Approaches towards optimising the gamma interferon assay for diagnosing Mycobacterium bovis infection in African buffalo (Syncerus caffer)

Abstract

The application of diagnostic tests for bovine tuberculosis in wildlife poses formidable technical difficulties and the use of the gamma interferon assay offers a simplified approach to testing wild animal species. We compared the performance of the gamma interferon assay in African buffalo (Syncerus caffer) under the recommended guidelines for interpretation of test results and found a high sensitivity (92.1%) at the cost of a greatly reduced specificity (68.3%). The optimised cut-off value for positive test results under local conditions was identified at an optical density of 0.385 at wavelength 450 nm as the preferred compromise between sensitivity and specificity. Additional optimisation approaches to improve test performance were examined and showed that the application of ‘a priori exclusions’ of test results on the basis of reactivity to fortuitum PPD (sensitin produced from Mycobacterium fortuitum) and to a lesser degree, avian PPD, increased specificity without losing sensitivity. The implications of these findings on a modified testing protocol adjusted to include measurement of immune responsiveness to fortuitum PPD and other interpretation schemes are discussed.

Introduction

The intradermal tuberculin test (IDT), or skin test, is still the most widely used method to diagnose bovine tuberculosis (BTB) in cattle in countries worldwide. Limitations of the IDT in cattle have been mainly described in developed countries and include aspects relating to test performance (Wood et al., 1992, Neill et al., 1992, Monaghan et al., 1994), source of tuberculin PPD (Cagiola et al., 2004) as well as to logistical drawbacks in terms of repeated handling of animals and the minimum testing interval (Radunz and Lepper, 1985).

Developing countries face a number of constraints in implementing and maintaining a bovine tuberculosis control scheme. In remote areas difficult accessibility, long travelling distances and large scattered herds are aggravating logistical constraints in addition to the frequent lack of veterinary capacity and handling facilities for cattle. In the communal farming systems of sub-Saharan Africa, BTB testing is typically performed at communal diptank stations where cattle owners in the district muster their herds in weekly or two-weekly intervals to receive general veterinary extension services. Failure of owners to present the injected cattle for test interpretation three days later is among the most common causes of the limited efficacy of BTB control in those areas. These factors constitute a high financial burden for governments and render BTB testing in developing countries less efficient and affordable.

The development of the gamma interferon (IFNγ) assay as an ancillary test for BTB diagnosis has improved the sensitivity of BTB testing (Wood et al., 1991). Cattle with early M. bovis infections are more readily detected by the IFNγ assay than the IDT (Neill et al., 1994) and parallel interpretation of both tests exceeded their individual diagnostic sensitivities (Whipple et al., 1995). The achieved specificity of approximately 96% was initially considered sufficient for BTB control purposes in cattle and could not be increased further without compromising the test's sensitivity (Wood et al., 1991, Buddle et al., 2001). As the incidence of M. bovis declined, a need for improved specificity arose and alternative antigens, i.e. early secreted antigenic target 6 kDa protein (ESAT-6) and culture filtrate protein (CFP)-10 were introduced in some laboratories to replace the tuberculin purified protein derivatives (PPDs), often at a loss in sensitivity (Buddle et al., 2009). As an objective, laboratory based test the IFNγ assay is not designed to consider differential interpretation for infected and uninfected herds, respectively.

Once bovine tuberculosis has established itself in an indigenous wildlife population it is difficult to control and probably impossible to eradicate. Despite its status as maintenance host for M. bovis the African buffalo is of high commercial and ecological value and diagnostic tools used towards BTB control are required to offer maximum sensitivity and specificity (Michel et al., 2006). We have previously observed false positive test results in free-ranging buffaloes when using the standard protocol for the gamma interferon assay. We have further established that false positive test results may be caused by sensitisation of the animals with environmental mycobacteria (Michel, 2008a). Subsequently the commercial assay was modified into a triple comparative test set-up. In addition to the standard test format based on stimulation of whole blood with bovine and avian tuberculin PPD, IFNγ produced in response to sensitin derived from M. fortuitum (fortuitum PPD), was explored. The results suggested that fortuitum PPD could be of potential value in detecting non-specific sensitisation in cattle and buffalo, hence possibly allowing improved test specificity in uninfected herds and populations. This modified IFNγ test protocol has been in use in the Kruger National Park (KNP) and other projects since 2000, but has to date not been formally validated. It was therefore the aim of this study to use data sets generated from the field application of the IFNγ assay in buffalo to determine measures to predict the BTB status and subsequently to improve test validity for African buffaloes by determining the most appropriate cut-off value(s) for the IFNγ test under local conditions. Our validation analyses furthermore include a comparison of the standard IFNγ test protocol with the modified protocol including fortuitum reactivity, and protocols optimized based on our findings in this study.