Ultrasonographic predictors of thyroid cancer in Bethesda III and IV thyroid nodules

Abstract

BackgroundBethesda III and IV thyroid nodules continue to be difficult to manage. Although molecular testing may assist in decision-making, it is expensive, not widely available, and not without pitfalls. The objective of this study is to assess whether certain thyroid ultrasonographic features may predict the risk of thyroid cancer in patients with Bethesda III and IV thyroid nodules and be used as additional decision-making tools to complement cytopathological results in deciding on diagnostic thyroidectomy.MethodsWe retrospectively evaluated the ultrasonographic features of Bethesda categories III and IV thyroid nodules in patients who underwent subsequent thyroidectomy. We used the final histopathological examination of the surgical specimens as the gold-standard test and analyzed individual preoperative ultrasonographic features as predictors of malignancy.ResultsOf the 278 patients who were diagnosed with Bethesda III and IV thyroid nodules on fine needle aspiration cytology (FNAC), 111 (39.9%) had thyroid cancer, and 167 (59.9%) exhibited benign nodules. The malignancy rate was higher in patients with Bethesda IV nodules (28/50, 56%) than those with Bethesda III nodules (83/228, 36.4%; p=0.016). In univariate analysis, hypoechogenicity (55.6% in malignant vs. 35.3% in benign, p=0.006) and calcifications (54.5 in malignant vs. 35.4% in benign, p=0.008) were significantly different between the benign and malignant pathology groups, whereas the size of the dominant nodule, number of nodules, irregular borders, taller-than-wide shape, and the presence of lymph nodes were comparable between the two groups. These two ultrasonographic features (hypoechogenicity and calcifications) remained significantly associated with the risk of malignancy in multivariate logistic regression analysis (for hypoechogenicity, p=0.014, odds ratio: 2.1, 95% CI:1.0–3.7 and for calcifications, p=0.019, odds ratio: 1.98, 95% CI:1.12–3.50). The sensitivity, specificity, positive and negative predictive values, and accuracy were 31.5%, 83%, 55.6%,64.7%, and 62.6%, for hypoechogenicity, respectively and 32.4%, 82%, 54.5%, 67.8%, and 62%, for calcification, respectively.ConclusionsHypoechogenicity and calcifications in Bethesda III and IV thyroid nodules are strong predictors of thyroid cancer and associated with a two-fold increased risk of malignancy.

Deep convolutional neural network model ResNeSt for discrimination of papillary thyroid carcinomas and benign nodules in thyroid nodules diagnosed as atypia of undetermined significance

“…The main clinical challenge is to rule out malignancy as most thyroid nodules are asymptomatic, benign, and do not require surgical treatment. Fine-needle aspiration (FNA) biopsy is currently the gold standard test for the preoperative diagnosis of malignancies ( 3 ). Although the cytologic features of PTC are well established ( 4 ), FNA is still unable to give definitive results in approximately 10–30% of thyroid nodules ( 5 ).…”