomarkers as indicators of antifibrotic effects of a drug candidate, we will need thorough documentation of the correlation of noninvasive endpoints with real outcomes, combined with histologic changes over a sufficient period of drug exposure. When such data are available, these noninvasive methods will certainly be useful for monitoring patients treated with approved drugs.
How do you evaluate and care for your patients who are at risk for fibrosis? I invite you to join the conversation on how data from the 2018 liver meeting will shape your practice.