Does plasma exchange work for long COVID, and is HELP apheresis abroad worth it?
Controlled trials have not shown a benefit. Six sessions of plasma exchange matched a sham procedure on every efficacy measure in a 50-person trial, and immunoadsorption failed against a sham in 40 patients. HELP apheresis, offered at clinics in Germany, had no published efficacy data as of 2023, so travelling for it means paying for an unproven treatment.
Nov 17, 2026Updated Nov 18, 20268 min read

The best controlled evidence says no. In a 2025 double-blind trial, 50 people with long COVID received six sessions of therapeutic plasma exchange or a sham procedure, and the groups did not differ on any efficacy measure. A sham-controlled immunoadsorption study of 40 patients was also negative. HELP apheresis, which some patients travel to Germany for, had no published efficacy data as of 2023.
What did the 2025 plasma exchange trial find?
It found no benefit. Researchers randomly assigned 50 people with post-COVID condition to six sessions of TPE or a sham plasma exchange and followed them for 90 days, and no efficacy measure differed between the groups.
In therapeutic plasma exchange (TPE), a machine separates your plasma, the liquid part of blood that carries antibodies, clotting proteins and other molecules. It discards that plasma and returns your blood cells with a replacement fluid. Post-COVID condition (PCC) is the clinical term for long COVID.
The study was a phase II, double-blind, placebo-controlled randomized trial, with participants randomized 1:1 between the two groups. Safety was the primary endpoint. Secondary endpoints covered functional status, symptoms, quality of life, neurocognitive symptoms, and blood measures of biochemistry, blood counts, clotting and inflammation.
The authors concluded: "Whereas TPE is safe, it did not lead to any discernible improvement of the PCC in this clinical trial." Put simply, TPE was safe, but it did not lead to any discernible improvement in long COVID.
The trial's limit is its scope. It is a single study of 50 people over 90 days, and it tested TPE only, so it cannot settle the question alone or speak for other apheresis methods. The paper discloses that one author "has received grants from Grifols", a plasma products company, and that other authors received honoraria from pharmaceutical companies.
Do immunoadsorption, HELP or INUSpheresis work any better?
None has shown a benefit for long COVID in a controlled trial. Immunoadsorption failed against a sham, HELP apheresis had no published efficacy data as of 2023, and no published controlled trial of INUSpheresis for long COVID was found.
These methods work differently from TPE. Immunoadsorption targets antibodies in plasma. HELP apheresis and INUSpheresis filter plasma and return it to the body instead of replacing it. The guide to INUSpheresis vs plasma exchange sets out the differences.
Immunoadsorption
The sham-controlled trial was published by Stortz and colleagues in The Lancet Regional Health Europe in 2026. Researchers at University Medical Center Rostock describe it as a randomised, sham-controlled, patient-blind study of 40 patients that "failed to demonstrate any clinically relevant improvements compared with the placebo group".
The Rostock team's own study had no comparison group. It treated 18 patients with elevated antibodies against adrenergic and/or muscarinic receptors with five consecutive immunoadsorption sessions. Antibody levels fell but returned to baseline within the following 3 to 6 months. The team concluded: "Our results indicate that immunoadsorption is not an effective approach for patients with post-COVID syndrome."
HELP apheresis
At a clinic in Mülheim, Germany, internist Beate Jaeger has suggested that HELP apheresis might work by removing tiny blood clots in small vessels. National Geographic reported that she described great success in some patients after one to three sessions, adding: "But this does not apply to all patients." As of that 2023 report, "Jaeger and other clinicians haven't published any data to show whether and to what extent this treatment is effective."
Jeffery Winters, a Mayo Clinic apheresis expert, said: "They have a hypothesis that this process is going to influence the disease … They may be right," but "we don't have any quantifiable proof that it is altering the disease outcome." Without a sham group, patient reports cannot separate a treatment effect from placebo response or natural recovery.
Where the evidence stands
| Method | Study type | Patients | Result |
|---|---|---|---|
| TPE | Double-blind, sham-controlled | 50 | No difference on any efficacy measure |
| Immunoadsorption | Patient-blind, sham-controlled | 40 | No difference from sham in symptom severity |
| Immunoadsorption | Uncontrolled | 18 | Antibodies fell, then returned within 3 to 6 months |
| HELP apheresis | None published | None | No published efficacy data as of 2023 |
- Established: the American Society for Apheresis (ASFA) grades apheresis disease by disease. Its tenth edition comprises 93 fact sheets and 183 graded and categorized indications. The FDA separately clears apheresis machines. The Spectra Optia, for example, holds 510(k) clearance as a therapeutic blood cell and plasma separator, with a 2007 decision date. That clearance covers the machine, not long COVID.
- Emerging and mostly negative: the two sham-controlled trials and the uncontrolled Rostock study above.
- Unproven: HELP apheresis, INUSpheresis and "microclot removal" for long COVID. Patient success stories are anecdotes.
Why do long COVID patients travel abroad for "blood washing"?
Private clinics in Cyprus, Germany and Switzerland have offered apheresis for long COVID outside clinical trials, and patients have paid out of pocket to try it. A 2022 BMJ investigation, reported by The Irish Times, described a blood filtering method normally used for lipid disorders that have not responded to drugs, combined with anti-clotting therapy.
Some patients spent approximately £7,000 (€8,300), including travel and accommodation. A BMJ statement said doctors and researchers worry that "desperate patients are spending life-changing sums on invasive, unproven treatments". Shamil Haroon, a researcher on the TLC trial, said such treatments should only be given in a clinical trial.
Patient accounts are mixed. Oonagh Carr, 47, said after two sessions: "It's not a magic pill." Pamela Bishop, treated at a clinic in Cyprus, said: "I'm very aware these gains will not sustain."
Is plasma exchange for long COVID safe?
In the 2025 trial, both study arms had a similarly favorable safety profile. Safe in a monitored trial is not the same as useful, and this trial showed no benefit.
For TPE in its established uses, the American College of Rheumatology says patients may feel cold, nauseated, or have mild tingling of the lips, hands or feet, and fatigue is common after a session. It notes that because the procedure uses a central line, there is a small risk of infection and bleeding. Plasma exchange can also remove some medications, including those given by IV, so timing matters if you take regular drugs.
For HELP-type apheresis, National Geographic reports that some patients experience fatigue, nausea, vomiting, low blood pressure, and chest pain caused by reduced blood flow to the heart, and that anticoagulants used with treatment raise bleeding risk.
How much does it cost, and will insurance pay?
Among the per-session prices listed below, checked November 18, 2026 in Northvane's clinic listings, general TPE runs from USD 3,500 for a low-volume (1 L) session to USD 12,000 per session, and some advanced multi-session protocols cost far more. RTHM reports that insurers classify these treatments for long COVID as experimental. These are general TPE prices, not long COVID protocols.
| Clinic | TPE price |
|---|---|
| Global Apheresis | USD 40,000 for a 6-session series |
| Neuroveda Health | USD 5,000 per session |
| Extension Health | USD 12,000 per session |
| Apeiron Center | From USD 3,500 for low-volume (1 L) TPE |
Biologicum Hamburg lists INUSpheresis at EUR 2,900 per session, checked on the same date. Another clinic, ELLANN Health, lists a two-treatment INUSpheresis course at EUR 5,400–5,700; two sessions at Biologicum's rate would be EUR 5,800 (arithmetic, not a quote).
Simple arithmetic, not a quote: six TPE sessions, the number tested in the 2025 trial, would come to USD 30,000 at USD 5,000 each and USD 72,000 at USD 12,000 each. The USD 3,500 low-volume session is a smaller exchange, so it is not a like-for-like comparison with the trial. Travel, lodging and lab work are extra. See what plasma exchange costs in 2026.
RTHM, which publishes a blog on long COVID treatments, estimates that a course of six to eight sessions "can easily exceed $15,000 to $30,000, not including the costs of travel". That is a clinic's estimate.
On insurance, National Geographic notes that most US health insurers cover the expense for approved usage. RTHM says insurers classify these treatments for long COVID as experimental and deny coverage. Ask your insurer in writing before you pay.
What should you ask a clinic before booking?
Ask what method you would receive, what evidence supports it for long COVID, and who manages complications once you are home. Bring the clinic's answers to a physician who knows your history.
- Which method exactly: TPE, immunoadsorption, HELP, INUSpheresis, or another filtration method?
- Which ASFA category, if any, applies to long COVID, and in which edition of the guidelines?
- What outcome data do you have for long COVID, and how do they compare with the two negative sham-controlled trials?
- What kind of venous access will you use, and who manages complications after I fly home?
- Will anticoagulants be used, and how will my current medications be timed around sessions?
- What is the full cost, including every session, lab tests and follow-up, and what is the refund policy?
- Is a clinical trial available instead, as some researchers recommend?
- Will you send records to my own physician before and after treatment?
Common questions
Is plasma exchange the same as convalescent plasma or plasma donation?
No. Plasma exchange removes your own plasma and replaces it. Convalescent plasma is plasma from people who recovered from COVID-19, given to a patient. Plasma donation means giving plasma for others' use. Young-donor plasma infusion is different again: the FDA says there is no proven clinical benefit of infusing plasma from young donors for conditions such as aging or memory loss.
How long does a HELP apheresis session take?
National Geographic reports that each HELP-type session can last two to four hours. The Mülheim clinic's internist described results after one to three sessions, but National Geographic reported that patients there averaged four to five sessions, so a trip abroad can mean several long appointments and follow-up with your own physician after you return.
Sources10 cited
- [1]Plasma exchange therapy for the post COVID-19 condition: a phase II, double-blind, placebo-controlled, randomized trial. · doi.org
- [2]Immunoadsorption Versus Sham Treatment for Post-COVID Syndrome: A Randomised Sham-Controlled Crossover Trial. · doi.org
- [3]Blood cleaning? Long COVID patients turn to unproven treatments | National Geographic · nationalgeographic.com
- [4]Impact of immunoadsorption on autoantibodies and physical performance in post-COVID patients: a prospective exploratory study of 18 participants - PMC · pmc.ncbi.nlm.nih.gov
- [5]tenth edition comprises 93 fact sheets and 183 graded and categorized indications · doi.org
- [6]510(k) Premarket Notification · accessdata.fda.gov
- [7]Some long Covid patients seek experimental ‘blood washing’ treatment abroad, investigation finds – The Irish Times · irishtimes.com
- [8]Plasma Exchange (Plasmapheresis) | American College of Rheumatology · rheumatology.org
- [9]Plasmapheresis for Long COVID (RTHM blog, March 10, 2026). The page did not return a title, so this is taken from the URL and topic. · rthm.com
- [10]Important Information about Young Donor Plasma Infusions for Profit | FDA · fda.gov
For information, not medical advice. Suggest a correction · Editorial policy


