A comparative study of eight serological methods shows that spike protein-based ELISAs are the most accurate tests for serodiagnosing SARS-CoV-2 infections in cats and dogs

INTRODUCTION: Coronavirus disease 2019 (COVID-19) is an infectious zoonotic disease caused by SARS-CoV-2. Monitoring the infection in pets is recommended for human disease surveillance, prevention, and control since the virus can spread from people to animals during close contact. Several diagnostic tests have been adapted from humans to animals, but limited data on the validation process are available. METHODS: Herein, the first comparative study of six “in house” and two commercial serological tests developed to monitor SARS-CoV-2 infection in pets was performed with a well-coded panel of sera (61 cat sera and 74 dog sera) with a conservative criterion (viral seroneutralisation and/or RT–qPCR results) as a reference. Four “in house” tests based on either the RBD fragment of the spike protein (RBD-S) or the N-terminal fragment of the nucleoprotein (N) were developed for the first time. The analytical specificity (ASp) of those tests that showed the best diagnostic performance was assessed. The validation included the analysis of a panel of sera obtained pre-pandemic from cats and dogs infected with other coronaviruses to determine the analytical Sp (17 cat sera and 41 dog sera). RESULTS AND DISCUSSION: ELISAS based on the S protein are recommended in serosurveillance studies for cats (RBD-S SALUVET ELISA, ELISA COVID UNIZAR and INgezim(®) COVID 19 S VET) and dogs (INgezim(®) COVID 19 S VET and RBD-S SALUVET ELISA). These tests showed higher diagnostic sensitivity (Se) and DSp in cats (>90%) than in dogs. When sera obtained prior to the pandemic and from animals infected with other coronaviruses were analyzed by RBD-S and N SALUVET ELISAs and INgezim(®) COVID 19 S VET, a few cross reactors or no cross reactions were detected when dog and cat sera were analyzed by tests based on the S protein, respectively. In contrast, the number of cross reactions increased when the test was based on the N protein. Thus, the use of tests based on the N protein was discarded for serodiagnosis purposes. The results obtained revealed the most accurate serological tests for each species. Further studies should attempt to improve the diagnostic performance of serological tests developed for dogs.