The "false-positive" conundrum: IgA reference level overestimates the seroprevalence of antibodies to SARS-CoV-2.

ABSTRACT

BACKGROUND: On 12 June 2020, Brazil reached the second position worldwide in the number of COVID-19 cases. Authorities increased the number of tests performed, including the identification of antibodies to SARS-CoV-2 (IgG, IgA, and IgM). There was an overflooding of the market with several tests, and the presence of possible false-positive results became a challenge. The purpose of this study was to describe the seroprevalence and immunoglobulin blood levels in a group of asymptomatic individuals using the reference levels provided by the manufacturer. METHODS: Levels of IgG and IgA antibodies to SARS-CoV-2 were determined in blood serum by the same ELISA (enzyme-linked immunoassay) test. Patients must be free of symptoms. RESULTS: From 20 to 22 May 2020, 938 individuals were tested. There were 441 (47%) men, age 53 years (interquartile range (IQR) = 39-63.2). The sample included 335 (35.7%) subjects aged ≥60 years old. Subjects with a positive test were 54 (5.8%) for IgG and 96 (10.2%) for IgA and 42 (4.5%) for both IgG and IgA. The prevalence of IgG and IgA positive test was not different in men and women and not different in individuals under 60 and over 60 years of age. Conversely, analysing only individuals with positive tests, the levels of IgG in positive subjects were significantly higher than those with an IgA positive test, 3.00 (IQR = 1.68-5.65), and 1.95 (IQR = 1.40-3.38), respectively; P = 0.017. Additionally, individuals with isolated IgA positive tests had significantly lower levels of IgA than those with both IgA and IgG positive tests 1.95 (IQR = 1.60-2.40) and 3.15 (IQR = 2.20-3.90), respectively, P = 0.005. These latter data suggest that IgA shows a deviation of the distribution to the left in comparison to IgG distribution data. Indeed, many subjects reported as IgA positive had immunoglobulin levels slightly elevated. CONCLUSIONS: In conclusion, we strongly suggest caution in the interpretation of IgA test results. This recommendation is more important for those with positive IgA just above the reference level.