Amyloid Microclots: What the Research Shows So Far

This is a research summary for patients and clinicians who want to know what's actually been published about amyloid microclots, which the papers call fibrin amyloid or "anomalous amyloid" microclots. It covers how they form, what the main studies found, where the evidence is thin, and how researchers detect them.

What makes amyloid microclots different?

They're small clot particles made of fibrin that has folded into an amyloid-like structure. Normal fibrin forms an open mesh that plasmin dissolves within hours to days. In the amyloid form, the protein packs into dense, beta-sheet-rich material that binds the dye Thioflavin T and resists that breakdown.

Kell and Pretorius call these particles "fibrinaloid" microclots. They range from a few micrometres to well over 100 micrometres across, which puts many of them in the size range of capillaries (Kell et al., 2022). For a biophysical exploration of the four structural stages and platelet activation, see our guide on anomalous amyloid microclots and phosphatidylserine, or start with our primer on what microclots are.

Key findings from the published studies

Three findings come up again and again across the published work. All of them come from small observational cohorts or lab experiments, so treat them as strong leads rather than settled clinical facts.

More microclots in Long COVID plasma

Using fluorescence microscopy with Thioflavin T, Pretorius and colleagues found that plasma from people with Long COVID still contained large anomalous amyloid deposits, while plasma from healthy controls showed far less (Pretorius et al., 2021). Some later cohorts reported that higher microclot counts tended to go with heavier symptom burden, though those were correlations in small groups.

Resistance to breakdown

The same 2021 study found that the microclots in acute COVID-19 and Long COVID samples resisted fibrinolysis and survived a first round of trypsin digestion. Only a second digestion dissolved them. That stubbornness is why D-dimer, which measures breakdown fragments, may not reflect them, as explained in D-dimer vs microclot testing.

Trapped inflammatory proteins

When researchers dissolved the clots and ran proteomics, they found inflammatory molecules packed inside. These included α2-antiplasmin, von Willebrand factor and serum amyloid A (Kruger et al., 2022). Trapping a plasmin inhibitor inside the clot offers one plausible reason the clots don't dissolve on their own.

How fibrin amyloid microclots form

Researchers think inflammation, rather than a single trigger, pushes fibrinogen into the amyloid form. Bacterial fragments like lipopolysaccharide, high blood sugar and the SARS-CoV-2 spike protein have all induced amyloid fibrin in lab experiments. The spike work is the reason this topic overlaps with post-vaccination microclots, since the same protein is involved.

There's also a cross-seeding angle. Once one protein misfolds into amyloid, it may template other proteins to do the same, a process covered in spike amyloid cross-seeding.

Next step

Curious what your own sample shows?

Our at-home amyloid fibrin test gives you microscopy data you can take to your clinician.

See the at-home test

How could microclots cause Long COVID symptoms?

The microclot hypothesis says that enough of these particles could slow blood flow through capillaries and cut oxygen delivery to tissues. Kell and Pretorius have also suggested that repeated cycles of poor flow and reperfusion could add oxidative damage, which might help explain post-exertional crashes (Kell and Pretorius, 2022).

Why don't standard blood tests detect microclots?

Routine coagulation tests measure how fast plasma clots (PT, aPTT) or how much breakdown product is circulating (D-dimer). None of them looks at the shape or structure of clot particles. That's why the research uses fluorescence microscopy instead, a method we explain in the amyloid fibrin test guide. For a side-by-side of every testing route, see how to test for microclots.

The dye side of that method, including what Thioflavin T does and doesn't bind, is covered in the amyloid fibril test guide.

Treatments under study

The evidence here is early and mixed. Kell, Pretorius and clinical collaborators have described several approaches, mostly in case series and reviews rather than controlled trials (Kell et al., 2022).

• Combination anticoagulant and antiplatelet regimens, sometimes called triple therapy for long COVID, used under specialist supervision because of bleeding risk.
• Therapeutic apheresis, which filters plasma outside the body; a 2025 sham-controlled trial is covered in plasmapheresis for long COVID. Our guide to microclot testing labs and treatment centers covers what to ask before booking.
• Fibrinolytic enzymes such as nattokinase and lumbrokinase. Lab data and its limits are reviewed in nattokinase for microclots.

What are the limitations of microclot research?

Critics raise four main points, and they're fair ones.

• Causation: microclots could drive symptoms, or they could be a downstream sign of inflammation.
• Specificity: they also appear in diabetes and other inflammatory diseases, so they aren't unique to Long COVID.
• Method standardization: sample handling, staining and scoring vary between labs, and there's no agreed reference range.
• Treatment evidence: there are no large randomized controlled trials showing that clearing microclots improves outcomes.

Some haematologists have argued that the wider evidence doesn't yet support clotting as the main driver of Long COVID (Hunt et al., 2024). That debate is healthy, and it's the reason results from any microclot test should be read as observational data.

Where testing fits for patients

If you want to see whether your own sample shows ThT-positive fibrin material, our at-home amyloid test uses the same staining principle as the published studies. You collect a finger-prick dried blood sample at home and receive a Videographic Microscopy Report. It isn't a diagnosis, and a clinician should interpret it alongside your history.