Detailed Study Finds Accurate Antibody “Fingerprint” for COVID-19

Serum antibody testing for COVID-19 has been plagued by high incidence of false-negative and false-positive results. Using a complex technology that identifies antibodies to the proteins produced by all known pathogenic human viruses (>400), researchers determined which viral proteins were targeted by antibodies in the blood of 232 U.S. COVID-19 patients and in blood from 190 people obtained before the emergence of COVID-19.

Using artificial intelligence techniques, the investigators identified antibodies that uniquely react to SARS-CoV-2 (the coronavirus that causes COVID-19), those that react to other individual human coronaviruses, and those that cross react to several coronaviruses. A small number of antibodies directed against particular parts of SARS-CoV-2 accurately distinguished COVID-19 patients from pre-COVID-19 subjects. The investigators created a simple and relatively inexpensive test for COVID-19, based on three specific antibodies, that accurately identified SARS-CoV-2 infection, with a 10% false-negative and a 5% false-positive rate. The study also revealed that patients with more-severe COVID-19 mount stronger antibody responses to SARS-CoV-2 and weaker antibody responses to other common respiratory viruses (including common coronaviruses) and also are more likely to harbor cytomegalovirus and herpes simplex virus 1. Men had higher SARS-CoV-2 antibody responses than did women.