A simplified four-tier classification for thyroid core needle biopsy.

Abstract

Purpose: To propose a simplified histological classification for core-needle biopsy (CNB) of thyroid nodules with four diagnostic categories (DC) and provide the risk of malignancy (ROM) and the expected incidence for each DC. There is no uniform scheme for categorizing CNB specimens, except for a Korean diagnostic classification similar to the Bethesda system for FNAC.

Methods: Data from a single institution using CNB as a routine diagnostic tool for thyroid nodules. Biopsies were classified as non-diagnostic, benign, follicular tumour (FT) or malignant. The frequency of each DC and the correlation with surgical pathology of nodules undergoing surgery after CNB were evaluated.

Results: Of 6284 CNBs on 5782 nodules [195 (3.1%) non-diagnostic, 5043 (80.3%) benign, 435 (6.9%) FT and 611 (9.7%) malignant], 1914 nodules (33.1%) underwent surgery after 2016 CNBs: 48 non-diagnostic, 1035 benign, 355 FT and 578 malignant. Malignancy was diagnosed after surgery in 11 non-diagnostic (ROM: 22.9%), 23 benign (ROM: 2.2%), 44 FT (ROM: 12.4%, 11.5% excluding low-grade malignancy) and 568 malignant CNBs (ROM: 98.3%, 93.8% excluding low-grade malignancy). Sensitivity and positive predictive value for malignancy of FT or malignant CNB were 94.7% and 65.6%, and for thyroid neoplasm (adenoma or carcinoma) were 93.2% and 92.6%.

Conclusions: Our diagnostic classification for CNB of thyroid nodules has a high diagnostic accuracy with a low rate of indeterminate categories. This classification, applied in a Western practice, shows a low ROM for nodules classified as follicular tumours, which could be improved with immunohistochemical studies.