Ultrasound and surgery agreement in dogs with parathyroid tumors

Abstract

OBJECTIVE: To investigate (1a) agreement of ultrasonographic, surgical, and histopathologic findings in the diagnosis of a neoplastic etiology underlying primary hyperparathyroidism (PHPT), (1b) the ability of ultrasonographically determined parathyroid gland size to distinguish between malignant (carcinoma) and non-malignant (hyperplasia, adenoma) pathology, and (2) variables associated with postoperative hypocalcemia in dogs undergoing surgical treatment of PHPT. STUDY DESIGN: Ambidirectional cohort study. ANIMALS: Forty-seven client owned dogs with PHPT (34 retrospective; 13 prospective). METHODS: Data were extracted from medical records. Method agreements were explored using Cohen's Kappa statistic. A receiver operating characteristic curve (ROC) was used to determine a cut-off separating parathyroid pathologies. Univariable and multivariable models assessed associations between postoperative hypocalcemia and potential risk factors. RESULTS: Agreement of ultrasound and surgery for number and side of affected glands was 31/47 (65.9%) and 34/47 (72.3%), respectively. In 37/47 (78.7%) cases, parathyroid tissue was correctly assessed as pathologic by the surgeon. An ultrasonographic cut-off of ≥8.0 mm (ROC AUC = 0.82) best distinguished malignant from benign pathologies. Dogs with a preoperative serum ionized calcium (iCa) concentration ≥1.75 mEq/L had 7.5 times greater odds of becoming hypocalcemic postoperatively. CONCLUSION: A fair agreement existed between ultrasonographic and surgical findings in dogs with PHPT. A parathyroid mass ≥8.0 mm on ultrasonographic examination was suggestive of malignancy, while dogs with a preoperative serum iCa concentration ≥1.75 mEq/L were at increased risk for postoperative hypocalcemia in this study. CLINICAL SIGNIFICANCE: This study supports the use of bilateral cervical surgical exploration to identify abnormal parathyroid glands for the treatment of PHPT.