To the Editor:

A dog lacerated its abdomen and developed a tuberculous abscess while walking in the Irish countryside in an area frequented by deer and badgers and where Mycobacterium bovis is known to be present in cattle and incidental animal hosts (1). The dog was being administered immunosuppressants (cyclosporine and prednisolone) for chronic enteropathy. Microbiological assessment of the abscess revealed M. bovis, and the dog was killed. One of the canine’s companions, a cat, became unwell 6 months later, with a cough and weight loss, after licking the dog’s wound. Ultrasonography and thoracic radiographs revealed pulmonary and hepatic nodules confirmed postmortem as tuberculosis (TB). Whole genome sequencing (WGS) of the two animals’ infections confirmed that the two M. bovis isolates were identical. This strain was derived from WGS isolates seen in a database of M. bovis locally in animal hosts (cattle, deer, and badgers). A second cat in the home was subsequently screened for TB but had no postmortem evidence of TB.

The pets’ owners underwent screening (Household 1). All members of the pets’ household were tested with interferon-γ release assays and a tuberculin skin test. All but one of the pet owner’s family members, and a house-sitting family member (Household 2), had a positive result. All screened humans had normal chest radiographs and were subsequently treated for asymptomatic latent TB infection with rifampicin.

Two years later, a family relative (Household 3) who had visited Household 1 became sick with cough and night sweats. Chest radiography showed an upper lobe cavity, and sputum cultures were positive for M. bovis. WGS confirmed that the species of M. bovis identified in this patient was identical to that identified in both pets. Likely onward transmission from this human M. bovis case to their partner was documented by interferon-γ release assays, and the partner was treated for latent TB infection. Neither person in Household 3 had any other known exposure to TB.

This case demonstrates zoonotic transmission of M. bovis from incidental wild animal hosts to the environment to an immunosuppressed household pet, then to another pet, then to a human (as confirmed with WGS) and another human, as illustrated in Figure 1.

Figure 1. Schematic illustration of transmission of Mycobacterium bovis from a household dog to a household cat to the pets’ owners and two other households. TB = tuberculosis.

Download Figure | Download Powerpoint

[More] [Minimize]

Even though M. bovis has long been associated with zoonotic transmission, it has most frequently been associated with transmission from cows to humans, as well as other larger mammals, including deer and badgers (2). Isolated cases have documented single cross-species transmission from household pets, including cats and dogs, to humans (3, 4).

WGS of mycobacterial disease was first described in 1998 but has become more widely available in clinical practice since 2017, providing crucial information in determining drug susceptibility and epidemiology and in research settings (5). This unusual case of transmission of M. bovis from the environment to canine to feline to human to human, as confirmed through WGS, demonstrates the utility of WGS in an epidemiological setting. To our knowledge, the present case is the first description of confirmed spread of M. bovis among three species in one outbreak.

This outbreak also highlights two other areas of interest with regard to the transmission of mycobacteria from the environment and the potential risks that household pets, particularly immunocompromised pets, provide in onward transmission of disease.

The dog that became sick had no direct contact with wild or farmed animals, but instead developed a local M. bovis infection after cutting her abdomen on a barbed-wire fence outdoors. This supports the concept that mycobacteria are able to survive on surfaces without a host and still prove to be transmissible, as has been demonstrated in the study of transmission of M. bovis among wild deer (6).

Although there is abundant literature regarding the caution that immunocompromised patients should exercise around pets, there is scarce discussion about the risks that immunocompromised pets may pose to healthy pet owners. There are many conditions for which household dogs are iatrogenically immunocompromised, with modern treatments for inflammatory conditions in pets including corticosteroids, as well as chemotherapeutic agents and biologic agents such as tumor necrosis factor-α blockers (7). This outbreak shows how an immunocompromised pet can be the conduit for a zoonotic infection in cohabiting humans. It raises the issue that immunocompromised pets may potentially pose a risk to the health of their owners and other close contacts.