Computed tomographic assessment of vascular invasion and resectability of mediastinal masses in dogs and a cat.

Aims: To assess the sensitivity of non-angiographic contrast-enhanced computed tomography (CT) to determine the presence of vascular invasion of cranial mediastinal masses in dogs and a cat, and to evaluate the association between vascular invasion and peri-operative mortality.

Methods: A retrospective study was conducted on 25 dogs and one cat. CT scans were completed with slices ranging from 2 to 10 mm. CT images were evaluated by a board-certified radiologist blinded to previous diagnoses and surgical findings. Each CT study was evaluated for vascular invasion, defined as disruption of the vessel wall and extension of the mass into the vessel lumen. Data retrieved from the surgery reports included surgical approach, whether vascular invasion was present, the surgeon's decision on operability, and post-operative complications.

Results: Computed tomographic evaluation revealed 25/26 masses had no evidence of vascular invasion. During surgical exploration, 10/26 masses were found to invade major regional vasculature; the cranial vena cava (CVC) was the vessel most commonly invaded (7/10 animals), and 4/7 (57%) patients with invasion of the CVC were euthanised or died in the peri-operative period, from surgical or disease-related problems, which was significantly higher than patients without vascular invasion (p=0.045).

Conclusions: Non-angiographic contrast-enhanced CT was significantly less sensitive for detecting vascular invasion of cranial mediastinal masses when compared with surgical evaluation. If the CVC was invaded by a tumour there was a significant risk of death peri-operatively when compared with non-invasive cases.

Clinical Relevance: Due to the significantly higher mortality risk associated with invasion of the CVC, a more sensitive method than CT should be investigated to determine vascular invasion of mediastinal masses pre-operatively.