Skip to main content
Premium Trial:

Request an Annual Quote

Weill Cornell Using $1.5M NIH Grant to Predict Severity of COVID-19 Cases


CHICAGO – Weill Cornell Medicine has received a two-year, $1.5 million grant from the National Institutes of Health to study ethnic, social, and biological factors — including a heavy dose of genomics — that might predict severity and outcomes of big-city patients with COVID-19. The New York City academic medical center will seek to uncover the role of genetics and social determinants of health in disparities in COVID-19 progression and mortality.

Researchers from the Clinical and Translational Science Center (CTSC) at Weill Cornell will be looking for polymorphisms that may indicate a predisposition to more severe cases of the respiratory disease caused by the novel SARS-CoV-2 coronavirus.

They are trying to determine if certain polymorphisms might be more prevalent in African American and Hispanic populations, or if higher rates of COVID-19 mortality and morbidity in these populations are mostly the result of socioeconomic conditions. They also will be measuring the prevalence of certain immunological phenotypes indicating susceptibility to severe COVID-19 in nonwhite groups.

"The diversity and strength of the immune system, shaped by exposure to prior infection, environmental factors, and poor nutrition may also explain differences in severity with an increased prevalence among ethnic groups. Environmentally induced immunological factors, such as latent inflammation occurring with variable frequencies across different populations, can perhaps further explain why disease severity is higher in some racial/ethnic groups," the researchers said in their NIH grant proposal.

They also noted that social determinants, including holding "essential" jobs serving the public, lack of access to healthcare services and healthy food, and inadequate or overcrowded housing, might be important risk factors in the susceptibility to SARS-CoV-2 infection and severe cases of COVID-19.

Weill Cornell noted that African Americans in New York have accounted for 28 percent of COVID-19-related deaths, despite making up just 22 percent of the population. Hispanics, who are 29 percent of the city's residents, have accounted for 34 percent of deaths.

The disparity is even more pronounced in other areas of the country that were hit hard by the initial wave of the pandemic. In Michigan, Louisiana, and the city of Chicago, African Americans have died from COVID-19 at more than twice the rate than their share of the population might suggest, the researchers said.

Olivier Elemento, director of the Englander Institute for Precision Medicine (EIPM) at Weill Cornell, said that it still is a scientific mystery why these communities are disproportionally affected by COVID-19, though theories abound. "There are a lot of hypotheses, but what we really want to do with this grant is to test these hypotheses," he said.

Elemento expects to find a combination of factors, including comorbidities, genetics, and social determinants of health. "The weight of the different aspects is unclear," he said.

The grant will fund immunophenotyping as well as genotyping. "We're hoping that we'll be able to do a lot of immunological investigation using the money that we got from NIH," Elemento said.

He would like to be able to quantify the role that prior exposure to microbes plays in COVID-19 outcomes.

"I think that the plan is to try to understand across multiple populations whether we can define the signatures of severe disease now and then get a sense whether the signature will be, let's say, more pronounced or more commonly found in some populations," Elemento said.

"Patients who are at high risk of blood clots because of some genetic aspect may be more likely to get blood clots as a result of the [SARS-CoV-2] infection because the infection enhances this baseline predisposition for blood clots," Elemento said.

Elemento said that it also may be important to look at risk scores for cardiovascular disease because COVID-19 can cause permanent heart damage. "It's possible that patients at risk of certain types of cardiovascular disease may also have a higher baseline for severe [COVID-19] disease," he said.

In developing the grant proposal, Weill Cornell assessed 669 COVID-19 patients from the publicly available UK Biobank dataset. In that experiment, Black patients were more than three times more likely than white participants, even after adjusting for relative risk. "These results illustrate feasibility of conducting genome-wide or focused association studies on cases with COVID-19," the researchers wrote in their proposal.

By conducting a genome-wide association study, the researchers plan on defining a genetic signature of COVID-19. They said that they expect that the Duffy-null antigen polymorphism might be part of the signature, which they will confirm by looking for DARC (Duffy antigen/receptor for chemokines) gene expression on erythrocytes. The team also will test whether a polygenic risk score they have already developed for thrombosis correlates with severity of COVID-19.

"We want to see whether this score is predictive of severe [coronavirus] disease because the patients with severe disease tend to have also coagulopathy," Elemento said. "We're thinking that maybe patients who have a high risk of blood clots, for example, may also have a high risk of developing severe disease."

The Weill Cornell team has developed several hypotheses, including that social determinants of health are largely to blame for higher rates of hospitalization, admissions to intensive care units, and death from COVID-19. On this front, they will compare the experiences of minority and low-income communities in New York City to those in affluent parts of the city, then correlate their findings with biological and clinical determinants of COVID-19 outcomes that they are also trying to ascertain.

On the biological front, Weill Cornell's CTSC will perform genetic and immunological analysis of patients admitted to NewYork Presbyterian Hospital (NYP) with COVID-19 symptoms, in hopes of defining a genetic signature that can predict disease severity. This testing will include mass cytometry-based immunophenotyping on Fluidigm Helios instruments and blood proteomics on Olink Proteomics hardware among 500 African Americans, 500 Hispanics, and, as a control group, 500 patients of European ancestry.

They are drawing samples from Weill Cornell's COVID-19 biobank, which has collected specimens from patients at NYP-affiliated hospitals in the New York City boroughs of Manhattan, Brooklyn, and Queens. Queens and Brooklyn in particular have been COVID-19 hotspots since the early days of the outbreak in the US when New York was the nation's epicenter.

The researchers also will be testing the hypothesis that there are genetic polymorphisms that occur more frequently both in those who develop severe COVID-19 and among racial or ethnic groups that have poorer outcomes with this disease. There have been no genetic polymorphisms yet correlated with COVID-19 severity, according to Elemento.

The CTSC will genotype the 1,500 subjects by running the Illumina Infinium Global Screening Array-24 version 3.0 BeadChip SNP array biobank samples collected at NYP hospitals. Half of each group will have had mild COVID-19, while the other half will have had more severe cases. The Englander Institute for Precision Medicine will perform the genotyping on Weill Cornell's genomics core technology infrastructure.

Elemento expects the study pool to include some who have died as well as some who have recovered, as long as the institution is able to obtain the proper consent.

Computationally, bioinformaticians led by Elemento will build supervised machine learning models that leverage artificial intelligence to determine the interplay between community-level social determinants of health, genetic, and clinical factors in the development of more severe cases of COVID-19 and in mortality. This technology will include predictive models to stratify patients according to risk of severe disease.

With machine learning, the researchers hope to identify the most influential community-level social, clinical, and biologic determinants, and develop predictive models that clinicians might use to assess hospitalized COVID-19 patients.

"We hope that this will lead to some actionable information," Elemento said. "I think the signatures could potentially give rise to some kind of risk score that could be used ahead of time to identify populations that are at extremely high risk."

However, he said that this work probably will require extensive validation from third parties before any risk scores could be brought into clinical settings.

The CTSC's Architecture for Research Computing in Health (ARCH) program will help the computational biologists and immunologists at Weill Cornell collect clinical data on patients with suspected or confirmed coronavirus infections. These records, which largely come from the institution's electronic health records system, include patient demographics, vital signs, diagnoses, progress notes, test results, medication lists, and biospecimens.

From an immunological perspective, previous research has shown that lymphopenia and increased neutrophil-lymphocyte ratio is present in at about 80 percent of patients who have tested positive for the SARS-CoV-2 virus, and these changes have been more stark in the most severe cases of the new disease. Both helper and suppressor T-cells have been observed in lower concentrations than normal in COVID-19 patients, and the number seems to decrease as disease severity increases, the researchers said, based on research from China.

However, it is unclear whether these and other signatures observed in Asian populations are present in those from different ethnic and racial groups, or if such markers might explain disparities.

The idea to look at whether the Duffy-null antigen polymorphism might be part of the genetic signature of severe COVID-19 comes from molecular geneticist Melissa Davis, scientific director of the International Center for the Study of Breast Cancer Subtypes at Weill Cornell. She has long been studying genetic risk factors for breast cancer in populations with West African ancestry, and is involved in this COVID-19 research.

In her breast cancer work, Davis has found that the Duffy-null T-46C polymorphism in the promoter region of the DARC gene, which is nearly exclusive to those of African descent, has been linked to higher mortality from acute lung injury. Davis herself has reported that DARC expression may be related to higher breast cancer survival rates, particularly for triple-negative breast cancer, a form that is more prevalent in those with West African ancestry.

This particular allele is known to cause lung inflammation, one symptom of COVID-19, according to Elemento. This new NIH-funded research will explore whether a DARC SNP known as rs2814778 could be a marker for severe COVID-19, which might explain disparities in Black populations.

Davis noted that genetics have been shown to improve the health of minority populations in sickle cell disease; states that require newborn genetic screening for that condition sometimes recommend it only for African Americans.

"This really comes down to understanding the functional consequences of genetic variation," particularly variants related to specific ethnic and racial groups that have often been excluded from research that has defined gene function and disease risk, Davis said in an email.

She said that if geneticists had not taken the steps to identify the sickle hemoglobin variant that causes sickle cell disease, there would be "greater stigma and suffering in not knowing" the genetic link to a potentially lethal condition.

Davis said there is always the possibility of discrimination — perhaps in the form of racist social policies — against populations who are genetically predisposed to certain diseases or outcomes related those diseases. She has tried to fight that perception with her previous work.

"As with all population health issues, marginalization of minorities has created a web of risks leading to poor outcomes," Davis said. "However, if we contextualize this work in the concept that we are defining the breadth of the disease burden from a precision medicine perspective, perhaps it will be received as inclusion instead of isolation/segregation."

If the Weill Cornell hypothesis about genetic predisposition to severe COVID-19 is correct, Davis believes that clinicians will be able to apply that knowledge to prevent complications and save lives. "In my opinion it's discriminatory not to address it," she wrote. "I'd like to reclaim ancestry differences as the beauty in biological diversity and cancel the narrative that genetic differences are a source of inferiority."

Elemento said that any polymorphisms they might find in a gene signature could inform targeted drug development in the future, helping to diversify the potential pool of patients for clinical trials.

The Eurocentric homogeneity itself of many clinical trials has created disparities as well.

"I think we are not paying enough attention to other populations," Elemento said. "I think this kind of research is also meant to demonstrate that there are differences in terms of severe disease that are due to some biological differences, but I think that this is something that is meant to be a good thing in the sense that we want to use this information as a way to develop more targeted therapies."

"What is most interesting to me is the biological understanding that would come out of this," Elemento said. He is excited about the possibility of creating actionable strategies based on any gene signature the research discovers.

He did temper his enthusiasm somewhat, though. "I think we need to keep in mind that genetics will probably most likely play a reasonably mild role" compared to social determinants of health in explaining disparities, Elemento said. "If I had to bet, I think it would be likely to explain the severity of a disease in a more profound way."