Diagnostic accuracy of ultrasound criteria for solid thyroid nodules

TOPIC: Ultrasound of thyroid nodules

Title: Benign and malignant thyroid nodules: US differentiation – A multicenter retrospective study.

Authors: Moon WJ, Jung SL, Lee JH, Na DG, Baek JH, Lee YH, Kim J, Kim HS, Byun JS, Lee DH; & the Thyroid Study Group, the Korean Society of Neuro- and Head and Neck Radiology.

Reference: Radiology 247: 762-770, 2008

Summary

Background

Thyroid ultrasonography is increasingly used in thyroid practice. Data on accuracy of differentiating benign from malignant nodules are conflicting.

Purpose

To evaluate retrospectively the diagnostic accuracy of ultrasonographic (US) criteria for the depiction of benign and malignant thyroid nodules by using tissue diagnosis as the reference standard.

Material & Methods

Present study had institutional board review approval and informed consent was waived. From January 2003 through June 2003, 8.024 consecutive patients had undergone US at 9 affiliated hospitals. A total of 831 patients (716 women & 115 men; mean age: 49.5 ± 13.8 [SD] years) with 849 nodules (360 malignant, 489 benign) that were diagnosed at surgery or biopsy were included in the study. Three radiologists retrospectively evaluated the following characteristics on US images: nodule size, presence of spongiform appearance, shape, margin, echotexture, echogenicity, and presence of microcalcification, macrocalcification, or rim calcification. A χ 2 test and multiple regression analysis were performed. Sensitivity, specificity, and positive and negative predictive values were obtained.

Results

Statistically significant (P<0.05) findings of malignancy were a taller-than-wide shape (sensitivity: 40.0%; specificity: 91.4%), a spiculated margin (sensitivity 48.3%; specificity 98.1%), marked hypoechogenicity (sensitivity 41.4%; specificity 92.2%), microcalcification (sensitivity 44.2%; specificity 90.8%), and macrocalcification (sensitivity 9.7%; specificity 96.1%). The US findings benign nodules were isoechogenicity (sensitivity 56.6%; specificity 88.1%; P<0.001) and a spongiform appearance (sensitivity 10.4%; specificity 99.7%; P<0.001). The presence of at least one malignant US finding had a sensitivity of 83.3%, a specificity of 74.0%, and a diagnostic accuracy of 78.0%. For thyroid nodules with a diameter of 1 cm or less, the sensitivity of microcalcifications was lower than that in larger nodules (36.6% versus 51.4%, P<0.05).

Conclusions

Shape, margin, echogenicity, and presence of calcification are helpful criteria for the discrimination of malignant from benign nodules. The diagnostic accuracy of US criteria is dependent on tumor size.

Commentary

In recent years, increasingly sensitive US machines have been used to detect, diagnose, and manage thyroid nodules. The utility of US and US-guided FNA biopsy in current thyroid practice is confirmed by numerous reports. Also clear is that the overlap in US appearance of benign and malignant nodules is significant enough that US-FNA is usually required to make a definitive diagnosis. Additionally, the sensitivity of US application is clearly influenced by the operator’s skills and experience.

In the report by Moon et al., a diagnostic accuracy of 78% was achieved when the presence of one of the malignant US findings was applied for a diagnosis of malignant thyroid nodules. In this report (as in previous similar studies), most thyroid nodules are predominantly solid, whether benign or malignant. In this study, a taller-than-wide shape was very specific for malignant lesions; hypoechogenicity, ill-defined margins, and microcalcifications were all important US findings of malignancy. The authors separated micro- from macrocalcifications by using a cut-off value of 1 mm. The authors found that a spongiform appearance and isoechogenicity are US indicators for benign nodules. The false-positive US results were greater for nodules <1 cm. In this study, the overall sensitivity of thyroid US for depicting a malignant nodule was 83.3%. The limitations of this study include retrospective analysis, selection bias, and the inability to evaluate US findings in real time. The study, however, confirms that when thyroid US is performed by experienced physicians, separation of benign from malignant nodules achieves a high degree of sensitivity. Nevertheless, it is also important to note that current US criteria do not replace FNA in establishing a definitive diagnosis in thyroid nodules. The report also emphasizes that nodules of less than 1 cm are more difficult to biopsy and more likely to yield non diagnostic or false-positive results. Summary and commentary prepared by Mahmood Gharib & Hossein Gharib (Related to Chapters 6[c] & 18 of TDM)