Distinct diagnostic criteria for ultrasonographic examination of papillary thyroid carcinoma: a multicenter study

Thyroid. 2005 Mar;15(3):251-8. doi: 10.1089/thy.2005.15.251.

Abstract

Recent advances permitting high-resolution ultrasonography have made ultrasonographic examination of nodular thyroid diseases an accessible examination for routine practice. However, diagnostic criteria for ultrasonographic examination of thyroid nodules are not surely established. To identify the optimal strategy for well standardized differential diagnosis of papillary thyroid carcinoma and benign nodules, we evaluated the significance of individual ultrasonographic characteristics of thyroid nodules in a multicenter study. Ten characteristics in ultrasonograms from 53 patients scored by 17 investigators from 15 centers were analyzed by t tests and logistic regression analyses. Between benign and papillary thyroid cancer groups, all characteristics but not size or multiplicity of strong echoes, which suggest calcifications, were significant parameters. Logistic regression analyses showed that border character, shape, and internal echo level are highly significant parameters (p < 0.0005). A multiple logistic regression showed to be the most important predictors of pathologic diagnosis. The diagnostic criterion with border character and internal echo level yielded 93% sensitivity and 92% specificity. In conclusion, univariate and multivariate analyses identified border character, shape, internal echo level, but not strong echoes (calcifications), as important characteristics in differentiating papillary thyroid carcinoma from benign nodules. These results will contribute to standardization of accurate ultrasonographic diagnosis of papillary thyroid carcinoma.

Publication types

  • Multicenter Study

MeSH terms

  • Adenocarcinoma, Papillary / diagnostic imaging*
  • Humans
  • Retrospective Studies
  • Sensitivity and Specificity
  • Thyroid Diseases / diagnostic imaging
  • Thyroid Neoplasms / diagnostic imaging*
  • Ultrasonography, Doppler / methods*