A collaborative study of an experimental kit for rapid rabies enzyme immunodiagnosis (RREID)

Bulletin of the World Health Organization, 65 (4): 489-493 (1987) ( World Health Organization 1987 A collaborative study of an experimental kit for rapid rabies enzyme immunodiagnosis (RREID) P. PERRIN' & P. SUREAU' Six laboratories took part in a study to assess an experimental kitfor the diagnosis of rabies using the rapid rabies enzyme immunodiagnosis (RREID) technique. The test is based on the immunocapture of rabies antigens present in homogenized brain specimens, followed by enzyme immunoassay. A total of 1253 specimens from various geographical locations and 27 animal species were tested with the RREID technique, and also with the fluorescent antibody test (FAT), which was used as a reference method. For 1220 specimens the results in RREID and FAT were the same (651 positive and 569 negative- concordance: 97.4%). However, the RREID technique appeared to be less sensitive, since 22 (3 %) ofthe 673 specimens that were positive with FAT were negative with RREID. The RREID test is therefore specific and convenient and is a useful tool for epidemiological studies and for laboratories not equipped with an ultraviolet microscope. Since 1958 when Goldwasser & Kissling (1) identified rabies antigens in the tissue of infected animals, the fluorescent antibody test (FAT), as described by Dean & Abelseth (2), has become the recommended procedure for routine diagnosis of the disease. Nevertheless, to carry out the test a micro- scope equipped with well-maintained ultraviolet (UV) accessories as well as considerable experience are prerequisites. Epidemiological studies under field conditions are therefore not easily carried out using FAT. However, many convenient enzyme immuno- assays have been described (3-5), and we have recently reported a new method for diagnosing rabies in which rabies antigens in infected brain specimens are detected by enzyme immunoassay (rapid rabies enzyme immunodiagnosis, RREID) (6). Since the method appeared to offer promise for the routine diagnosis of rabies, we have developed an experi- mental kit for RREID. Six laboratories agreed to evaluate the test, and we report here the results of the study. MATERIALS AND METHODS Principle of the RREID technique The RREID technique is an enzyme-linked im- munosorbent assay (ELISA), based on the following From the Rabies Unit, Institut Pasteur, WHO Collaborating Centre for Reference and Research on Rabies, 25 rue du Dr Roux, 75724 Paris Cedex 15, France. Requests for reprints should be sent to Dr Perrin at this address. principle (sandwich technique): rabies antigens that have been solubilized from brain specimens are specifically bound to immobilized rabies antinucleo- capsid antibodies and subsequently revealed using antinucleocapsid antibodies conjugated to horse- radish peroxidase. Description and use of the RREID experimental kit The RREID experimental kit' contains all the reagents necessary for the diagnosis of rabies in brain specimens: -ELISA plates sensitized with rabbit antirabies nucleocapsid IgGs; -washing solution (phosphate-buffered saline, pH=7.2); - peroxidase conjugate (rabbit antirabies nucleo- capsid IgGs conjugated with horseradish peroxidase); -buffer containing the substrate (hydrogen peroxide) for the enzymatic reaction; -tablets of o-phenylenediamine (chromogen); -stopping solution (2 mol/l sulfuric acid); - positive control antigen (inactivated and lyophil- ized brain suspension collected from mice infected with the CVS-strain of the rabies virus); and - a negative control antigen (suspension of inacti- vated and lyophilized brain from uninfected mice). The immunoassay using sensitized ELISA plates was performed following the procedure described in " Diagnostics Pasteur, 92430 Marnes la Coquette, France. 4808 -489- P. PERRIN & P. SUREAU the kit instructions and as previously reported (6), except that the enzymatic coloration was allowed to develop for 30 minutes instead of 5 minutes. Preparation of specimens Specimens of brain were homogenized in four volumes of washing solution and centrifuged (800 g for 30 minutes) to eliminate particles of debris. The clear supernatant was then analysed for the presence of rabies antigen using the experimental kit. Participating laboratories Six laboratories (see Annex) participated in the collaborative study and these were coded A-F (the letters do not correspond to the order in the list reported in the Annex). Assay methods Each laboratory received experimental kits and staff were requested to carry out the RREID test using the instructions supplied with the kits. The results were compared with those obtained using the refer- ence technique (FAT). Although the RREID kit permits detection of rabies antigens with the naked eye, for the purposes of the study the absorbance of solutions was determined quantitatively using a spectrophotometer. The absorb- ance of specimens at X=492 nm was compared with that of a negative control antigen; samples were considered positive if their absorbance was greater than or equal to 0.05 absorbance units above that of the control. RESULTS AND DISCUSSION The results obtained by the six laboratories are summarized in Table 1. Of the 1253 specimens analysed in the trial, 651 were positive in both the FAT and the RREID tests, 569 were negative in both tests, 22 were positive in the FAT but negative in the RREID test, while 11 were negative in the FAT but positive in the RREID test. The same results were obtained for the FAT and RREID tests for 1220 specimens (concordance: 97.4%). Of the 684 specimens that were positive in one or other of the tests, 651 (95.2 %) were positive in both, 22 (3.2%) were positive in only the FAT, and 11 (1.6%) were positive in only the RREID test. Furthermore, ofthe 602 specimens that were negative in one or other of the tests, 569 (94.5%) were negative in both, 11 (1.8%) were negative in only the FAT, while 22 (3.6%) were negative in only the RREID test. The concordance varied from 89.7% to 99.2%, depending on the laboratory. For laboratories B, C, D, and F, where 84.5% of the total number of specimens were analysed, the concordance was greater than 96%. Laboratories B, C, and D reported that, of the 11 specimens that were negative in the FAT but Table 1. Correlation between the results of the fluorescent antibody test (FAT) and rapid rabies enzyme immuno- diagnosis (RREID) for the six laboratories in the study No. of specimens + FAT/ - FAT/ + FAT/ - FAT/ Concordanceb Laboratory Total + RREID -RREID -RREID + RREID (%) A 155 79 66 10 0 93.3 B 130 39 89 1 1 98.5 C 474 115 352 4 3 98.5 D 199 150 41 1 7 96.0 E 39 34 1 4 0 89.7 F 256 234 20 2 0 99.2 Total 1253 651 569 22 1 1 97.4 ' Results are expressed as the number of specimens that were positive in the FAT and RREID tests (+FAT/+ RREID); negative in both (-FAT/-RREID); positive in FAT but negative in RREID (+FAT/-RREID); and negative in FAT but positive in RREID (-FAT/+RREID). (No. positive in both FAT and RREID + No. negative in both FAT and RREID) No. of specimens tested 490 RAPID RABIES ENZYME IMMUNODIAGNOSIS Table 2. Correlation between the results of the fluorescent antibody test immunodiagnosis (RREID) for five laboratories according to type of animal (FAT) and rapid rabies enzyme No. of specimensa + FAT/ -FAT/ + FAT/ -FAT/ Concordanceb Type of animal Total +RREID -RREID -RREID +RREID (%) Cat 197 30 164 0 3 98.5 Dog 130 23 104 2 1 97.7 Cattle 85 48 35 1 1 97.6 Sheep 50 30 19 1 0 98.0 Goat 3 1 2 0 0 100 Horse 11 5 5 1 0 91.0 Pig 3 2 1 0 0 100 Fox 367 331 28 6 2 97.8 Skunk 18 17 1 0 0 100 Coyote 5 5 0 0 0 100 Raccoon 2 2 0 0 0 100 Groundhog 2 2 0 0 0 100 Badger 16 12 2 0 2 87.5 Marten 19 7 12 0 0 100 Weasel 3 1 2 0 0 100 Ferret 1 0 0 0 1 0 Rat 9 0 9 0 0 100 Muskrat 2 0 2 0 0 100 Dormouse 3 0 3 0 0 100 Wild mouse 2 0 2 0 0 100 Laboratory mouse" 7 6 1 0 0 100 Squirrel 6 0 6 0 0 100 Rabbit 4 0 4 0 0 100 Hedgehog 2 0 2 0 0 100 Deer 11 5 6 0 0 100 Bat 7 5 2 0 0 100 Bird 1 0 1 0 0 100 Total 966 531 414 11 10 97.8 See footnote a to Table 1. See footnote b to Table 1. Inoculated with wild rabies isolates. positive in the RREID test, some (if not all) were decomposed upon receipt. Studies are currently under way to determine the specificity of this positive RREID response: in two cases, the RREID response was inhibited by a previous incubation with antirabies nucleocapsid serum, but no inhibition occurred with normal rabbit serum. On the other hand, the 22 specimens that were positive in the FAT but negative in the RREID test indicate that the latter technique may be less sensitive. Nevertheless, if the high concordance between the two methods is taken into account, the findings indicate that the RREID test is specific and sensitive. The results obtained from five of the laboratories (A, C, D, E, and F) were also analysed according to animal species (Table 2). In general, for large samples, there was no difference between the results of the RREID and FAT tests for the various animal 491 P. PERRIN & P. SUREAU species in the study. For example, for species for which more than 18 animals were tested (cat, cattle, dog, fox, marten, sheep, and skunk), in 866 out of a total of 966 (90%) specimens the concordance was greater than 97.6%. The purified antinucleocapsid immunoglobulins used in the RREID test kits were obtained from rabbits that had been immunized with antigens prepared with the Pasteur strain of rabies virus, and the test is recommended for diagnosis of lyssaviruses of serotype 1. The specificity of the antibodies used in the kit was the same as that of those prepared in the laboratory and marketed by Diagnostics Pasteur for the diagnosis of rabies by direct immunofluorescence. Use of these antibodies in the FAT test permitted detection also ofantigens ofother lyssaviruses (Lagos Bat, Mokola, and Duvenhage) but the sensitivity was lower than for the serotype 1. Nevertheless, with the RREID experimental kit several of the laboratories observed a weak reaction with Mokola, Lagos Bat, and Duvenhage viruses. In view of this, we have prepared a new batch of horseradish peroxidase conjugate using the same antinucleocapsid immuno- globulins as before, which appears to be more sensitive. An improved RREID experimental kit for the detection of the various serotypes of lyssaviruses will soon be available. In conclusion, if a laboratory has a good UV microscope and the number of specimens to be analysed for rabies antigens is not too large, use ofthe RREID test instead of the FAT is not recommended. However, for laboratories that have no fluorescence microscope or for epidemiological studies that involve a large number of specimens, the RREID technique is a useful tool for the rapid and specific diagnosis of rabies. ACKNOWLEDGEMENTS We wish to thank Dr J. F. Delagneau (Diagnostics Pasteur) for skilful preparation and presentation of the experimental RREID kit. We also thank P. Versmisse, S. Favre, and E. Lecocq for their excellent technical assistance. RESUME ETUDE CONCERTEE D'UNE TROUSSE DE REACTIFS DESTINEE A L'IMMUNODIAGNOSTIC ENZYMATIQUE RAPIDE DE LA RAGE (RREID) A la demande de laboratoires de diagnostic de la rage nous avons recemment mis au point une methode permettant de detecter les antigenes rabiques dans des broyats de sub- stances cerebrales. Cette m6thode est fond6e sur le principe d'une immuno-capture des antigenes rabiques (nucl6o- capsides) dans des microplaques sensibilis&es (anticorps antinucleocapsides), suivie d'une revelation de l'antigene fixe specifiquement par le meme anticorps (antinucl6o- capside) conjugue a la peroxydase et d'une r6action enzymatique en presence d'un chromogene. Dans ces conditions, les prelevements rabiques generent une coloration jaune-brun apres mise en ceuvre du RREID. Cette nouvelle technique s'etant r&elee sensible et specifique, nous en avons poursuivi le developpement et propose une trousse de reactifs prets a l'emploi pour le depistage des antigenes rabiques. Plusieurs trousses ont 6te fournies a six laboratoires differents ayant accepte de participer a une etude concertee sur la faisabilit6 et l'interet de l'utilisation du RREID comme technique de diagnostic de la rage. Nous rapportons ici les resultats de cette etude. Six laboratoires ont teste 1253 prelevements provenant de 27 especes animales et de 10 origines g6ographiques differentes. Les r6sultats obtenus avec le RREID ont ete compares a ceux obtenus par immunofluorescence (FAT). Les deux techniques ont donne le meme resultat pour 1220 prelevements (651 positifs et 569 negatifs), ce qui correspond a 97,4% du total des prelevements test6s. Le RREID semble cependant legerement moins sensible que le FAT puisque, sur l'ensemble des prelevements positifs avec le FAT, 22 soit 3% se sont reveles negatifs avec le RREID. Toutefois, I'etude concertee confirme que le RREID est une technique facile a mettre en euvre, rapide et specifique, qui peut etre consideree comme tres utile pour des etudes epidemiologiques et pour des laboratoires non equipes en immunofluorescence. 492 RAPID RABIES ENZYME IMMUNODIAGNOSIS 493 REFERENCES 1. GOLDWASSER, R. & KISSLING, R. Fluorescent antibody staining of street and fixed rabies virus antigens. Proceedings of the Society for Experimental Biology, 98: 219-229 (1958). 2. DEAN, D. J. & ABELSETH, M. K. The fluorescent anti- body test. In: Kaplan, M. M. & Koprowski, H., ed. Laboratory techniques in rabies, 3rd edition. Geneva, World Health Organization, 1973, pp. 73-83 (WHO Monograph Series No. 23). 3. KURSTAK, E. ET AL. Enzyme immunoassays and related procedures in diagnostic medical virology. Bulletin of the World Health Organization, 64: 465-479 (1986). 4. ATANASIU, P. ET AL. Identification des glycoprot6ine et nucidocapside rabiques des cellules infectees par des immunoglobulines specifiques marquees. Comptes rendus des seances de l'Academie des Sciences. Serie D: Sciences naturelles, 281: 1913-1915 (1975). 5. ATANASIU, P., ET AL. Immunofluorescence and immunoperoxydase in the diagnosis of rabies. In: Kurstak, E. & Morisset, R., ed. Viral immunodiag- nosis. New York, Academic Press, 1974, pp. 141-155. 6. PERRIN, P. ET AL. A rapid rabies enzyme immuno- diagnostic (RREID): a useful and simple technique for the routine diagnosis of rabies. Journal of biological standardization, 14: 2 17-222 (1986). Annex List of participating laboratories and investigators Dr J. Blancou & Dr J. Barrat Centre National d'Etudes sur la Rage et la Pathologie des Animaux Sauvages, B.P. No. 9, 54220 Malzeville, France. Dr R. Camero & Dr M. Picard Laboratoire Central de Recherches Vdterinaires, 22 rue P. Curie, B.P. No. 67, 94703 Maison-Alfort Cedex, France. Dr G. A. Casey & Dr W. A. Webster Rabies Unit, Animal Diseases Research Institute, P.O. Box 11300, Station "H", Nepean, Ontario K2H 8P9, Canada. Dr P. E. Rollin & Dr H. Bourhy Centre National de Reference pour la Rage, 28 rue du Docteur Roux, 75724 Paris Cedex 15, France. Dr L. Schneider & Dr J. H. Cox WHO Collaborating Centre for Rabies Surveillance and Research, Federal Institute for Animal Virus Diseases, D 7400 Tubingen, Postfach 1149, Federal Republic of Germany. Dr M. Tollis WHO Collaborating Centre for Reference and Research on Rabies, Rabies Unit, The Wistar Institute, 36th Street at Spruce, Philadelphia, PA 191044268, USA.