Reviewed by: Dr. Michael Payne, MD; London Health Sciences Centre and Apara Dave, MD; Exeter Hospital
Two recent studies evaluated the performance of commercial COVID-19 serological assays. Overall, there was a relatively wide range of sensitivity/specificity results. While potentially useful to determine seroprevalence at the population level, serological assays should be applied with caution for “immunity passports” or vaccine prioritization. Spike-based immunoassays appear to be better correlates of neutralizing antibodies activity, with implications for convalescent plasma treatment.
A multicenter study (10 clinical laboratories), published in the Journal of Clinical Microbiology, evaluated 10 commercial high throughput COVID-19 immunoassays. A sensitivity panel of 176 samples from PCR confirmed cases was utilized. These samples were stratified by date after symptom onset. A subset of these (67) were used for neutralization activity assessment. Specificity was assessed using samples collected from patients with alternative diagnoses (respiratory viruses, etc.), those with autoimmune conditions and also some samples collected before the COVID-19 pandemic. Health Canada recommends sensitivity and specificity performance of assays be higher than or equal to 95% and 98%, respectively, for samples collected 2 weeks or more after the onset of symptoms. Seven immunoassays examined in this study were shown to have excellent specificity (98 to 100%) and good to excellent positive predictive values (82 to 100%) when used in a low (5%)-seroprevalence setting. At 2 weeks, sensitivity of assays ranged from 74-95%, improving to 79-100% at greater than 5 weeks. As expected, sensitivity at <2 weeks from symptom onset was poor. Sensitivity improves at greater than 2 weeks and 5 weeks from symptom onset, which confirms previous studies showing sensitivity improving at more than 3 weeks from infection. Spike-based assays showed the highest coefficient of association between neutralizing antibodies and binding antibodies. COVID-19 serology is useful to determine population seroprevalence, evaluate vaccine response and assess the immune status of convalescent plasma. The data shown in this study confirms commercial assays are not useful for diagnosis of acute COVID infections.
Taking a different approach, a study from Germany recently published in the Journal of Infectious Diseases conducted the first side-by-side comparison of four different commercial serological assays for SARS-CoV-2. Samples were taken from 119 patients with confirmed prior positive pharyngeal SARS-CoV-2 PCT testing; these individuals presented as potential convalescent plasma donors. Of note, these were all patients with prior mild to moderate COVID-19, not severe disease as was the case with most prior studies of serological assays. Median duration was 48 days (range 18-170) between symptom onset and serology and 35 days (range 0-143) between convalescence and serology. The control cohort was 110 healthy individuals. The four commercial assays included in the study were Roche, Abbott, Euroimmun, and Snibe/Medac. All 229 samples were tested on Roche, Abbott, and Euroimmun, and a smaller number on Snibe/Medac given testing resources. Specificity was highest in the Roche and Abbott tests at 100% for both, followed by 97.2% for Snibe/Medac and and 90.9% for Euroimmun. Sensitivity was highest in Euroimmun at 97.5%, followed by Roche at 95%, Abbott at 81.5%, and Snibe/Medac at 60.3%. Serologic testing with Roche and Euroimmun had the highest sensitivity, with potential use in screening high-risk populations, while Roche and Abbot had the highest specificity, offering more utility in the general, average risk population. Notably, the individuals in this study had mild-moderate COVID-19, instead of severe cases or hospitalization, allowing for broader applicability of these results.