Mycobacterial species capable of causing clinical tuberculosis (TB) belong to the highly phylogenetically conserved Mycobacterium tuberculosis complex (MTBC) and are some of the oldest recorded zoonotic diseases known to both human and veterinary medicine, with evidence of tuberculous disease found in mummified human remains dated to circa 2400BC.
By contrast, recognition of companion animal tuberculosis has been a relatively recent phenomenon. Canine tuberculosis caused by the human adapted Mycobacterium tuberculosis species first received significant attention in the UK in the late 1920s and early 1930s when it was shown that companion dogs can act as both spill-over hosts and a sentinel species for human infections. However, with declining human incidence during the 20th century following the introduction of pasteurisation and other public health control measures the recognition of canine disease also declined and is now limited to sporadic individual cases. The first significant recognition of feline tuberculosis as a clinical entity in the UK only entered the veterinary literature in the late 1990s with a discussion of 19 cases. Since then, clinical awareness of feline TB has dramatically increased and UK studies have shown that approximately 1% of feline histopathology submissions show changes indicative of mycobacteriosis and culture results from the Animal and Plant Health Agency (APHA) have shown that, of UK cases where culture was successful, nearly 75% were organisms belonging to the MTBC. In practice, about a third of feline mycobacteriosis cases are caused by MTBC pathogens with the disparity with culture results being accounted for by the bias for MTBC detection by routine mycobacterial culture protocols.
Causative agents and clinical signs
Nine species of mycobacteria comprise the MTBC, those that cause disease in companion animals are Mycobacterium (M.) tuberculosis, Mycobacterium (M.) bovis and Mycobacterium (M.) microti. Tuberculous dogs are most frequently infected with M. bovis and M. tuberculosis with only sporadic and small numbers of confirmed M. microti cases (which have been limited to France to date). Clinical signs are frequently severe and multi-organ system including generalised lymphadenomegaly, chronic coughing, weight loss, vomiting and diarrhoea and polyuria with polydipsia.
Feline tuberculosis, in contrast to canine tuberculosis, is most frequently caused (in rank order) by M. microti and M. bovis, with cats showing strong natural resistance to infection with M. tuberculosis. Cases most typically present with single or multifocal, well demarcated nodular skin masses, frequently around the face, legs and tail base (‘fight and bite sites’). Local lymph nodes are consistently palpably enlarged or the lymphadenomegaly may be generalised. It is not uncommon for cats to present with an incomplete primary complex whereby there is no obvious granulomatous lesion on clinical examination but there is a reactive lymphadenomegaly, most commonly the submandibular or popliteal lymph nodes. Feline tuberculosis most often disseminates to the lungs, putatively haematogenously, to give rise to interstitial lung pathology which is detectable radiographically (see below). However, cats do not show clinical signs of respiratory involvement until advanced stages of disease. Rarer presentations of feline TB include primary gastrointestinal and primary pulmonary forms resulting from ingestion and inhalation of infectious particles respectively.
Routine haematology and serum biochemistry show no pathognomonic changes during tuberculous infections. Counter to clinical reasoning, cases are very rarely found to have a leucocytosis with both lymphocyte and monocyte counts both generally remaining within reference intervals; it is hypothesised that this results from chemotaxis of these cells out of the circulating pool and into tissues where they contribute to granuloma formation. The most frequently identified abnormality on serum biochemical analysis is hypercalcaemia (both total and ionised); this appears to correlate with more severe and disseminated disease states and is believed to result from the activation of vitamin D within macrophages which allows them to raise intracellular ionised calcium concentrations sufficiently to kill engulfed mycobacteria. Previous studies have shown that serum hypovitaminosis D also correlates with disease severity in cats, but how this relates to elevated circulating calcium levels remains unclear.
Thoracic radiography is always indicated in confirmed cases of tuberculosis or where there is a high index of clinical suspicion in order to stage disease progression. Observed pulmonary changes are most frequently a diffuse interstitial, alveolar or mixed pattern, with enlargement of the associated lymph nodes (most commonly sternal lymphadenomegaly). Less frequently, dystrophic mineralisation of the lung parenchyma can occur and bronchial involvement is only seen in the most advanced cases. Computed tomography (CT) has been shown to be more sensitive at detecting early pulmonary infiltration than standard radiography, but its use is currently limited due to the availability of specialist equipment. Thoracic dissemination of tuberculosis is an indication to extend the typical course of treatment.
Abdominal ultrasonography can reveal gastrointestinal involvement of disease (either primary disease form or dissemination); intestinal thickening, mesenteric and/or generalised abdominal lymphadenomegaly are the most routinely identified pathological findings. Gastrointestinal involvement is considered a poor prognostic indicator.
The majority of cases seen in clinical practice will have the diagnosis made by histopathological analysis of a granulomatous lesion; the histological appearance of a typical skin lesion would be large numbers of inflammatory cells within the dermis and subcutis, often associated with ulceration of the overlying epidermis. Locally extensive and present within associated lymph nodes, where they typically form multifocal to coalescing clusters and aggregates, effacing much of the normal architecture of the lymph node.
Special staining with Ziehl-Neelsen (ZN) stains on such histological sections may reveal positive-staining (dark pink) intracytoplasmic bacterial organisms within the cytoplasm of macrophages. Numbers of acid-fast organisms varies widely between cases however, and cannot be used to speciate the causative agent.
Identification of the described features, with or without the presence of acid-fast bacilli, by an experienced histopathologist can be considered diagnostic of mycobacterial often coalescing areas of these inflammatory cells efface parts of the normal tissues, and comprise large numbers of macrophages, some with large nuclei and copious, sometimes vacuolated, cytoplasm. In some cases, macrophages will form small aggregates or clusters and these may themselves have small central areas of necrosis and/ or neutrophils present. In other cases, there may be more extensive areas of necrosis and variable numbers of neutrophils scattered throughout or forming small clusters. Aggregates of lymphocytes and plasma cells may also be present, most often towards the periphery of the lesions.
Similar populations of large macrophages may also be disease (but not which species) in companion animals in the UK.
Culture and PCR
Identification of the causative species is important to discern whether a mycobacterial infection is tuberculous or non-tuberculous. Within the MTBC, the exact species can define the zoonotic risk (see below).
Culture is currently considered to be the ‘gold standard’ diagnostic test, although it fails in ~50 per cent of attempts, even when ZN positive organisms are present histopathologically. When it is successful it can take a long time e.g. M. microti requires a minimum of 12 weeks to culture, during which time treatment must be instigated based on a presumptive diagnosis. Until recently culture was undertaken free of charge by the Animal and Plant Health Agency (APHA) in the UK but continuing financial restrictions have meant that this service is no longer always available depending on geographical area. As an alternative, rapid molecular methods such as PCR diagnosis can provide definitive speciation. PCR diagnosis can be performed on a formalin fixed paraffin embedded (FFPE) tissue section or a fresh tissue submission, and is more frequently successful with the latter. However, even PCR is not always successful, putatively due to the paucibacillary nature of many of the samples submitted and is currently limited in companion animals to the use of commercially available tests used in human patients.
The interferon gamma (IFN-γ) release assay (IGRA) is based on the principle of quantitatively evaluating the amount of IFN-γ produced upon in vitro stimulation from peripherally circulating specific effector T-memory cells in order to aid the diagnosis of both active and latent tuberculosis. It has been widely adopted by human clinicians as it has greater sensitivity and specificity than the tuberculin skin test.
An adapted form of the IGRA test has been validated for use in cats and has anecdotally been useful in dogs; however, test sensitivity currently depends on subjective interpretation of test data and can therefore range from 70-100%. When used to diagnose MTBC only (i.e. non-tuberculous mycobacterial cases were excluded from data analyses) the assay has been shown to have a sensitivity of 100% which coupled with the rapid turnaround time of five days, make this a very useful ante-mortem test.
As with tuberculosis infections in other animals, post-mortem examination to identify granulomatous lesions can confirm the results of ante-mortem testing and allows for multiple samples to be collected and submitted to a range of tests, providing a much greater diagnostic sensitivity. M. bovis infection in the carcass of any non-bovine animal in the UK is Notifiable, and Veterinary Officers of the APHA can request post-mortem examinations to be performed. As of March 2016, the Royal (Dick) School of Veterinary Science & Roslin Institute team has the correct facilities to conduct such post-mortems free of charge and we liaise with the APHA to share the findings of such investigations.
Different species of the MTBC present variable zoonotic risks; M. tuberculosis infection in any companion animal would present a significant zoonotic risk to owners and treatment should not be attempted; the animals should be euthanased.
Currently approximately 1 per cent of human TB cases diagnosed in the UK are caused by M. bovis infection, indicating zoonotic potential. However, as of September 2015, Public Health England, Public Health Wales and Health Protection Scotland all consider the risks to humans to be ‘very low’ and so treatment of companion animals can be considered.
The risk posed to humans from M. microti is lower than that of M. bovis with a very limited number of human cases having been reported and, to date, a lack of any epidemiological evidence linking human infections with companion animals.
Information and advice
The Royal (Dick) School of Veterinary Studies, University of Edinburgh, has a growing team of vets and scientists actively treating and researching companion animal mycobacterial diseases.
The team is headed by Professor Danièlle Gunn-Moore who has extensive experience and expertise in this field. It is able to provide clinical advice on the potential treatment of tuberculosis infections to any vet dealing with a companion animal tuberculosis case, as well as post-mortem examinations of infected carcasses free of charge.
Companion Animal Mycobacterial Disease Referral and Advice Service Royal (Dick) School of Veterinary Science & the Roslin Institute Easter Bush Campus Edinburgh, Midlothian EH25 9RG
Direct email: companion.animalTB@ed.ac.uk