A scanning electron microscope image taken by physiologist Resia Pretorius showing red blood cells covered with what she says are microclots, from a sample prepared from whole blood smears of people with long Covid.

Bad Blood? The Uncertainty Around Microclots and Long Covid


Asad Khan first developed Covid-19 in late 2020. A pulmonologist in Manchester, England, Khan had spent most of the year working on packed hospital wards full of acutely ill Covid patients. After falling ill with the disease himself, Khan spent what he described as a dreadful month at home waiting for the symptoms to subside. But they never did. Instead, Khan fell into the grips of long Covid, a post-viral syndrome that can last months to years after a SARS-CoV-2 infection. By the following September, he was sequestered in a darkened room, wearing earplugs and a blindfold. Khan suffered from relentless nausea and couldn’t stand the presence of other people, even his own children. The symptoms were so intolerable, he said, that he would have taken drastic measures to end them.

Then Khan heard about an experimental treatment offered by Beate Jaeger, an internist in Germany. Jaeger’s approach drew from preliminary evidence connecting long Covid with microscopic blood clots that can potentially deprive tissues and organs of sufficient oxygen. To remove these so-called microclots, Jaeger was using a procedure called apheresis, whereby a patient’s blood is removed, filtered, and then returned to the body. Apheresis is typically used to treat certain blood disorders or cancers such as leukemia or lymphoma. To Khan, the blood-washing strategy made “physiological sense.” So, after taking a one-hour flight to Germany, Khan met with Jaeger at her clinic. Apheresis treatments can run in the thousands of dollars, and Khan’s first session was scheduled for the following day. But Jaeger “took one look at me,” Khan recalled, and said, ‘You’re having it today.’”

A study supported by the National Institutes of Health in the United States found that nearly one in 10 people will experience long-lasting symptoms after a SARS-CoV-2 infection, which include loss of sense of smell, muscle pain, dizziness, brain fog, shortness of breath, and fatigue. For some, the symptoms are so severe that they are unable to work, attend school, or go about their daily lives. Long Covid is more likely among people who were hospitalized during the acute phase of their illness. But it also occurs among those who were never severely sick with Covid. Women have a higher risk than men, as do people with asthma and other underlying illnesses.

Pulmonologist Asad Khan and his doctor Beate Jaeger discuss Khan’s debilitating experience with long Covid and the treatment that he credits with his recovery. Khan traveled from England to Germany to undergo HELP apheresis, which removes the patient’s blood, filters it, and returns it to the body in an effort to remove so-called microclots. Visual: BBC News/YouTube

Research agencies have spent vast sums and effort investigating long Covid, but there’s still no consensus regarding its definition, its underlying mechanisms, or how best to manage it. Long Covid has hundreds of potential symptoms. The syndrome “likely has no single cause,” said Roy Silverstein, a hematologist who studies the effects of inflammation on blood clotting at the Medical College of Wisconsin. Some scientists attribute the condition to a mysterious immune system activation, possibly triggered by viral fragments lingering in the body. In this theory of long Covid’s origin, those viral bits and pieces inflame blood vessels, potentially leading to microclots that clog veins and capillaries and deprive tissues of oxygen. For patients frustrated by the lack of therapeutic alternatives, clot-busting approaches have become “a beacon of hope,” Silverstein said. That’s because unlike other purported mechanisms in long Covid, clotting is readily treatable with existing therapies.

But the link between microclots and long Covid has also proven controversial. While clotting is clearly documented in patients who become severely ill with a Covid infection, the notion that similar mechanisms drive long Covid symptoms hasn’t been fully accepted by the scientific community. Some experts say the hypothesis rests on questionable data generated using lab methods that are difficult to interpret and reproduce, and they note that researchers have yet to show evidence that microclots circulate in long Covid patients and trigger persistent effects. Microclots could plausibly explain long-lasting symptoms in some patients, Silverstein said, “but the whole story needs to go through rigorous peer review before we jump to major conclusions about it.”

Beverley Hunt, a professor of thrombosis and hemostasis at King’s College, London, co-authored a 2023 Cochrane Review of five available studies on a type of apheresis called plasmapheresis in long Covid, which concluded that evidence in support of the hypothesis is insufficient. (Cochrane Reviews analyze and synthesize available evidence to inform health decision making on particular topics.) Hunt agreed that there is “a lot of conjecture” surrounding microclots and strongly advises long Covid patients against spending thousands of dollars out of pocket on what she described as unproven microclot-directed therapies with potential harms. Complications from apheresis are rare and can include bleeding, infections, and allergic reactions.

Beverley Hunt, a professor of thrombosis and hemostasis at King’s College, London, co-authored a Cochrane review last year on plasmapheresis in long Covid. She is skeptical of the microclot hypothesis for long Covid, suggesting that more evidence is needed to support it — for example imaging of such clots in the small vessels of patients. Visual: Lewis Khan for Undark

Some doctors also treat long Covid with “triple therapy,” which is a potent cocktail of three blood-thinners that’s typically reserved for heart attack patients, people in need of coronary stents, and people at high risk of stroke. Triple therapy poses a significant risk of dangerous bleeding, so people who get it need careful monitoring, cautioned Amitava Banerjee, a cardiologist and professor of clinical data science at University College London. Banerjee leads the only active clinical trial investigating blood thinning in long Covid: It’s part of a broader U.K.-based effort called Stimulate-ICP that involves more than 30 organizations evaluating current care and working to study drugs that show some evidence of benefitting patients with the condition. Researchers should publish evidence that triple therapy can actually work in long Covid patients, otherwise “we’re just giving them bleeding risks,” Banerjee said. “And with triple anti-coagulation therapy,” he noted, that’s significant.

But some clinicians who give these therapies disagree. They say that the blood of their long Covid patients is so sticky and hypercoagulated that triple therapy merely returns it to normal without the bleeds one might expect in other patients. Jaeger has not seen any major bleeding complications among the thousand or more patients she has treated so far, she told Undark in an email.

Meanwhile, severe long Covid patients like Khan say they have little choice but to seek treatment from doctors willing to push clinical boundaries. “In response to specialists suggesting that patients should wait for trials, I say this — you have not walked in our shoes,” wrote Khan in a 2022 editorial. “There are so many people suffering,” he said during an interview with Undark. “We don’t have the luxury of waiting for long, longitudinal randomized control trials. So we need to use the evidence that we have available.”

The microclot hypothesis starts with the endothelium, which is a single-cell layer lining the interior surfaces of blood vessels. The endothelium normally releases anti-coagulating substances that keep blood liquid and flowing. But when inflamed by wounds or invading pathogens, it tilts towards a defensive posture: Endothelial cells activate small cell fragments called platelets, which combine with a protein called fibrinogen to initiate clotting. During that process, fibrinogen is converted into a tough fibrous mesh called fibrin that holds a clot in place. The body relies on blood clots to staunch bleeding and defend itself against infections. But during severe Covid, endothelial inflammation and clotting can go into overdrive. Autopsies of people who died from Covid reveal blood clots plugging blood vessels, along with evidence of multi-organ failure. That research gave rise to a view that Covid is, in many ways, a endothelial disease.

Evidence that microclots might similarly drive persistent symptoms in long Covid emerged from a collaboration between two scientists: Resia Pretorius, a physiologist at Stellenbosch University, in South Africa, and Douglas Kell, a biochemist at the University of Liverpool, in the U.K. The pair have long been studying how microbes can cause inflammation, which in turn, they say, pushes fibrinogen and other plasma proteins to adopt abnormal shapes they call amyloids. Scientists ordinarily use the term amyloid in reference to misshapen proteins that accumulate in organs and tissues of patients with certain diseases, including Alzheimer’s. Pretorius and Kell, however, claimed to have found them in blood samples from Covid patients obtained with the help of a local doctor in Stellenbosch, Jaco Laubscher.

“The whole story needs to go through rigorous peer review before we jump to major conclusions about it.”

Pretorius was studying blood samples from people who were severely ill with Covid around May 2020 when, she said, she first discovered the odd structures. She had stained the samples with a dye called thioflavin T that glows brightly when it binds to amyloid structures. Pretorius then looked at the samples through a fluorescence microscope. “The whole screen just lit up like Christmas trees,” she said. By the end of 2020, Pretorius was detecting amyloid fibrin deposits — or microclots, as she and Kell began calling them — in blood samples from patients with long Covid symptoms as well.

In the body, blood clots ordinarily dissolve over time. But in Pretorius’ test tubes, the amyloid clots resisted this normal degradative process. Pretorius and her research team made this discovery while comparing plasma samples from pre-treatment Covid patients and acute or long Covid patients with those obtained from healthy controls. First, they spun the samples down in a centrifuge. Then, they exposed all the samples to trypsin, a digestive enzyme that breaks plasma proteins into chains of amino acids. In the test tubes containing normal plasma, the proteins were fully digested. By contrast, proteins in the Covid samples had clumped into amyloid pellets made visible when stained with dye.

Physiologist Resia Pretorius in her laboratory at Stellenbosch University in South Africa. Along with biochemist Douglas Kell, Pretorius has been studying what the pair calls microclots in blood samples of Covid patients. On the lab bench behind Pretorius sits an instrument provided by Martin Krater of the Max Planck Institute, which has been used to study the microfluidics of whole blood samples containing such microclots. Visual: Kent Andreasen for Undark
A blood sample in Pretorius’ lab. Around May 2020 she first discovered odd structures while analyzing blood from severely ill Covid patients. Visual: Kent Andreasen for Undark
An array of test tubes; some used to collect blood, others used to balance the centrifuge when spinning down samples. Visual: Kent Andreasen for Undark
Microscope slides in Pretorius’ lab containing processed blood to search for microclots (right) and to be analyzed for platelets (left). Visual: Kent Andreasen for Undark
From Pretorius’ lab, fluorescence microscope images of a control sample (left) taken from a healthy individual compared to a sample taken from a person with long Covid (right). When the blood plasma is processed and stained with thioflavin T, Pretorius says microclots light up in green under the microscope. Visual: Courtesy of Resia Pretorius

In another study, Pretorius, Kell, and other researchers reported that adding part of the spike protein from SARS-CoV-2 triggers amyloid clotting in plasma samples from healthy people. Pretorius and her collaborators later performed studies showing that clots in blood plasma samples from people with long Covid contain a variety of pro-inflammatory molecules. Based on these test tube studies, Pretorius believes that microclots function as “little garbage bags” that roll through blood vessels, entrapping various proteins and bits of molecular debris.

In patients who recover normally from Covid, “microclots will be dissolved by the normal physiological processes, and then the endothelial layers will be fine,” she said. But in a subset of the population, she speculated that spike proteins studding the SARS-CoV-2 surface will linger in circulation such that patients experience an “onslaught of permanent, ongoing microclot formation that doesn’t break down; platelet hyper-activation; and widespread vascular damage.”

Kell agreed and took that speculation further: “When you’ve got these essentially insoluble bits of gunk sloshing around, they can do bad stuff like block blood flow,” he said. “That can cause almost all the consequences that you see, including fatigue, post-exertional malaise, brain fog, auto-antibodies.” In Kell’s view, microclots are at the core of what’s driving long Covid symptoms in millions of people around the world.

“We don’t have the luxury of waiting for long, longitudinal randomized control trials. So we need to use the evidence that we have available.”

But other scientists who spoke with Undark said that conclusion is premature given that Kell and Pretorius’ results haven’t been replicated or supported with follow-up studies. Hunt also questioned whether long Covid patients have endothelial cell activation — the inflammatory reaction that triggers clotting. “I’m very happy to believe there’s endothelial cell activation going on,” she said. “I’d just like more substantial evidence to support it.” Also, Hunt said that researchers have yet to actually find blood clots in the small vessels of patients who have long Covid, explaining that they could be easily found through imaging or biopsies. “I’m not saying it doesn’t happen,” she said. But “where is the evidence?”

Indeed, a recent study examining post-exertional malaise, the symptom of extreme fatigue experienced after a bout of exercise, used thioflavin T to look for amyloids in muscle biopsies of long Covid patients. The researchers found amyloid-containing deposits in muscle tissue but not in capillaries, leading them to conclude that microclots weren’t blocking blood vessels as hypothesized. What the clots are doing in the muscle and whether they lead to any symptoms, though, was unclear, said study co-author Rob C. I.  Wüst of Vrije Universiteit Amsterdam, in an email.

At the same time that Pretorius and Kell were delving into amyloid clots in the lab, Laubscher, was independently treating his severe Covid patients with triple therapy. He was using a tool called thromboelastography to measure their blood’s ability to clot, and then prescribing blood thinners based on the diagnostic results. Laubscher never published any data with his acute Covid patients in scientific journals. In videos posted to YouTube, he has described his use of triple therapy and claimed that most of his patients got better and never needed to be placed on ventilators. Clinical trials, though, have been conducted. One showed that anticoagulants improve outcomes for those with severe Covid, specifically in hospitalized patients who don’t yet require intensive care.

Toward the end of 2020, Pretorius started referring people with long Covid to Laubscher. Following treatment with anticoagulants, the long Covid patients “started to recover,” Pretorius said. Laubscher has since posted one of the few available articles describing triple therapy outcomes in long Covid patients. It appears on Research Square, an online server for papers that haven’t been peer-reviewed, but has not yet been accepted to a journal, according to Khan, who is one of the co-authors.

In a recent study, researchers used thioflavin T to look for amyloids in muscle biopsies of long Covid patients. Shown here, fluorescence microscope images reveal amyloids (green) in skeletal muscle, but located adjacent to, rather than inside endothelial cells (pink). The researchers concluded that amyloids were not present within capillaries. Visual: Appelman et al., Nature Communications, 2024

During the study, also co-authored by Pretorius and Kell, Laubscher treated 91 people with a triple therapy cocktail of two antiplatelet drugs — clopidogrel and aspirin — combined with apixaban, an anticoagulant. The report claimed that symptoms resolved in most patients, along with the severity of amyloid microclotting in their blood. In a video posted in December 2022, Laubscher claimed that out of the 373 long Covid patients he had treated with triple therapy, two had experienced major bleeds, one requiring hospitalization. Bruising was common among the rest, he said. The treatment typically worked faster and was more effective in patients who had suffered long Covid symptoms for less than six months. But without a control group, it’s impossible to know if the patients improved because of the treatment or because of the passage of time.

Pretorius emphatically said she doesn’t endorse any particular microclot treatment. “I’m not a clinician,” she said. “But I do support the efforts of the clinicians that want to try different things to try to help patients.” Pretorius’ role in that regard has mainly been to share the analytical methods for detecting microclots with scientists who plan to use it for research. Commercial users — doctors who want to use the method for diagnostic purposes and therapy — must first sign a licensing agreement with Pretorius’ startup company, Biocode, which assumes responsibility for quality control.

David Putrino, a professor at the Icahn School of Medicine at Mount Sinai who has a doctorate in neuroscience, is now trying to improve the specificity of the diagnostic approach. Putrino directs a center at Mount Sinai that offers specialized rehabilitative treatment for people with complex chronic illnesses such as Lyme, long Covid, and myalgic encephalomyelitis/chronic fatigue syndrome. To Putrino, the microclots hypothesis sounded plausible given that SARS-CoV-2 enters cells after binding with their ACE-2 receptors, which he described as being over-represented on blood vessels. If the virus circulates in the bloodstream, he reasoned, then that might go a long way towards explaining long Covid’s systemic symptoms. So Putrino reached out to Pretorius and asked if she would teach him how to look for amyloid clots in patient blood samples.

Doctors who want to use the method for diagnostic purposes and therapy must first sign a licensing agreement with Pretorius’ startup company.

Pretorius wound up coming to Mount Sinai and teaching him the method personally. “Sure enough, when we started testing folks, we would see evidence of these microclots,” and hyper-activated platelets, Putrino said. But the method is also limited in that it can only reveal if microclots and activated platelets are present in a given blood sample, and not how their numbers vary based on symptom severity, he said. So, Putrino’s team is now using machine learning and computer vision tools to derive quantitative microclot scores that, he said, might reveal how well particular treatments are working.

Putrino corresponds routinely with Pretorius and Kell, and he said their findings helped to show that long Covid patients actually have biological evidence of disease. Complex chronic conditions that now include long Covid, he said, have experienced years of neglect in part because doctors can’t find anything physically wrong with people who suffer from them. “We have the gall to call these illnesses ‘invisible illnesses,’” Putrino said. But “they’re highly visible when you know where to look.”

“Resia taught us how to look for microclots and platelet hyperactivation,” he added. “In many cases, these things would correlate with symptom burden.”

Pretorius looks into the fluorescence microcope which she used to image microclots for the first time. Pretorius has gone on to teach colleagues, like professor David Putrino, the technique. Visual: Kent Andreasen for Undark

But the nature of the biological evidence remains contentious. Jeffrey Winters, a pathologist and chair of the division of transfusion medicine at the Mayo Clinic, said the way Pretorius and Kell describe amyloid clots is confusing to scientists who view amyloids in a different context. “I don’t understand how they are using this term,” he said. “I assume that what they are referring to is not what I refer to as amyloids,” which he describes as various proteins that deposit into organs and don’t ordinarily float around in the blood. “I don’t understand what they are testing, and I don’t understand what they are seeing.” The dye used in the test, thioflavin T, can and does stain other things that are not amyloids, Winters said in an email.

For researchers accustomed to working with fluorescence microscopy, the method for detecting microclots is simple, Pretorius said. When asked for a published description of the method, she directed Undark to a 2021 paper with Laubscher as lead author. Hunt’s response was that working off the written account, “I would be unable to reproduce the assay in my lab.” The method needs to be more clearly presented, Hunt said, so that another investigator can perform it “without having to go for a teaching lesson.” Furthermore, the method has yet to be publicly validated in studies with control patients who do not have long Covid, Winters pointed out. If confirmed by other researchers in peer-reviewed publications, then “I might be inclined to seek to offer the test and even license it for clinical use,” Winters wrote in an email. “But without peer-reviewed publications demonstrating utility, I am [loathe] to do so.”

The clinical evidence with microclot treatments is yet another point of contention, given the few available studies with treated patients. In May 2023, Jaeger published a case report describing the results of apheresis in long Covid. Jaeger uses a specialized form of the technology called heparin-mediated extracorporeal LDL precipitation, or HELP, apheresis that selectively removes fibrinogen and lipoproteins from blood plasma. (The equipment is not currently being sold in the U.S.) HELP differs from plasmapheresis, the more widespread alternative that replaces all of a patient’s plasma. Jaeger claims HELP apheresis filters microclots and SARS-CoV-2 spike proteins along with viral debris. Her report claims that 16 of 17 treated patients felt immediate improvement and 12 reached nearly full recovery. However, Winters, who is also editor in chief of the Journal of Clinical Apheresis, said he has yet to see published, peer-reviewed evidence showing that HELP apheresis removes spike protein or amyloid clots.

Hunt’s Cochrane Review, which was published in July 2023, couldn’t identify reliable research showing that fibrinogen particles contribute to long Covid or studies that investigated removing those particles with plasmapheresis. In the absence of a better therapeutic rationale, Hunt and her colleagues urged that plasmapheresis for long Covid should not be used outside of clinical trials. But in an email to Undark, Kell wrote that HELP and plasmapheresis are “completely different,” adding that “no one informed and of whom we are aware has ever suggested that plasmapheresis might be used to remove microclots, and we have never made any comments to that effect.”

For all of Undark’s coverage of the global Covid-19 pandemic, please visit our extensive coronavirus archive.

According to Winters, European clinics offering HELP apheresis are generally doing “cash on the barrelhead,” meaning that patients pay upfront for the service along with “a lot of other alternatives that you can add on, such as hyperbaric oxygen and vitamin infusions.” The associated costs range from roughly $10,000 to $20,000 for three to five weeks of therapy, including transportation, accommodation, and laboratory expenses. Jaeger is medical adviser to the Apheresis Center in Larnaca, Cyprus, which posts glowing testimonials from its treated patients, many of whom claim improvements to their symptoms, and also markets its services directly to consumers. But Winters emphasized that, short of a clinical trial, “even a large case-series in a peer-reviewed publication would be useful for me to get a sense of what they are seeing,” and comments posted to long Covid forums reveal that some patients do worse after treatment.

Triple therapy for long Covid also attracts skepticism from scientists who say it isn’t adequately supported by published evidence. Michael Putman, an assistant professor and rheumatologist at the Medical College of Wisconsin, said that given the bleeding risk, treating long Covid with triple therapy outside the context of a clinical trial is “dangerous and irresponsible.” Putrino said he believes there’s a rationale for triple therapy in long Covid, but that clinicians at Mount Sinai are holding off on the treatment until supporting data become available. “I would challenge people who are strong proponents of this therapy to run the necessary trial and do the necessary work to validate it,” he said.

Cost of treatment with HELP apheresis ranges from roughly $10,000 to $20,000 for three to five weeks of therapy, including transportation, accommodation, and laboratory expenses.

Jordan Vaughn, an internist and chief executive officer of MedHelp Clinics, a private practice that employs about 20 physicians at five locations in the Birmingham, Alabama area, claims to have treated roughly 1,600 long Covid patients with blood thinners in various combinations — along with over-the-counter anti-clotting supplements (nattokinase and serrapeptase) and other approaches — but has never published a case series describing their clinical outcomes. “I am kind of the independent guy out there trying to help people,” Vaughn said, adding that he lacks the resources to conduct clinical research and publish his results.

During the pandemic, he noticed that acute Covid patients often had coagulation problems, which he diagnosed in part by measuring blood levels of D-dimer, a clot-associated protein. If the levels were high, then Vaughn treated with antiplatelets and anticoagulants, claiming they were “incredibly effective.” Later, when patients started showing up with long Covid, Vaughn dug into the medical literature and concluded that the “theories that Resia and Doug had, it made a lot of sense.” Vaughn bought a used fluorescence microscope for around $50,000, and then Pretorius taught him remotely how to look for amyloid clots in blood samples. He said his long Covid patients trusted him when he suggested blood thinners. “These patients were miserable,” he said. They were willing to undertake experimental treatment with drugs that were already approved by the U.S. Food and Drug Administration — meaning they weren’t Emergency Use Authorization “like the other crap,” he added, alluding to Covid vaccines and other treatments that received accelerated approval in the early days of the pandemic.

Vaughn said that he might launch a clinical trial with other collaborators, mentioning Rae Duncan, a consultant cardiologist with the U.K.’s National Health Service.

Academic and government-funded institutions, meanwhile, have yet to conduct any trials with either triple therapy or HELP apheresis, according to Banerjee and Winters. Winters said he has “great deal of respect” for some of Jaeger’s findings, even if her 2023 study “smacks a bit of too much enthusiasm for the treatment and not as independently as I would have liked.”

The blood thinner arm of the U.K.’s Stimulate-ICP study is testing a single drug — rivaroxaban — rather than the triple therapy administered by Vaughn and other doctors. “You can’t just jump into any study and go straight to triple therapy,” said Banerjee, the study’s principal investigator. Safety needs to be shown first “with single-agent, randomized evidence.” Banerjee acknowledged that the pace of research isn’t ideal for patients. “But we’d rather have science done properly than shoddily and running out things that might not work,” he said.

Amitava Banerjee, a cardiologist and professor of clinical data science at University College London, leads the only active clinical trial investigating blood thinning in long Covid. “We’d rather have science done properly than shoddily and running out things that might not work,” he said. Visual: Lewis Khan for Undark

Hannah Davis, a data scientist and co-founder of the Patient-Led Research Collaborative, which advocates for patient involvement in long Covid research, expects that clinical trials will yield some results on treatments this year. She added that blood-based treatments might result in short-term improvements for some, or potential cures for others, if given early enough. “But the more likely scenario is that something is happening upstream of microclots that is causing them,” she said. “And that is what would need to be treated to solve the full illness.”

For Khan and other long Covid patients, effective treatments can’t come soon enough. “In the U.K., we’ve got maybe 2 million people affected, according to the most reliable current estimates, and at least a third of them are severely disabled, as in they’re unable to function in society,” he said. “This is a huge issue and it’s not going to go away”.

Khan wound up having 15 sessions of HELP apheresis and six sessions of plasmapheresis, and nearly three years later, he remains on anticoagulants, which he has tried to discontinue multiple times. He credits the treatments with improving his heart palpations, cognition, sleep, and other problems, adding that no one symptom is cured. The microclots hypothesis “is an important area for research,” he said. A few committed clinicians and researchers “are trying their very best to get this research out there, to get the studies done properly,” he added. But “there’s no will on the part of responsible government bodies to go down that route.”

UPDATE: A previous version of this piece described David Putrino as a physical therapist and professor at the Icahn School of Medicine at Mount Sinai. Contrary to Putrino’s public profile on the Mount Sinai website, he is not licensed as a physical therapist in New York. As state law requires licensure to use the title, it has been removed.


Charles Schmidt is a senior contributor to Undark and has also written for Science, Nature Biotechnology, Scientific American, Discover Magazine, and The Washington Post, among other publications.