Estimation of receiver-operating characteristic curves to determine accuracy of a competitive enzyme-linked immunosorbent assay for the serodiagnosis of Brucella infection in domestic water buffalo (Bubalus bubalis) and cattle

OBJECTIVE: To estimate receiver-operating characteristic (ROC) curves for a competitive ELISA (c-ELISA) that is used in serodiagnosis of brucellosis in water buffalo and cattle, to determine the most appropriate positive cutoff value for the c-ELISA in confirmation of infection, and to evaluate species differences in c-ELISA function. SAMPLE POPULATION: Sera from 4 herds of cattle (n = 391) and 4 herds of water buffalo (381). PROCEDURE: Serum samples were evaluated for Brucella-specific antibodies by use of a c-ELISA. On the basis of previous serologic test results, iterative simulation modeling was used to classify animals as positive or negative for Brucella infection without the use of a gold standard. Accuracy of c-ELISA for diagnosis of infection was compared between cattle and water buffalo by comparison of areas under ROC curves. RESULTS: A positive cutoff value of 30% inhibition for c-ELISA yielded sensitivity and specificity estimates, respectively, of 83.9 and 92.6% for cattle and 91.4 and 95.4% for water buffalo. A positive cutoff value of 35% inhibition yielded sensitivity and specificity estimates, respectively, of 83.9 and 96.2% for cattle and 88.0 and 974% for water buffalo. Areas under ROC curves were 0.94 and 0.98 for cattle and water buffalo, respectively. CONCLUSIONS AND CLINICAL RELEVANCE: ROC curves can be estimated by use of iterative simulation methods to determine optimal cutoff values for diagnostic tests with quantitative outcomes. A cutoff value of 35% inhibition for the c-ELISA was found to be most appropriate for confirmation of Brucella infection in cattle and water buffalo.     

Development of a monoclonal antibody based competitive-ELISA for detection and titration of antibodies to peste des petits ruminants (PPR) virus

Peste des petits ruminants (PPR) is an acute febrile, viral, disease of small ruminants with great economic importance. A competitive-ELISA (c-ELISA) test was developed for detection of antibodies to PPR virus in the sera samples of goats and sheep. The test uses monoclonal antibody to a neutralizing epitope of haemagglutinin protein of the virus. Based on the distribution of known negative sera samples (n=933) in respect of PPR virus antibodies in the test, a cut-off value was set as 38%. This value was the result of mean of negative population added with two times the standard deviations. A total of 1668 sera samples from goat and sheep and 32 sera from cattle were screened by c-ELISA and virus neutralization test (VNT). Efficacy of c-ELISA compared very well with VNT having high relative specificity (98.4%) and sensitivity (92.4%). The sensitivity of c-ELISA for PPR sero-surveillance could further be increased (95.4%), if the target population is non-vaccinated. c-ELISA test correlated well with VNT (r=0.845) for end-point titration of PPR virus antibody in 64 goat sera samples. It could clearly separate infected population from uninfected in field sera. Using c-ELISA test paired sera samples from 13 goats provided a clear diagnosis of PPR virus infection. Furthermore, antibodies to PPR virus could be successfully detected during 1 year after vaccination in four goats inoculated with an experimental PPR vaccine. Findings suggest that the c-ELISA test developed can easily replace VNT for sero-surveillance, sero-monitoring, diagnosis from paired sera samples and end-point titration of PPR virus antibodies.     

Development and application of a competitive enzyme-linked immunosorbent assay for the detection of serum antibodies to porcine circovirus type 2.

We report the development of a competitive enzyme-linked immunosorbent assay (c-ELISA) for the detection of antibodies to porcine circovirus type 2 (PCV2), the agent associated with the recently described postweaning multisystemic wasting syndrome in pigs. At present, no method has been published describing a c-ELISA for the detection of antibodies to PCV2, and currently employed tests are impractical for use in some laboratories. The assay described here uses a cell culture isolate of porcine circovirus type 2 as antigen and a PCV2-specific monoclonal antibody as the competing reagent. Evaluation of the ELISA was performed by comparison with results obtained using an indirect immunofluorescent test on 484 sera from pig herds in the United Kingdom, Canada, France, and the USA and serial bleeds from pigs experimentally infected with porcine circoviruses. The sensitivity and specificity of the ELISA were determined as 99.58% and 97.14%, respectively, at 2 standard deviations (SD) from the mean or 95.81% and 100% at 3 SD from the mean. Using this ELISA, a serologic survey of 461 sera collected from commercial pig herds in Northern Ireland between 1973 and 1999 was undertaken. Analysis of the results of this survey demonstrated that the number of ELISA-positive sera detected in an individual year during this period ranged from 55% to 100%. This c-ELISA has applications for large-scale rapid diagnosis of PCV2 infection in pig populations worldwide and for immunoscreening of sera from other species for antibodies to PCV2.     

A comparison of serologic tests for the detection of serum antibodies to whole-cell and recombinant Borrelia burgdorferi antigens in cattle

Serum samples from healthy dairy and beef cattle, living in tick-infested areas of Connecticut, USA, were analyzed by polyvalent enzyme-linked immunosorbent assays (ELISA), indirect fluorescent antibody (IFA) staining methods, or Western blot procedures to detect antibodies to tick-borne agents. Of the 80 sera tested by ELISA with whole-cell or 10 separate recombinant antigens (fusion proteins) of Borrelia burgdorferi sensu stricto, 57 (71%) were positive to 1 or more antigens, while 36 (45%) reacted to whole-cell antigens by IFA staining methods. Three (4%) of 80 samples had antibodies to Anaplasma phagocytophilum. There were antibodies to outer surface protein (Osp) A, OspB, OspC, OspE, OspF, protein (p) 41-G, p35, p37, and VlsE antigens of B. burgdorferi, but there was no reactivity to the p39 antigen by ELISA. Western immunoblots of a subset of 9 sera verified antibody presence in all samples and showed distinct reactivities to multiple proteins having molecular masses of about 31 kilodaltons (kDa), 34 kDa, 35 kDa, 41 kDa, and 83/93 kDa. High specificity (97%) was noted when 16 cattle sera containing antibodies to Leptospira interrogans serovars, Brucella sp., Anaplasma marginale, or A. phagocytophilum were tested by ELISA with separate whole-cell or recombinant B. burgdorferi antigens. An ELISA and Western blot analyses can be used to confirm the exposure of cattle to B. burgdorferi.     

Comparison of complement fixation test,immunoblotting, indirect ELISA,and competitive ELISA for detecting antibodies to Mycoplasma mycoides subspecies mycoides small colony (SC) in naturally infected cattle from the 1995 outbreak in Botswana

Four serological tests were compared in order to evaluate their efficacies in detecting antibodies to M. mycoides subspecies mycoides SC in cattle sera sampled in 1995 from herds affected with contagious bovine pleuropneumonia (CBPP) in the north-western part of Botswana. Tests that were compared included immunoblotting test (IBT), indirect enzyme-linked immunosorbent assay (i-ELISA), competitive enzyme-linked immunosorbent assay (c-ELISA) and complement fixation test (CFT). The percentages of seropositive samples in the iELISA (48%) and in the c-ELISA (48%) were similar but were comparatively lower than those obtained by the IBT (57%) and CFT (61%). The percentages of positive sera in the IBT and CFT were also similar and overall the efficacy of these tests was better than that of the two ELISA tests. There was 95.5% agreement between the IBT and CFT, 85% agreement between the IBT and c-ELISA, 90.9% agreement between the IBT and i-ELISA, 88.6% agreement between the i-ELISA and CFT, 80% agreement between the c-ELISA and CFT and 90% agreement between the two ELISA tests. It became clearly evident from this comparative study that no single serological test was capable of detecting all animals affected by CBPP under natural field conditions of infection.     

Assessment of performance of selected serological tests for diagnosing brucellosis in pigs

Swine brucellosis due to Brucella suis is considered an emerging zoonotic disease whose control is based on serological testing and the subsequent culling of seropositive animals or the full depopulation of affected flocks. Here we assessed the performance of several serological tests (Rose Bengal Test [RBT], indirect ELISA [i-ELISA], blocking ELISA [b-ELISA], and two competitive ELISAs [c-ELISA]) for diagnosing swine brucellosis caused by B. suis biovar 2. Both frequentistic and Bayesian statistical inference were used. A frequentistic analysis, using sera from known gold standard (GS) populations (i.e., from truly infected or brucellosis free animals), resulted in maximum (100%) diagnostic sensitivity (Se) and specificity (Sp) in the RBT, i-ELISA and b-ELISA tests. However, c-ELISAs resulted in lower diagnostic Se (ranging from 68.5% to 92.6%, according to the different cut-offs selected). A Bayesian analysis of tests yielding the best diagnostic performance with GS sera (RBT, i-ELISA and b-ELISA), but using a large collection of field sera, resulted in similar Se among tests but markedly lower (≈ 80%) than that resulting from the frequentistic analysis using the GS serum populations. By contrast, the estimated Sp in the Bayesian analysis was only slightly lower than 100%, thus similar to that obtained frequentistically. Our results show that adequate diagnostic tests for brucellosis in swine are available, but also emphasize the need for more extensive validation studies before applying these tests under field conditions.     

小反刍兽疫病毒核蛋白单克隆抗体的制备与初步应用

针对小反刍兽疫病毒核蛋白制备特异性的单克隆抗体,并对其进行生物学特性鉴定和初步应用。以纯化的Bacmid-PPRV-N重组蛋白为抗原免疫BALB/c小鼠,取免疫小鼠的致敏脾细胞与SP2/0骨髓瘤细胞在PEG作用下融合,获得单克隆抗体,并通过染色体技术等方法研究其生物学特性,将其作为竞争单抗,Bacmid-PPRV-N重组蛋白作为检测抗原建立竞争ELISA检测方法。结果表明:经克隆和间接ELISA筛选,获得了2株能稳定分泌抗小反刍兽疫病毒N蛋白抗体的杂交瘤细胞株,分别命名为5B11和3H10-3B8。生物学特性鉴定试验表明:5B11和3H10-3B8抗体类型和亚类均为IgG2b;5B11单抗腹水的效价达1∶819 200,3H10-3B8达1∶12 800;血清学试验证明2株单抗均能与Bacmid-PPRV-N重组蛋白抗原结合,具有高度的特异性;相加ELISA试验结果显示,5B11和3H10-3B8 2株单克隆抗体分别识别N蛋白上不同的抗原位点;2株杂交瘤细胞的染色体均为99～104。应用建立的c-ELISA检测方法对222份血清样品进行PPRV抗体的检测,与参考试剂盒比较得到98.20%的符合率。本研究获得了2株能稳定分泌抗PPRV N蛋白单克隆抗体的杂交瘤细胞株,以单抗5B11作为竞争抗体建立了PPRV的c-ELISA检测方法。     

Comparison of the enzyme-linked immunosorbent assay and complement fixation test for detecting Brucella ovis antibodies in sheep

A solid-phase, indirect, enzyme-linked immunosorbent assay (ELISA) was compared with the microtitre complement fixation test for detecting Brucella ovis antibodies in 220 ram sera. The ELISA was more sensitive than the complement fixation test; it demonstrated antibodies in 11 sera from known infected or vaccinated rams that were complement fixation test negative. No false positives were recorded with the ELISA and, in 36 sera positive to both tests, the ELISA titres were consistently higher than the corresponding complement fixation test titres.     

Discrimination between sheep antibodies to Brucella melitensis and to Brucella ovis

When preparations containing smooth Brucella abortus lipopolysaccharide (LPS) were used as antigens in an ELISA, strong positive reactions were obtained with sera from sheep infected with Brucella melitensis or with Brucella ovis. Oxidation of the LPS with sodium metaperiodate greatly reduced the extent of the cross-reactions with antisera to B. ovis, with little effect on the reactions with antisera to smooth B. melitensis. Periodate oxidation of hot saline extract (HSX) antigen of B. ovis markedly reduced its reactivity in ELISA with anti-B. ovis sera and eliminated cross-reactivity with anti-B. melitensis sera. The reactivity of HSX was maintained after treatment with proteinase K.A simple ELISA system, in which replicate samples from a single serum dilution were tested in parallel against both B. ovis HSX antigen and periodate-oxidised smooth phase B. abortus LPS, was evaluated. It was found to discriminate well between antibodies induced by vaccination or virulent infection with B. melitensis strains and those induced by infection with B. ovis.     

The ELISA for the examination of hare sera for anti-Brucella antibodies

Hare brucellosis is caused primarily by Brucella suis biovar 2. Hares along with wild boars are the natural reservoir of this microorganism. In view of restriction of applicability of traditional serological methods the work aimed to develop the ELISA to examine hare sera for the presence of anti-Brucella antibodies. Lipopolysaccharide (LPS) antigen obtained from the strain S19 of Brucella abortus and the conjugate of antibodies against rabbit immunoglobulin with horseradish peroxidase were used in the test. Hares' sera positive and negative in the CFT were used as controls of the ELISA. The sera collected from 9 hares suspected to be infected with Brucella organisms, positive in CFT (in this number 7 hares revealed clinical symptoms or anathomopathological lesions characteristic of brucellosis), 6 sera from hares showing no symptoms of the disease, negative in CFT and 520 sera from hares monitored for brucellosis were tested. All serum samples from hares suspected for Brucella infection were positive in ELISA and 2 of them were negative in RBPT. Additionally among the samples from hares monitored 12 sera were positive in ELISA and CFT, whereas 9 sera from 12 ones were also positive in the RBPT. The obtained results indicated that the ELISA developed in our laboratory proved to be equivalent in specificity to CFT. In addition, ELISA proved to be more sensitive than RBPT for the diagnosis of Brucella infection in hares.     

Antibodies to the CP24 protein of Brucella melitensis lack diagnostic usefulness in ovine brucellosis

The potential diagnostic usefulness of antibodies to the ribosome recycling factor of Brucella melitensis (CP24) was assessed in sheep by an indirect ELISA with purified recombinant CP24. Sera from uninfected animals from the UK (n=44) and from local flocks (n=42), from sheep naturally infected with B. melitensis (n=12) or B. ovis (n=12), and from lambs (n=7) or pregnant ewes (n=6) vaccinated with B. melitensis Rev-1, were assayed. High specific optical densities (OD(with antigen) - OD(without antigen)) were obtained with both the groups of normal sera, which resulted in high cut-off values (1.414 and 1.267, respectively). Only two infected sheep yielded specific OD higher than these cut-off values. No significant difference was found between mean specific OD from B. melitensis- or B. ovis-infected sheep (0.574 and 0.472, respectively), those from vaccinated animals (0.396 and 0.400 for pregnant ewes and lambs, respectively), and those from Brucella-free animals. An inhibition ELISA with soluble CP24 confirmed the specificity of the antibodies detected in normal sera by the indirect ELISA; these antibodies belonged to the IgG class as revealed by the use of a specific conjugate. Sera from infected sheep were all positive for antibodies against lipopolysaccharides and lumazine synthase from Brucella. These results show that anti-CP24 antibodies have no diagnostic role in ovine brucellosis.     

Enzyme-linked immunosorbent assay for detecting antibodies to Brucella abortus in bovine milk and serum

An enzyme-linked immunosorbent assay (ELISA) method was evaluated for the detection of antibodies to Brucella abortus in cows milk. Milk samples from seropositive or -negative cows were sed to determine the distribution of absorbance values to classify milk as ELISA positive or ELISA negative. Brucella abortus was isolated from milk samples from 10 (45%) of the 22 cows whose milk and serum were ELISA positive. The ELISA was evaluated and determined to be an appropriate method for detecting antibodies to B abortus in bovine milk.     

Comparison of complement fixation test,competitive ELISA and LppQ ELISA with post-mortem findings in the diagnosis of contagious bovine pleuropneumonia (CBPP)

The complement fixation test (CFT), the c-ELISA and an indirect LppQ ELISA were compared to post-mortem (PM) inspection for the diagnosis of contagious bovine pleuropneumonia (CBPP). Sera from 797 cattle in the CBPP affected area of Kazungula, Zambia and 202 sera from Lusaka, Zambia, a CBPP-free area were used. The clinical history of CBPP was recorded and all the cattle from Kazungula were slaughtered and PM inspections conducted. The prevalence of CBPP in Kazungula was 67.5% (95%CI 67.2%, 70.8%), 52.6% (95%CI 49.2%, 56.2%), 59.0% (95%CI 55.5%, 62.4%) and 44.4% (95%CI 41.0%, 47.9%) using PM inspection, CFT, c-ELISA and LppQ ELISA, respectively. Three of the 202 negative control animals tested positive on the c-ELISA although they were from a known CBPP negative zone. In this study, the c-ELISA was more sensitive in detecting cattle with lesions in the chronic stage than any other test whilst the CFT detected more during the onset stage. No single serological test could detect all stages of CBPP infection, therefore the use of more than one test is advised.     

Comparison of results from five serologic methods used for detecting Brucella abortus antibody activity in coyote sera

A total of 423 serum samples representing 94 coyotes which were wild trapped in east Texas were used to compare the serologic results from five different methods for detecting antibodies to Brucella abortus. The sera were tested for Brucella spp. antibody activity by the Card (CARD), rivanol precipitation (RIV), standard agglutination tube (SAT), cold complement fixation test (CF), and enzyme linked immunosorbent assay (ELISA) methods. Each serum sample selected for this comparison demonstrated antibody activity by one or more of the five serologic methods. When the serologic results of the five different methods were compared, 143 sera were positive according to the CF test and agreement was 67.1-70.6% with CARD, RIV and SAT. The maximum agreement for CF positive was with CARD (70.6%) and the lowest agreement fro CF negative was also with CARD (56.4%). Agreement among the serologic methods for the SAT positive ranged from 69.1% (CARD) to 72.7% (RIV). Agreement between SAT and ELISA was poor with only 38.1% agreement for SAT positive and 11.3% agreement for SAT negative. Agreement between methods for CARD positive sera was poor, with a low of 43% for both SAT and ELISA, and a high of 55.6% for RIV. Agreement between methods for 149 RIV positive sera was 83.2% for CARD, 67.8% for SAT, 64.4% for CF and only 50.3% for ELISA. Agreement between methods for ELISA positive results ranged from 49.0% for RIV to 62.7% for CARD.(ABSTRACT TRUNCATED AT 250 WORDS)     

Validation of the Fluorescence Polarisation Assay (FPA) for the serological diagnosis of brucellosis

In order to evaluate suitability of Fluorescence Polarisation Assay (FPA) for serological Brucella diagnostic, 1739 samples of sera from cattle, pigs, sheep and goats (65 Brucella-positive, 960-negative and 714 false-positive sera) were investigated at a dilution of 1:10. The cut-off was adjusted by means of ROC analysis. Furthermore, the serum samples were examined for Brucella antibodies using SAT, CFT and ELISA and the results were evaluated regarding sensitivity and specificity. FPA, SAT, CFT and ELISA attained a sensitivity of 92.3, 98.5, 84.6 and 86.2%. In comparison, specificity varied with 87.8, 72.6, 92.5 and 85.8%, respectively. Accordingly, FPA is a suitable test for serodiagnosis of brucellosis.     

First serological evidence of West Nile virus infection in wild birds in Northern Algeria

While the epidemiology of Flaviviruses has been extensively studied in most of the Mediterranean basin, little is known about the current situation in Algeria. In order to detect the circulation of West Nile (WNV) and Usutu viruses (USUV) in Kabylia, 165 sera were collected from two wild birds species, namely the long distance migrant Turdus philomelos (song thrush) (n = 92) and the resident Passer domesticus (house sparrow) (n = 73). A total of 154 sera were first analyzed by commercial competition ELISA. WNV and USUV micro-neutralization tests were performed on all c-ELISA positive sera and all samples with poor volume. Overall, 7.8 % (CI95 %: 3.5–11.9) were positive by c-ELISA. Positive results were detected in 12.5 % (CI95 %:5.6–19.4) of song thrushes and 1.5 % (CI95 %: 0.0–4.5) for sparrow.Micro-neutralization tests revealed an overall seroprevalence of 6.7 % for WNV (CI95 %: 2.9–10.3), Neutralizing antibodies were found in 8.7 % (CI95 %: 3.0–14.4) for song thrushes and in 4.1 % (CI95 %: 0.0–8.7) of sparrows. The current study demonstrates significant seroprevalence of WNV antibodies in wild birds in Algeria.     

Bovine brucellosis: evaluation of field sera by a competitive and superimposable ELISA utilising a monoclonal antibody against Brucella abortus lipopolysaccharide

With the aid of a horseradish peroxidase (HRP) tagged monoclonal antibody against smooth lipopolysaccharide from Brucella abortus (Bruce 1), a competitive and superimposable ELISA test procedure for bovine brucellosis has been evaluated for its ability to discriminate between Strain 19-vaccinated (S19-Vacc) and Biotype 1-infected (B1-Inf) cattle. In the competitive assay, all sera from S19-Vacc animals competed effectively against HRP-Bruce 1 (low HRP activity), while 10 out of 40 B1-Inf animals competed less effectively with Bruce 1 (high HRP activity). Successful competition by cattle antibodies would result in an increased proportion of cattle Igs binding to the assay antigen. This was confirmed by superimposing an alkaline phosphatase conjugated rabbit anti-cattle Ig after the competitive ELISA had been completed. With the superimposable assay, alkaline phosphatase activity was correspondingly high for S19-Vacc animals, and low for 36 out of 40 B1-Inf animals. The superimposable ELISA had therefore improved the discriminatory capabilities of the assay procedure from 75% to 90%.     

Seroprevalence and risk factors for bovine brucellosis in Jordan

We investigated the seroprevalence and risk factors for Brucella seropositivity in cattle in Jordan. The sera from 671 cows were randomly collected from 62 herds. The antibodies against Brucella were detected using a Rose Bengal plate test and indirect ELISA. A structured questionnaire was used to collect information on the cattle herds'' health and management. A multiple logistic regression model was constructed to identify the risk factors for Brucella seropositivity. The true prevalence of antibodies against Brucella in individual cows and cattle herds was 6.5% and 23%, respectively. The seroprevalence of brucellosis in cows older than 4 years of age was significantly higher than that in the younger cows. The seroprevalence of brucellosis in cows located in the Mafraq, Zarqa and Ma''an governorates was significantly higher than that of the other studied governorates. The multiple logistic regression model revealed that a larger herd size (odd ratio = 1.3; 95% CI: 1.1, 2.6) and mixed farming (OR = 2.0; 95% CI: 1.7, 3.7) were risk factors for cattle seropositivity to Brucella antigens. On the other hand, the use of disinfectants (OR = 1.9; 95% CI: 1.1, 2.1) and the presence of adequate veterinary services (OR = 1.6; 95% CI: 1.2, 3.2) were identified as protective factors.     

Seroprevalence of antibodies against Borrelia burgdorferi and Anaplasma phagocytophilum in cats

OBJECTIVE: To determine whether cats in the northeastern United States develop serum antibodies against antigens of Borrelia burgdorferi and Anaplasma phagocytophilum and whether coinfection with the 2 organisms occurs. SAMPLE POPULATION: Serum samples from 84 healthy cats and 9 cats with lameness, fever, anorexia, or fatigue. PROCEDURE: Serum antibodies against B. burgdorferi and A. phagocytophilum were measured with an ELISA incorporating a whole-cell preparation or purified recombinant antigens, by means of Western blot analysis, or indirect fluorescent antibody (IFA) staining. RESULTS: ELISA results indicated that 44 of 93 (47%) sera contained antibodies against > or = 3 B. burgdorferi antigens, whereas 43 (46%) were reactive to whole-cell B. burgdorferi. Serum reactivity to protein 35, VlsE, and outer surface proteins A and F was most common. Seropositivity to > or = 3 antigens occurred at the same rate (5/9) in the 9 ill cats as in the 84 healthy cats (46% [39/84]). Of 13 sera reactive to recombinant antigens, 9 were seropositive as measured by Western blot testing with whole-cell antigen. Seropositivity rates of 30% and 38% were detected for antibodies against A phagocytophilum via IFA and ELISA testing, respectively. Fifteen (16%) sera had antibodies against both pathogens. CONCLUSIONS AND CLINICAL RELEVANCE: Cats living in areas infested by Ixodes scapularis ticks are exposed to B. burgdorferi and A. phagocytophilum and, in some instances, may be coinfected. Most cats appeared healthy. An ELISA incorporating specific recombinant antigens may be used adjunctively with Western blot and other assays to confirm B. burgdorferi and A. phagocytophilum infection in cats.     

Comparison of sensitivity and specificity in three commercial foot-and-mouth disease virus non-structural protein ELISA kits with swine sera in Taiwan

Three commercialized ELISA kits for the detection of antibodies to the non-structural proteins (NSPs) of FMD virus were compared, using sera from uninfected, vaccinated, challenged and naturally infected pigs. The kinetics of the antibody response to NSPs was compared on sequential serum samples in swine from challenge studies and outbreaks. The results showed that ELISA A (UBI) and ELISA B (CEDI) had better sensitivity than that of the 3ABC recombinant protein-based ELISA C (Chekit). The peak for detection of antibodies to NSPs in ELISA C was significantly delayed in sera from natural infection and challenged swine as compared to the ELISA A and B. The sensitivity of the three ELISAs gradually declined during the 6-month post-infection as antibodies to NSP decline. ELISA kits A and B detected NSP antibody in 50% of challenged pigs by the 9-10th-day and 7-8th-day post-challenge, respectively. ELISA B and C had better specificity than ELISA A on sequential serum samples obtained from swine immunized with a type O FMD vaccine commercially available in Taiwan. Antibody to NSPs before vaccination was not detected in swine not exposed to FMD virus, however, antibody to NSPs was found in sera of some pigs after vaccination. All assays had significantly lower specificity when testing sera from repeatedly vaccinated sows and finishers in 1997 that were tested after the 1997 FMD outbreak. However, when testing sera from repeatedly vaccinated sows or finishers in 2003-2004, the specificity for ELISAs A, B and C were significantly better than those in 1997. This effect was less marked for ELISA A. The ELISA B was the best test in terms of the highest sensitivity and specificity and the lowest reactivity with residual NSP in vaccinates.