Objective: Thyroid nodules are common within the general population. Cytological analysis of fine needle aspirates (FNAs) of these lesions allows for identification of those that require further surgery. A numerical classification system is in place to streamline reporting. The 3a category is used for lesions that are neither benign nor malignant but show atypia of undetermined significance. We reviewed our use and clinical outcomes of Thy3a over a 4-year period. Methods: All thyroid FNAs performed at this institute from January 2012 to December 2015 were identified from our laboratory information system using SNOMED codes. Cytology was correlated with histology. Results: Of the 1,259 FNAs reported at this institute, Thy3a constituted only 1.2% (n = 16) of all cases, with a malignancy rate of 7%. Five Thy3a cases had a repeat FNA that was reported as Thy2 (benign), 1 as Thy1c (cyst), 1 as Thy3f (follicular lesion), and 1 as Thy5 (malignant). Six cases without repeat FNA were follicular adenomas at resection. Two cases were lost to follow-up. Within all thyroid cytology categories in this 4-year period, we had a false-positive rate of 1.9% and a false-negative rate of 0.3%. Conclusions: The Thy3a subclassification has varied diagnostic criteria and lacks reproducibility. Despite the rare use of the Thy3a category at our centre, our diagnostic accuracy remained high. At this time, further Thy3a cohort studies are required to assess the real benefits of this category.

Objectives: The British Thy system is a widely used classification system for reporting thyroid fine-needle aspiration (FNA) cytology. The Royal College of Pathologists in 2009 recommended the subdivision of the Thy-3 (indeterminate) category into Thy-3a (atypia) and Thy-3f (follicular neoplasm). Our objective was to examine the malignancy rates of Thy-3a and Thy-3f cases at our institution and to investigate whether the risk of malignancy in Thy-3a cases is reduced by FNA on a different occasion showing benign cytology. Methods: This is a retrospective study of 748 thyroid nodules undergoing 1,032 FNAs, with indeterminate (Thy-3) cytology subdivided into Thy-3a and Thy-3f. Cases were correlated with final histology in surgical cases. Incidental carcinomas occurring outside the biopsied nodule were discounted. Results: A total of 109 nodules had a final cytological diagnosis of Thy-3a, of which 67 underwent surgery, with an incidence of malignancy of 13.4% (9/67); 90 nodules had a final cytological diagnosis of Thy-3f, of which 84 underwent surgery, with an incidence of malignancy of 17.9% (15/84). The difference in malignancy rates was not significant (p = 0.51). The incidence of malignancy in nodules with benign and Thy-3a cytology on separate occasions was not significantly different from cases with a single Thy-3a cytology. Conclusions: Thyroid nodules with Thy-3a cytology have a slightly lower risk of malignancy than Thy-3f cases. However, the difference is not significant and does not appear to be reduced by FNA on a separate occasion showing benign cytology. Management decisions for patients with Thy-3a cytology should be taken carefully to avoid missing cancers.