Scans and blood scores can now stage liver scarring well enough to stand in for a biopsy needle, but not one of them could reliably detect the liver inflammation that decides who qualifies for the new fatty liver drugs. That is the finding of the LITMUS Imaging Study, published July 24 in Nature Medicine, which tested imaging and blood markers head to head against centrally read liver biopsy in 357 people with metabolic dysfunction-associated steatotic liver disease (MASLD).
Key takeaways
- For cirrhosis, MR elastography, FibroScan and the Agile scores all beat the study’s prespecified accuracy bar, so a scan can answer “how scarred is my liver.”
- For steatohepatitis (MASH), the inflammation itself, nothing cleared the bar. The best blood marker reached 0.83 and the best scan only 0.71.
- These were patients already referred to liver specialists, so the numbers do not transfer to general screening.
What the study found
The prospective, multicenter study compared imaging (MR elastography, or MRE, and FibroScan vibration-controlled transient elastography), blood markers such as NIS2+, and composite scores such as Agile 3+ and Agile 4 against a biopsy read by a central panel.
Among the 357 participants, all stages of scarring were represented: F0 in 12%, F1 in 16%, F2 in 25%, F3 in 32% and F4 in 15%. At-risk MASH means MASH with at least stage 2 scarring.
For cirrhosis, several markers cleared the study’s minimum acceptable performance criterion: MRE (AUC 0.91), FibroScan liver stiffness (0.87), Agile 3+ (0.89) and Agile 4 (0.88), each with less than a 1 in 100 chance the result was coincidence. For advanced fibrosis, MRE (0.91) and Agile 3+ (0.84) also cleared it.
Inflammation was a different story. NIS2+ was the best blood marker for MASH (AUC 0.83) and at-risk MASH (0.82), but neither result beat the bar by more than chance could explain (roughly a 15 in 100 and 19 in 100 chance of coincidence). The best imaging marker for inflammation, MRE, reached only 0.71 for MASH and 0.75 for at-risk MASH, both below the bar.
Dr. Kumar’s take
Patients think of this as one question. It is two.
“How scarred is my liver” is now answerable with a scan. “How inflamed is my liver” is not, and inflammation is the part that decides drug eligibility. The first drug approved for pre-cirrhotic MASH is licensed for stage 2 to 3 fibrosis with active steatohepatitis, and this study says no scan can confirm that second half.
AUC is worth translating. It is a measure of how well a test separates the people who have the disease from the people who do not, on a scale where higher is better. In a clinic, 0.71 is not a diagnosis. It is a hint.
The moment this gets summarized as “no more liver biopsies,” someone will be handed a FibroScan number and a prescription decision that this trial specifically did not validate. The correct reading is narrower and still useful: fewer needles for staging scarring, same uncertainty about inflammation.
One more limit the authors flag themselves. These were people already referred to specialist liver care, where disease is common. Nearly half had stage 3 or 4 scarring. Run the same tests in a primary care waiting room, where advanced disease is rare, and a positive result means much less.
Biopsy is not harmless, and that is why staging without a needle matters.
What it means for you
If you have fatty liver disease and your doctor orders a FibroScan or an MR elastogram, that is a reasonable way to find out how much scarring you have, and a normal result is reassuring.
If the conversation moves to starting a MASH drug, ask a direct question: are we treating confirmed steatohepatitis, or scarring plus an assumption? A scan cannot tell the two apart yet, and a biopsy may still be the honest answer.
Weight, blood sugar and alcohol remain the levers you control. A new blood test that detects liver disease before symptoms appear is moving in the same direction as this work, and habits as ordinary as coffee, which lowers the risk of liver disease, cancer and death, still show up in the liver data.