What non-invasive test is worth using to track hepatic fibrosis?

I have three data points across nine months, which I hope is enough to see a trend.

The comparison I want does not seem to exist anywhere in a form I can evaluate.

I have read the arguments for each and they do not engage with each other.

So which one, and on what grounds?

Answer first: resolution of steatohepatitis without worsening of fibrosis is the endpoint the field uses, and it is a histological endpoint read by a pathologist on a biopsy.

Fibrosis regression by one stage is the usual secondary endpoint. It is a slower process than inflammation resolution and a two-year trial is at the short end for detecting it.

Relative to absolute, worked

The last two rows describe the same finding. Only one of them is used in headlines.

The relevant detail is that biopsy sampling variability is a live methodological problem: two cores from the same liver can differ by a fibrosis stage, which puts a floor under how precise any histological trial can be.

The current endpoint definitions for this indication come from regulatory guidance requiring histological resolution without fibrosis worsening, or fibrosis improvement without steatohepatitis worsening.

Read the endpoint definition before the result. In this field it is doing more work than usual.

edited 18 Dec 2024 by low_dead_space — added the placebo-arm figures

The short version: meaningful histological improvement in the trials that biopsied, larger for inflammation than for fibrosis, and the fibrosis component is the harder one.

FIB-4 and ELF are risk-stratification tools rather than endpoints. They are appropriate for deciding who should be investigated further and inappropriate as evidence that a treatment worked.

Liver fat fraction by MRI-PDFF falls quickly and substantially with weight loss in this class, often by half or more, and that is a real change that does not by itself predict a fibrosis outcome.

The caveat is that liver disease is a clinical matter with staging, monitoring and comorbidity questions, and none of that is answerable here.

The nomenclature change matters when you are comparing papers across it.

Put another way, the relevant trials in this class report biopsy-confirmed endpoints, which is why they are small, slow and expensive compared with weight trials.

The ESSENCE programme evaluated semaglutide in biopsy-confirmed MASH with fibrosis, reporting resolution of steatohepatitis without worsening of fibrosis in a substantially larger fraction of the treated arm than of placebo. The fibrosis-improvement component was the more modest of the two.

More usefully, survodutide, a glucagon and GLP-1 receptor co-agonist, was studied in phase 2 in MASH and reported improvement on histology; the glucagon limb is the mechanistic reason that class is being examined here specifically.

Cite the biopsy trials for histological claims and the imaging studies for fat-fraction claims, and do not mix them.

Non-invasive markers are useful for tracking and are not what the endpoint is defined on. That gap explains most confusion in this area.

Placebo response in MASH histology trials is not small — resolution in the placebo arm commonly runs to a fifth or more of participants, partly because biopsy reading is variable and partly because trial participation changes behaviour.

The renaming from NASH to MASH followed a multi-society consensus that also introduced cardiometabolic criteria into the diagnosis, so prevalence figures across the change are not directly comparable.

Nothing in this answer is medical advice. Elevated liver enzymes with an unclear cause need a clinician, not a forum.

Fat fraction falls fast, inflammation resolves next, fibrosis moves last if it moves at all. That order explains most of the literature.

Weight loss itself improves this condition substantially, which makes the attribution question — drug effect versus weight effect — genuinely hard rather than rhetorically hard.

Where weight loss and histological improvement track closely, the honest reading is that some of the effect is mediated by weight — which does not make it less real, only less specific.

Improvement in a liver-fat number is not the same as improvement in disease stage, and the two are routinely conflated in summaries.

If a claim about fibrosis rests on a blood-based score, it is a hypothesis rather than a finding.