Objective differential classification of thyroid lesions by nuclear quantitative assessment

Histol Histopathol. 1997 Apr;12(2):425-31.

Abstract

Quantitative nuclear parameters estimated by morphometric and stereological methods in combination with discriminant analysis were used in order to evaluate the diagnostic efficiency of thyroid lesions. This study includes 55 patients with thyroid pathology. Samples of follicular adenomas, follicular carcinomas, and papillary carcinomas were examined by image analysis to obtain size and form nuclear parameters. Stepwise discriminant analyses were performed. There was an increase in nuclear size from follicular adenomas to follicular carcinomas, and a greater increase from follicular carcinomas to papillary carcinomas. The increase was more significant when the three-dimensional estimates of the volume-weighted mean nuclear volume were assessed. No significant differences between follicular adenomas and follicular carcinomas were found with respect to the nuclear form factors, however, a significant increase in nuclear elongation and irregularity was registered between follicular and papillary tumors (p < 0.01). The overall accuracy rate of discrimination was 75% when the three lesions were included in the analysis, with an efficiency of 85% for papillary carcinoma samples. These percentages increased when two lesion discrimination was performed. The worst discrimination (69% of efficiency) was found between follicular adenomas and follicular carcinomas.

MeSH terms

  • Adenocarcinoma, Follicular / classification
  • Adenocarcinoma, Follicular / pathology
  • Analysis of Variance
  • Carcinoma, Papillary, Follicular / classification
  • Carcinoma, Papillary, Follicular / pathology
  • Cell Nucleus / ultrastructure*
  • Data Interpretation, Statistical
  • Discriminant Analysis
  • Humans
  • Paraffin Embedding
  • Thyroid Gland / pathology
  • Thyroid Gland / ultrastructure
  • Thyroid Neoplasms / classification*
  • Thyroid Neoplasms / pathology*