Mycobacterium avium subsp. paratuberculosis (MAP) causes granulomatous enteritis in ruminants that can lead to weight loss and emaciation, as well as diarrhoea in certain species . MAP has been detected in wild ruminant species worldwide, and it is thought that all ruminant species are susceptible to MAP infection . Although clinical disease, known as Johne’s disease, has only occasionally been described in wildlife, its impact may be under-estimated due to incomplete knowledge of the clinical progression of MAP infection in these species, non-specific clinical signs, and limited testing in these populations.
Determining the infection status of wildlife is important to reduce the disease risks associated with translocation, to certify infection status for importation/exportation purposes, to limit sources of infection for humans and livestock, and to establish baseline values for future monitoring or surveillance . For the detection of current or prior MAP infection in wild ruminants, serological methods such as enzyme-linked immunosorbent assay (ELISA) may be preferred over alternative methods such as faecal culture, as ELISA is less expensive, fast and easy to perform, and collected serum can be used to screen for other infections.
Serum ELISA kits marketed for the detection of MAP-specific antibodies in domestic ruminant species are generally not validated for use with sera from other species . It is unlikely that these tests give equivalent results for other, even closely related species . Inter-species variations in test outcomes are due in part to the specificity of certain reagents included in commercial kits; for example, the labelled secondary antibodies or conjugates may have varying capacity to bind immunoglobulin G (IgG) from different species . The level of non-specific binding of serum proteins to components of the ELISA assay may also be difficult to predict, and can potentially result in a reduction of the signal-to-noise ratio, which is the ratio of the optic density (OD) of a positive control and a negative control at a given sample and conjugate concentration . Kits generally recommend cut-off values to classify samples as ‘negative’, ‘positive’ or ‘suspect’. Cut-off values are selected based on a set of parameters, in particular the species being tested, the target condition (i.e. infected, infectious, clinically diseased), and the testing objectives (e.g. demonstrating freedom from infection, estimating population-level prevalence, etc.) . When using a commercial kit for wildlife samples, changes in these parameters need to be reflected by adapting the cut-off values.
Guidelines for conducting a complete diagnostic validation have been thoroughly outlined [6, 7], and detailed recommendations for the design and reporting of diagnostic evaluation studies for chronic diseases have been developed . However, it is not always possible or appropriate to undertake such a full evaluation in wildlife studies due to time and budget constraints as well as sample availability, in particular accessing known positive and negative controls. In these cases, certain modification and evaluation steps may reduce the level of uncertainty in the test results if a complete validation is impossible.
The objectives of this paper were first, to assess current practices in testing and reporting of MAP serum ELISA for wildlife samples, and second, to modify and evaluate a commercial ELISA kit (IDEXX Mycobacterium paratuberculosis Antibody Testa, hereafter referred to as the IDEXX kit) for detecting current or prior MAP infection in elk (Cervus elaphus), bison (Bison bison) and caribou (Rangifer tarandus).