Vet. Res. 31 (2000) 73-74

Detection of antibodies to porcine reproductive and respiratory syndrome virus (PRRSV) in blended sera or in blood samples collected on filter papers

E. Hutet^a, S. Chevallier^a, M. Eloit^b, A. Touratier^c, P. Blanquefort^d and E. Albina<sup>a

a</sup>  Agence Française de Sécurité Sanitaire des Aliments (AFSSA), Unité de Virologie et Immunologie Porcines, BP 53, 22440 Ploufragan, France
^b  École Nationale Vétérinaire d'Alfort (ENVA), Unité de Génétique Moléculaire, Génétique Virale, 7 avenue du Général de Gaulle, 94704 Maisons-Alfort, France
^c  Fédération Nationale des Groupements de Défense Sanitaire du Bétail (FNGDSB), 149 rue de Bercy, 75595 Paris Cedex 12, France
^d  Fédération Régionale des Groupements de Défense Sanitaire (FRGDS), La Quantinière, 49800 Trélazé, France

Abstract - The detection of serum antibodies against PRRSV is easily done with commercial ELISA kits. These diagnostic tools are reliable and now routinely used. The kits are, however, generally designed for individual sera. Our objective was to validate the assays for blended sera and blood samples collected on filter papers in order to reduce diagnostic costs. Two ELISA kits, commercialised in France, were used. In a first step, the feasibility of detecting specific antibodies in blood eluted from filter papers was assessed. Blood was collected on filter papers from small skin scarifications of the ear or the tail. When dried, the filter papers were sent to the laboratory, and there, eluted in the sample buffer provided with the ELISA kits. The best protocol for testing blood eluted from the filter papers was defined for each kit. Then, by testing around 100 filter papers collected from non-infected pigs and 5 filter papers collected from 5 low sero-positive pigs (as assessed in the immunoperoxidase monolayer assay, IPMA), we determined the positive thresholds for the filter papers. In a second step, the amount of antibodies detectable by the two ELISAs was evaluated for the sera and filter papers by titrations of paired samples collected from around 70 infected pigs. For both kits, the geometric mean titres of specific antibodies in sera or filter papers were nearly identical. From these titre determinations, the probability for detecting one positive sample (one serum or one filter paper) mixed with increasing number of negative samples was determined. The results indicated that blends of up to 5 samples still gave a good probability of detection. Sensitivity and specificity of the detection of PRRS infection with kit 1 were then determined on 200 paired serum-paper samples, using randomly blended samples in comparison to individual sera. With blends of 5 sera, the sensitivity and specificity compared to the analysis of individual sera were 71% and 100%, respectively, whereas, the sensitivity and specificity of blends of 5 filter papers were 79% and 86%, respectively. For kit 2, the sensitivity and specificity of PRRS detection were only determined on individual sera and individual filter papers in comparison with the results of kit 1: the results were very similar, thus validating the performances of kit 2. The specificities of the two kits were subsequently estimated under routine use in a Regional Veterinary Laboratory. Around 120 blended filter papers were tested. From these tests, the specificity was statistically estimated: for both kits, specificity was above 97.5% (p = 0.05). The validation of the two kits with blended filter papers was finally achieved by a running interlaboratory test which involved 8 Regional Veterinary Laboratories and a set of 10 blends of five filter papers (1 negative, 2 low positive, 3 high positive blends and 4 repetitions). All but one laboratory obtained results in agreement with the expected results. After analysis, it turned out that the laboratory which had failed did not correctly store the samples before testing. The results of this study allow to propose a test using three random blends of 5 filter papers by ELISA, in order to diagnose PRRSV infection on farms. This protocol has been evaluated as sensitive as 10 individual sera for PRRSV infection detection. This study was granted by the Fédération Régionale des Groupements de Défense Sanitaire des Pays de la Loire and the French Ministry of Agriculture.

Corresponding author: E. Albina Tel.: (33) 2 96 01 62 90; fax (33) 2 96 01 62 53;
    e-mail: e.albina@ploufragan.afssa.fr

© INRA, EDP Sciences 2000