Investigation of two methods for assessing thyroid-lobe asymmetry during pertechnetate scintigraphy in suspected hyperthyroid cats.

Abstract

Our aim was to investigate thyroid:thyroid (T:T) ratio and visual inspection for assessing thyroid-lobe asymmetry in suspected hyperthyroid cats. Although thyroid-salivary asymmetry is a preferred test, inherent thyroid symmetry may assist image interpretation. Association was determined using a scatter plot and Spearman's rank correlation. Agreement was assessed using the kappa (K) statistic. Accuracy was assessed by sensitivity and specificity. Hyperthyroidism was diagnosed in 33/48 (69%) cats based on elevated serum total thyroxine level. Using two Wilcoxan rank-sum tests, a significant difference (P < 0.0001) was detected between cats with and without hyperthyroidism for both methods of assessing thyroid symmetry. For the 18 cats with T:T ratios < or = 1.5, there was poor correlation between the two methods (r(s) = 0.39). Using a cut-point of 1.5 for the T:T ratio, the test accurately predicted hyperthyroidism in 28/33 cats (sensitivity, 85%; 95% confidence interval (CI), 71-99%) and correctly predicted that hyperthyroidism was absent in 14/15 cats (specificity, 93%; CI, 77-100%). For visual inspection, agreement for diagnosing hyperthyroidism was excellent between methods (kappa = 0.82), within the same examiner (weighted kappa = 0.85) and between examiners (weighted kappa = 0.89). Considering cats with only definitely asymmetric thyroid lobes as positive, visual inspection accurately predicted hyperthyroidism in 28/33 cats (sensitivity, 85%; CI, 71-99%) and correctly predicted that hyperthyroidism was absent in 11/15 cats (specificity, 73%; CI, 48-99%). Thyroid-lobe asymmetry occurs more frequently in hyperthyroid than in euthyroid cats but caution should be exercised because some euthyroid cats have asymmetric thyroid glands.