Abstract
This study aimed to establish a comprehensive and accurate numerical chest X-ray radiograph (CXR) scoring system in cynomolgus macaques by using image intensity values from healthy, tuberculosis (TB)-free animals as references. The CXRs were obtained in both dorsoventral and lateral postures from 90 macaques and viewed by the RadiAnt DICOM Viewer software version 2023.1. The mean and maximum intensity values were analyzed and showed significant differences between sex (male and female) and age class (juvenile and subadult/adult), varying based on body sizes. The cutoff values were, therefore, set separately and were tested for accuracy in detecting TB status in 18 naturally Mycobacterium tuberculosis-infected macaques, which were assessed for active tuberculosis infection (ATBI) using Xpert MTB/RIF Ultra at least once during a 12-month follow-up. Only the cutoff values of maximum lateral image intensity (MLIs) correctly identified TB infection in 100% of cases. Thus, the MLIs were selected to follow up on the development of TB lesions in those 18 Mycobacterium tuberculosis-infected macaques. The lateral digital radiograph was divided further into 9 areas, and the MLIs can predict the progression of TB lesions, which were most likely located in the dorsal part of the cranial lung lobe between thoracic vertebrae 1 (T1) to T4. Finally, the CXR results of another group of 8 Mycobacterium tuberculosis-exposed macaques, whose TB status was either uninfected, latent, or ATBI, were compared between a blind test by an expert radiologist and our established CXR scoring system. The blind test results showed 62.5% (5/8) agreement with our scoring system. This suggests that the CXR-MLI scoring system can be used as a supplementary tool for TB diagnosis in cynomolgus macaques.