Indeterminate Thyroid Nodules With RAS Mutations Have Higher Rates of Malignancy When Multiple Non-Cystic Nodules or Irregular Borders Are Present

Background: There is heterogeneity in positive predictive value for cancer in RAS-mutated cytologically indeterminate nodules and paucity of data with regards to specific ultrasound features that are associated with malignancy in these nodules. The goal of this study is to assess the ultrasonographic characteristics, in relation to clinical and histopathologic outcomes of thyroid nodules known to have a RAS mutation. Design and Methods: Cases were identified using our institutional Afirma® Genetic Sequence Classifier (GSC) database, and a retrospective review of electronic medical records for thyroid nodules biopsied between January 2018 and August 2020. Nodules categorized as Bethesda III or IV and harboring a RAS mutation by Xpression Atlas® were included. Thyroid ultrasound images were reviewed by the authors and were risk stratified according to the 2015 American Thyroid Association Thyroid Nodule and Differentiated Thyroid Cancer Guidelines (ATA Guidelines) and the 2017 ACR TIRADS system (ACR Guidelines). The nodules were divided into benign or malignant categories based on surgical pathology. Noninvasive follicular thyroid neoplasms with papillary like nuclear features (NIFTP) were categorized as benign. Results: A total of 22 nodules were identified to have a RAS mutation. NRAS mutated nodules, all with the same point mutation (pQ61R c.182A>G), were most common 14/22 (63.6%). There was no significant difference in clinical features, ultrasonographic appearance or histopathologic outcomes between NRAS- and HRAS-mutated nodules. 12/22 (54.4%) were low risk by ATA Guidelines and 11/22 (50%) were TIRADS 4 (moderately suspicious) by ACR Guidelines. There was no significant difference in predictive value of ATA Guidelines vs ACR Guidelines. The prevalence of malignancy was 45.4% (only slightly lower than the general risk for a suspicious GSC). Invasive follicular variant papillary thyroid cancer (FVPTC), was the most common malignancy, 4/10 (40%). 6/10 (60%) were classified as low risk of recurrence post-operatively. All malignant RAS-mutated nodules (10/10) had at least one other non-cystic nodule present on ultrasonography whereas only 4/9 (44%) of RAS-mutated benign nodules did [P=.006]. RAS-mutated malignant nodules had significantly more nodules with irregular borders compared to RAS-mutated benign nodules (4/10 and 0/10, 40% and 0% respectively) [P=.03]. Conclusions: This is the first study to observe higher rates of malignancy in RAS-mutated indeterminate nodules when other non-cystic nodules are present. A lobectomy is the preferred surgical approach for RAS-mutated nodules, however a total thyroidectomy may be considered in patients with other non-cystic nodules or irregular nodules borders. Overall, RAS-mutated nodules have a low risk of recurrence post-operatively.