|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
||||||||||||||||
|
|
|
|
|
|
|||||||||||
|
||||||||||||||||
AABP News |
Johne’s
Disease, Article Four, part two
Johne’s Disease Diagnostic Tests - the ELISA Prepared and edited by Christine Rossiter and Don Hansen From the National Johne’s Working Group
Current tests are good tools Diagnostic testing is a tool to help producers make decisions and reach their goals for prevention or control of Johne’s disease in their herd. No single test is sufficient for all purposes, but reasonably accurate and cost-effective tests are available for the different diagnosis and control needs. Deciding which tests to use beyond confirmation of clinical suspects should be carefully considered. To choose among available tests, it is important to have a clear understanding of: ·Stages
of the infection
Detection of bacteria verses antibody, fecal culture and the ELISA The two most frequently used tests are the fecal culture and the absorbed ELISA. Fecal culture directly detects the intracellular microbe of Johne’s disease and its presence indicates infection has occurred. Direct detection tests like fecal culture are considered the most definitive. Antibody assays like the ELISA are less definitive in the individual, but can provide useful information in assessing status of herds and some individuals. These two tests measure infection differently but the magnitude of antibody or organisms cultured tend to correlate with advancing stages of infection. The Absorbed ELISA An important step in absorbed ELISA is the addition, before testing, of an extract of Mycobacterium phlei to the serum. This reduces cross-reacting antibodies that may produce false-positive results. The absorbed ELISA test method is illustrated in Figure 1. In this partially automated system, antibody in the serum sample, binds to M paratuberculosis antigen attached to walls in plastic plates. In subsequent steps, sample antibody is bound by conjugated anti-antibody that colorizes when the final substrate is added. The amount of color produced is measured at a precise time and wavelength. This numeric value, Optical Density (OD), reflects the amount of antibody in the sample. The OD value is used to define the sample as positive or negative relative to a cutoff value pre-determined by the manufacturer. ELISA results Recall that the accuracy of a diagnostic test is determined by performing the test on known infected animals for the sensitivity (Se) and known uninfected animals for the specificity (Sp). The OD values produced by the ELISA range from zero to +10.00 requiring a value to be determined that separates positive from negative results, i.e., a cutoff value. During validation, the manufacturer defines a cutoff point that is "optimal". In the Johne’s IDEXX test, the cutoff is reported by the manufacturer to have a sensitivity of 60% and specificity of greater than 97% in the reference populations. Since the most laboratories performs the IDEXX ELISA, it is worth reviewing how results are provided. OD value for each sample is first "adjusted" then compared to the cutoff value. At least one negative (NC) and positive (PC) control sample is run for each plate. At the end of a valid run, the negative control OD value is subtracted from each sample OD value and from the positive control OD value. The result is then reported as a sample to positive ratio (S/P). (S/P = S - NC / PC – NC) The sample S/P ratio is then compared to the pre-determined cutoff, which is 0.25. If the S/P is greater than 0.25, the manufacturer recommends the sample be interpreted as positive. Thinking beyond the single cutoff More information is available from the ELISA than the positive-negative interpretation. Practitioners should obtain the S/P, PC, and NC values from the laboratory and see how sample S/P ratios distribute between the negative control and the cutoff, and around the cutoff. With Johne’s disease, infected animal ratios will be spread out on both sides of the cutoff. Non-infected animal ratios tend to cluster around the NC value, but some will overlap with infected animal ratios. Animals with ratios below the cutoff, but well above the NC, have a higher likelihood of infection than those with values similar to the NC. Some of these animals may be good candidates for closer evaluation, additional testing, e.g., fecal culture or making management decisions where the cost of a false positive is relatively inexpensive, i.e., rejecting colostrum. It is important to recall that changing the cutoff value, i.e. accepting a lower ratio as a positive result, changes the sensitivity and specificity values. Since Se and Sp are inversely related, lowering the cutoff increases the Se, decreases the Sp and includes more true positives and false positives. Raising the cutoff lowers the Se, increases the Sp and reduces the chance of false positive results, but will miss some truly positive animals. Michael Collins, University of Wisconsin, evaluated the current (1998) IDEXX Johne’s ELISA in reference serum samples and determined the Se and Sp for several S/P ratios. The Se and Sp were used to determine a Likelihood Ratio (LR) for each S/P. The LR is the odds that a given result value would be expected in an animal with the disease as opposed to without. The LR for a positive test result is Se/(1-Sp). Using the LR values, he developed a multiple cutoff interpretation using five categories of increasing S/P ratios. The multiple cutoff interpretation uses more information from the ELISA and reflects the fact that higher S/P ratios occur more frequently in cattle with higher numbers of bacteria shed in their feces (late Stage II through Stage IV). ELISA Interpretation Chart
for S/P Ratios in Cattle
Reproduced by permission of the author.
To be used only in the interpretation of the absorbed ELISA
Accuracy of the absorbed ELISA The accuracy of the ELISA reported by IDEXX is 60% Se and 97% Sp. However, evaluations in other reference populations illustrate how sensitivity varies with the distribution of infection stages in the tested population. Sweeney demonstrated increasing sensitivity with increasing stages of infection. The ELISA sensitivity was the lowest in subclinical light shedding cattle, 15% + 6.6% and the highest in clinical cases, 87% + 8.4%. Overall at the IDEXX cutoff, sensitivity was 45%. ELISA specificity has been reported between 97% and 99%. There is a chance of cross-reaction occurrence with this test. Practitioners should be aware of this potential. Occasionally individuals with clinical Johne’s disease, are fecal culture, AGID positive and ELISA negative. Why antibody is not detected is unknown. The ELISA is used in herds that have a blend of infection stages. To standardize interpretation of the Johne’s absorbed ELISA for the National Johne’s Herd Status Program, an expert subcommittee of the National Johne’s Working Group came agreed on the estimates for Se and Sp values for the ELISA. They recommended a sensitivity of 25%, that accounts for early infections not represented in many reference populations, and a specificity of 98%. The influence of prevalence on interpretation of results The chance that a test result correctly reflects the animal’s true infection status depends upon the Johne’s herd-prevalence or your presumptive diagnosis. This reality emphasizes that tests are tools to be used in the context of other relevant information on infection risk. The following scenarios are important to recognize when interpreting ELISA results for mature cattle. 1. In a low prevalence environment, i.e., 1%, there is only a 15%-16% chance that a positive ELISA result (S/P ratio = 0.25) is correct and a 99% chance that a negative ELISA result is correct. 2. In a high prevalence situation, i.e., 30%, there is an 88% to 90% chance that a positive ELISA result (S/P ratio = 0.25) is correct and a 75% to 85% chance that a negative ELISA result is correct. Note: There is little difference in predictive values between a sensitivity of 25% and 60%. See Article Four, Part 1 Table 1 for predictive values at different prevalences. Use of the ELISA Since the sensitivity of the ELISA is not 100%, a single negative test-result should be interpreted as "antibodies not detected" versus "completely absent". The ELISA is not recommended as a stand-alone test for individuals. The ELISA is recommended for assessing herd status. As more information is gained about herd status, the test becomes more useful in measuring individuals. Repeat testing on an individual may also be helpful. Most Johne’s disease experts concur that culture should be used to confirm infection in ELISA positive animals in the following circumstances: 2. the pretest prevalence is uncertain 3. the cost of a false positive is high
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
