Risk Stratification in Follicular Neoplasm

BACKGROUND: The 2007 Bethesda classification for thyroid cytology defines follicular neoplasm as a category of cases with cellular specimens demonstrating abundant follicular cells arranged in a microfollicular pattern with little or no colloid. The current recommendation for the management of these cases is diagnostic lobectomy. There has been great difficulty and variability in triaging and reporting follicular neoplasm. To increase diagnostic accuracy, at the study institution, this category is subclassified further into 3 categories: 1) microfollicular-patterned neoplasm (MN); 2) Hurthle cell neoplasm (HN); and 3) follicular lesion with some features suggestive of but not diagnostic of the follicular variant of papillary thyroid carcinoma (FL). The authors reviewed the cases of follicular neoplasm observed over a period of 5 years to document the follow-up trend using this modified classification. METHODS: A search of the cytology records was performed for the period between January 2008 and December 2012. All thyroid fine-needle aspiration cases were reviewed and those with a diagnosis of follicular neoplasm (including Hurthle cell neoplasm) were identified. Correlating follow-up surgical pathology reports were reviewed. RESULTS: A total of 399 cases of follicular neoplasm with surgical follow-up were identified. Malignancy was identified in 32% of all cases of follicular neoplasm and was found to be disproportionately higher in the FL category (73%). A cytological diagnosis of FL is more likely to be called malignant (73%) than benign neoplastic (9%) or benign nonneoplastic (18%). A cytological diagnosis of MN or HN is more likely to be benign neoplastic (46% and 46%, respectively) than malignant (29% and 26%, respectively) or benign nonneoplastic (25% and 28%, respectively). Of the cytological features examined, 2 (nuclear enlargement and nuclear grooves) were significantly associated with the follicular variant of papillary thyroid carcinoma. CONCLUSIONS: The results of the current study clearly indicate that follicular lesions with even subtle nuclear atypia have a high positive predictive value for malignancy and therefore should be distinguished from other follicular lesions because these cases require more aggressive surgical management. The current study also raises an important issue concerning the current thyroid classification based on the 2007 Bethesda classification for thyroid cytology. Future thyroid fine-needle aspiration classification schemes should consider subclassifying follicular neoplasms for the purpose of risk stratification. (c) 2014 American Cancer Society.