COMPARISON OF PERFORMANCES BETWEEN THREE COMMERCIAL ELISA KITS FOR DETECTION OF ANTIBODIES AGAINST PORCINE REPRODUCTIVE AND RESPIRATORY SYNDROME VIRUS (PRRSV) IN SWINE SERA SAMPLES

Porcine Reproductive and Respiratory Syndrome (PRRS) is one of the most economically important diseases in pigs, worldwide. In the USA alone, the total cost to the swine industry has been estimated at $664 million per year. Th erefore, the continuous and reliable monitoring of the PRRS status of a pig herd is required in order to prevent and reduce the costs caused by this infection. Nowadays, commonly used methods for laboratory diagnosis of PRRS infection are serological (ELISA) and molecular (PCR) ones. Th is study aims to assess the sensitivity and specifi city of three diff erent commercially available ELISA kits for detection of antibodies against PRRSV (IDEXX PRRS X3 Ab Test (IDEXX, USA), INgezim PRRS Universal (Ingenasa, Spain), and Pigtype PRRSV Ab (Qiagen, Germany)) using 91 blood serum samples collected from pigs in Serbia. Our study showed a certain level of diff erences in specifi city and sensitivity between three commercially available ELISA kits. However, IDEXX ELISA proved to be a more reliable kit for detecting antibodies against PRRSV with sensitivity of 97.4% and specifi city of 98.1%, compared to INgezim and Qiagen kits specifi city of 92.5% and 83%, respectively, and sensitivity of 94.7% for both kits. In order to achieve maximal reliability of the obtained results, ELISA diagnostic protocol for laboratory diagnosis of PRRS infection should be complemented with additional tests such as PCR and virus neutralization test.


INTRODUCTION
Porcine reproductive and respiratory syndrome (PRRS) is a contagious viral infection and one of the most common infectious diseases of swine globally, responsible for signifi cant economic losses worldwide. Th e infection was fi rst recognized in the USA in 1987, while the fi rst cases in Serbia occurred in 2001, aft er illegal import of boar semen (Petrović et al., 2011).
Th e infection is caused by a single stranded RNA virus which belongs to the Arteriviridae family and the Arterivirus genus. Th e virus is biologically, antigenically and genetically heterogenic (Meng, 2000). Currently, PRRS virus is divided into two genotypes: PRRS-1 and PRRS-2 (ICTV -International Committee on Taxonomy of Viruses, 2018). Both genotypes are globally spread and enzootic in many countries. PRRSV-1 predominates in Europe while PRRSV-2 is most prevalent in Americas and Asia (Brar et al., 2015;Balka et al., 2018).
PRRS aff ects all categories of pigs. Clinical signs of PRRS vary greatly, from respiratory symptoms to reproductive failure in breeding herds, such as premature parturitions, late abortions and farrowing of stillborn and non-viable piglets. Clinical signs of PRRS are not characteristic and the course of PRRS infection can be subclinical, enabling the persistence of infection for a longer period in the herd until diagnosed, causing signifi cant economic losses. In adult pigs, seroconversion may be the only indication that infection with PRRSV has occurred (Bojkovski et al., 2012). All of the above mentioned indicates that extensive surveillance programs are necessary for control of the infection in order to minimize losses caused by PRRS, as well as to improve animal welfare. Disease control nowadays oft en involves vaccination (Savić et al., 2018). Th e laboratory diagnosis of PRRS infection is sometimes very complex, due to signifi cant antigenic diversity of fi eld isolates (Milićević et al., 2020).
Most commonly used methods for detection of PRRSV are polymerase chain reaction (PCR) and enzyme-linked immunosorbent assay (ELISA). Th ere are several commercial ELISA kits, usually used as a cost-eff ective method for detection of antibodies against PRRSV.
One of the most cited kits is IDEXX PRRS X3 Ab Test (IDEXX, Westbrook, USA) (Zimmerman et al., 2006) which stands out with great performances such as high sensitivity and specifi city, easy protocol and reproducibility (Ferrin et al., 2004), though the occurrence of false-positive reactions is noted at a rate of 0.5-2.0% (Zimmerman et al., 2006). ELISA results are sometimes indefi nite and require additional tests (Antunes et al., 2015).
Th e aim of this study was to compare diagnostic sensitivity and specifi city of three commercial PRRSV ELISA kits performed on 91 pig blood serum samples in order to determine the optimal tool for characterization of herd immunity.

MATERIAL AND METHODS
A total of 91 pig blood serum samples were used to evaluate the performances of three commercially available ELISA kits. For this purpose, serum samples were selected from the serum bank of the Institute of Veterinary Medicine of Serbia, collected during 2018 and 2019. Serum samples were collected individually from pigs located on diff erent farms in Serbia with unknown PRRSV status and were brought to the Institute of Veterinary Medicine of Serbia for diagnostic purposes. Serum samples represent heterogeneous group of pigs, referring to diff erent age of animals originating from diff erent farms. All the samples were centrifuged prior to testing and were stored at the temperature of -80 °C.
All the samples were analyzed using three diff erent, commercially available indirect ELISAs: INgezim PRRS Universal (Ingenasa, Madrid, Spain) -in the following text referred to as INgezim, IDEXX PRRS X3 Ab t est -in the following text referred to as IDEXX, pigtype® PRRSV Ab (QIAGEN, Leipzig, Germany) -referred to as Qiagen in the following text. All ELISAs detect antibodies to both PRRSV genotypes. Th e assays were performed with no modifi cations, according to the manufacturers' recommendations. In all three ELISAs, (S/P) cut-off value of positive samples is set at 0.4. Th e optical density (OD) was measured by the ELISA reader (Multiscan, Labsystem). As the true PRRS infection status of the sampled animals was unknown, the diagnostic sensitivity and specifi city of the three ELISAs were assessed based on the comparison of the proportion of the samples that reacted positively i.e. negatively. Th e samples that reacted positively in at least two applied kits were considered as positive, i.e. negative samples were marked as negative when at least two applied kits showed a negative result. False positive and false negative samples were regarded as samples that on one of the performed tests reacted ly i.e. negatively, while other two performed tests showed negative i.e. positive result, respectively. Sensitivity and specifi city values for all three ELISAs were calculated according to the following formula (sensitivity = number of true positives / number of true positives + number of false negative, specifi city = number of true negatives / number of true negatives + number of false positives).
Th e statistical evaluation was carried out using Chi Square test, with statistical signifi cance at the level of P<0.05.

RESULTS
Th e results obtained from three commercially available indirect ELISAs for the detection of PRRSV antibodies are given in Table 1. Sensitivity and specifi city values for all three ELISAs are given in Table 2.  ., 2015). In the present study, the performances of three commercially available ELISAs were compared between each other using pig serum samples collected from the fi eld. Our results have shown that IDEXX kit with sensitivity of 97.4% and specifi city of 98.1% has higher sensitivity and specifi city in comparison to other two ELISAs. Th is result is in agreement with results of the earlier studies that also indicated excellent performance of this kit  ) have calculated that the specifi city of Qiagen kit is 98.1%, while our results showed that this value was at the point of 83%. Although some authors state high sensitivity of INgezim and Qiagen kits (Sattler et al., 2015), in our research, sensitivity of these two tests were equal (94.7%) and are not signifi cantly lower than sensitivity of IDEXX kit (97.4%). Th e same authors (Sattler et al., 2015) cited that IDEXX distinguishes itself with a particularly high specifi city, while the INgezim and Qiagen ELISAs stand out with a high sensitivity. Regarding the results of other authors (Sattler et al., 2014), we have also demonstrated that IDEXX and INgezim ELISA kits are reliable for the anti-PRRSV antibodies detection. However, despite their reliability, sometimes the specifi city of the kits has been challenged by unexpected false positive results (Seo et al., 2016).
Furthermore, the results obtained with Qiagen ELISA kit, regardless of its high sensitivity (94.7%), but compromised with low specifi city (83%), should be interpreted with caution and confi rmed by another method due to the high percentage of false positive results. Th is is of paramount importance when testing herds free from PRRS infection either to maintain or prove their freedom. All the above mentioned leads to the conclusion that the selection of test should depend on the goal that is expected to be achieved and specifi c purpose of use.
Another point that should be taken into account is that validation of all tests should be performed with blood sera samples of local animals before their use in practice. Th e samples used in our study represent samples collected from the fi eld, obtained from animals with unknown PRRS status, unlike other researches that used the herds with well-known PRRS status -experimentally infected or vaccinated animals (Diaz et al., 2012; Sattler et al., 2014).
Th e diff erences in the obtained results between the three kits may originate from diff erent viral antigens used in ELISA kits, method of its preparation, the heterogeneity of local virus strains circulating in our country, etc. Furthermore, a relatively small sample size can be the reason for the obtained results and non-signifi cant diff erences in performances of the tested kits. Our results are in agreement with reviewed publications and they show a good accordance between PRRSV ELISAs in general. Taking into account the heterogeneity of recently isolated PRRSV strains in diff erent European countries, the defi nition of an adequate gold standard may be diffi cult (Karniychuk and Nauwynck, 2014). Regardless of the choice of applied ELISA kit for detection of antibodies against PRRSV infection, all assays should be complemented with virus neutralization test, that is internationally regarded as gold standard for fi nal identifi cation of PRRSV antibody positive individuals, or in any case with RT-PCR for detection of virus presence.

CONCLUSION
Our study showed a certain level of diff erences in specifi city and sensitivity between three commercially available ELISA kits. However, IDEXX ELISA proved to be highly sensitive and highly specifi c. PCR diagnosis or virus neutralization test should complement ELISA diagnostic protocol to ensure the maximal reliability of obtained results.
In any case, the choice of an applied test should be in accordance with specifi c situation and purposes specifi ed in the beginning.