Tuberculosis in elephants: a reverse zoonosis

  • Infectious Disease News, December 2008
    Arnon Shimshony, DVM

I recently had a few suggestions to write a “Zoonotic Infections” column addressing diseases of humans that are transmissible to animals. When I subsequently discussed the appropriate terminology for such diseases with a peer veterinary epidemiologist, he said, “Reverse zoonosis is fine by me but only because the medics insist that all zoonoses go from animals to man.”

Indeed, WHO defines a zoonosis as an infection “shared in nature by man and animals.” This is simple and elegant and implies that there is no need to coin terms to define the direction of infection. However, for the sake of clarity, the term “reverse zoonoses” has been applied for the current column about diseases of humans that infect animals.

It is well known, especially among primatologists, that several human infectious diseases are transmitted with some frequency to nonhuman primates – in particular, measles and human respiratory viral infections.

The following list includes selected reverse zoonoses in various animal species:

Arnon Shimshony, DVM
Arnon Shimshony
  • Mumps virus may cause parotiditis in dogs.
  • Infectious hepatitis A virus may cause hepatitis in non-human primates.
  • Corynebacterium diphtheria may cause ulcers on teats and mastitis in cattle.
  • Staphylococcus aureus may cause furunculosis and mastitis in cattle.
  • Streptococcus pyogenes may cause mastitis in cattle.
  • Entamoeba histolytica of human origin may infect cats, causing hemorrhagic diarrhea.
  • Influenza virus (strains A & B) may cause severe respiratory disease in pet rabbits and ferrets.
  • Herpes virus simplex may infect marmosets and tamarins.
  • Salmonella from infected human patients may infect farm and pet animals.
  • Helicobacter has been mentioned as a likely pathogen for cats, mostly subclinical.
  • Brucella melitensis was suspected as possible reverse zoonosis by the exposure of susceptible sheep and goats to infected human urine.
  • Methicillin-resistant Staphylococcus aureus (MRSA) has been suspected, by several researchers, as a reverse zoonosis, mainly in companion animals (dogs, cats).
  • Mycobacterium bovis, though significantly less common in humans than in animals, has been incriminated as a possible reverse zoonosis.
  • Mycobacterium tuberculosis may infect several animal species, mainly deer, dogs and elephants; cat infection is possible but extremely rare.

Tuberculosis in elephants

For this review, we will deal with Mycobacterium tuberculosis in elephants. Sanskrit documents from about 2,000 years ago describe a disease in elephants that is clearly TB. The pathogen has also been implicated in the extinction of the mastodon (Mammut americanum); in one study of mastodon skeletons, 59 of 113 (52%) had bone lesions diagnostic for TB.

The first zoo elephant reportedly affected by TB was an Indian elephant that died at the London Zoo in 1875. Additional cases appeared in the literature in the early 20th Century. The first case of mycobacterial infection in an African elephant was reported in the 1960s.

TB became a disease of health concern for elephants and humans in the United States following the published death of two elephants from a privately owned traveling herd in 1996. In response, the USDA formed an advisory panel to establish diagnostic and treatment protocols for the remaining elephants in the herd. The following year, five new elephant cases were identified, prompting the development of protocols for the surveillance of all elephants. In 1998, the USDA and National Tuberculosis Working Group for Zoo and Wildlife Species released the first Guidelines for the Control of TB in Elephants, establishing culture as the recommended test for the diagnosis of TB in elephants. The guidelines have been revised since.

Between 1994 and 2005, there were 34 confirmed cases of tuberculosis in elephants in the United States: 31 of them in Asian elephants and three in African elephants. In 33 of the cases, the etiologic agent was found to be M. tuberculosis and in one, M. bovis. Twenty-two of the 34 cases were diagnosed antemortem, on the basis of cultures obtained by trunk washes or swabs and 12 cases were diagnosed postmortem. The majority of elephants, including the 12 cases diagnosed at post-mortem, did not show clinical signs suggestive of TB. Of the 34 cases, 19 have died or were euthanized. In some cases, TB was an incidental finding at necropsy and was not considered the cause of death.

Spread of TB

The zoonotic potential for the spread of TB between elephants and humans has been documented in one case. Other reports have demonstrated epizootic spread of the same strain of M. tuberculosis between elephants and other zoo mammals.

As of 2008, about 13% (one of every eight elephants) in the few well-studied populations of elephants in the United States, have been found to be TB-infected.

Alarmingly, at least one case of multidrug-resistant-TB (MDR-TB) has occurred in an elephant in the United States. Because MDR-TB has increased in human populations and in view of the fact that some elephants are being inadequately treated (typically with single-drug therapy), MDR-TB is expected to be found in captive range country elephants. TB-infected elephants are a potential reservoir that could infect humans and wildlife, perhaps even with more virulent and novel forms of TB.

Close-up of a Mycobacterium tuberculosis culture
Close-up of a Mycobacterium tuberculosis culture revealing the organism’s colonial morphology.
Source: CDC/Dr. George Kubica

Elephants in the United States are not exceptional in their susceptibility to TB. In a recent study in southern India involving 387 captive elephants in four states, 59 of the elephants had TB. The largest percent of TB affected elephants were in temples, where 16 of the 63 animals were found infected. Of 160 privately owned elephants, 24 had TB. Of the 164 owned by forest departments of various state governments, 19 had TB. Though an attempt to treat such animals was under study, experts warned that, in view of the fact that the disease is zoonotic in nature – transmitted from man to animal and vice versa – such treatment, replacing the traditionally prescribed culling of the infected animals, deserves cautious consideration.

Clinical signs in elephants

As in humans, TB in elephants – which get infected aerogenically – can be a chronic, debilitating disease. Signs may include weakness, weight loss, exercise intolerance and coughing. Elephants may not show signs until the disease is quite advanced. Even in the absence of clinical signs, infected elephants can shed TB bacteria and infect others.

The current method used to diagnose TB in elephants is to culture samples obtained by a trunk wash procedure, similar to obtaining a sputum sample in humans. Sterile saline is placed in the trunk, the trunk is elevated, and the elephant is instructed to forcibly exhale into a collection device. Barring cross-contamination or laboratory error, a positive culture is an accurate indication of TB disease, but false-negatives do occur. The trunk-wash sample may be from the end of the trunk instead of from deeper in the respiratory tract as intended. Because elephants use their trunks for many purposes including eating and dust bathing, contamination of the sample is common, leading sometimes to delayed or indecisive results.

In a TB outbreak in Sweden, five elephants were affected. Of 189 trunk wash samples collected, only seven were positive from the five elephants that were confirmed (on postmortem) to be infected.

The skin test (a common screening test in humans) has been evaluated and is not accurate in elephants.

In August 2007, the USDA licensed a commercially available kit as a screening serological test, to be confirmed by the Multi-Antigen Print Immuno Assay (MAPIA). This test has reportedly been shown to be positive months to years in advance of a positive culture.


A number of elephants in the United States have been treated, few of them twice. The exact number is not known because some facilities are reluctant to share information. The same drugs that are used in humans are also used in elephants. The main or “first line” drugs are isoniazid (INH), rifampin (RIF), ethambutol (ETH), and pyrazinamide (PZA). PZA is not effective against M. bovis and can only be used if the causative agent is M.tb. A precise (scale) weight is essential to determine accurate drug doses. Isoniazid and PZA can be given orally or rectally; rifampin and ethambutol must be given orally. The duration of the treatment is, generally, one year – longer than in humans while still being experimental.

For more information:
  • Arnon Shimshony, DVM, is Associate Professor at the Koret School of Veterinary Medicine Hebrew University of Jerusalem, Rehovot, and is the ProMED-mail Animal Diseases and Zoonoses Moderator. Dr. Shimshony was Chief Veterinary Officer, State of Israel, from 1974 to 1999.
  • Additional information from the Elephant Tuberculosis Research Workshop is available online at