Diagnostic Tests & Testing Strategies for Neospora caninum
ACVIM 2004
Paul Baillargeon, DVM, MSc; Gilles Fecteau, DMV, DACVIM
St-Louis de Gonzagues, QC, Canada; St-Hyacinthe, QC, Canada
INTRODUCTION
Since its first identification in 1984 in a litter of puppies6 and the recognition of its important role in cattle abortion in 1988, N. caninum has been reported in 26 countries all over the world13. Studies have estimated that N. caninum was responsible of 12.5% of abortions in England 11, 15 to 20% in Holland33and as much as 43% in California1,31. The herd sero-prevalence level for most countries ranges between 64% and 83%17,16,21,26,23 and individual sero-prevalence ranges between 10% and 39%3.
Abortion is the dominant clinical feature for animal infected with N. caninum. Despite the fact that the first report of cattle abortion by N. caninum was of the epidemic type (>10% of the females at risk aborting), it is well accepted that a sporadic pattern is more common in endemic herds. Abortion storms however have been observed in endemic herds and are believed to have been induced by factors causing recrudescence of N. caninum in chronically infected animals32. Infected females may abort more than once but this risk is less than 5% at the following pregnancy 14. Hence, this figure does not represent the lifelong risk of re-abortion for an infected female. The presence of serum antibodies does not provide protection even if the risk of re-abortion for an infected animal decreases with the number of pregnancies29.
The majority of new bovine infections are acquired transplacentally in utero30. Up to 93% of the calves born to infected females have developed antibodies at birth28. It has been demonstrated that vertical transmission can maintain the prevalence of infection from generation to generation7. The first evidence of the horizontal transmission (HT) of the parasite was the demonstration, in Québec dairy herds, of the association between farm dogs and the sero-prevalence to N. caninum23. The possibility of HT was further reinforced in 1998 with the demonstration that dogs are a definitive host for N. caninum20. The natural mechanism for HT from dogs to susceptible species has not been clarified yet.
DIAGNOSIS
The identification of N. caninum has triggered an important research effort during the nineties: using the keyword Neospora generates a list of close to 700 scientific papers from the Pubmed journal databank. Diagnostic methods to identify N. caninum in cases of bovine abortions were developed in the early '70. In 1994 and 1995, the development of 2 ELISA tests by Dutch and Californian workers has opened the field of sero-epidemiology. These tests have also helped to refine our understanding of the transmission and of the life cycle of the parasite. Over the last few years, research has focused on the understanding of the pathogenesis and immune mechanisms involved in this infection.
The demonstration of the parasite in the target tissues of an infected animal or foetus is still the only way of confirming an infection by N. caninum12. Tissue sections staining by immunohistochemistry (IHC) and detection of specific antibodies by indirect fluorescent antibody testing (IFAT) have increased both the sensitivity and the specificity of this diagnostic procedure10,5,24. The presence of specific antibodies to N. caninum in the blood of an animal confirms exposure to the parasite. Nevertheless, Dubey has stated that even a low IFAT titer in fetal fluids is specific of the infection12. It is now widely accepted that a high pre-colostral titer is strongly indicative of an in utero acquired infection.
Serological tests have allowed screening of large groups of individuals in order to estimate the prevalence of the disease. Serological testing for N. caninum has been reviewed in 2 different papers in 1998 and 20002,9. An important feature of serology is that test sensitivity(Se) and specificity(Sp) can be adjusted to specific needs by using different cut-offs for interpretation27. Clinicians and epidemiologists are more interested into the predictive value of the results which takes into account the prevalence of the infection in population and the test characteristics. Three types of serologic tests have been used either commercially or for research purposes.
Indirect fluorescent antibody test (IFAT): First serological test used and still used as the reference test to validate other tests. In this procedure, intact tachyzoites are exposed to the sample antibodies and then to fluorescein label-antibodies directed against the animal species specific immunoglobulins of the sample. The evaluation of the results under a fluorescent microscope requires training and experience and, hence may vary from one laboratory to another. Interest in IFAT was related originally to its high specificity (>99%), its cross-reactivity with Toxoplasma gondii and other coccidians being very low.
Direct Agglutination test (DAT): This test is a modification of the commercial test used for the detection of T.gondii22. Sensitivity and specificity of the test, now commercially available, compared well with IFAT in that study. The interest of the test, in comparison with IFAT and ELISA, is that it can be used across a large number of animal species without modification.
Enzyme linked immunosorbent assay (ELISA) testing has become the most widely used diagnostic method because it can be automated, hence allowing for screening of large number of samples and for a more objective interpretation of the test results. Five variations of the ELISA tests have been developed: 1. Crude antigen (whole tachyzoites) lysates by sonication or detergent solubilisation 2. Fixed tachyzoites 3. Immuno Stimulating Complex (ISCOM): an IgG avidity ELISA has been developed in 19998 4. Recombinant protein5.Antigen Capture.Commercial kits available have reported Se ranging between 87% and 90% and Sp ranging between 92 and 100%.2
Several diagnostic kits are available commercially and results of different laboratories can not be compared easily. All serological tests perform well with serum containing high antibody levels. At low levels, the ELISA results present more variability and discrepancies may even be observed2. Serological testing works at its best with groups making it the instrument of choice for the control of the infection into populations. Individual testing remains a challenge for results interpretation. Adjusting cutoff levels to the purpose of the investigation would help but is not expected to become routinely used by veterinary practitioners.
COMPARISON OF RESULTS BEFORE AND AFTER MODIFICATIONS TO A COMMERCIAL CRUDE ANTIGEN ELISA DETECTION KIT FOR N. caninum
The serum bank
In order to compare the test results of a modified commercial ELISA kit for N. caninum antibody detection with results obtained before the modification, the serum bank built during the course of a research project on the efficacy of embryo transfer for the prevention of vertical transmission of N. caninum was used.4 This bank was composed of the monthly serum samples of 93 cows and heifers used as recipients for the embryos implanted during the trial. Eighty-seven females were included in the study results4 and 6 were not included because their pre-implantation test result was higher than 0,40 and lower than 0,80. The commercial ELISA kit for the detection of N. caninum antibodies used for these tests has a reported Se of 88.4% and Sp of 99% at a cut-off of 0.6 by the manufacturer (Biovet, St-Hyacinthe, Qc).
The infectious status for N. caninum of these recipients was determined by the serologic profile of their monthly serum tests and by the ELISA result of a pre-colostral serum sample on a live calf or the tissue analysis results of the aborted fetuses or stillborn calves. Eight of the 93 recipients were serologically positive before implantation and, using the 0.6 S/P cutoff ratio, consistently maintained this status along pregnancy. Two of these recipients gave birth to calves that did not present signs of infection at birth or abortion (one was serologically negative on its pre-colostral serum sampling and one was IHC negative at tissue analysis). All other females (n=79) included in the study results, consistently maintained their serological negative status, using the 0.6 cutoff level, and gave birth to either sero-negative on the pre-colostral serum sample or IHC negative calves at birth or abortion. For the 6 females that were not included in the final study results because they did not meet with inclusion criteria, 6 of the 54 monthly serum samples from 3 different animals, using the same cutoff level, were inconsistent with their original negative serological status. Nevertheless, they gave birth to 5 sero-negative calves on a pre-colostral serum sample and one IHC negative stillborn calf at tissue analysis.
The serum samples (n=207) from 25 Neospora negative recipients were randomly selected from the serum bank and resubmitted to the same laboratory who did the original testing. All sera from Neospora positive recipients (n=61) and from the 6 Neospora negative recipients (n=53) that were excluded from the study results were also resubmitted. These samples (n=321) were tested with the modified commercial ELISA kit for N. caninum antibody detection.
Results of the retest
Confidence intervals (CI 95%) for ELISA S/P ratios of the pregnancy monthly serum samples are presented in Table 1 for the non infected and infected groups. The 2003 monthly results are significantly lower for the non infected group of females when compared with the results of the 2000 test. In addition, 11 of the 261 samples (4%) in 2000 would have misclassified these non infected animals in comparison of none in the retest of 2003. These results are in agreement with the limitations reported for ELISA serology with samples containing low levels of antibodies. Bjorkman et al., in a review on serological diagnosis of N. caninum raises the point that "the demonstrable test result of the antigen-antibody reaction is the summation of several molecular interactions." Secondary antibodies may have different affinity for a particular immunoglobulin antigen with sizable effect on test results9. This effect is also known as "background noise" and is best illustrated by testing pre-colostral serum samples. In the embryo transfer study, 60 calves (83%) from the non infected dams had no detectable level of antibody before colostral intake. All of the other calves (n=12) were below 0.20 on their pre-colostral serum3.
There is no published report on the cause of this background noise. Cross-reactivity with other coccidians, particularly T.gondii, is negligible13. Antibody levels for infected animals vary with stage of pregnancy25. Anecdotal reports have mentioned current vaccination programs as a likely cause of temporarily non specific serologic reactions for N.caninum negative animals. The results of the 2003 retest of our serum bank sampling demonstrates the improved ability of the ELISA test used to interact more specifically with specific N.caninum antibodies.
There does not appear to be any significant change in the ability of the ELISA test to detect N. caninum antibodies in the group of infected recipients and this is in agreement with the observations already published2,9.
Table 1. Confidence Intervals (CI 95%) for ELISA s/p ratios on 322 sera collected monthly during the pregnancy of 39 recipients
Click on the table to see a larger view
Strategies for interpretation of N. caninum ELISA test results
Despite this significant improvement in the serological diagnosis of N. caninum, clinicians will occasionally be puzzled when interpreting serology results. Retesting is not always an option and when performed, may still be influenced by the background noise effect. Two strategies will assist in the interpretation of suspicious or contradictory results.
Pre-colostral serum sampling is an effective way of accurately determining serological status of individual animals (foetus and dam). As already mentioned, more than 80% of the calves from non infected dams are born without detectable levels of N. caninum antibodies. These pre-colostral serum samples were not retested with the test kit used in 2003. We postulate that the proportion of calves without detectable levels would be increased with this new test. On the other hand, calves from infected dams all had S/P ratios above 1.00 except for one. One limitation though, is that up to 8% of infected calves may be born with a low antibody level to serologically positive dams and will eventually seroconvert during the first month of their life. This is apparently caused by late pregnancy fetal infections that do not have time to seroconvert before birth15.
The high level of efficiency of vertical transmission of N. caninum, greater than 90% according to many reports, may be used to predict the serological status of a given individual with a Se and Sp similar to serological testing. Building the familial flowchart of a suspicious animal as in cases presented in Figure 1. helps at identification of discrepancies and will eventually confirm infectious status from progeny or ancestors testing. In this case, cow 265 was not infected and it can be predicted that heifer 807 is not infected neither. Cow 626 in the second flow chart had always been classified as infected in this herd since her March 1999 test. It was obvious when serological status of her progeny became available that her own status needed to be reevaluated. Retesting in March 2003 confirmed that she was indeed serologically negative to N.caninum. This case exemplifies a potentially significant cause for contradictory results on a laboratory report. It is more than likely that misclassification of this animal was an error in sample handling or an improper identification of the animal at the time of the first test.
Click on the table to see a larger view Figure 1. Familial flowchart for assessment of N. caninum serological status
CONCLUSION
The cornerstone of a neosporosis control program in a herd is the serological testing of all animals of unknown infectious status. Herd eradication of N.caninum can be achieved with the assistance of a "Test and Cull" strategy. Recent improvements in detection kits provide results that improve our ability to properly classify individual animals as the results presented in this article tend to demonstrate.
Culling of infected animals early after birth is a more economical way for herd control of N. caninum. Pre-colostral blood sampling and familial flow chart analysis will improve overall predictive value of serological testing in such programs.
Embryo transfer using the washing and trypsin treatment protocol recommended by IETS remains the only proven way to prevent vertical transmission between N. caninum infected dams and their progeny4,18. Biosecurity strategy should include minimizing access to feed bunks and feed storage areas to dogs and wild canids. Serological testing of all new entries in the herd is highly recommended biosecurity measure since prevalence appears to be higher in purchased animals than in the general population19,3.
REFERENCES
1. Anderson, M. L. et al, J.Am.Vet.Med.Assoc.,1995;207,(9):1206-1210
2. Atkinson, R. et al, Parasitology Today,2000;16,(3):110-114;
3. Baillargeon, P.,Thesis, U. de Montréal,2000;
4. Baillargeon, P. et al, J.Am.Vet.Med.Assoc.,2001;218,(11):1803-1806;
5. Barr, B. C. et al, J.Am.Vet.Med.Assoc.,1993;202,(1):113-117;
6. Bjerkas, I. et al, Zestschrift fur Parasitenkunde,1984;70,():271-274;
7. Bjorkman, C. et al, J.Am.Vet.Med.Assoc.,1996;208,(9):1441-1444;
8. Bjorkman, C. et al, J Vet Diag Invest,1999;11,(1):41-44;
9. Bjorkman, C. et al, Int J Parasitol,1998;29,(10):1497-1507;
10. Boger, L. A. et al, Vet.Parasitol.,2003;113,(1):1-6;
11. Davison, H. C. et al, Int J Parasitol,1999;29,(8):1189-1194;
12. Dubey, J. P., Vet Parasitol,1999;84,(3-4):349-367;
13. Dubey, J. P., Korean J.Parasitol., 2003;41,(1):1-16;
14. Dubey, J. P. et al, Vet Parasitol,1996;67,():1-59;
15. Hietala, S. K. et al, Int J Parasitol, 1999;29,(10):1669-1676;
16. Jensen, A. M. et al, Preventive Veterinary Medicine,1999;40,(3-4):151-163;
17. Keefe, G. P. et al, Can.Vet.J.,2000;41,(11):864-866;
18. Landmann, J. K. et al, Aust.Vet J,2002;80,(8):502-503;
19. Mainar-Jaime, R. C. et al, Vet Rec,1999;145,(3):72-75;
20. McAllister, M. M. et al, Int J Parasitol,1998;28,(9):1473-1478;
21. Ould-Amrouche, A. et al, Veterinary Research,1999;30,(5):531-538;
22. Packham, A. E. et al, Clin.Diagn.Lab Immunol.,1998;5,(4):467-473;
23. Paré, J. et al, J.Am.Vet.Med.Assoc.,1998;213,(11):1595-1598;
24. Paré, J. et al, J Vet Diag Invest,1995;7,():273-275;
25. Paré, J. et al, J Parasitol,1997;83,(1):82-87;
26. Quintanilla-Gozalo, A. et al, Int J Parasitol,1999;29,(8):1201-1208;
27. Reichel, M. P. et al, Vet.Parasitol.,2002;107,(3):197-207;
28. Schares, G. et al, Vet Parasitol,1998;80,(2):87-98;
29. Thurmond, M. C. et al, Am J Vet Res,1997;58,(12):1381-1385;
30. Thurmond, M. C. et al, Current Vet. Therap, 1999;4,(9):425-431;
31. Trees, A. J. et al, Int J Parasitol,1999;29,(8):1195-1200;
32. Wouda, W. et al, Theriogenol, 1999;52,():233-245;
33. Wouda, W. et al, J Vet Diag Invest,1997;9,(2):180-185.
Speaker Information
(click the speaker's name to view other papers and abstracts submitted by this speaker)
Paul Baillargeon, DVM, MSc
Clinique Veterinaire St-Louis-Embryobec
St-Louis De Gonzague
CO. Beauharnois, QUE., CANADA
Gilles Fecteau, DMV, DACVIM