The author can be contacted at the FAO/WHO Centre for Brucellosis Reference and Research, OIE Brucellosis Reference Laboratory, Central Veterinary Laboratory, New Haw, Weybridge, Surrey, KT15 3NB, UK. Tel.: +44 1932 357216; fax: +44 1932 347046.
La brucellose est une maladie contagieuse spécifique de l'homme et des
animaux transmise par des bactéries du groupe Brucella. Cette maladie est considérée
par la FAO, l'OMS et l'OIE comme la zoonose la plus répandue dans le monde. Sa gravité
est due à la fois à son incidence économique sur l'industrie animale, ayant un effet
négatif sur les disponibilités protéiques totales d'origine animale, et aux graves
dangers qu'elle présente pour la santé humaine, soit par contact direct avec les animaux
infectés ou, plus fréquemment, par la consommation de lait et de produits laitiers
contaminés.
Parmi les différents tests sérologiques permettant le diagnostic et la surveillance de
la brucellose, la coloration sur plaque au Rose bengale (RBT) est largement utilisée
comme test de dépistage pour déterminer la prévalence dans les troupeaux. La présente
étude a permis de comparer la valeur de lots d'antigènes RBT d'origines différentes et
de proposer des critères pour la normalisation de ces antigènes.
La brucelosis es una enfermedad contagiosa específica del hombre y los
animales, transmitida por bacterias del grupo Brucella. La FAO, la OMS y la OIE la
consideran como la zoonosis más difundida del mundo. La gravedad de esta enfermedad
reside, por un lado, en sus repercusiones económicas en la industria ganadera, que van en
detrimento de la disponibilidad total de proteínas de origen animal, y por otro en el
serio riesgo que comporta para la salud de las personas, que pueden contraerla por
contacto directo con los animales infectados o, con más frecuencia, por el consumo de
leche y productos lácteos contaminados.
Entre las distintas pruebas serológicas de que se dispone para el diagnóstico y
vigilancia de la brucelosis es muy común el uso de la prueba con placa de tinte rosa
bengala (RBT) para el cribado de las muestras con miras a determinar la prevalencia de la
enfermedad en los rebaños. El presente estudio ha permitido comparar el valor de lotes de
antígenos RBT de orígenes diferentes y proponer criterios para su normalización.
Brucellosis is a specific contagious disease of humans and animals caused by bacteria of the Brucella group. The disease is considered by FAO, WHO and OIE as the most widespread zoonosis in the world. The importance of brucellosis is due both to its economic impact on the animal industry, where it has an adverse effect on total animal protein supplies, and to the severe hazard it represents to human health, through either direct contact with infected animals or, more frequently, the consumption of contaminated milk and dairy products.
Brucella melitensis, one of the six members of the Brucella group,
is encountered primarily in sheep and goats. However, owing to the extended presence of
the infection in small ruminants in a number of countries, especially around the
Mediterranean, there is growing evidence of an increased prevalence of B. melitensis
in cattle and sometimes in camels. On the other hand, B. melitensis is the most
pathogenic Brucella for humans.
There are several serological tests available for brucellosis diagnosis and surveillance.
Among these tests, the Rose Bengal plate test (RBT) is the recommended method for the
screening of samples to determine herd and flock prevalence. The value of this test
varies, however, depending essentially on the source of sera (e.g. cattle versus small
ruminants) and the batch of the RBT antigen used in the test. The aim of the present
article is to study the variability of the test in small ruminants and to propose criteria
for standard RBT antigens.
A variety of tests have been developed for diagnosis of brucellosis in cattle, and the
performance of these tests has been well established and validated (MacMillan, 1991). In
particular, RBT, the complement fixation test (CFT) and the indirect enzyme-linked
immunosorbent assay (iELISA) test are recommended by the International Office of
Epizootics (OIE) for use in this species (Corbel and MacMillan, 1995). These tests are
widely used for diagnosis in small ruminants, largely based on their effectiveness in
cattle, but they have not been sufficiently evaluated in sheep and goats. There is
increasing evidence that the sensitivity of RBT antigens obtained from different sources
may vary considerably, but that this is only observed when used for testing sera from
animals in flocks of low prevalence (Blasco et al., 1994).
The purpose of this study is to compare the sensitivity and specificity of the RBT using
antigens obtained from different sources and to relate this to the level of
standardization against the second International Standard anti-B. abortus Serum
(ISABS). Samples for this evaluation were selected by testing a large number of samples
from infected animals and choosing those that gave a negative reaction with an insensitive
antigen, then using this subset of "borderline" samples to test a range of RBT
antigens from a variety of sources and different levels of standardization.
For experiment 1, serologically positive samples were obtained from
animals in flocks in the field from which
B. melitensis had been isolated and also from animals experimentally infected with B.
melitensis. Samples were also obtained from animals in brucellosis-free flocks,
collected as part of routine monitoring (Table 1).
For experiment 2, the following subset of these samples was selected on the basis of their lack of RBT reaction (Table 2).
1
Experiment 1
Expérience 1
Experimento 1
Infected flocks |
Brucellosis-free flocks |
||
Sheep |
Field-infected |
114 |
5 872 |
Experimentally infected |
307 |
||
Goats |
Field-infected |
67 |
433 |
Experimentally infected |
211 |
||
Total |
699 |
6 305 |
2
Experiment 2
Expérience 2
Experimento 2
Infected flocks |
Brucellosis-free flocks |
||
Sheep |
Field-infected |
6 |
1 112 |
Experimentally infected |
21 |
||
Goats |
Field-infected |
49 |
100 |
Experimentally infected |
35 |
||
Total |
111 |
1 212 |
Antigen source. A number of laboratories supplied antigens for the trial. Nine antigen preparations were used (Table 3).
3
Antigen source
Sources d'antigènes
Fuentes de antígenos
Source |
Standardization |
PCV |
|
1. |
China (Beijing) |
+/20 |
4 |
2. |
Portugal |
+/45 -/50 |
8 |
3. |
France |
+/40 -/50 |
5 |
4. |
China (Nanning City) |
+/20 -/40 |
5 |
5. |
UK (CVL) |
3 |
20 |
6. |
Italy |
? |
5 |
7. |
UK (CVL) |
+/47.5 -/50 |
21 |
8. |
UK (CVL) |
ND |
10 |
9. |
UK (CVL) |
ND |
5 |
1 Producers standardized their antigens to give a positive reaction at one dilution and a negative reaction at a higher dilution.
Titration of ISABS. A primary 1/10 dilution of ISABS was made by adding 1 ml to 9 ml of phenol saline. Further dilutions were made in phenol saline to give dilutions of 1/35, 1/40, 1/45, 1/47.5, 1/50, 1/55, 1/60 and 1/80. All nine antigen preparations were tested against each dilution and the degree of agglutination, either + or ++ recorded.
The RBT was carried out according to the method specified in the OIE manual (Corbel and MacMillan, 1995). Reactions that were immediately obvious at the end of the four-minute incubation were recorded as ++ reactions. If close examination of the test was required to observe fine agglutination, the reaction was recorded as +.
The iELISA method was as described by Corbel and MacMillan (1995). The cELISA method was as described by Greiser-Wilke, MacMillan and Moenig (1991).
Experiment 1. All samples were tested by the RBT using a batch of antigen prepared at the Central Veterinary Laboratory (CVL), preparation 5. They were also tested by iELISA and cELISA.
Experiment 2. All samples from infected animals that did not give ++ agglutination to the RBT in experiment 1 were selected and tested by the RBT using antigens from a number of sources, as detailed above.
Level of standardization against the ISABS. The results of titrating the ISABS and testing the antigen preparations in experiment 2 are shown in Table 4.
4
Levels of standardization against the ISABS
Niveaux de normalisation par rapport aux ISABS
Niveles de estandarización respecto al ISABS
Antigen |
1/35 |
1/40 |
1/45 |
1/47.5 |
1/50 |
1/55 |
1/60 |
1/80 |
1 |
++ |
++ |
++ |
+ |
- |
- |
- |
- |
2 |
++ |
++ |
++ |
++ |
- |
- |
- |
- |
3 |
++ |
++ |
+ |
- |
- |
- |
- |
- |
4 |
++ |
++ |
++ |
++ |
- |
- |
- |
- |
5 |
++ |
++ |
+ |
- |
- |
- |
- |
- |
6 |
++ |
++ |
+ |
- |
- |
- |
- |
- |
7 |
++ |
++ |
++ |
+ |
- |
- |
- |
- |
8 |
++ |
++ |
++ |
++ |
++ |
++ |
++ |
+ |
9 |
++ |
++ |
++ |
++ |
++ |
++ |
++ |
++ |
Test results of experiment 1. Of the 699 samples from infected sheep and goats tested, 587 were positive to the RBT, reading a ++ reaction, 602 were positive to the iELISA and 578 to the cELISA. All samples taken from 5 872 sheep and 433 goats in brucellosis-free flocks were negative to all tests (Table 5).
5
Test results of experiment 1
Résultats des essais de l'expérience 1
Resultados de los ensayos del experimento 1
Total tested |
Number positive |
Sensitivity |
Specificity (percentage) |
|
RBT |
699 |
587 |
84.0 |
100 |
iELISA |
699 |
602 |
86.1 |
100 |
cELISA |
699 |
578 |
82.7 |
100 |
Test results of experiment 2. One hundred and eleven samples from infected animals, which did not give a ++ reaction with the "insensitive" antigen (preparation 5) in experiment 1, were tested by the nine RBT antigens; 1 212 samples from brucellosis-free sheep and goats were also tested (Table 6).
6
Test results of experiment 2
Résultats des essais de l'expérience 2
Resultados de los ensayos del experimento 2
Antigen |
No. of infected animals positive |
Reciprocal titre of ISABS (++) |
Reciprocal titre of ISABS (+) |
Relative sensitivity (+) |
Specificity (+) |
1 |
14 |
45 |
47.5 |
42.4 |
100 |
2 |
17 |
47.5 |
47.5 |
51.5 |
100 |
3 |
11 |
40 |
45 |
33.3 |
100 |
4 |
17 |
47.5 |
47.5 |
51.5 |
100 |
5 |
10 |
40 |
45 |
30.3 |
100 |
6 |
10 |
40 |
45 |
30.3 |
100 |
7 |
13 |
45 |
47.5 |
39.4 |
100 |
8 |
22 |
60 |
80 |
66.7 |
100 |
9 |
33 |
80 |
>80 |
100 |
100 |
iELISA1 |
21 |
- |
- |
63.6 |
100 |
cELISA2 |
4 |
- |
- |
12.1 |
100 |
1 i= indirect.
2 c= competitive.
Note: There was no correlation between the antigen's PCV and its sensitivity.
The RBT is rightly recognized as a quick, cheap and effective test for the diagnosis of brucellosis. It can be carried out with the minimum of equipment, and the end result is read by the naked eye. Because of its apparent simplicity, the need for stringent standardization of antigen and accuracy of reading is often overlooked. This problem has been highlighted by this study.
Previous studies have shown that there is a marked variation in the
levels of standardization of antigens in use around the world. In addition, there is
disagreement between laboratories so that, for example, different results were obtained
for antigens tested both in this study at CVL, in the study of Blasco et al.,
(1994) and in the level of standardization quoted by the source institute. If antigens are
standardized in one laboratory and used for routine testing in another, quite different
results may be obtained. For example, experience has shown that many factors affect RBT
reactions and their reading. The room temperature and efforts to ensure that reagents have
sufficient time to reach the antigen before use will have an effect, as will the method of
making the dilutions of the ISABS. There is no doubt that different people are able to see
finer agglutination than others and this is an important cause of variation. Thus, an
antigen standardized at a laboratory where these factors maximize agglutination and then
used at another laboratory where the opposite was true would create the illusion that the
antigen is incorrectly standardized and insensitive. There would appear to be a need for
the development of a method to ensure greater uniformity of reading both for
standardization and testing, possibly photometrically.
This study has shown that even the most sensitive antigen preparations demonstrated 100
percent specificity with brucellosis-free, unvaccinated sheep and goats. Thus, the
opportunity exists to change the definition of the standardization of RBT antigens to be
used for the testing of sheep and goats. There would be a considerable gain in diagnostic
sensitivity, especially in flocks with a low prevalence, with no loss in specificity.
The level of sensitivity that was achieved using the antigens standardized to the 1/60 or
1/80 dilutions of the ISABS detected more than three times the number of
"borderline" samples from infected animals than did some conventionally
standardized animals. It is impossible to tell what difference this would make in a
diagnostic situation, as it would depend on the disease prevalence. However, it is very
clear that the use of antigens standardized to this level seems to offer a significant
improvement over antigens currently in use. If it were assumed that all the samples found
to be positive to the insensitive antigen when tested in experiment 1 were also positive
to the preparations in Table 7, then it can be seen that the most sensitive antigen is
nearly 5 percent more sensitive than the least sensitive antigen tested.
If this were borne out in appropriate field trials, the OIE recommendation of the level of
standardization for RBT antigens for the diagnosis of brucellosis in sheep and goats
should be altered.
7
Extrapolation of experiment 1 tests
Extrapolation des essais de l'expérience 1
Extrapolación de los ensayos del experimento 1
Antigen |
No. of infected animals positive |
Sensitivity |
Specificity (+) |
1 |
601 |
86.0 |
100 |
2 |
604 |
86.4 |
100 |
3 |
598 |
85.6 |
100 |
4 |
604 |
86.4 |
100 |
5 |
597 |
85.4 |
100 |
6 |
597 |
85.4 |
100 |
7 |
600 |
85.8 |
100 |
8 |
619 |
88.5 |
100 |
9 |
630 |
90.1 |
100 |
iELISA |
602 |
86.1 |
100 |
cELISA |
578 |
82.6 |
100 |
Corbel, M.J. & MacMillan, A.P. 1995. OIE manual of standards
for diagnostic tests and vaccines. Paris. OIE.
Blasco, J.M., Garin-Bastuji, B., Marin, C.M., Gergier, G., Fanlo, J., Jiminez de
Bagues, M.P. & Cau, C. 1994. Vet. Rec., 133.
Greiser-Wilke, I., MacMillan, A.P. & Moenig, V.A. 1991. A competition enzyme
immunoassay with monoclonal antibodies for the analysis of sera from cattle of two herds
with suspected brucellosis. Tierarzliche Praxis., 19(2): 131-134.
MacMillan, A.P. 1991. In J.R. Nielsen and K. Duncan, eds. Animal
brucellosis. Boca Raton, Fla., USA, CRC Press.