The value of second opinion in thyroid cytology: A review
2014
BACKGROUND
Second-opinion diagnosis (SOD) on pathological material is an accepted practice before definitive therapy is considered for referred patients. The thyroid gland is an anatomical site prone to diagnostic disagreement between pathologists. We performed a review of the literature that addressed the role of interinstitutional SOD on thyroid fine-needle aspirations (FNAs).
METHODS
Nine studies comprising second opinions on thyroid FNAs were selected. The parameters analyzed included: discordances between the initial diagnoses (IDs) and SODs; cytohistologic correlation; changes in the clinical management of the patients with thyroid nodules after SOD. The same parameters were applied to the “indeterminate” diagnostic category comprising cases initially reported as “atypia,” “atypia of undetermined significance/follicular lesion of undetermined significance,” “suspicious for a follicular neoplasm,” “follicular neoplasm,” “suspicious,” and “suspicious for malignancy.”
RESULTS
A total of 7154 thyroid FNAs were retrieved, showing an overall discordance rate between ID and SOD of 28.6%. In general, SOD was better supported by clinical follow-up and histological diagnosis, showing higher diagnostic accuracy in comparison with ID. Almost one-third (30.4%) of the discordant cases resulted in changes in the clinical management of patients with thyroid nodules. Numerous thyroid FNAs initially categorized as “indeterminate” were definitively classified as benign or malignant by SOD, with an overall diagnostic resolution rate of 42.5%, sensitivity of 97.9%, and diagnostic accuracy of 73.7%.
CONCLUSIONS
Second-opinion review of thyroid FNA improves diagnostic accuracy and potentially changes clinical management. SOD also demonstrates a significative rate of diagnostic resolution for thyroid FNAs originally diagnosed as “indeterminate.” Cancer (Cancer Cytopathol) 2014;122:611–619. © 2014 American Cancer Society.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
38
References
23
Citations
NaN
KQI