ETIOLOGY

The TB complex includes Mycobacterium tuberculosis, M. bovis, M. africanum, M. microti, and M. pinnipedii.^11,¹⁴ M. tuberculosis is the predominant cause of TB in humans and elephants, whereas M. bovis is the most common cause of TB in domestic animals and wild mammals.³⁰ M. microti is primarily found in small rodents (voles) and hyraxes but has also been isolated from llamas, pig, and ferrets. M. africanum is a rare cause of TB in humans, cattle, and pigs.

Mycobacterial classification has typically relied on biochemical and phenotypic characteristics of the organisms. These bacteria are slow growing and take up to 8 weeks to appear on Löwenstein-Jensen media cultured aerobically at 37°C. Culture morphology varies from coccoid to filamentous, and microscopically the rod-shaped bacteria are 0.2 to 0.6 μm × 1.5 to −3.0 μm. Presumptive identification of mycobacteria may be made by demonstrating acid-fast staining characteristics using Ziehl-Neelsen or the Kinyoun staining techniques with carbolfuchsin. In addition to biochemical differentiation, deoxyribonucleic acid (DNA)–specific probes have been developed to provide speciation.^1,³⁹ Strains have also been identified within species using restriction fragment length polymorphism (RFLP) of identified sequences, spoligotyping, and DNA sequencing.^30,³³

DIAGNOSTIC TESTS

No antemortem test is 100% reliable for detecting TB in zoo animals. The approach to routine screening and clinical examination of suspect cases requires application of multiple testing modalities. It is important to realize that most tests are not validated in zoo animal species, and those based on immunologic responses especially may show significant variability among species. As technology and knowledge expand, the ability to interpret these tests will increase, but until then the clinician using these diagnostic methods is advised to use caution and understand the potential limitations of each test. A brief synopsis of current diagnostic test modalities follows; the reader is advised to refer to more extensive literature reviews on the subject.

Testing Based on Detection of Mycobacterial Organisms

Diagnostic tests that identify the mycobacterial organism, or components, are the most definitive method of detecting infection. Culture and speciation is considered the “gold standard” and also takes the longest to obtain results (up to 8 weeks, or more for speciation). Even in human cases, infection is only demonstrated in 50% of adult cases by proof of bacilli in biologic samples.³⁸ Site of infection, intermittent shedding, and difficulty of obtaining samples from some species may lead to decreased recovery of organisms. Laboratories with expertise in mycobacterial culture should be chosen when submitting samples. If treatment is being considered in highly valuable or endangered individuals, culture is necessary for identification and antibiotic sensitivity testing. Improved culture methods, such as at BACTEC, Septi-Chek, MB/BacT systems, and mycobacterial growth indicator tubes (MGITs), have the potential to decrease time to detection of growth and increase rate of recovery.³¹

Direct staining of sample material may provide presumptive identification as acid-fast bacteria, but there are also nonmycobacterial organisms, such as Nocardia, that may stain positive. Immunohistochemical staining of tissues is also useful for antemortem diagnosis in limited cases in which biopsy or other relevant samples (e.g., lymph node) may be available. Labeled monoclonal antibodies may confirm acid-fast organisms in tissues as being mycobacteria.

Amplified M. tuberculosis direct test (MTD) and multiplex polymerase chain reaction (PCR) assays may provide rapid results by detecting nucleic acid from the organism in clinical samples.^33,³⁹ Gene probes are used for rapid identification of mycobacterial isolates, whereas the gene amplification methods such as PCR are used to aid in identification of species as well as to test culture-negative samples.^30,³⁹ By choosing the appropriate primers, PCR tests may distinguish between M. tuberculosis complex and M. avium.

PCR may also be performed on postmortem samples, including formalin-fixed tissues.³⁹ A combination of techniques was compared for postmortem detection of M. bovis in white-tailed deer (Odocoileus virginianus). Histopathology had a positive predictive value (PPV) of 94%, acid-fast staining had a PPV of 99%, and application of an M. tuberculosis group-specific genetic probe had 100% PPV compared with mycobacterial culture.²⁰

Secreted antigens from proliferating mycobacteria have been the focus of recent diagnostic research. Antigen 85 (Ag85), produced during active infection, has been detected in sera using dot blot immunoassay. Nyala (Tragelaphus angasi) with pulmonary granulomatous lesions had elevated values of Ag85 compared to those with no history of exposure to M. bovis.³⁶ However, similar tests on orangutans showed equivocal results.³² Serum Ag85 could be used as an adjunct test but appears to require further validation in each species.

Testing Based on Immunologic Response to Mycobacteria

Cell-Mediated Immunologic Tests

The most common diagnostic test for TB in mammals is the intradermal test, based on in vivo, delayed-type hypersensitivity response to tuberculin antigens. Purified protein derivative (PPD) tuberculins prepared from M. bovis and M. avium are used for single and comparative testing, particularly of ungulate species.³⁰ The standard dose is 0.1 mL (5000 tuberculin units) in mammals, injected intradermally, usually in the caudal tail fold, skin of the cervical region, or upper eyelid of primates. Other sites used include the lateral thorax, axillary region, abdomen, and ear. Old tuberculin (OT), prepared from either M. tuberculosis or M. bovis, has historically been used in primates and zoo ungulates but has been phased out because it is more difficult to standardize between lots and is less specific. Currently, most PPD tuberculin is produced at a protein concentration of 1 mg/mL.³⁰ Ideally, injection sites are measured with calipers at initial injection and again after 48 hours in nonhuman primates and swine or after 72 hours in ungulates. Specific criteria for “negative” and “suspect” have been developed only for a few nondomestic species, including some cervids. If swelling is present, additional diagnostic testing, including a comparative cervical test (CCT), is warranted. Ancillary tests, such as the interferon-gamma (IFN-γ) test, have been approved in the U.S. federal eradication program for domestic cattle to replace or augment the results of CCT. The basis of the CCT is that there will be a differential response to M. avium and M. bovis PPD based on whether the animal is infected with M. tuberculosis complex or has had a transitory sensitization from nontuberculous mycobacteria.

Intradermal testing is fraught with problems, including anergic responses in individuals with fulminant disease, species and individual variability in response, and false-positive and false-negative reactions. Even in humans, the positive predictive value for tuberculin skin test varies with infection prevalence in the tested population, with at least a PPV greater than 75% in which infection prevalence was above 10%, but decreased PPV in populations with lower prevalence.³ Certain zoo species are known to have an increased likelihood of nonspecific reactions, including tapirs, bongo antelope, reindeer, and orangutans. To address these issues, the use of purified antigens in vivo and in vitro is being investigated in a variety of species.

Diagnostic tests based on in vitro cell-mediated immune responses to mycobacteria include lymphocyte transformation, cytokine production (i.e., IFN-γ, interleukin-2), and other indirect measures of immunologic stimulation, such as cytokine ribonucleic acid (RNA) assays. Lymphocyte transformation (LT) tests are performed by stimulating mononuclear cells with specific antigens and then incubating the proliferating cells with a radioisotope-labeled nucleotide. The amount of label incorporated is correlated with the degree of proliferation and is an indicator of previous exposure and immune recognition of the specific antigen. The LT assay was part of the blood tuberculosis (BTb) test developed to overcome the problems associated with skin testing and was used as an ancillary test for U.S. deer in the 1990s.¹² A similar comparative lymphocyte stimulation test developed for M. bovis–infected Eurasian badgers (Meles meles) using bovine and avian tuberculins showed 87.5% sensitivity and 84.6% specificity.¹⁷

Assays that measure cytokine production, such as IFN-γ and interleukin-2 (IL-2), appear to be more sensitive than skin tests. Cytokines are generally more conserved between species, so detection methods may be more widely applicable. For example, the immunoassay developed for human IFN-γ was able to detect chimpanzee, orangutan, gibbon, and squirrel monkey IFN-γ and correlated with in vivo tuberculin skin reactivity.¹⁹ This test was commercially available as Primagam (CSL Veterinary, Australia) for use in gorilla, orangutan, chimpanzee, gibbon, guereza, mandrill, squirrel monkey, marmoset, and baboon. A similar assay was produced for cattle (Bovigam), deer (Cervigam), and humans (Quantiferon). The IFN-γ test has been used with African buffaloes (Syncerus caffer) to aid in a test and cull program for bovine TB in Kruger National Park, South Africa.²⁴ Necropsy and culture results were used to confirm field cases, and the specificity of the IFN-γ test was shown to be 99.3%. Recent research investigating other cytokine production (e.g., IL-2) or cytokine RNA may provide additional in vitro methods of assessing response to mycobacterial infection across a range of species.⁴⁴ Difficulties associated with using these assays include (1) specific culture parameters need to be developed for each species, and (2) whole blood needs to be properly handled for accurate test results. Many of these tests are not currently available on a commercial basis.

Serologic Tests

Enzyme-linked immunosorbent assay (ELISA) has been the most frequently used serologic test for TB diagnosis. These assays incorporate various forms of mycobacterial antigens for detection of antibodies in the test sample and also are a component of the BTb test. In one study of 12 cervid herds, the specificity and sensitivity of a five-antigen ELISA were 78.6% and 70.0%, respectively.²¹ The ability to diagnose TB increased if ELISA and tuberculin skin test results were used in parallel, rather than using either test alone.

ELISA has been used to evaluate M. bovis infection in brushtail possums (Trichosurus vulpecula) in field tests.⁶ The sensitivity and specificity of the assay using M. bovis culture filtrate was 45% and 96%, respectively, and the results were 21% and 98% when the antigen was MPB70. Further study showed that M. bovis– infected possums develop antibody late in the course of disease that may affect the sensitivity of serologic diagnostic tests for this species. This underscores the importance of understanding the immunologic response to TB in each species and the potential limitations of serologic assays.

With the development of purified, recombinant, and fusion proteins, tailored antigen panels may be developed to change specificity and sensitivity of serologic tests. In addition, other methods may be employed, such as Western blot (immunoblot), thin-layer immunochromatography, and multiantigen print immunoassay (MAPIA). Immunoblot has been demonstrated to be a sensitive method to detect and monitor development of serologic response to specific mycobacterial protein antigens in a variety of species.⁴⁹ Immunodominant antigens may be identified and used for development in other serologic assays, such as ELISA or immunoblot. MAPIA entails application of antigens to nitrocellulose membranes, followed by incubation with test sera and detection using standard chromogenic immunodevelopment.³⁵ MAPIA has been useful in choosing antigens appropriate for a rapid test that utilizes thin-layer immunochromatography and may provide a diagnostic screening test for field situations.²³

In a study comparing serologic and cell-mediated responses to M. bovis in reindeer, antibody could be detected as early as 4 weeks after experimental infection.⁴⁹ Animals tested positive using multiple serologic tests but showed individual variation in antigen recognition at different time points. MAPIA appeared to be most sensitive and detected antibodies earliest after infection at 4 weeks, immunoblot at 8 weeks, and ELISA at 15 weeks. When compared with IFN-γ and skin test responses, all the infected reindeer tested positive by CCT at 3 and 8 months after infection, but no correlation was found between skin test reaction and level of antibody. Similarly, there was no correlation between antibody levels and IFN-γ response. This study shows the potential diagnostic value of serologic tests in a species that has a low prevalence of disease and a high number of nonspecific reactions with skin testing.