MON-569 Performance of American Thyroid Association Sonographic Risk and Gene Expression Classifier in the Diagnosis of Thyroid Nodules with Indeterminate Fine Needle Aspiration Cytology

Introduction: The 2015 American Thyroid Association (ATA) guidelines recommend decision making for nodules with indeterminate cytology to include consideration of sonographic features prior to performing molecular testing. However, only a few studies have evaluated how ATA sonographic risk categories predict risk of malignancy along with gene expression classifier (GEC). It has been suggested that in ATA high suspicion nodules, GEC may not add much value. Objective: Evaluate, among indeterminate nodules, how ATA sonographic risk categories perform in comparison to GEC and if the combination of these tools may increase the predictability of malignancy. Methods: A retrospective single center study included Bethesda III thyroid nodules that had undergone evaluation by GEC between 2012-2015. Each nodule was reviewed by 3 radiologists and classified using the 2015 ATA risk categories based on sonographic features: a) high suspicion (HS), b) intermediate suspicion (IS), c) low suspicion (LS), and d) very low suspicion (VLS) of malignancy or e) benign. Nodules were determined to be benign or malignant based on surgical pathology or minimum 2 year follow up if available. Positive predictive values (PPV) and negative predictive values (NPV) were calculated. Overall PPV and NPV for ATA classification were calculated using HS and IS nodules as a positive test and LS and VLS nodules as a negative test. Results: 97 nodules with Bethesda III cytology that had undergone GEC testing were included. Of these, 72 were deemed benign or malignant (17 of 25 nodules that lost follow up had benign GEC). One nodule with suspicious GEC and malignant pathology had no agreement between the readers for ATA sonographic risk category. Prevalence of malignancy was 29%. Overall PPV for malignancy of GEC was calculated at 40% and NPV at 92% (95% CI 0.75-0.98). The overall PPV of ATA classification was 39% and NPV was 82%. The ATA HS nodules (n=31) had 65% suspicious GEC, a PPV of 43% alone, and a combined PPV and NPV with GEC, of 59% and 100%, respectively. The IS nodules (n=21) had 57% suspicious GEC, a PPV of 29%, and combined PPV and NPV with GEC of 44% and 100%. The LS nodules (n=37) had 49% suspicious GEC, a PPV of 18%, and combined PPV and NPV with GEC of 19% and 83% (95% CI 0.55-0.95). The VLS nodules (n=7) had 57% suspicious GEC, a PPV of 17%, and combined PPV and NPV with GEC of 25% and 100%. Conclusion: Among thyroid nodules with Bethesda III cytology that underwent GEC, overall PPV based on ATA classification is comparable to the PPV of GEC, but NPV was higher with GEC. The HS category had the highest rate of suspicious GEC. Within each ATA risk category, the PPV increased moving from VLS to HS categories. Combining ATA sonographic risk with GEC resulted in improved predictive values within each category. Overall, the NPV across all ATA categories was very high and suggests that it is a reliable rule-out test.