Tuberculosis in Elephants—A Reemergent Disease

Abstract

Tuberculosis (TB) in elephants has been described since ancient times. However, it was not until 1996 when infection with Mycobacterium tuberculosis was identified in a herd of circus elephants that significant research into this disease began. The epidemiology and natural history of TB were unknown in elephants since there had been no comprehensive screening programs, and diagnostic techniques developed for cervidae and bovidae were of unknown value. And, while precepts of test and slaughter were the norm for cattle and deer, this was considered untenable for an endangered species. With no precedent for the treatment of TB in animals, treatment regimens for elephants were extrapolated from human protocols, which guided changes to the Guidelines for the Control of Tuberculosis in Elephants. In the absence of diagnostic testing to confirm cure in elephants, the efficacy of these treatment regimens is only beginning to be understood as treated elephants die and are examined postmortem. However, because of pressures arising from public relations related to elephant husbandry and the added considerations of TB infection in animals (whether real or imagined), sharing of information to aid in research and treatment has been problematic. Here we review the challenges and successes of the diagnosis of tuberculosis in elephants and discuss the natural history of the disease to put the work of Landolfi et al on the immunological response to tuberculosis in elephants in perspective.

Tuberculosis (TB) in elephants has been well described since a case report was published in the transactions of the Zoological Society of London, ³ with more ancient references to a disease consistent with tuberculosis appearing in Vedic literature. ⁶ However, it was not until 1996 when infection with Mycobacterium tuberculosis was identified in a herd of circus elephants ² that significant research into this disease was begun. The epidemiology and natural history of TB were unknown in elephants since there had been no comprehensive screening programs, and diagnostic techniques developed for cervids and bovids were of unknown value. And, while precepts of test and slaughter were the norm for cattle and deer, this was considered untenable for an endangered species. With no precedent for the treatment of TB in animals, treatment regimens for elephants were extrapolated from human protocols, ^11,12,21,27 which guided changes to the Guidelines for the Control of Tuberculosis in Elephants (2010). ²⁵ In the absence of diagnostic testing to confirm cure in elephants, the efficacy of these treatment regimens is only beginning to be understood as treated elephants die and are examined postmortem. However, because of pressures arising from public relations related to elephant husbandry and the added considerations of TB infection in animals (whether real or imagined), sharing of information to aid in research and treatment has been problematic.

Here we review the challenges and successes of the diagnosis of tuberculosis in elephants and discuss the natural history of the disease to put the work of Landolfi et al ⁸ on the immunological response to tuberculosis in elephants in perspective.

Epidemiology of Infection

Over the 19-year period between 1994 and 2013, there were 57 culture-confirmed cases of TB in elephants in the United States, and additional animals yielded positive tests on an array of commercial and investigative methods. Mycobacterium tuberculosis was the causative agent in all but 1 case. These cases comprise a lifetime prevalence (defined as the proportion of a population that at some point in their life have experienced the condition) of 12.4% among a population of approximately 446 elephants. And, among the subpopulation of Asian elephants for the years 1994 to 2011, the lifetime prevalence was estimated to be 16.4% (45/247), a rate approximately 6 times greater than that for African elephants. ¹⁹ In comparison, the incidence of TB among humans has declined from 7.9 to 3.0 cases per 100 000 population from 1996 to 2013 (http://www.cdc.gov/tb/statistics/) for an imputed lifetime prevalence of less than 1%.

Analysis of the “index herd” (ie, the herd of circus animals that sparked the current focus on elephant TB) is of interest. In 1996, 2 animals were diagnosed with TB at necropsy, and a third elephant from the same herd was found infected on a culture of trunk wash. ¹⁴ Genetic fingerprinting of the infecting strains showed these to be the same and also matched that of an elephant from the same herd that died in 1994, as well as matching the strain cultured from an elephant handler who was diagnosed with active TB. ¹³ An elephant from the same herd was reported to have died of TB a decade earlier. ²³ Interestingly, the cumulative attack rate among the index herd has now exceeded 50%, with 4 animals dying of or with active TB.

Subsequently, 5 culture-confirmed cases were diagnosed at 4 additional US facilities the following year. And, over the next 2 years, more generalized testing of elephants in captivity across the United States identified 5 additional elephant herds infected with strains of M. tuberculosis. ¹⁴

International screening programs have since identified infection across a total of 4 continents—the only exceptions being Antarctica, South America, and Africa. In Europe, disease was first documented at the Kolmarden Zoo in Stockholm. ^9,15 In Asia, sero-surveys have been conducted in India, Nepal, Laos, Thailand, and Myanmar. Culture-confirmed cases have been identified in India, Nepal, and Thailand, and all cases to date have been due to M. tuberculosis. In areas of Asia where captive and wild elephants intermingle, the threat of introducing TB into fragmented, endangered populations is a serious concern. In Australia, TB occurred in an Asian elephant imported from Thailand and subsequently spread to chimps housed 110 meters away. ²⁴

The first case of M. tuberculosis in an ex-captive wild African elephant was found in Kenya. ¹⁷ M. tuberculosis has also been isolated from at least 2 wild Asian elephants in India. ²⁶

The ultimate source of infection is presumed to originate from humans who care for elephants; however, elephant-to-elephant spread is certainly possible, and the role of fomites is unknown. For the initial descriptive epidemiologic study, ¹⁴ genotypic analysis of infecting isolates demonstrated that animals within each herd shared common strains and also that there was animal-to-animal spread of infection between California herds, whereas other herds were infected with genetically distinct strains of M. tuberculosis. ^13,20 Similar findings were demonstrated for European herds as all of the animals in the Stockholm zoo were infected with an identical strain. ⁹

Zoonotic Considerations

Zoonotic spread of TB had been postulated from a variety of animals (including elephants) at the Los Angeles zoo based on evidence of high rates of skin test reactivity among staff with greater levels of direct animal contact. ¹⁸ However, testing of staff occurred 2 years following identification of TB in the zoo animals and, coupled with the lack of baseline PPD (purified protein derivative) testing, limits conclusive proof of causation. ¹⁸ Outbreak analysis of staff caring for the Illinois index herd was performed within a few months of the identification of TB in elephants and repeat testing performed 3 months later. ¹³ Initial testing demonstrated a high rate of skin test reactivity among animal trainers consistent with a number of staff who had emigrated from countries with high basal TB rates and/or had other epidemiologic risks for TB. Repeat intradermal testing showed that 3 had skin-test conversions consistent with TB transmission along with the 8 staff who were PPD positive at baseline. ¹³ In addition, 1 elephant trainer was found with early active pulmonary culture-positive tuberculosis with a strain that was genetically indistinguishable from the elephant strains at the facility, although the lack of preculture skin test data in the infected trainer precluded definitive determination of the direction of transmission—elephant to human vs human to elephant. ¹³ Definitive evidence of elephant-to-human transmission of infection was demonstrated by epidemiologic analysis at the Elephant Sanctuary in Tennessee that housed many of the index herd animals. Prior to transfer of the infected herd, all staff had baseline PPD skin testing. Retesting after relocation of the animals to the sanctuary showed skin test conversions from negative to positive in 9 staff. The risk of a positive test correlated with increased animal contact and time in close, enclosed quarters with infected animals or in office areas that shared airspace with the barn housing the infected animals. ¹⁶ The authors are aware anecdotally that many if not most institutions with TB elephants have also had human PPD conversions, although most of these cases have not been published.

Diagnosis

Compared with humans, the options to diagnose TB in elephants are limited. Clinical signs are typically absent until the disease is well advanced, chest radiographs are not feasible in adult elephants, and the results of intradermal skin testing has not correlated with culture. ^4,9,14,15 While a positive culture can confirm disease, a negative culture does not rule it out. The trunk wash, the elephant equivalent of a sputum sample, has poor sensitivity—due to inefficient sample collection, contamination of samples by microbial contaminants in the trunk, or low and/or intermittent bacterial shedding. An illustrative example is that for the outbreak in Sweden; only 7 of 189 trunk wash samples sequentially collected from 5 elephants diagnosed with TB were culture positive. ¹⁵ Similar findings have been reported from Thailand ¹ and from the authors’ own observations (unpublished data).

Interestingly, elephants develop unusually robust antibody responses to M. tuberculosis, particularly in the later stages of disease, and the sensitivity of serology to diagnose TB in elephants is significantly higher than for any other mammalian host. ^5,10 Sero-diagnostic tests (the ElephantTB STAT-PAK, Dual Path Platform [DPP] VetTB, and multiantigen print immunoassay [MAPIA], ChemBio Diagnostic Systems, Inc., Medford, NY) have thus proven valuable in the diagnosis of TB in elephants. ^5,10

TB Politics

Worldwide, captive elephants are used by humans—as draft animals, for processions, in temples, for entertainment in circuses, and for education and conservation in zoos. The innate stigma and fear associated with a diagnosis of TB has implications for the owners of these venues. On learning that his elephant has TB, a private owner may sell his elephant to an unsuspecting buyer. Commercial operations like circuses may experience a loss of revenue from public fear and may lobby against TB-related regulations. Perhaps the worst-case scenario would be for infected elephants in range countries to be released into the wild and spread TB to wild elephants or other endangered species such as rhinos.

Immunology of TB and Immunologic Correlates of Disease in Elephants

Control of TB in humans is reliant on a robust Th1 immune response with granuloma formation dependent on sufficient CD4+ T-cell numbers and activity and expression of inflammatory cytokines, including tumor necrosis factor α (TNF-α) and interferon γ (IFN-γ). ²² Conditions and treatments associated with CD4+ T cell depletion and decreased cell-mediated immunity, such as AIDS, corticosteroid treatment, and TNF-α inhibitors used for the treatment of rheumatologic disease, are associated with a significant risk of reactivation disease and dissemination. Significantly less is known regarding the immune correlates of infection for elephants.

Elephants have been shown to develop robust B-cell responses associated with infection with TB. ^5,10 Primary responsiveness to the M. tuberculosis–specific antigens ESAT-6 and CFP10 proteins was recognized by elephants during active disease ^5,10 and demonstrated 100% sensitivity and 95% to 100% specificity in diagnosis. ⁵ Moreover, sero-reactivity was used as part of the commercial assay to diagnose and predict subsequent disease in the Stockholm Zoo outbreak. ⁹ However, reactivity during infection has an unknown correlation to the pathogenesis of infection.

Landolfi and colleagues ⁷ examined systemic cytokine levels in cryopreserved serum samples from 106 Elephant TB STAT-PAK seropositive and seronegative animals. Seropositive animals had statistically higher levels of TNF-α and transforming growth factor beta (TGF-β) with nonsignificant trends toward higher levels of IFN-γ and interleukin (IL)–4 with slightly lower levels of IL-10 and IL-12.

In the accompanying article, Landolfi and colleagues ⁸ have provided additional information regarding the local immune response to infection with M. tuberculosis in Asian elephants. They compared granulomatous lung lesions from 9 culture-positive animals with lung tissue from 5 culture-negative animals that died of other causes. Among the 9 culture-positive animals, 6 demonstrated poorly formed granulomas and 3 had well-defined granulomas and/or granulomatous responses with significant central necrosis. Poorly formed granulomas were characterized by the presence of foamy macrophages and few multinucleated giant cells within the granulomas. Lymphocyte populations were predominately CD20 positive (B cells), with small fractions of CD3-positive cells (T cells) in poorly formed granulomas and among necrotic lesions. In contrast, well-formed granulomas were characterized by greater fractions of CD20 and CD3+ cells. In situ hybridization for TNF-α, IFN-γ, IL-4, IL-10, and TGF-β demonstrated expression associated with granulomatous lesions in a third of cases for both TNF-α and IFN-γ and half of cases for IL-4, but had little correlation to the histologic appearance of the granuloma, although the 1 case with well-formed granulomas had the most robust TNF-α staining. Detection of cytokines in samples, however, was not consistent as most cases were positive for only 1 or 2 cytokines. Hybridization did not detect cytokine presence in non-TB specimens. Thus, these data may provide some basis for the serologic responsiveness seen in elephants and suggests that immunologic containment of TB in elephants requires a Th1 response similar to that observed in humans. However, as the sample size is small, further study is needed to fully delineate the natural history of infection in relation to the immunologic profile and whether loss of cytokine induction is associated with loss of control of containment and reactivation. Landolfi et al ⁸ have, however, provided significant new species-specific immunological tools with which to build a fuller understanding of this devastating disease affecting a keystone species in conservation.