Companion animal tuberculosis: clinical presentations, outbreak investigations, improved diagnostics and the early macrophage response

Abstract

Tuberculosis caused by the Mycobacterium (M.) tuberculosis-complex (MTBC) of organisms remains one of the most prevalent and deadly infectious diseases of man and other animals. The mycobacteria responsible are a highly conserved group of pleomorphic acid-fast bacilli which cause chronic granulomatous infections. Tuberculous infections in humans and cattle often remain latent for prolonged periods of time before progressing to disease that has severe, negative consequences for the health and welfare of the infected host. Some of the organisms within the MTBC are highly specialised and limited to just a single or small number of host species whereas others, such as M. bovis, can infect a broad range of mammals including humans. Companion animals are susceptible to MTBC infections and understanding of the significance and frequency of these infections has grown in recent years. Cats and dogs share unrivalled proximity to their owners and therefore pose a small but real risk for the zoonotic transmission of tuberculous infections. Despite the high frequency of mycobacterial infections observed in companion animals, diagnostic tests to identify the commonly encountered mycobacterial species are lacking. The first aim of this work was therefore to improve on the currently available diagnostic test methodologies for companion animals. A diagnostic PCR assay was developed and applied to 380 histologically confirmed feline and eight canine mycobacteriosis samples. This novel assay specifically targeted the mycobacterial species most frequently identified by mycobacterial culture (M. bovis and M. microti) and was optimised for use with formalin-fixed tissue; a prerequisite for the safe handling of tuberculous tissue from companion animals by UK laboratories. The assay was suitable for both feline and canine tissue with a significantly quicker turnaround time, higher rate of test positive results and a significant increase in the proportion of M. microti diagnoses compared to culture results. Since evaluation of cytokines has shown diagnostic potential in other species, this project explored the potential of cytokine profiling in cats for the rapid and sensitive detection of mycobacteriosis. By evaluating serum/plasma from 116 naturally infected cats, this study demonstrated a consistent elevation in the cytokines associated with macrophage activation and antigenic stimulation compared to control cats. Sub-group analysis showed that elevations in PDGF-BB were specifically associated with M. microti infections whereas elevated TNF-α sensitively identified cats infected with M. bovis. Investigation of an unprecedented outbreak of M. bovis associated with the ingestion of a putatively contaminated raw food product led to the use of the interferon-γ release assay (IGRA) as a screening test for clinically healthy cats, a purpose for which it had not previously been evaluated. Nearly a third of clinically normal IGRA test-positive cats were subsequently found to have structural disease detected by diagnostic imaging. Though this raises questions regarding the specificity of the IGRA in clinically normal cats, it was an invaluable diagnostic tool to evaluate individual cats involved in the outbreak. The work on feline TB was complemented by similar investigations of canine TB. When an outbreak of M. bovis tuberculosis occurred in a kennel of 164 working Foxhounds, a testing strategy to successfully bring the outbreak under control and investigate the cause was developed. Collaborative work undertaken to screen at risk humans exposed to the hounds identified a latently infected person, highlighting the zoonotic risk posed by M. bovis infections in companion animals. Eight novel and existing diagnostic testing methodologies were evaluated for use in dogs of which a cell-based IGRA, and three serological tests comprising a novel comparative peptide ELISA, the Chembio DPP VetTB assay and the Idexx M. bovis Ab ELISA showed diagnostic potential for canine TB. Additional analysis employing a Bayesian latent class modelling approach revealed that the IGRA developed herein was as sensitive and specific as comparable tests in other species. The serological assays were shown to have markedly lower sensitivity than the IGRA but had higher specificities. All four tests had positive and negative predictive value estimates which indicate that these tests can be informative to clinicians who suspect cases of canine TB. A review of 1012 cases of canine TB highlighted an apparently lower incidence of infections in dogs compared to cats, but an increased severity of clinical signs when disease occurred. To investigate this discrepancy a protocol to derive macrophages from canine and feline bone marrow was developed. These cells acquired cell surface molecules indicative of a macrophage phenotype during ten days of culture with recombinant CSF-1. The response of primary macrophages and the DH82 canine histiocytic cell line was assessed following stimulation with LPS, infection with M. bovis Bacille Calmette Guerin, M. bovis AF2122/97 (the reference strain) or a clinical isolate of M. bovis. These investigations consistently revealed that DH82 cells do not accurately represent primary canine macrophage biology which was associated with altered morphology, lack of nitrite production and significantly reduced secretion of the pro-inflammatory cytokines IL-6 and TNF-α. Overall, the work presented in this thesis demonstrates novel advancement of the diagnostic methodologies for identifying cases of companion animal mycobacteriosis and in particular cases of TB. It further begins to explore the immunological basis for the clinical differences seen between species that may further contribute to novel testing and treatment strategies in the future.