Kirsty Officer ^1,2* Natalie Webster ³ Alana Rosenblatt ⁴ Phorn Sorphea ² Kris Warren ⁵ Bethany Jackson ⁶

• ¹ School of Veterinary Medicine, Murdoch University, Perth, Australia
• ² Free the Bears, Phnom Penh, Phnom Penh Municipality, Cambodia
• ³ Melbourne Animal Specialist Hospital, Melbourne, Australia
• ⁴ School of Veterinary Science, The University of Queensland, Brisbane, Queensland, Australia
• ⁵ Centre for Terrestrial Ecosystem Science and Sustainability, Harry Butler Institute, Murdoch University, Perth, Australia
• ⁶ Centre for Biosecurity and One Health, Harry Butler Institute, Murdoch University, Perth, Australia

Early and accurate diagnosis of pulmonary tuberculosis (TB) is key to effective outbreak management, and in humans thoracic radiography is used extensively for screening purposes. In wildlife TB radiography is a relatively accessible diagnostic tool, particularly in under-resourced settings, however its use is limited by body size. Sun bears are susceptible to human-associated TB, and their small body size makes thoracic radiography feasible. However, there are no established guidelines on normal thoracic radiographs or radiographic manifestations of TB in this species. We provide a first description of thoracic radiographs from healthy and TB affected sun bears at a bear rescue sanctuary, including correlation with postmortem results for a subset of bears. Findings of two veterinary radiologists, blinded to clinical information, revealed high agreement on broad categorisation of radiographic studies as normal, abnormal, or needing correlation with further information. Agreement was lower for the presence of specific lung patterns, reflecting inherent subjectivity when classifying these features. Very few studies were identified as definitively normal, however definitively abnormal studies were significantly associated with TB-cases. Diffuse bronchial and/or bronchointerstitital lung patterns were commonly reported, with a high proportion needing correlation with age and/or clinical signs to further interpret. Interstitial, interstitial-to-alveolar, alveolar and nodular lung patterns, along with radiographic signs of lymphadenomegaly and pleural fluid, were almost exclusively found in TB cases, however the sensitivity of the presence of any of these changes for detecting TB was below 70%. Radiographic reporting of thoracic lymph node enlargement detected at postmortem was low (4/17; 23%), and aortic outflow tract dilation and positional atelectasis were differential diagnoses for radiographic changes that could also represent TB. Together these findings demonstrate the importance of developing species-specific criteria for interpretation to differentiate between common findings and manifestations of TB, and to highlight areas where radiographic techniques can be optimised to assist this. Given TB remains a global health challenge in humans and other animals, and detection is key to control, we recommend development of standardised approaches to radiographic studies and their interpretation to bolster diagnostic pathways for detecting TB in sun bears, and other novel or understudied hosts.