Searchable abstracts of presentations at key conferences in endocrinology
Endocrine Abstracts (2015) 37 GP23.07 | DOI: 10.1530/endoabs.37.GP.23.07

1Servicio de Endocrinología y Nutrición, Hospital Germans Trias i Pujol, Badalona, Catalunya, Spain; 2Instituto de Investigación en Salud Germans Trias i Pujol, Badalona, Catalunya, Spain; 3Instituto de Investigación Biomédica de Lleida, Universidad de Lleida, Lleida, Catalunya, Spain; 4Servicio de Patología y Genética Molecular, Hospital Universitario Arnau de Vilanova, Lleida, Catalunya, Spain; 5Centro Nacional de Investigaciones Oncológicas, Madrid, Catalunya, Spain.


Introduction: Follicular tumours of the thyroid gland include benign follicular adenomas, follicular carcinomas and the follicular variant of papillary carcinoma. Although, molecular markers are a promising area of research to differentiate thyroid neoplasms, there is ample room for improvement in the research and clinical applications in this field. Furthermore, the final diagnosis requires thyroid lobectomy. The present study examined whether a differential protein expression could provide an accurate tool to improve diagnosis in these tumours.

Design: The expression of 19 proteins was evaluated in 125 surgically removed thyroid tumours including 81 follicular adenomas (FA) and 44 malignant tumours (nine follicular carcinomas (FTC) and 35 follicular variant of papillary carcinomas (FVPTC)). Classification trees and scoring systems were constructed based on protein expression to generate models that could allow a differential diagnosis between benign and malignant tumours. Forty hyperplastic nodules (HN) were used to test the goodness of fitness of the generated models.

Results: The classification tree that better discriminated between benign and malignant nodules was APLP2 (n=0), RRM2 (n>5), PRC1c (n≥40), APLP2c (n≤105), and SIAH1 (n≤5), with a sensitivity of 100% and a specificity of 98.8%. In HN, the sensitivity was 94.8%. The punctuation system model 4×(APLP2 (n=0))+2×(RRM2 (n>5))+1×(PRC1c (n≥40)) reached a sensitivity of 100% and a specificity of 71.3%.

Conclusion: Our findings support the potential of these models to improve not only the diagnosis of follicular thyroid tumours but also, the management and wellness of patients by reducing the number of exploratory surgeries. We are currently validating these results in an independent series.

Disclosure: This work was financed for Institute Carlos III (FIS PI11/1354).

Article tools

My recent searches

No recent searches.