TB or not TB?

Abstract

We all know that mycobacterial infections are a global health concern, both in humans and animals (Glaziou et al. 2009, Lobue et al. 2010). Globally, one third of all humans are infected with Mycobacterium tuberculosis (WHO 2010), and Mycobacterium bovis infection is a significant concern in many countries; in cattle and badger in Great Britain (GB) and Ireland (Gallagher and Clifton-Hadley 2000, Abernethy et al. 2013); in deer in parts of the US (Walter et al. 2012); in common brushtail possums in New Zealand (de Lisle et al. 2002); and in cattle and many endangered species in Africa (Renwick et al. 2007). In GB, tuberculosis (TB) in animals is not confined to cattle and badger; it is also a serious concern in deer, South American camelids (Crawshaw et al. 2013) and cats (Gunn-Moore et al. 2011a). Cats can be infected with many different mycobacteria. These infections are generally divided into three different syndromes: (1) TB, being caused by M. bovis and Mycobacterium microti (the vole bacillus – found commonly in voles and mice, and cats in GB [Cavanagh et al. 2002, Burthe et al. 2008, Smith et al. 2009, Gunn-Moore et al. 2011a]); (2) feline leprosy syndrome (FLS), which is caused by mycobacteria that are difficult to grow so their environmental niche is unclear e.g. M. leprimurium; and (3) non-tuberculous mycobacteriosis (NTM), which is caused by facultatively pathogenic opportunistic saprophytes, e.g. Mycobacterium avium (Gunn-Moore 2010). However, regardless of which mycobacteria are involved, the majority of infected cats present with cutaneous disease (nodules, draining tracts and/or ulceration) and/or regional lymphadenopathy, commonly submandibular lymphadenopathy (Gunn-Moore et al. 2011a). While disease can sometimes generalise secondary to skin inoculation (typically spreading to the lungs, and resulting in dyspnoea and sometimes a soft cough) (Bennett et al. 2012), primary systemic disease is seen only occasionally (Gunn-Moore et al. 2010). The prevalence of the three syndromes varies in different countries. All three occur in GB, with approximately 1% of feline tissue samples submitted to diagnostic laboratories for routine histopathology having changes consistent with mycobacteriosis (Gunn-Moore et al. 2013). When samples from GB are cultured M. microti is identified in 19%, M. bovis in 15%, M. avium in 7%, Mycobacterium malmonoense in 1% and unclassified mycobacteria in 4%. Unfortunately, depending on the culture system used, up to 50% cannot be grown, making the identification of these mycobacteria (e.g. Mycobacterium leprimurium, which is one of the causes of FLS) particularly frustrating. In addition, even when they do grow, identification can take months (e.g. up to 3 months for M. microti [Smith et al. 2009]). While TB is identified frequently in GB, Ireland and sometimes New Zealand (Greene and Gunn-Moore 2012, de Lisle et al. 2002), NTM and FLS are identified more commonly in Australia and in parts of the US (Kunkle et al. 1983, Malik et al. 2000, 2002, Jang and Hirsh 2002, Malik et al. 2006a,b, Horne and Kunkle 2009, Malik et al. 2013). Identifying exactly which mycobacterial species is involved is essential in determining the potential source of infection, possible zoonotic risks, whether a notifiable infection is present (M. bovis infection is notifiable in GB [DEFRA 2013]), best treatment options, and likely prognosis. Importantly, until the species has been identified it is not possible to tell whether the cat has TB, FLS or NTM. In GB, TB is likely in 34% of cases, with any particular case having an approximate 15% chance of being M. bovis. However, this is very geographically dependent, with most cases of M. bovis infection being seen in cats from Wales and the south-west of England (Gunn-Moore et al. 2011a), i.e. in the same areas where M. bovis is found in cattle and badger (Gallagher and Clifton-Hadley 2000). M. bovis-infected cats are rare elsewhere. M. microti is found in cats from the south of Scotland, the north of England and the south-east of England (Gunn-Moore et al. 2011a). It is interesting that the two infections occur in such distinctly separate areas. It is currently thought that most cats are infected with M. bovis or M. microti when they are bitten by small infected rodents during hunting (Gunn-Moore 2010). These cats are typically hunters (Gunn-Moore et al. 1996, Greene and Gunn-Moore 2012), and small rodents can be infected with M. bovis (Delahay et al. 2002, 2007) or M. microti (Cavanagh et al. 2002, Burthe et al. 2008). Rodents may become infected with M. bovis when visiting contaminated badger setts: the M. bovis spoligotypes found in cats are the same as those in cattle, badger and small rodents in any particular area (Monies et al. 2006, Delahay et al. 2007). Cats could also be infected with M. bovis via environmental contamination from infected badgers as under some circumstances M. bovis can survive for extended periods in the environment (Wray 1975, Morris et al. 1994). While it is possible that badgers could directly infect cats, this seems unlikely given that they rarely interact closely. Drinking tuberculous cow's milk is an unlikely source of infection as most milk is pasteurised in GB and infected cattle are removed early in infection (when detected as reactors) so tuberculous mastitis is rare. M. bovis can also be endemically present in many other species of free-ranging wildlife, so the risk of feline infection may vary in each country depending on the likely interaction between domestic cats and that species (Morris et al. 1994, de Lisle 1992), e.g. M. bovis is endemic in common brushtail possums in New Zealand and these have been found to infect cats (de Lisle et al. 2002). There have been a small number of nosocomial infections, where cats naturally infected with M. bovis have infected other cats via contamination within a veterinary practice, typically during routine neutering (de Lisle et al. 1990, unpublished data). Animals with TB pose a potential, if rare, zoonotic risk. Recently, a number of owners of cats and dogs diagnosed with TB (due to M. bovis) have been found positive for Mantoux test (indicating TB exposure), and a young boy who lived with a dog that was found to be ill due to M. bovis was found positive for Mantoux test and interferon gamma test (indicating active infection), necessitating treatment (unpublished data). M. microti appears to have a much lower zoonotic risk, and no cases of cat to human transmission have yet been identified. Identifying animals infected with TB as quickly as possible is therefore essential if we are to decide the best way forward in any particular case. Unfortunately, diagnosis is challenging. TB, FLS and NTM can all present with similar clinical signs, there are no pathognomonic histopathological changes (Davies et al. 2006, Gunn-Moore et al. 2011b), and serology and intradermal skin tests are not reliable (Greene and Gunn-Moore 2012). The interferon gamma test is proving helpful (Rhodes et al. 2011), but what we really need are rapid, sensitive, specific and, ideally, inexpensive molecular diagnostic tests by which to identify these bacteria. The article in this month's JSAP, titled Detection and identification of mycobacteria in fixed stained smears and formalin-fixed paraffin-embedded tissues using PCR, by Malik et al., describes a technique where mycobacterial species are identified using PCR after being extracted from stained cytology slides or histopathology slides following formalin fixation. In that article, they show that this can be done, with apparent ease, so long as the slides have not been prepared with Ziehl Neelsen (ZN) stain. When this technique becomes widely available it will revolutionise the diagnosis of these globally important infections. In the case of feline infections it will mean that when those approximate 1% of feline tissue samples that are submitted for routine histopathology are found to have changes consistent with a mycobacterial infection (Gunn-Moore et al. 2013), the stained slide or a section from the formalin-fixed block can be sent straight for mycobacterial identification using this new technique. That the same method can also be used on fixed cytology smears is wonderful; many of these cats present with submandibular or other regional lymphadenopathy, and being suspected as lymphoma initially they need fine needle aspiration. This technique would allow the identification of the causal mycobacteria, without necessitating anaesthesia and biopsy. Importantly, this methodology could be used to diagnose all mycobacterial species, not just TB. Being able to identify FLS and NTM is also really important as we currently find these infections very difficult to deal with, partly because the right treatment is dependent on correct identification (Malik et al. 2006a,b, 2013), and this is so often not possible. Using this methodology we could diagnose their presence more readily, identify them and design appropriate treatment plans. As these infections occur all over the world (Kunkle et al. 1983, Malik et al. 2000, 2002, Jang and Hirsh 2002, Malik et al. 2006a,b, Fyfe et al. 2008, Horne and Kunkle 2009), are being increasingly recognised (Malik et al. 2013), and are responsible for over 50% of mycobacterial infections in cats in GB (Gunn-Moore et al. 2011a), it would be beneficial to study this enigmatic group of mycobacteria, and understand more about their aetiopathogenesis, treatment and prognosis. Danièlle Gunn-Moore graduated with distinction from the R(D)SVS, University of Edinburgh, in 1991. After a year in small animal practice she joined The Feline Centre, University of Bristol, initially as the Feline Advisory Bureau Scholar, then as the Duphar Feline Fellow, and completed a PhD study into Feline Infectious Peritonitis in 1997. After a short period as Lecturer in Veterinary Pathology, University of Bristol, she returned to Edinburgh to establish the Feline Clinic and is a Professor of Feline Medicine. She is interested in all aspects of feline medicine; she is an internationally recognised expert in her area, lectures extensively and her work has been published widely. In 2009, she was awarded the BSAVA Woodrow Award for outstanding contribution in the field of small animal veterinary medicine; in 2011, she was awarded the International Society for Feline Medicine/Hill's award for outstanding contributions to Feline Medicine; and in 2012, the R(D)SVS students voted her ‘The Clinician I would most like to be’.