If we do not get better at spotting liver scarring early, a lot of people with MASH will keep sliding toward cirrhosis while routine clinic visits politely shrug at them.
That is the background hum behind this new Gut paper on thrombospondin-2, or TSP2 - a blood biomarker the authors say can flag at-risk MASH and advanced fibrosis better than several standard noninvasive tools in a large European cohort [1]. Which, if it holds up, is a big deal. Liver biopsy is still the reference standard, but it is invasive, imperfect, and about as scalable as sending everyone to space camp.
The Problem: MASH Loves Hiding in Plain Sight
MASH - short for metabolic dysfunction-associated steatohepatitis - is the angry version of fatty liver disease. Fat piles up, inflammation kicks in, cells get damaged, and over time the liver can scar. That scarring is fibrosis. Enough fibrosis, and now you are playing with cirrhosis, liver failure, cancer, or all the other prizes nobody wants.
Clinicians already use noninvasive tools like FIB-4, transient elastography and composite scores like FAST or Agile 3+ [2,3]. These help, but they are not magic. Some are better at ruling people out than ruling them in. Some wobble in the messy middle. Some are great until you try to use them in a different population and the accuracy suddenly gets stage fright.
This is why biomarker papers keep arriving like contestants on a reality show called America’s Next Top Liver Test.
What This Study Actually Did
The authors looked at 469 patients from three European centers who had liver biopsies for suspected MASLD. They tested four blood proteins - TSP2, IGFBP7, GDF15, and CD163 - because prior liver transcriptomics linked them to disease severity. Then they compared those markers against more familiar tools like VCTE, FIB-4, Agile3, and FAST [1].
Two outcomes mattered:
- At-risk MASH (ARM): active steatohepatitis plus fibrosis stages F2-F4
- Advanced fibrosis (AF): fibrosis stages F3-F4
TSP2 came out on top. Its AUROC was 0.812 for advanced fibrosis and 0.812 again for at-risk MASH. That beat the other candidate biomarkers and outperformed FIB-4 and VCTE in this cohort. When the team combined TSP2 with common clinical variables - AST, age, platelets, albumin, and GGT - performance improved to about 0.845-0.846 [1].
In plain English: this blood marker did a pretty solid job separating the higher-risk patients from the lower-risk ones.
Not perfect. Not crystal-ball territory. But better than “let’s squint at the lab panel and hope the liver tells on itself.”
Why TSP2 Is Interesting
TSP2 is not just some random molecule dredged up by a statistics net. It belongs to the thrombospondin family, proteins involved in the extracellular matrix - basically the scaffolding and signaling environment around cells [4]. Since fibrosis is, at heart, a bad remodeling project where the liver keeps laying down scar tissue like an overcaffeinated contractor, a matrix-related biomarker has at least a plausible biological story.
That matters because a lot of older scores are built from indirect clues. Useful clues, yes. But still clues. TSP2 is trying to be closer to the crime scene.
This also lands at a practical moment. The treatment landscape is finally moving. The FDA approved resmetirom for MASH in March 2024, and semaglutide gained an FDA MASH indication on August 15, 2025 for certain patients with moderate-to-advanced fibrosis [5,6]. Once treatments exist, finding the right patients stops being an academic hobby and starts becoming an actual clinic problem.
Before We Start Engraving TSP2 on a Trophy
Now for the fine print, because the fine print is where medicine hides the land mines.
This was a retrospective study. The cohort was decent-sized, but still not gigantic. The validation used repeated train-test splits inside the overall dataset, which is better than nothing, but not the same as proving the test works prospectively in totally separate real-world populations. Also, biopsy itself is an imperfect benchmark because the tiny tissue sample may not represent the whole liver. Your “gold standard” sometimes behaves like a gold-plated mood ring.
There is also a broader issue: MASH already has a growing pile of promising noninvasive tests. Reviews and guidelines keep emphasizing that combinations of blood markers and imaging often work better than any single test alone, and that implementation in practice is still the hard part [2,3]. Another 0.84 AUROC does not automatically solve reimbursement, assay standardization, referral pathways, or what to do with patients stuck in the diagnostic gray zone - which, in liver disease, is a very crowded zip code.
So the skeptical take is this: TSP2 looks genuinely promising, but “promising” is not the same thing as “ready to run the clinic.” An editorial published after this paper said basically that in more polite journal language [7].
Still, the signal here looks real enough to pay attention. If future studies confirm it, TSP2 could help doctors find the patients who most need follow-up, treatment, or biopsy - while sparing some lower-risk patients from extra testing. In MASH, that would be useful. Also refreshing. Also a mild miracle, because liver diagnostics has spent years acting like every solution needs either a biopsy needle or a spreadsheet with anger issues.
References
- Ratziu V, Surabattula R, Bugianesi E, et al. Thrombospondin-2 is a performant biomarker of at-risk MASH and advanced MASH fibrosis in a large multicentre European cohort. Gut. 2025. DOI: 10.1136/gutjnl-2025-335598. PubMed: PMID 40738743
- EASL-EASD-EASO Clinical Practice Guidelines on the management of metabolic dysfunction-associated steatotic liver disease (MASLD). J Hepatol. 2024. DOI: 10.1016/j.jhep.2024.04.031. PubMed: PMID 38851997
- Boon-Yasidhi P, Karnsakul W. Non-Invasive Biomarkers and Breath Tests for Diagnosis and Monitoring of Chronic Liver Diseases. Diagnostics (Basel). 2025;15(1):68. DOI: 10.3390/diagnostics15010068. PMCID: PMC11720471
- Wikipedia contributors. Thrombospondin. Wikipedia. Accessed May 14, 2026. https://en.wikipedia.org/wiki/Thrombospondin
- Harrison SA, Bedossa P, Guy CD, et al. A Phase 3, Randomized, Controlled Trial of Resmetirom in NASH with Liver Fibrosis. N Engl J Med. 2024;390:497-509. DOI: 10.1056/NEJMoa2309000
- U.S. Food and Drug Administration. FDA approves treatment for serious liver disease known as ‘MASH’. August 15, 2025. https://www.fda.gov/drugs/news-events-human-drugs/fda-approves-treatment-serious-liver-disease-known-mash
- Castera L. Thrombospondin-2: a promising but unproven biomarker for MASH. Gut. 2025. DOI: 10.1136/gutjnl-2025-337225. PubMed: PMID 41224643
Disclaimer: This blog post is a simplified summary of published research for educational purposes. The accompanying illustration is artistic and does not depict actual model architectures, data, or experimental results. Always refer to the original paper for technical details.