Aim: The purpose of this study was to assess the positive predictive value of the suspicious sonographic features of solid nodules of the thyroid. Materials and methods: After approval by our institutional review board, we evaluated 594 sonographically detected nodules on which fine needle aspiration biopsy had been performed from January to December 2005. Among these, pure cystic lesions and inadequate pathologic results were excluded. The remaining 530 solid thyroid nodules were analyzed by two radiologists. Each lesion was classified based on four sonographic features that suggested malignancy: microcalcifications; an irregular or microlobulated margin; marked hypoechogenicity; and a shape that was taller than wide. The sensitivity, specificity, positive predictive value, and negative predictive value of the each sonographic feature were calculated. Results: Final pathologic results showed that 97 (18.3%) of 530 nodules were malignant. The positive predictive values for malignancy in each sonographic feature were microcalcifications, 38.6% (39/101); an irregular or microlobulated margin, 28.2% (70/248); marked hypoechogenecity, 49.4% (38/77); taller than wide shape, 59.8% (49/82). In terms of relative risk, microcalcification [P<.01, relative risk (RR)=3.115, 95% CI: 1.724-5.628], hypoechogenecity (P<.01, RR=2.510, 95% CI: 1.290-4.881). The shape of nodule which was taller than wide (P<.01, RR=7.624, 95% CI: 4.156-13.986) revealed the highest predictive sonographic finding suggesting malignancy. However, margin is the least significant feature of detection of thyroid malignancy (P=.27, RR=1.395, 95% CI: 0.777-2.505). Conclusion: The three sonographic features of solid thyroid nodule, that is, microcalcifications, marked hypoechogenecity, and a taller than wide shape are meaningful findings in the diagnosis of thyroid malignancy. The shape that was taller than wide was the most reliable sonographic feature for predicting malignancy.
All Science Journal Classification (ASJC) codes
- Radiology Nuclear Medicine and imaging