- Email: [email protected]
Evaluation of an indirect ELISA for the diagnosis of bovine brucellosis in milk and serum samples in dairy cattle in Argentina
Preventive Veterinary Medicine 36 (1998) 211±217
Evaluation of an indirect ELISA for the diagnosis of bovine brucellosis in milk and serum samples in dairy cattle in Argentina V.R. Vanzinia,*, N. Aguirreb, C.I. Lugaresic, S.T. de Echaidea, V.G. de Canavesioa, A.A. Guglielmonea, M.D. Marchesinod, K. Nielsene a
Instituto Nacional de TecnologõÂa Agropecuaria, EstacioÂn Experimental Agropecuaria Rafaela, CC 22, CP 2300 Rafaela (Santa Fe), Argentina b Universidad Nacional del Litoral, Facultad de AgronomõÂa y Veterinaria de Esperanza, Rdo. Padre L. Kreder 2805, CP 3080 Esperanza (Santa Fe), Argentina c Alberdi 406, CP 2300 Rafaela (Santa Fe), Argentina d Bvd. J.A. Roca 883, CP 2300 Rafaela (Santa Fe), Argentina e Canadian Food Inspection Agency, Animal Diseases Research Institute, 3851 Fallowfield Road, Nepean, Ontario, Canada K2H 8P9 Accepted 17 April 1998
Abstract An indirect enzyme-linked immunosorbent assay (ELISA) for Brucella abortus antibodies detection in bovine milk and serum samples was validated. The assays use B. abortus smooth lipopolysaccharide as antigen, immobilized on a polystyrene matrix; milk diluted 1:2 or serum diluted 1:50, in a buffer containing divalent cation chelating agents EDTA and EGTA (ethyleneglycol-bis-aminoether-N,N,N0 ,N0 -tetraacetic acid) to reduce non-specific reactions; and a mouse monoclonal antibody specific for an epitope of bovine IgG1, conjugated with horseradish peroxidase. A total of 2646 sera and 2119 milk samples from cows older than 24 months were obtained from 12 brucellosis-free herds for at least the previous 5 years. Milk samples were obtained in parallel with serum samples. The remaining 527 serum samples were from dry cows. All cattle were vaccinated with B. abortus strain 19 between 3±10 months of age. Five hundred and fifty-two milk samples and 562 serum samples were obtained from 6 infected herds with abortions where B. abortus was isolated at least once no more than 6 months before sampling. The complement-fixation test (CFT) on serum samples was considered the gold standard. Serum samples were also tested with the official screening test: the buffered plate antigen (BPA) test. The cut-off point was determined using receiver-operating characteristic (ROC) analysis. For milk * Corresponding author. Fax: +54 49 240114; e-mail: [email protected] 0167-5877/98/$ ± see front matter # 1998 Elsevier Science B.V. All rights reserved PII: S 0 1 6 7 - 5 8 7 7 ( 9 8 ) 0 0 0 8 0 - 4
212
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
samples, it was fixed at 36 percent positivity (PP) giving a sensitivity of 99.6% with a 95% confidence interval (CI) of 98.6±99.9%. The specificity was 99.1% (CI 98.9±99.4%). For serum samples, the cut-off was fixed at 53 PP giving a sensitivity of 99.6% (CI 98.6±99.9%) and a specificity 98.6% (CI 98±99%). The BPA test showed a relative sensitivity of 99.6% (CI 98.6± 99.9%) and a relative specificity of 98.6% (CI 98.1±99%). Our results indicate that the indirect ELISA is a highly sensitive and specific test and can be adapted to process a large number of samples. # 1998 Elsevier Science B.V. All rights reserved. Keywords: Cattle-microbiological diseases; Brucella abortus diagnosis-bacteria; ELISA; Argentina
1. Introduction Several tests have been developed to detect antibodies against B. abortus in cattle. Positive results from confirmatory serological tests (which vary amongst countries) are generally considered as unequivocal evidence of brucellosis. The Argentinean Servicio Nacional de Sanidad Animal (SENASA) recognizes the buffered plate antigen (BPA) test as a screening test, while the tube agglutination test (SAT) and the 2-mercaptoetanol agglutination test (2-Me) are used as confirmatory tests and the complement-fixation test (CFT) as the reference test (SENASA, 1995). The current Argentinean brucellosis control programme also involves the compulsory vaccination of 3±10-month-old female calves with B. abortus, strain 19 (SENASA, Res. 1269/93). Additionally, an area of Santa Fe province (where dairy cattle are predominant) is under an eradication scheme based on the detection of antibodies in blood serum in female cattle older than 18 months by any of the above tests and sending reactor cattle to the slaughterhouse (SENASA, Res. 043/94). A test for milk samples with similar sensitivity and specificity as those applied to serum samples is economically advantageous for cattle milk producing areas, since the sampling cost is very low. Moreover, it will be useful in situations such as the eradication programme carried out in the milk-producing area of Argentina as a tool for epidemiological surveillance. There is a considerable literature related to levels of immunoglobulins in bovine blood and mammary secretions (Penhale and Christie, 1969; Butler et al., 1972). Concentration of IgG in milk correlates well with serum levels (Caffin et al., 1983; Caffin and Poutrel, 1988); therefore, it is reasonable to assume that milk reflects the immunological status of the animal. In addition to several enzyme-linked immunosorbent assays (ELISA) developed for serum antibodies to B. abortus, a number of ELISAs to detect Brucella antibodies in milk has been reported (Thoen et al., 1979, 1983; Heck et al., 1980). Recently, a highly sensitive and specific indirect ELISA for milk antibody to B. abortus has been reported (Nielsen et al., 1996a). This test uses a B. abortus smooth lipopolysaccharide as antigen; a monoclonal antibody for a specific epitope of bovine IgG1 (conjugated with HRPO) for detection; and divalent cation chelating agents EDTA and EGTA to reduce non-specific reactions. This ELISA was evaluated using known negative milk samples from Canada (where cattle are not vaccinated against brucellosis)
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
213
and positive milk samples from Argentina and Chile giving a sensitivity of 95.2% and 98.7%, respectively. We hypothesized that the detection of antibodies by an indirect ELISA in milk samples is suitable to identify dairy cows infected with B. abortus. To address this hypothesis, milk and serum samples were obtained in parallel from dairy cows to determine the cutoff point of this ELISA for cattle vaccinated with B. abortus strain 19. The results were compared with those from CF and BPA tests. 2. Materials and methods 2.1. Negative samples Serum and milk samples were obtained from cows vaccinated with B. abortus strain 19 when 3±10 months-old and belonging to 12 Holstein Friesian herds serologically free of B. abortus for at least the previous 5 years. Additionally, all serum samples were tested using the BPA and CFT to confirm negative status. In this trial, 2646 sera and 2119 milk samples from all cows older than 24 months were used. All milk samples were obtained in parallel with serum samples from the same cows; the remaining 527 serum samples were taken from dry cows in the same herds. 2.2. Positive samples Serum and milk samples were obtained from 6 Holstein Friesian herds where B. abortus had been isolated from aborted fetuses from at least one cow in each herd no more than 6 months before sampling and within-herd sero-prevalences were 11±26% according to BPA and CFT. All sera which reacted in the CFT were considered as positive. Sera from 562 cows and milk from 552 cows greater than 24 months old positive to CFT were used. The samples were, obtained in parallel except for serum from 10 dry cows. The CFT was performed following Hill's methods (Alton et al., 1975) using the second International Standard Anti-Brucella abortus Serum (ISAbS) as reference and considering as positive all sera which gave 50% haemolysis at the dilution equal or higher than 1:8. 2.3. ELISA procedure The ELISA for serum and milk B. abortus antibodies was performed using reagents and test protocols provided by the Animal Disease Research Institute (ADRI), Agriculture Canada (Nielsen et al., 1996a, b). Briefly, 100 ml of a hot water/hot phenol-extracted B. abortus smooth lipopolysaccharide (SLPS) at a dilution of 1.0 mg/ml in a 0.06 M carbonate/bicarbonate buffer pH 9.6 (coating buffer) was adsorbed passively to 96-well flat-bottom polystyrene microplates (NUNC, 269620), overnight at 4.08C. The following day the plates were frozen at ÿ208C. For use, the plates were thawed at 378C for 45 min, then washed four times with 0.01 M phosphate buffer containing 0.015 M sodium chloride and 0.05% Tween 20, pH 7.2 (PBS-T). To reduce the background, 50 ml/
214
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
well of EDTA and EGTA (each 0.015 M) in PBS-T, pH 6.3 was added. Serum samples were added diluted 1:25 in PBS-T to achieve a final dilution of 1:50 and milk samples were added undiluted for a final dilution 1:2. The plates were incubated for 30 min at room temperature (20±308C) with continuous shaking (Micro Shaker II, Dynatech). All sera and milk samples were tested in duplicate. A strong positive serum (C), a weak positive serum (C), a negative serum (Cÿ) and PBS-T as a conjugate control (Cc) were included in each plate in quadruplicate. After four washings, 100 ml of an appropriate dilution of monoclonal mouse anti-bovine IgG1 conjugated to horseradish peroxidase was added to all the wells followed by 1 h of incubation at room temperature with continuous shaking. Finally, after another four washings, 100 ml/well of a substrate/chromogen solution containing 4 mM 2,20 -azino-bis (3-ethylbenzthiazoline-6-sulfonic acid) and 1 mM hydrogen peroxide in 0.05 M citrate buffer (pH 4.5). Reactions were not stopped before reading. When serum samples were tested, plates were read after 10 min incubation at room temperature with continuous shaking. For milk samples, the C control was used to time color development. An initial reading was made at 4 min of incubation and the target time for the final optical density (OD) reading of 1.0 for the C control was calculated by the computer (Wright et al., 1985). Color development was assessed using an ELISA reader Labsystems Multiskan1 Plus (type 314) linked to a personal computer. For sera, the BREIA 1.02 version software program was used (#W. Kelly, Agriculture Canada, 1992); for milk samples, the ELISA 2.20, Pre-release version (#W. Kelly, Agriculture Canada, 1993) was used. Absorbance values were determined at 405 nm. The mean OD values were expressed as percentage positivity (PP) related to the C control serum, with lower and upper limits of OD acceptance for the C control of 0.777±1.220 and 0.570±1.260 for milk and serum, respectively (Nielsen et al., 1996b). The absorbance values obtained for the C control within the range of OD, for each plate run, were set as 100% positivity. 2.4. Determination of seronegative/seropositive threshold PP value (cut-off point) for the ELISA The PP threshold was determined by plotting the sensitivity and specificity for various cut-off values by means of Receiver Operating Characteristic (ROC) analysis to achieve the best relationship between sensitivity and specificity (Metz, 1978; Nielsen et al., 1996b). Sensitivity was defined as the proportion of infected cattle correctly identified as positive by the test and specificity as the proportion of uninfected cattle correctly identified as negative by the test. All serum samples negative to CFT from brucellosisfree herds were considered uninfected and all samples positive to CFT from the six infected herds were considered infected. 2.5. Statistical analysis The ROC analysis, confidence interval (CI) at 95% confidence level, sensitivity and specificity were calculated using MedCalc software (Schoonjans et al., 1995).
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
215
Fig. 1. Percent positivity distribution of 552 positive and 2119 negative dairy cow milk samples from Brucella abortus-infected and uninfected herds, respectively; using an indirect ELISA at 36% cut-off point.
3. Results Analysis of the data for milk samples gave a PP threshold of 36% (Fig. 1). The area under the ROC curve was 0.999 with a Cl of 0.997±1.0 (data not shown). The mean for negative samples was 8.39%6.86% (SD). There were only 19 false positive and 2 false negative. The relative sensitivity compared with CFT was 99.6% (Cl 98.6±99.9%) and the relative specificity 99.1% (Cl 98.6±99.4%). The PP cut-off point for serum samples was 53% (Fig. 2). The area under the ROC curve was 0.999 with a Cl of 0.997±1.0 (data not shown). The mean for negative samples was 17.85%11.6% (SD). Thirty-eight samples were false-positive and 2 were falsenegative. The relative sensitivity to the CFT was 99.6% (Cl 98.6±99.9%) and the relative specificity was 98.6% (Cl 98±99%). The BPA test showed a relative sensitivity of 99.6% to the CFT (Cl 98.6±99.9%) and a relative specificity of 98.6% (Cl 98.1±99%). 4. Discussion The indirect ELISA for antibodies to B. abortus in milk samples evaluated in this study gave high sensitivity and specificity. The sensitivity was higher than the one observed in milk samples from Argentina by Nielsen et al. (1996a) but the specificity was slightly lower. The sensitivity of the indirect ELISA in milk samples is similar to the one observed for BPA in serum samples but the specificity was slightly higher. The relative
216
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
Fig. 2. Percent positivity distribution of 562 positive and 2646 negative dairy cow serum samples from Brucella abortus-infected and uninfected herds, respectively; using an indirect ELISA at 53% cut-off point.
sensitivity and specificity for BPA were comparable to those reported by Stemshorn et al. (1985); Dohoo et al. (1986). We acknowledge that CFT-negative samples from the positive herds were discarded; had they been included in the analysis, the apparent accuracy might have been lower. Testing milk samples has some advantages over testing serum samples. Sampling is not invasive; therefore, the accidental transmission of diseases by needle and reduction of production due to stress can be reduced. Since the monitoring of the monthly individual productivity of dairy cows is an increasing practice in this area, sampling to test brucellosis in milk could be done without extra cost. Sample condition is not critical and it can be stored frozen or refrigerated with a preservative (Nielsen et al., 1996a). The ELISA can be applied to a large number of samples and the assessment of positivity is performed electronically (i.e. not subjective). Although the sensitivity and specificity of the indirect ELISA applied to serum samples were similar to the BPA, its capacity to be adapted to test a large number of samples, the objectivity of the assessment of positivity and the possibility to adjust the cut-off point according to different epidemiological situations (improving sensitivity or specificity) are advantages not provided by the BPA. Acknowledgements Supported by Suc. de Alfredo Williner S.A., Bvd. J.A. Roca 883, 2300-Rafaela (Santa Fe), Argentina.
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
217
References Alton, G.G., Jones, L.M., Pietz, D.E., 1975. Serological methods in: Laboratory techniques in brucellosis, Chapter 2. FAO and WHO, Geneva, 64±124. Butler, J.E., Maxwell, C.F., Pierce, C.S., Hilton, M.B., Asofky, R., Kiddy, C.A., 1972. Studies on the relative synthesis and distribution of IgA and IgG1 in various tissues and body fluids of the cow. J. Immunol. 109, 38±46. Caffin, J.P., Poutrel, B., Rainard, P., 1983. Physiological and pathological factors influencing bovine immunoglobulin G1 concentration in milk. J. Dairy Sci. 66, 2161±2166. Caffin, J.P., Poutrel, B., 1988. Physiological and pathological factors influencing bovine immunoglobulin G2 concentration in milk. J. Dairy Sci. 71, 2035±2043. Dohoo, I.R., Wright, P.F., Ruckerhauer, G.M., Samagh, B.S., Robertson, F.J., Forbes, L.B., 1986. A comparison of five serological tests for bovine brucellosis. Can. Vet. Res. 50, 485±493. Heck, F.C., Williams, J.D., Pruett, J., Sanders, R., Zink, D.L., 1980. Enzyme-linked immunosorbent assay for detecting antibodies to Brucella abortus in bovine milk and serum. Am. J. Vet. Res. 41, 2082±2084. Metz, C.E., 1978. Basic principles of ROC analysis. Seminar Nucl. Med. 8, 283±298. Nielsen, K., Smith, P., Gall, D., Perez, B., Cosma, C., Mueller, P., Trottier, J., Cote, G., Boag, L., Bosse, J., 1996a. Development and validation of an indirect enzyme immunoassay for detection of antibody to Brucella abortus in milk. Vet. Microbiol. 52, 165±173. Nielsen, K., Gall, D., Kelly, W., Vigliocco, A., Henning, D., Garcia, M., 1996b. Immunoassay development. Application to enzyme immunoassay for the diagnosis of brucellosis. Agriculture and Agri-Food Canada, ISBN 0-662-24163-0, mimeo, 246 pp. Penhale, W.J., Christie, G., 1969. Quantitative studies on bovine immunoglobulins. I. Adult plasma and colostrum levels. Res. Vet. Sci. 10, 493±501. Schoonjans, F., Zalata, A., Depuydt, C., Comhaire, F., 1995. MedCalc: A new computer program for medical statistics. Comput. Methods Programs Biomed. 48, 257±262. SENASA, 1995. Manual de procedimientos. DiagnoÂstico seroloÂgico de la brucelosis bovina. (Res. 1269/93). mimeo, 66 pp. Stemshorn, B.W., Forbes, L.B., Eaglesome, M.D., Nielsen, K.H., Robertson, F.J., Samagh, B.S., 1985. A comparison of standard serological tests for the diagnosis of bovine brucellosis in Canada. Can. J. Comp. Med. 49, 391±394. Thoen, C.O., Pietz, D.E., Amrbrust, A.L., Harrington Jr., R., 1979. Enzyme immunoassay for detecting brucella antibodies in cow's milk. J. Clin. Microbiol. 10, 222±225. Thoen, C.O., Bruner, J.A., Luchsinger, D.W., Pietz, D.E., 1983. Detection of Brucella antibodies of different immunoglobulin classes in cow's milk by enzyme linked immunoassay. Am. J. Vet. Res. 14, 306±308. Wright, P.F., Kelly, W.A., Gall, D.G., 1985. Application of a timing protocol to the reduction of inter-plate variability in the indirect enzyme immunoassay for detection of anti-Brucella antibody. J. Immunoassay 6, 189±205.
Evaluation of an indirect ELISA for the diagnosis of bovine brucellosis in milk and serum samples in dairy cattle in Argentina V.R. Vanzinia,*, N. Aguirreb, C.I. Lugaresic, S.T. de Echaidea, V.G. de Canavesioa, A.A. Guglielmonea, M.D. Marchesinod, K. Nielsene a
Instituto Nacional de TecnologõÂa Agropecuaria, EstacioÂn Experimental Agropecuaria Rafaela, CC 22, CP 2300 Rafaela (Santa Fe), Argentina b Universidad Nacional del Litoral, Facultad de AgronomõÂa y Veterinaria de Esperanza, Rdo. Padre L. Kreder 2805, CP 3080 Esperanza (Santa Fe), Argentina c Alberdi 406, CP 2300 Rafaela (Santa Fe), Argentina d Bvd. J.A. Roca 883, CP 2300 Rafaela (Santa Fe), Argentina e Canadian Food Inspection Agency, Animal Diseases Research Institute, 3851 Fallowfield Road, Nepean, Ontario, Canada K2H 8P9 Accepted 17 April 1998
Abstract An indirect enzyme-linked immunosorbent assay (ELISA) for Brucella abortus antibodies detection in bovine milk and serum samples was validated. The assays use B. abortus smooth lipopolysaccharide as antigen, immobilized on a polystyrene matrix; milk diluted 1:2 or serum diluted 1:50, in a buffer containing divalent cation chelating agents EDTA and EGTA (ethyleneglycol-bis-aminoether-N,N,N0 ,N0 -tetraacetic acid) to reduce non-specific reactions; and a mouse monoclonal antibody specific for an epitope of bovine IgG1, conjugated with horseradish peroxidase. A total of 2646 sera and 2119 milk samples from cows older than 24 months were obtained from 12 brucellosis-free herds for at least the previous 5 years. Milk samples were obtained in parallel with serum samples. The remaining 527 serum samples were from dry cows. All cattle were vaccinated with B. abortus strain 19 between 3±10 months of age. Five hundred and fifty-two milk samples and 562 serum samples were obtained from 6 infected herds with abortions where B. abortus was isolated at least once no more than 6 months before sampling. The complement-fixation test (CFT) on serum samples was considered the gold standard. Serum samples were also tested with the official screening test: the buffered plate antigen (BPA) test. The cut-off point was determined using receiver-operating characteristic (ROC) analysis. For milk * Corresponding author. Fax: +54 49 240114; e-mail: [email protected] 0167-5877/98/$ ± see front matter # 1998 Elsevier Science B.V. All rights reserved PII: S 0 1 6 7 - 5 8 7 7 ( 9 8 ) 0 0 0 8 0 - 4
212
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
samples, it was fixed at 36 percent positivity (PP) giving a sensitivity of 99.6% with a 95% confidence interval (CI) of 98.6±99.9%. The specificity was 99.1% (CI 98.9±99.4%). For serum samples, the cut-off was fixed at 53 PP giving a sensitivity of 99.6% (CI 98.6±99.9%) and a specificity 98.6% (CI 98±99%). The BPA test showed a relative sensitivity of 99.6% (CI 98.6± 99.9%) and a relative specificity of 98.6% (CI 98.1±99%). Our results indicate that the indirect ELISA is a highly sensitive and specific test and can be adapted to process a large number of samples. # 1998 Elsevier Science B.V. All rights reserved. Keywords: Cattle-microbiological diseases; Brucella abortus diagnosis-bacteria; ELISA; Argentina
1. Introduction Several tests have been developed to detect antibodies against B. abortus in cattle. Positive results from confirmatory serological tests (which vary amongst countries) are generally considered as unequivocal evidence of brucellosis. The Argentinean Servicio Nacional de Sanidad Animal (SENASA) recognizes the buffered plate antigen (BPA) test as a screening test, while the tube agglutination test (SAT) and the 2-mercaptoetanol agglutination test (2-Me) are used as confirmatory tests and the complement-fixation test (CFT) as the reference test (SENASA, 1995). The current Argentinean brucellosis control programme also involves the compulsory vaccination of 3±10-month-old female calves with B. abortus, strain 19 (SENASA, Res. 1269/93). Additionally, an area of Santa Fe province (where dairy cattle are predominant) is under an eradication scheme based on the detection of antibodies in blood serum in female cattle older than 18 months by any of the above tests and sending reactor cattle to the slaughterhouse (SENASA, Res. 043/94). A test for milk samples with similar sensitivity and specificity as those applied to serum samples is economically advantageous for cattle milk producing areas, since the sampling cost is very low. Moreover, it will be useful in situations such as the eradication programme carried out in the milk-producing area of Argentina as a tool for epidemiological surveillance. There is a considerable literature related to levels of immunoglobulins in bovine blood and mammary secretions (Penhale and Christie, 1969; Butler et al., 1972). Concentration of IgG in milk correlates well with serum levels (Caffin et al., 1983; Caffin and Poutrel, 1988); therefore, it is reasonable to assume that milk reflects the immunological status of the animal. In addition to several enzyme-linked immunosorbent assays (ELISA) developed for serum antibodies to B. abortus, a number of ELISAs to detect Brucella antibodies in milk has been reported (Thoen et al., 1979, 1983; Heck et al., 1980). Recently, a highly sensitive and specific indirect ELISA for milk antibody to B. abortus has been reported (Nielsen et al., 1996a). This test uses a B. abortus smooth lipopolysaccharide as antigen; a monoclonal antibody for a specific epitope of bovine IgG1 (conjugated with HRPO) for detection; and divalent cation chelating agents EDTA and EGTA to reduce non-specific reactions. This ELISA was evaluated using known negative milk samples from Canada (where cattle are not vaccinated against brucellosis)
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
213
and positive milk samples from Argentina and Chile giving a sensitivity of 95.2% and 98.7%, respectively. We hypothesized that the detection of antibodies by an indirect ELISA in milk samples is suitable to identify dairy cows infected with B. abortus. To address this hypothesis, milk and serum samples were obtained in parallel from dairy cows to determine the cutoff point of this ELISA for cattle vaccinated with B. abortus strain 19. The results were compared with those from CF and BPA tests. 2. Materials and methods 2.1. Negative samples Serum and milk samples were obtained from cows vaccinated with B. abortus strain 19 when 3±10 months-old and belonging to 12 Holstein Friesian herds serologically free of B. abortus for at least the previous 5 years. Additionally, all serum samples were tested using the BPA and CFT to confirm negative status. In this trial, 2646 sera and 2119 milk samples from all cows older than 24 months were used. All milk samples were obtained in parallel with serum samples from the same cows; the remaining 527 serum samples were taken from dry cows in the same herds. 2.2. Positive samples Serum and milk samples were obtained from 6 Holstein Friesian herds where B. abortus had been isolated from aborted fetuses from at least one cow in each herd no more than 6 months before sampling and within-herd sero-prevalences were 11±26% according to BPA and CFT. All sera which reacted in the CFT were considered as positive. Sera from 562 cows and milk from 552 cows greater than 24 months old positive to CFT were used. The samples were, obtained in parallel except for serum from 10 dry cows. The CFT was performed following Hill's methods (Alton et al., 1975) using the second International Standard Anti-Brucella abortus Serum (ISAbS) as reference and considering as positive all sera which gave 50% haemolysis at the dilution equal or higher than 1:8. 2.3. ELISA procedure The ELISA for serum and milk B. abortus antibodies was performed using reagents and test protocols provided by the Animal Disease Research Institute (ADRI), Agriculture Canada (Nielsen et al., 1996a, b). Briefly, 100 ml of a hot water/hot phenol-extracted B. abortus smooth lipopolysaccharide (SLPS) at a dilution of 1.0 mg/ml in a 0.06 M carbonate/bicarbonate buffer pH 9.6 (coating buffer) was adsorbed passively to 96-well flat-bottom polystyrene microplates (NUNC, 269620), overnight at 4.08C. The following day the plates were frozen at ÿ208C. For use, the plates were thawed at 378C for 45 min, then washed four times with 0.01 M phosphate buffer containing 0.015 M sodium chloride and 0.05% Tween 20, pH 7.2 (PBS-T). To reduce the background, 50 ml/
214
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
well of EDTA and EGTA (each 0.015 M) in PBS-T, pH 6.3 was added. Serum samples were added diluted 1:25 in PBS-T to achieve a final dilution of 1:50 and milk samples were added undiluted for a final dilution 1:2. The plates were incubated for 30 min at room temperature (20±308C) with continuous shaking (Micro Shaker II, Dynatech). All sera and milk samples were tested in duplicate. A strong positive serum (C), a weak positive serum (C), a negative serum (Cÿ) and PBS-T as a conjugate control (Cc) were included in each plate in quadruplicate. After four washings, 100 ml of an appropriate dilution of monoclonal mouse anti-bovine IgG1 conjugated to horseradish peroxidase was added to all the wells followed by 1 h of incubation at room temperature with continuous shaking. Finally, after another four washings, 100 ml/well of a substrate/chromogen solution containing 4 mM 2,20 -azino-bis (3-ethylbenzthiazoline-6-sulfonic acid) and 1 mM hydrogen peroxide in 0.05 M citrate buffer (pH 4.5). Reactions were not stopped before reading. When serum samples were tested, plates were read after 10 min incubation at room temperature with continuous shaking. For milk samples, the C control was used to time color development. An initial reading was made at 4 min of incubation and the target time for the final optical density (OD) reading of 1.0 for the C control was calculated by the computer (Wright et al., 1985). Color development was assessed using an ELISA reader Labsystems Multiskan1 Plus (type 314) linked to a personal computer. For sera, the BREIA 1.02 version software program was used (#W. Kelly, Agriculture Canada, 1992); for milk samples, the ELISA 2.20, Pre-release version (#W. Kelly, Agriculture Canada, 1993) was used. Absorbance values were determined at 405 nm. The mean OD values were expressed as percentage positivity (PP) related to the C control serum, with lower and upper limits of OD acceptance for the C control of 0.777±1.220 and 0.570±1.260 for milk and serum, respectively (Nielsen et al., 1996b). The absorbance values obtained for the C control within the range of OD, for each plate run, were set as 100% positivity. 2.4. Determination of seronegative/seropositive threshold PP value (cut-off point) for the ELISA The PP threshold was determined by plotting the sensitivity and specificity for various cut-off values by means of Receiver Operating Characteristic (ROC) analysis to achieve the best relationship between sensitivity and specificity (Metz, 1978; Nielsen et al., 1996b). Sensitivity was defined as the proportion of infected cattle correctly identified as positive by the test and specificity as the proportion of uninfected cattle correctly identified as negative by the test. All serum samples negative to CFT from brucellosisfree herds were considered uninfected and all samples positive to CFT from the six infected herds were considered infected. 2.5. Statistical analysis The ROC analysis, confidence interval (CI) at 95% confidence level, sensitivity and specificity were calculated using MedCalc software (Schoonjans et al., 1995).
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
215
Fig. 1. Percent positivity distribution of 552 positive and 2119 negative dairy cow milk samples from Brucella abortus-infected and uninfected herds, respectively; using an indirect ELISA at 36% cut-off point.
3. Results Analysis of the data for milk samples gave a PP threshold of 36% (Fig. 1). The area under the ROC curve was 0.999 with a Cl of 0.997±1.0 (data not shown). The mean for negative samples was 8.39%6.86% (SD). There were only 19 false positive and 2 false negative. The relative sensitivity compared with CFT was 99.6% (Cl 98.6±99.9%) and the relative specificity 99.1% (Cl 98.6±99.4%). The PP cut-off point for serum samples was 53% (Fig. 2). The area under the ROC curve was 0.999 with a Cl of 0.997±1.0 (data not shown). The mean for negative samples was 17.85%11.6% (SD). Thirty-eight samples were false-positive and 2 were falsenegative. The relative sensitivity to the CFT was 99.6% (Cl 98.6±99.9%) and the relative specificity was 98.6% (Cl 98±99%). The BPA test showed a relative sensitivity of 99.6% to the CFT (Cl 98.6±99.9%) and a relative specificity of 98.6% (Cl 98.1±99%). 4. Discussion The indirect ELISA for antibodies to B. abortus in milk samples evaluated in this study gave high sensitivity and specificity. The sensitivity was higher than the one observed in milk samples from Argentina by Nielsen et al. (1996a) but the specificity was slightly lower. The sensitivity of the indirect ELISA in milk samples is similar to the one observed for BPA in serum samples but the specificity was slightly higher. The relative
216
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
Fig. 2. Percent positivity distribution of 562 positive and 2646 negative dairy cow serum samples from Brucella abortus-infected and uninfected herds, respectively; using an indirect ELISA at 53% cut-off point.
sensitivity and specificity for BPA were comparable to those reported by Stemshorn et al. (1985); Dohoo et al. (1986). We acknowledge that CFT-negative samples from the positive herds were discarded; had they been included in the analysis, the apparent accuracy might have been lower. Testing milk samples has some advantages over testing serum samples. Sampling is not invasive; therefore, the accidental transmission of diseases by needle and reduction of production due to stress can be reduced. Since the monitoring of the monthly individual productivity of dairy cows is an increasing practice in this area, sampling to test brucellosis in milk could be done without extra cost. Sample condition is not critical and it can be stored frozen or refrigerated with a preservative (Nielsen et al., 1996a). The ELISA can be applied to a large number of samples and the assessment of positivity is performed electronically (i.e. not subjective). Although the sensitivity and specificity of the indirect ELISA applied to serum samples were similar to the BPA, its capacity to be adapted to test a large number of samples, the objectivity of the assessment of positivity and the possibility to adjust the cut-off point according to different epidemiological situations (improving sensitivity or specificity) are advantages not provided by the BPA. Acknowledgements Supported by Suc. de Alfredo Williner S.A., Bvd. J.A. Roca 883, 2300-Rafaela (Santa Fe), Argentina.
V.R. Vanzini et al. / Preventive Veterinary Medicine 36 (1998) 211±217
217
References Alton, G.G., Jones, L.M., Pietz, D.E., 1975. Serological methods in: Laboratory techniques in brucellosis, Chapter 2. FAO and WHO, Geneva, 64±124. Butler, J.E., Maxwell, C.F., Pierce, C.S., Hilton, M.B., Asofky, R., Kiddy, C.A., 1972. Studies on the relative synthesis and distribution of IgA and IgG1 in various tissues and body fluids of the cow. J. Immunol. 109, 38±46. Caffin, J.P., Poutrel, B., Rainard, P., 1983. Physiological and pathological factors influencing bovine immunoglobulin G1 concentration in milk. J. Dairy Sci. 66, 2161±2166. Caffin, J.P., Poutrel, B., 1988. Physiological and pathological factors influencing bovine immunoglobulin G2 concentration in milk. J. Dairy Sci. 71, 2035±2043. Dohoo, I.R., Wright, P.F., Ruckerhauer, G.M., Samagh, B.S., Robertson, F.J., Forbes, L.B., 1986. A comparison of five serological tests for bovine brucellosis. Can. Vet. Res. 50, 485±493. Heck, F.C., Williams, J.D., Pruett, J., Sanders, R., Zink, D.L., 1980. Enzyme-linked immunosorbent assay for detecting antibodies to Brucella abortus in bovine milk and serum. Am. J. Vet. Res. 41, 2082±2084. Metz, C.E., 1978. Basic principles of ROC analysis. Seminar Nucl. Med. 8, 283±298. Nielsen, K., Smith, P., Gall, D., Perez, B., Cosma, C., Mueller, P., Trottier, J., Cote, G., Boag, L., Bosse, J., 1996a. Development and validation of an indirect enzyme immunoassay for detection of antibody to Brucella abortus in milk. Vet. Microbiol. 52, 165±173. Nielsen, K., Gall, D., Kelly, W., Vigliocco, A., Henning, D., Garcia, M., 1996b. Immunoassay development. Application to enzyme immunoassay for the diagnosis of brucellosis. Agriculture and Agri-Food Canada, ISBN 0-662-24163-0, mimeo, 246 pp. Penhale, W.J., Christie, G., 1969. Quantitative studies on bovine immunoglobulins. I. Adult plasma and colostrum levels. Res. Vet. Sci. 10, 493±501. Schoonjans, F., Zalata, A., Depuydt, C., Comhaire, F., 1995. MedCalc: A new computer program for medical statistics. Comput. Methods Programs Biomed. 48, 257±262. SENASA, 1995. Manual de procedimientos. DiagnoÂstico seroloÂgico de la brucelosis bovina. (Res. 1269/93). mimeo, 66 pp. Stemshorn, B.W., Forbes, L.B., Eaglesome, M.D., Nielsen, K.H., Robertson, F.J., Samagh, B.S., 1985. A comparison of standard serological tests for the diagnosis of bovine brucellosis in Canada. Can. J. Comp. Med. 49, 391±394. Thoen, C.O., Pietz, D.E., Amrbrust, A.L., Harrington Jr., R., 1979. Enzyme immunoassay for detecting brucella antibodies in cow's milk. J. Clin. Microbiol. 10, 222±225. Thoen, C.O., Bruner, J.A., Luchsinger, D.W., Pietz, D.E., 1983. Detection of Brucella antibodies of different immunoglobulin classes in cow's milk by enzyme linked immunoassay. Am. J. Vet. Res. 14, 306±308. Wright, P.F., Kelly, W.A., Gall, D.G., 1985. Application of a timing protocol to the reduction of inter-plate variability in the indirect enzyme immunoassay for detection of anti-Brucella antibody. J. Immunoassay 6, 189±205.