Integrated Basic Heart and Lung Ultrasound Examination for the Differentiation between Bacterial Pneumonia and Lung Neoplasm in Dogs—A New Diagnostic Algorithm

SIMPLE SUMMARY: Dyspnea is a highly alarming sign both for dog owners and veterinarians. Although its recognition is usually easy due to prominent suffering of an animal, finding its cause is challenging because many diseases of the heart, lungs, and airways may manifest themselves this way. Echocardiography and lung ultrasound allow for relatively quick and accurate identification of heart diseases. Dyspneic dogs without a heart and upper airway disease are usually suspected of either bacterial pneumonia or lung neoplasm. Although prognosis in these two conditions is diametrically different, differentiation between them is challenging. Chest radiography is performed in a lateral position, which is barely tolerated by a dyspneic dog, and intensive chest movements often make X-ray scans inconclusive. Computed tomography, although much more accurate, requires general anesthesia, which is difficult and potentially life-threating in a dyspneic dog. Therefore, lung ultrasound, which can be performed quickly in a conscious dog, standing or in sternal position, seems to be the method of choice. We develop and evaluate a diagnostic algorithm based on detection of three well-defined abnormalities in the lung ultrasound. The algorithm allows one to distinguish between bacterial pneumonia and lung neoplasm in a dyspneic dog with high probability of a conclusive result (91%) and high accuracy (>95%). ABSTRACT: The diagnostics of two of the most prevalent lung diseases in dogs, bacterial pneumonia (BP) and lung neoplasm (LN), are challenging as their clinical signs are identical and may also occur in extrapulmonary diseases. This study aims to identify ultrasonographic criteria and develop a lung ultrasound (LUS)-based diagnostic algorithm which could help distinguish between these two conditions. The study is carried out in 66 dyspneic dogs in which a heart disease was excluded using echocardiography. Based on imaging and laboratory diagnostic tests, as well as follow-up, the dogs are classified into LN (35 dogs) and BP (31 dogs) groups. LUS is performed at admission and the presence of seven lung abnormalities (pleural thickening, B-lines, subpleural consolidations, hepatization with or without aeration, nodule sign and mass classified together as a tumor, and free pleural fluid) and classification and regression trees are used to develop an LUS-based diagnostic algorithm. Distribution of all LUS abnormalities except for aerations differs significantly between groups; however, their individual differentiating potential is rather low. Therefore, we combine them in an algorithm which allows for definitive classification of 60 dogs (91%) (32 with LN and 28 with BP) with correct diagnosis of LN and BP in 31 dogs and 27 dogs, respectively.