Can an ultrasound-based nomogram have an impact in differentiating between malignant and benign thyroid nodules with peripheral calcification?
In a new retrospective study, recently published in Academic Radiology, researchers developed and assessed the nomogram for predicting malignant thyroid nodules with peripheral calcification. The total cohort was comprised of 202 patients with 206 benign nodules and 172 patients with 174 malignant nodules, according to the study. The study authors noted a malignancy rate of 49.1 percent in the 112-patient validation data set.
The researchers said the nomogram was built off six independent predictive factors for malignancy including: the halo sign, type of peripheral calcification, nodule margin, extrusion beyond calcification, nodule composition and internal echogenicity.
The study authors found that the ultrasound-based nomogram offered an 88.2 percent AUC for predicting malignancy in thyroid nodules with peripheral calcification. Researchers also noted a sensitivity of 90.9 percent, a specificity of 73.7 percent and a negative predictive value of 89.4 percent for the nomogram.
“These results showed that our nomogram can greatly increase the accuracy and efficacy of predicting malignant nodules with peripheral calcifications for clinicians,” wrote lead study author Song Bai, M.D., who is affiliated with the Department of Ultrasound at the Shenzhen People’s Hospital in Guangdong, China, and colleagues.
Three Key Takeaways
1. High predictive performance. The ultrasound-based nomogram demonstrated strong diagnostic accuracy in predicting malignant thyroid nodules with peripheral calcifications, achieving an AUC of 88.2 percent, sensitivity of 90.9 percent, and specificity of 73.7 percent.
2. Key predictive factors. Highlighting the clinical importance of detailed ultrasound features, the nomogram integrates six independent malignancy predictors, including halo sign, type of peripheral calcification, nodule margin, extrusion beyond calcification, composition, and internal echogenicity.
3. Echogenicity as a strong indicator: Hypoechoic and extremely hypoechoic nodules were strongly associated with malignancy with over 90 percent of malignant nodules in both training and validation cohorts showing hypoechogenicity, reinforcing its value as a critical diagnostic marker.
Emphasizing that internal echogenicity is one of the key indicators for predicting thyroid malignancy, the researchers noted that 91.6 percent of malignant nodules in the training cohort and 90.9 percent of malignant nodules in the validation group were hypoechoic. Multivariable analysis in the training cohort found a greater than threefold higher odds of malignancy with hypoechoic presentation.
“Hypoechoic and extremely hypoechoic echoes are strongly associated with an increased risk of thyroid malignancy, whereas isoechoic and hypoechoic nodules are associated with a lower risk of malignancy,” pointed out Bai and colleagues.
(Editor’s note: For related content, see “Study: AI Boosts Ultrasound AUC for Predicting Thyroid Malignancy Risk by 34 Percent Over TI-RADS,” “FDA Clears Enhanced 3D Ultrasound Platform for Thyroid Imaging” and “FDA Clears AI-Powered Software for Thyroid Ultrasound.”)
Beyond the inherent limitations of a single-center retrospective study, the authors acknowledged the lack of external validation and the need for multicenter studies to validate their findings.
