Plasmapheresis and DFPP for Long COVID: Results From Japan and Germany

Double filtration plasmapheresis (DFPP) is the treatment people in long COVID and vaccine-injury communities ask about most, especially since patients began sharing their results from Edogawa Hospital in Tokyo. This guide explains what DFPP removes, what the results from Japan and Germany show, and how it connects to the amyloid microclot and spike protein research covered across our Science Hub.

What is DFPP?

DFPP processes your blood outside the body in two steps. A first membrane separates plasma from blood cells. A second, much finer filter then catches large molecules in the plasma, and everything that passes through is returned to you along with your cells. It was introduced in Japan in 1980 and is used there for kidney, neurological and lipid disorders.

Why DFPP targets what our research describes

Fibrinogen and amyloid fibrin. Amyloid is a stacked, extremely stable protein shape. In long COVID plasma, fibrin folds into amyloid-like microclots that resist plasmin, the clot-dissolving enzyme, as summarized in amyloid microclots research. Fibrinogen is the raw material, and it's one of the largest molecules DFPP removes. H.E.L.P. apheresis was proposed for the same reason.

Inflammatory proteins and autoantibodies. Proteomics of dissolved microclots found α2-antiplasmin, von Willebrand factor, serum amyloid A and antibodies trapped inside. Autoantibodies against nerve and blood vessel receptors are another proposed driver. Both sit in the size range DFPP filters.

Spike protein and spike fibrils. Inflammation cuts spike protein into fragments that form amyloid fibrils, and fibrils from the Spike685 segment make fibrin clots resist breakdown. We cover this in spike protein amyloid fibrils and neutrophil elastase and spike protein. Spike bound to antibodies, spike on vesicles and amyloid aggregates are large, which is why DFPP's second filter is expected to retain them.

Immune debris. Neutrophil extracellular traps sit inside microclots in long COVID, as covered in microclots and NETs.

What the results show

Edogawa Hospital, Tokyo: the McCairn-Edogawa protocol

The program at Edogawa Hospital was developed with neuroscientist Kevin McCairn for long COVID and post-vaccine injury. It has two phases:

1. DFPP to remove large plasma components, including fibrinogen, antibodies and amyloid-like aggregates.
2. Stem cell growth factors: infusions of growth factors and vesicles from dental pulp stem cells (SHED). No live stem cells are given.

Each patient's plasma is examined with fluorescence microscopy before and after treatment. According to reporting on the program, 18 patients had completed the full course by mid-2026, the earliest 9 to 12 months out. The team reports reduced amyloid microclot burden after treatment and lower spike-related material in the filtered plasma. Several patients describe major recoveries, some within hours of the first session.

The program has not yet published its data in a peer-reviewed journal, and IRB-approved studies in the US are being planned. Until then, the numbers of responders, durability and adverse events can't be compared across patients. But it's the clearest clinical test so far of the idea that removing amyloid fibrin and spike-related material helps.

Dresden, Germany: biomarkers fall as patients improve

In a 2023 Molecular Psychiatry study, the Dresden University Clinic group measured biomarkers before and after therapeutic apheresis in long COVID patients with chronic fatigue. In patients who reported significant improvement after two cycles:

• neurotransmitter autoantibodies, lipids and inflammatory markers fell significantly;
• fibrinogen dropped by about 70%;
• red cell clumping (rouleaux) and fibrin fibers largely disappeared on dark-field microscopy.

That's the first peer-reviewed study tying a pattern of blood markers, including fibrin, to clinical improvement after apheresis.

Charité, Berlin: removing autoantibodies

The Charité group treated 20 post-COVID ME/CFS patients who had raised β2-adrenergic receptor autoantibodies with five sessions of immunoadsorption. IgG fell by 79% and the autoantibodies by 77%. Physical function improved by a mean of 17.75 points on the SF-36 scale. Seven patients who relapsed after an initial response were treated again.

Case reports

A 2022 case report from California described symptom improvement and immune changes after plasma exchange in a long-haul COVID patient. Our article on spike antibody vs antigen testing describes a public case of a patient whose spike antibody level stayed near 12,500 U/mL for five years before they went on to DFPP.

Next step

Considering DFPP? Measure before and after.

Our at-home amyloid fibrin test records Thioflavin T microscopy of your sample, so you and your doctor can compare microclot burden before and after treatment.

See the at-home test

What about the negative plasma exchange trial?

A 2025 Spanish trial randomized 50 people with post COVID-19 condition to six sessions of standard plasma exchange or a sham procedure. Both groups improved similarly. It's an important result, but it doesn't settle DFPP:

• Different procedure. TPE replaces plasma non-selectively. DFPP selectively removes large molecules such as fibrinogen and aggregates.
• Unselected patients. Participants weren't chosen for high microclot burden, autoantibodies or spike persistence, the groups Dresden, Charité and Edogawa focus on.
• No second phase. Edogawa pairs DFPP with growth factor support, and Charité retreats relapses.

Selective apheresis shows consistent positive signals in targeted patients, and it now needs a controlled trial in those same patients.

Why symptoms can come back

Relapse weeks or months later is one of the most common experiences people share. Fibrinogen is replaced within days and antibodies by memory cells, and if a driver like persistent spike or ongoing inflammation is still active, new microclot material can form. That's the rationale for Edogawa's second phase and Charité's retreatment cycles. Measuring microclot burden before and after, and again months later, shows whether the effect is holding.

Cost, travel and safety

DFPP for long COVID is almost always self-funded. Reports put programs abroad at roughly $20,000 or more plus travel, and a 2022 BMJ investigation documented high costs at some European clinics. Travel itself can trigger crashes in people with post-exertional malaise, so plan recovery time.

DFPP is generally well tolerated by experienced teams. Watch for low blood pressure, citrate tingling, access-site bruising or infection, and bleeding risk while fibrinogen is low, especially with blood thinners or enzymes covered in triple therapy and nattokinase for microclots.

Questions to ask any DFPP program

• What exactly does your second filter remove? Pore size determines what's taken out.
• What do you measure before and after? Fibrinogen, autoantibodies, inflammatory markers and microclot imaging give an objective baseline.
• Who follows me up at home, and how is relapse handled?
• What's your total cost, including access, labs and any second phase?

Our directory of microclot testing labs and treatment centers has more questions for comparing clinics.

Where to go next

To get a baseline before treatment, read how to test for microclots. If routine labs keep coming back normal, does long COVID show up on blood tests explains why.