Dr. Stuart Ray was repeatedly exposed to COVID-19 in the early days of the pandemic while treating patients flooding Johns Hopkins Hospital, but he never tested positive for the disease.
That left him wondering: Was it precautions, luck or some kind of superpower?
Ray and a group of graduate students took to their computers to investigate. They created a mathematical algorithm, among the nation’s first of its kind, to search data from people like him in hopes of seeing a pattern — one that might eventually lead to better preventive measures and treatments.
“A lot of us have encountered people heavily exposed and never infected,” said Ray, vice chair of medicine for data integrity and analytics and professor of medicine in the Hopkins School of Medicine.
“We wanted to see if we could use a large data set and machine learning to identify a subset of people resistant to COVID-19,” he said. “We went in not knowing what the subset may be.”
They tapped thousands of medical records from a patient registry at Hopkins. The data included people who reported being in close contact with others infected with COVID-19 and who were tested multiple times in the months before vaccines were available, from June to December 2020.
The data also included the people’s other conditions and medications, but specifics like names and addresses were removed for privacy.
Some common threads have emerged, including at least one that’s a bit counterintuitive. Cancer patients, for example, seemed to dodge infections. That could mean their bodies developed some sort of defense mechanism, or just that they were especially vigilant about safety measures. Others may have gone on to test positive outside the Hopkins system.
The scientists also are mindful that medical records are an imperfect data set, Ray said, because they rely on medical staff to input codes for someone’s conditions and other information.
To really see patterns, and confirm the model works, Ray said he’ll need millions of medical records from around the country from people tested before vaccines, which could mask any natural defenses. He and a new crop of students plan to seek such a data set, plus funding to pay for the more extensive work.
The scientists published their initial findings in the journal PLOS ONE last month.
Eventually, the scientists could go about getting community buy-in and institutional approvals to contact individuals to see if they would provide a blood or other sample. In the lab, they could look for specific mutated genes, proteins or other factors linked to the apparent COVID-19 resistance.
Other Hopkins scientists used a similar kind of algorithm to predict which COVID-19 patients were most likely to develop a severe infection — information helpful for planning and treatment.
The idea of finding a biological factor on which to base new treatments is not unprecedented, said Matthew Frieman, a longtime coronavirus researcher who was not involved in the Hopkins study.
Frieman, a professor in the University of Maryland School of Medicine’s Center for Pathogen Research, said there “could certainly be people resistant to COVID” and other infections.
“Mutations in a gene that change immune responses to viruses or other pathogens do exist, just like people with mutations in the receptor for HIV are highly resistant to HIV,” said Frieman.
“It will be interesting to see what they find and then can validate in laboratory experiments.”
While that’s a ways off, the idea of finding people able to dodge a COVID-19 infection interested Ray’s students studying “precision medicine,” or therapies directed to individuals.
“If we can identify which people are naturally able to avoid infection by SARS-CoV-2, we may be able to learn — in addition to societal and behavioral factors — which genetic and environmental differences influence their defense against the virus,” said a statement from lead study author Karen (Kai-Wen) Yang, a biomedical engineering graduate student in the Hopkins’ Translational Informatics Research and Innovation Lab.
“That insight,” Yang said, “could lead to new preventive measures and more highly targeted treatments.”