Estimating adrenal gland size from CT scans in dogs having adrenal

Abstract

Canine adrenal gland volume can be predicted based on body weight and computed tomography (CT) measurements using a validated algorithm. Use of this algorithm to detect adrenal pathology, including hyperplasia, hypoplasia and neoplasia, in clinical cases has not been described. The objective of this study was to illustrate application of the algorithm by estimating subject-specific adrenal gland volume in a historical cohort of dogs with known adrenal disease. Forty-six dogs that underwent CT and subsequent adrenalectomy were included. Clinical records and CT images from dogs that underwent adrenalectomy and histologic examination of the excised adrenal gland(s) were reviewed. Normal adrenal gland volumes for each dog were estimated using the algorithm, and compared with measured volumes of the affected glands. Linear measurement of the largest lesion diameter was also recorded. Fifty-eight adrenal glands were removed from 46 dogs, with pathology confirmed in all glands. Pathology included 28 adenomas, 13 carcinomas, 11 pheochromocytomas and 6 other benign pathologies. The volume of all removed adrenal glands was measured to be larger than the expected normal volume estimated by the algorithm, ranging from 1.1 to 212.9 times larger than estimated. Adrenal glands with malignant and benign pathology showed variable volumes with overlapping ranges recorded. Assessment of the dimensions of any focal lesion against a cut-off of 20 mm failed to discriminate malignancy. This study illustrates and supports the application of a validated volumetric algorithm for estimation of subject-specific adrenal gland volume to identify the presence of pathology and as a tool to assist clinical decision-making.