The accurate evaluation of sternal lymph nodes (StLNs) is critical for the staging of canine thoraco-abdominal tumours. Computed tomography (CT) provides a non-invasive means of assessing StLNs, but its diagnostic accuracy for identifying metastases is unclear. In this retrospective cross-sectional study, we assessed the diagnostic power of various CT measurements. Fifty-seven dogs that underwent concurrent CT and cytological examination of the StLNs were enrolled retrospectively. The size, shape, X-ray attenuation and uniformity of the StLNs were assessed. The dogs were divided into metastasis-negative (n = 21) and metastasis-positive (n = 36) groups. Logistic regression analysis showed that the size (StLN-to-second sternebra ratio [ratio-size]) and precontrast attenuation were significantly different between groups. Combining these parameters achieved a specificity and positive predictive value of 100% (cut-off values: 1.0, 37.5 Hounsfield units, respectively). This suggests that the combination of ratio-size and precontrast attenuation is effective for differentiating metastasis to the StLNs on CT.
Keywords: CT; Hounsfield unit; diagnostic accuracy; dog; lymphadenopathy; metastasis.
© 2017 John Wiley & Sons Ltd.