Home| Journals | Statistics Online Expert | About Us | Contact Us

[Abstract] [PDF] [HTML] [Linked References]

International Journal of Recent Trends in Science and Technology, ISSN 2277-2812 E-ISSN: 2249-8109

Volume 11, Issue 2, June 2014 pp 173-175

Research Article

Follicular lesions of thyroid- A diagnostic dilemma in cytopathology

Archana Bhat¹, Hilda Fernandes², Sumanth S. D.³, Nisha Marla³, Leena Priya⁴, Namitha Hegdekatte¹

Introduction: Follicular lesions of thyroid fall in the gray zone of cytopathology. The bulk of cytohistologic discordant cases belong to this gray zone due to the presence of overlapping features. Aims/objectives: To find out the diagnostic accuracy of thyroid FNA in diagnosis of follicular lesions and to critically evaluate theses cases for causes of discordance. Materials and Methods: 1091 patients underwent thyroid FNA in a period of 2 years in our institute. Histopathological diagnosis was available in 313 cases. Data of these cases was collected and statistically analyzed considering histopathology as the Gold Standard. Observation/results: Of the 313 cases with histopathology, 56 cases were follicular lesions cytologically, which were broadly grouped as adenomatous lesions (33) and follicular neoplasms (23). Of 33 adenomatous lesions, 7 cases did not correlate with histopathology. Of 23 follicular neoplasms, 6 cases were discordant. Overall diagnostic accuracy of FNAC for follicular lesions was 77%. We then analysed the possible causes for discordance. Too much emphasis on cellularity and presence of focal subtle nuclear features and less emphasis on the amount of colloid were the most common reasons for discordance. Hashimotos thyroiditis and adenomatous hyperplasia were the most common false negative diagnoses in the group of follicular neoplasms. Conclusions: Since the follicular lesions of thyroid constitute good number of discordant cases due to the focal presence of subtle nuclear features, there is a need to formulate sensitive cytologic scoring criteria. Establishing of criteria will also minimize inter-observer variations.