46.08 Gene Expression Classification has Limited Utility in the Evaluation of AUS/FLUS Thyroid Nodules

K. D. Klingbeil1, R. L. Deitz1, M. L. Mao1, J. C. Farrá1, J. I. Lew1  1University Of Miami Miller School Of Medicine,Division Of Endocrine Surgery, DeWitt Daughtry Family Department Of Surgery,Miami, FL, USA

Introduction: The current management for thyroid nodules remains a challenge for physicians due to the underlying risk of malignancy. With the Bethesda System for Reporting Thyroid Cytopathology (BSRTC), atypia of undetermined significance/follicular lesion of undetermined significance (AUS/FLUS), known as Bethesda III, has an unclear rate of malignancy. Gene expression classification (GEC) testing was developed to further stratify patients with AUS/FLUS nodules. Given its known variability between institutions, this study examines the utility of GEC testing in predicting malignancy in patients with AUS/FLUS thyroid nodules. 

Methods:  A retrospective review of prospectively collected data for patients with index thyroid nodules who underwent FNA and thyroidectomy at a single institution was performed. GEC testing utilized in patients with AUS/FLUS by FNA was reported as benign or suspicious for malignancy. Patients with AUS/FLUS nodules underwent initial thyroid lobectomy for definitive diagnosis unless there was a history of known risk factors and/or patient preference for which total thyroidectomy was performed. AUS/FLUS nodules were subdivided into malignant or benign groups based on final pathology. Among patients who underwent GEC testing, final pathology was compared to initial GEC results.

Results: Of 863 patients who underwent FNA and thyroidectomy, 224 patients (26%) were found to have AUS/FLUS nodules. Following thyroidectomy, 120 patients (54%) were shown to have thyroid cancer (Papillary, n=110; Follicular, n=8; Medullary, n=2) on final pathology. The remaining 104 patients had benign final pathology, most commonly presenting as multinodular hyperplasia, n=31. GEC testing was performed in 102 patients with AUS/FLUS testing, of which 96 had suspicious results. The rate of malignancy for patients with AUS/FLUS nodules and suspicious GEC results was 51% (49/96) whereas the rate of malignancy for AUS/FLUS nodules without GEC testing was 55% (67/122). Of the 6 patients with benign GEC results, 4 were found to be malignant (66%). The positive predictive value (PPV) for GEC testing in AUS/FLUS nodules was 51%.  

Conclusion: Surgical patients with AUS/FLUS nodules had a high malignancy rate compared to the general population. GEC testing demonstrated a high frequency of suspicious results in AUS/FLUS thyroid nodules, yet had limited utility of predicting malignancy. There was no significant difference in malignancy rates when comparing AUS/FLUS nodules without GEC testing to those with suspicious GEC results. In addition, GEC testing failed to rule out malignancy in AUS/FLUS nodules with benign results. Thus, the application of GEC testing has limited utility in surgical decision-making. Surgeons should assess their local institutional experience to determine if there is added utility of GEC testing for AUS/FLUS nodules in their everyday clinical practice.