Clues About Who Will Get Long COVID

(Inside Science) -- A significant proportion of people who contract COVID-19 -- around one-third, according to most estimates -- will go on to experience symptoms that can linger for months. Little is known about this post-acute COVID-19 syndrome (PACS), more commonly known as “long COVID,” but one group of researchers has discovered a simple blood test that, when combined with other factors, could help to predict who is most at risk.

Long COVID is defined as any symptom that lasts for more than four weeks after a confirmed infection. It can vary widely, but some of the most common lasting symptoms are pain, fatigue, problems with memory and concentration, difficulty breathing, and a persistent cough. A variety of explanations have been suggested, including a lingering infection, a deconditioning of respiratory muscles, and immune dysfunction, but the cause remains unknown. And while long COVID mostly seems to affect people who had a more severe initial infection, those who had only mild cases can also experience extended symptoms.

Onur Boyman, an immunologist at the University of Zurich in Switzerland, and his colleagues are one of the many teams around the world trying to get to the bottom of this mystery. “We want to know who will be affected, and why,” he said.

Having reliable biomarkers that can help predict the risk of long COVID would be very useful said Manali Mukherjee, who studies lung disease at McMaster University in Hamilton, Canada, and who has been struggling with long COVID herself for more than a year. Even though no specific treatments exist, biomarkers could help doctors understand the disease better and offer patients more answers. “There are lots of people with long symptoms and the medical community has no answers for them,” she said. “That confusion leads to massive social impact in terms of anxiety and mistrust of medicine.”

Boyman’s team followed a group of more than 100 COVID-19 patients for a year, to track who went on to have long-lasting symptoms. They found that people with long COVID were generally older, had a more severe initial infection, showed five or more symptoms during their initial infection, were more likely to have been hospitalized, and were more likely to have a history of lung disease, particularly asthma -- all similar to other studies of the syndrome. But they also looked at antibody levels in the patients’ blood and found that two particular kinds of antibodies -- known as IgM and IgG3 -- also seemed to be involved. People with lower levels of these subsets of antibodies after their initial COVID-19 infection were at higher risk of developing long COVID.

When the antibody data was combined with the other factors the team identified, they were able to predict who would go on to have long COVID with about 63% accuracy in another group of almost 400 patients. Among hospitalized patients it was even more effective, rising to more than 90% accuracy. “It appears to be a very useful score for predicting PACS, particularly in those who were hospitalized,” said Boyman. The work was published today in the journal Nature Communications.

Konstantinos Tselios, a rheumatologist at McMaster who is working with Mukherjee on research into long COVID, says it makes sense that those particular kinds of antibodies would be involved. Other studies have shown that COVID-19 patients with higher levels of IgG3 have lower levels of the virus and less severe disease, for example. And while IgM and IgG3 make up just 12% of the pool of antibodies in the body, they are responsible for 80% of our defense against the virus that causes COVID-19, known as SARS-CoV-2. “These antibodies seem to be really significant,” he said. “If you don’t have enough of them not only will you have a worse case of COVID, but also a higher risk of long COVID.”

The fact that these antibodies can be measured with a simple blood test that is widely available would make them a highly useful biomarker, Tselios added. But both he and Mukherjee want to see the research confirmed with larger and more diverse groups of patients before it is put into clinical practice. Tselios would also like to see a clearer cutoff in how a low level of the antibodies is defined, while Mukherjee wants to know more about the mechanism that links antibody levels to long COVID.

Boyman said he and his colleagues have already started using the biomarkers with their own patients to help predict their risk of long COVID, or to see if those who already have it also have the antibody signature. And they have created a website with a calculator in which doctors can enter a patient’s information, and it will give a risk prediction based on the team’s model. “These are estimates, that don’t replace clinicians’ own judgement,” he said. “But it is good to have a guide to the risk.”