Nattokinase for Microclots: What Studies Show About Enzyme Therapy

This review is for people with Long COVID or post-viral symptoms who've heard that nattokinase or lumbrokinase can dissolve microclots. It covers what the studies actually tested, how the enzymes are thought to work, the safety issues, and where the evidence runs out.

Fibrinolytic enzymes in brief

Fibrinolytic enzymes are proteases, meaning they cut proteins, and they can break down fibrin, the main structural protein in blood clots. Your body makes its own, chiefly plasmin. The idea behind supplements is that if fibrin amyloid microclots resist plasmin, an enzyme with different properties might get through.

• Nattokinase is a serine protease from Bacillus subtilis var. natto, the bacterium used to ferment soybeans into natto. It was purified and named in 1987 and has been studied mostly for cardiovascular effects (Chen et al., 2018).
• Lumbrokinase is a group of enzymes from earthworms (Lumbricus rubellus). It's sold as a prescription product for clotting conditions in some countries.
• Serrapeptase (serratiopeptidase) comes from Serratia bacteria and is mainly studied for inflammation and, in lab work, for breaking down amyloid aggregates (Metkar et al., 2024).

Does nattokinase dissolve microclots?

In a dish, there's some evidence it can. In people, nobody knows yet.

The 2024 nattokinase study

Grixti, Kell, Pretorius and colleagues built an automated microscopy method to count fibrinaloid microclots stained with Thioflavin T (Grixti et al., 2024). They formed the microclots in vitro by mixing purified fibrinogen with thrombin and bacterial lipopolysaccharide (LPS). They didn't use patient blood.

Without nattokinase, microclot counts rose for about an hour and then levelled off. With recombinant nattokinase, counts fell after the first time point and reached their lowest level at around two hours, with the higher dose producing a bigger drop. The paper is a preprint, and the authors present it as adding to evidence that's otherwise largely anecdotal.

What this does and doesn't show

It shows that nattokinase can break down amyloid fibrin made in a test tube. It doesn't show that swallowing nattokinase reaches the blood in active form, clears microclots in the body or improves symptoms. That gap matters, because gut digestion and absorption could change everything.

What about lumbrokinase and serrapeptase?

Direct microclot data is thin for both.

Lumbrokinase

Lumbrokinase breaks down fibrin directly and has a longer clinical history for clotting conditions in parts of Asia, though not for Long COVID. Reviews also describe it as one of several serine proteases that can break down amyloid aggregates in lab conditions (Metkar et al., 2024).

The 2025 ultrasound study

A lab-on-chip study tested rtPA, lumbrokinase and nattokinase on plasma-derived amyloid microclots, with and without ultrasound (Rasouli et al., 2025). Ultrasound alone did most of the work: at 150 kHz it cut mean clot diameter by over 60% and large-clot counts by more than 80%.

Adding enzymes made no statistically significant difference at 150 to 300 kHz. Only at 500 kHz, where ultrasound alone was weaker, did the enzymes modestly improve large-clot reduction. So this study is better evidence for ultrasound than for enzymes, and it isn't a clinical treatment yet.

Serrapeptase

Serrapeptase has shown anti-amyloid activity in lab studies, including against amyloid-beta, and has some fibrinolytic activity. There's no published microclot study for it that we could verify, so any microclot benefit is a hypothesis.

Clinical next step

Want data before and after a treatment?

Our at-home amyloid fibrin test gives you microscopy footage to review with your clinician.

Start screening

How these enzymes might work when plasmin doesn't

Fibrin amyloid microclots resist normal breakdown for a few reasons. Their dense beta-sheet structure limits enzyme access, and they trap α2-antiplasmin, the body's main plasmin inhibitor (Pretorius et al., 2021). The structural side of this is covered in spike protein amyloid fibrils and fibrinolysis.

Because these enzymes cut fibrin at their own sites, they might reach parts of a clot that plasmin can't. That's plausible, but it hasn't been demonstrated in microclots inside the body.

Some researchers think the DNA scaffolding from neutrophil extracellular traps also stabilizes microclots, which would need a different enzyme class entirely. That idea is explored in DNases and NET clearance and microclots and NETs in Long COVID.

Are fibrinolytic enzymes safe?

They're generally tolerated by healthy adults at moderate doses, but they're not harmless.

• Bleeding risk: all of them can raise bleeding risk, especially combined with anticoagulants, antiplatelet drugs or NSAIDs.
• Surgery: clinicians usually advise stopping them before procedures.
• Quality: supplement potency, usually listed in fibrinolytic units (FU), varies between brands and batches.
• Who should avoid them: people with bleeding disorders, active bleeding or on blood thinners should only use them under supervision. Safety data in pregnancy, children and people with kidney or liver disease is lacking.

Any enzyme plan should be part of a wider care plan with a clinician who can monitor you. If you're comparing clinicians or centers, our guide to microclot testing labs and treatment centers lists questions to ask.

Limits of the enzyme evidence

• Lab vs body: the best data comes from microclots made or studied outside the body.
• No outcome trials: no randomized controlled trial shows enzymes improve Long COVID symptoms, quality of life or recovery time.
• Absorption questions: it's still debated how much active enzyme survives digestion and reaches the bloodstream.
• Standardization: there's no agreed therapeutic dose, and label claims may not match real activity.
• Commercial interest: much of the enthusiasm comes from supplement sellers, and independent research is limited.

Anecdotal improvement can reflect natural recovery, placebo effects or other changes made at the same time. That's why tracking symptoms carefully matters if you try anything.

Can testing show whether enzymes are working?

Some clinicians compare microclot microscopy before and after a treatment. That approach hasn't been validated in trials, so a change in your results is a data point, not proof that a treatment worked.

If you want that kind of data, our at-home amyloid test uses Thioflavin T fluorescence microscopy on a finger-prick dried blood sample, the same staining principle as the published research. The method is explained in the amyloid fibrin test guide. Results are observational data for you and your clinician, not a diagnosis.

To understand why routine bloods may not reflect microclots at all, read D-dimer vs microclot testing. For the wider research picture, see fibrin amyloid microclots research.