Skip to main content
Premium Trial:

Request an Annual Quote

Through $1.15B NIH Project, Researchers Look for Markers of Long COVID for Potential Dx Tests


The story previously said that 40 pathobiology studies  received funding due to an error on the RECOVER website. The correct number is 27 projects.

NEW YORK – Hoping to crack the code behind long COVID's myriad ailments, more than 400 researchers are coordinating on a federal project that could identify biomarkers of the disease and lead to reliable diagnostics tests.

The National Institutes of Health's four-year, $1.15 billion longitudinal study program, Researching COVID to Enhance Recovery (RECOVER), is on track to recruit about 18,000 adults and 20,000 children to study the causes, effects, and treatments of long COVID, including 27 pathobiology studies that could help reveal signatures of the disease. The project started in May 2021 and runs through May 2025.

"One of the unique aspects about RECOVER is that we are studying a disease that doesn't yet have an internationally or even nationally agreed definition," said Stuart Katz, principal investigator of the RECOVER Clinical Science Core and director of the New York University Langone Health Heart Failure Program. While long COVID, also known as post-acute sequelae of COVID-19 (PASC), is commonly associated with fatigue, pain, cognitive difficulties, and breathing difficulties, Katz said, patients have also reported symptoms affecting every organ of the body, often in unpredictable combinations.

Through about 200 institutions working under RECOVER, the NIH is funding efforts to pin down which symptoms are more common among people with virus exposure versus uninfected controls, the durations of those symptoms, and what new symptoms emerge during the disease's progression. The participants agree to return surveys every three months and provide blood samples every six, with that data analyzed alongside imaging, electrocardiograms, pulmonary function tests, exercise tests, and sleep studies.

The RECOVER researchers are simultaneously trying to define the disease, identify potential biomarkers, and learn how they could improve treatments.

"We are hoping that this is all going to come together and, once we have a critical mass of data, we'll be able to define the syndrome and determine whether or not we've discovered any biomarkers," Katz said.

Five institutions are overseeing the research consortium: NYU Langone Health is convening researchers and coordinating clinical science studies, while Massachusetts General Hospital is managing access to data such as research results and health records. The Mayo Clinic is managing and distributing biospecimens obtained through RECOVER studies, and the Duke University Clinical Research Institute is coordinating clinical trials data, while the nonprofit Research Triangle Institute International is providing administrative oversight.

Multiple phenotypes may require multiple tests

The Institute for Health Metrics and Evaluation at the University of Washington School of Medicine, which is not involved in RECOVER, said in September that its research suggests about 145 million people worldwide suffered from long COVID symptoms during the first two years of the pandemic. In October, IHME researchers published a study in the Journal of the American Medical Association that indicated about 6 percent of people with SARS-CoV-2 infections experienced long COVID symptoms for at least three months in 2020 and 2021 and about 1 percent had symptoms for at least one year. Women were about twice as likely to develop long COVID than men.

Meanwhile, the US Centers for Disease Control and Prevention estimated in June that about 8 percent of US adults had long COVID symptoms that started during or after COVID-19 and lasted at least three months.

Long COVID remains a mystery, however, with many unanswered questions. For example, researchers from the US National Institute of Allergy and Infectious Diseases reported earlier this year they were unable to explain PASC through results of physical examinations, laboratory testing, questionnaires, and assessments of cognitive and cardiopulmonary function.

Even without an established signature of the disease, a few companies have already begun work on their own long COVID tests. QMC Health, for example, said last month it was developing a blood-based assay using biomarkers discovered at Lawson Health Research Institute in Ontario. Lawson officials separately announced in October that machine learning had identified elevated concentrations of the vascular transformation proteins ANG-1 and P-SEL as useful to classify long COVID with 96 percent accuracy.

University College London researchers also recently suggested plasma proteome profiling could help identify who is susceptible to long COVID, especially through an iron-sulfur cluster biogenesis protein known as HSCB or HSC20.

In August, IncellDx's test for long COVID received CE-IVD marking. The blood-based test is for measuring immune system signatures, and according to the San Carlos, California-based company, a validation study demonstrated the test had greater than 90 percent accuracy across SARS-CoV-2 strains.

To help solidify the causes or combination of causes, Katz said the coordinators of RECOVER's pathobiology research program are focusing on mechanistic studies that are organized under the themes of viral persistence, autoimmune disease, and altered immune system function.

"I anticipate that we're going to have diagnostic tests that are more tailored towards specific phenotypes," said Paul Utz, professor of immunology and rheumatology at Stanford University. "It becomes challenging, though, because it's not like we have clean phenotypes of patients."

Utz is leading a RECOVER pathobiology study that is analyzing PASC patients' levels of autoantibodies that are typically associated with connective tissue diseases, such as lupus or rheumatoid arthritis, as well as characterizing anti-cytokine antibodies, potential cross-reactions between antibodies and the body's own proteins, and G-protein coupled receptors that can be important elements of autonomic dysfunction.

"If we could identify even a small number of analytes to which there might be antibodies against them, then it would be pretty straightforward to then translate those to CLIA-level platforms," he said.

Because long COVID is heterogeneous in presentation, Utz predicts some patients with immune system defects will benefit from immunosuppressive therapies whereas others with lingering virus would benefit from antiviral treatments.

"In many ways, this reminds me of studying lupus, where no two patients are alike and it can be a real challenge to figure out what kind of diagnostic test to employ," he said.

Katz said it's unclear so far who will get to turn any research results into assays. The investigators participating in the consortium would likely develop any promising biomarkers into commercial products through their academic institutions and industry partners.

In another RECOVER-funded project, Rafick-Pierre Sékaly, professor and vice chair of translational medicine at the Emory University School of Medicine's Department of Pathology and Laboratory Medicine and longtime researcher into immune response to viral infections and vaccines, is trying to identify metabolic and epigenetic mechanisms that could help predict who will develop PASC and how it will manifest. Although he expects each patient will have a unique metabolite combination that is influenced by factors including comorbidities, gut microbiome populations, prescription drug use, and diet, better understanding of those influences could aid treatment decisions.

Matt Huentelman said his research team at the Phoenix-based nonprofit Translational Genomics Research Institute is working in a RECOVER-funded project to understand the interactions of neurological and inflammatory responses in pediatric long COVID patients ages 13 and under, with a goal of identifying who needs the most medical care and how doctors might be able to prevent the disease. He and Amy Salisbury at the Virginia Commonwealth University School of Nursing are co-principal investigators of the study.

Through that study, the researchers are collecting data on patient-reported brain functions such as cognitive skills and memory, imaging the brain to examine its structure and function, looking for markers in blood of an overly active inflammatory response, and characterizing patients' T cells and their concentrations in the bloodstream.

"We don't know which of these is going to be critical to measure, the brain side or the inflammatory/immune side," he said. "But we're kind of guessing that you need to look at both because we clearly know that these long COVID patients oftentimes have complaints that are related to the brain."

The search for treatment targets

In a Nov. 8 recording for the RECOVER Research Review video series, a series of researchers tied to RECOVER-funded projects detailed the importance of identifying biomarkers that can reveal treatment targets and provide measurements on whether those treatments are effective.

Michael Peluso, infectious disease clinician at the University of California, San Francisco and clinical lead for the university's Long-term Impact of Infection with Novel Coronavirus (LIINC) study, said in the video that he sees similarities between the early work to identify biomarkers of long COVID and the struggles of the 1980s and '90s to develop diagnostics and treatments for HIV/AIDS due to a lack of biomarkers. Studies so far have identified potential mechanistic biomarkers but not how they are related, though he expects increasingly complex studies over the next year will explore those relationships and help delineate which pathways to disrupt in long COVID treatment.

He cited as examples studies that identified higher concentrations of SARS-CoV-2 proteins in neuronal and astrocytic exosomes of patients with neurologic symptoms of long COVID and, in general, higher concentrations of inflammatory biomarkers such as interleukin 6, TNF alpha, and IP-10 among people with long COVID. He noted one study that linked elevated biomarkers of inflammation following COVID-19 with lower exercise capacity.

"We're beginning now to tie measurements that are made with biomarkers in the blood to both symptoms and also objective physiologic measurements in people experiencing long COVID," he said.

Mohamed Abdel-Mohsen, associate professor at the nonprofit Wistar Institute research organization located on the University of Pennsylvania campus, said in the video his research team identified possible links between severe COVID-19 and "leaky gut," or translocation of bacteria and fungi from the gut to the bloodstream. The team found that patients who developed PASC also had elevated concentrations of the protein zonulin, which can reflect increased digestive tract permeability, and the soluble fiber beta glucan, which can result from fungal translocation.

Though the team didn't find signs of bacterial translocation, Abdel-Mohsen said this may be because its test for lipopolysaccharide binding protein was not sensitive enough. He said it is unlikely fungal translocation occurred without bacterial translocation.

Grace McComsey, VP of research and associate CSO at the University Hospitals Health System in Cleveland and professor of pediatrics and medicine at Case Western Reserve University, said in the video it's unlikely that one biomarker is going to be predictive for all PASC phenotypes. But biomarkers are needed to understand the mechanisms and phenotypes of the disease, and studies are necessary to help predict who will develop PASC and to develop interventions to prevent it.

"We may need to combine multiple biomarkers or even have some scores with weighted clinical factors in addition to a combination of biomarkers that we can use in treatment studies," she said.

Long road to recovery

Katz knows firsthand the disruptive, even devastating effects, of long COVID. He developed severe fatigue, sleep disturbance, an altered sense of taste that made foods including coffee unbearable, hair loss, and intermittent quivering in his legs — symptoms that together took the course of a year to wane. He said at least tens of thousands of others have had worse experiences, and some have struggled with disabilities from long COVID since early 2020.

He described the ordeals of a mid-career physician who had no known health conditions before she contracted COVID-19 almost three years ago and has since been unable to return to work because of severe fatigue. Another woman he met had teeth and hair loss, was unable to focus, and, because she could no longer work, lacked health insurance.

"There are people suffering, and that's one of the things we struggle with in RECOVER because we're trying to get the answers," Katz said. "But getting answers takes time and, for people that are suffering, it's not fast enough."