Malignancy Risk in RAS-Mutated Cytologically Indeterminate Thyroid Nodules: Real-World Clinical Experience

Abstract

Introduction: Fine needle biopsy (FNB) is the standard procedure for thyroid nodules meeting criteria for biopsy according to current national guidelines (ATA and ACR). However, 15-25% of biopsies are categorized as indeterminate in the Bethesda Categories III (Atypical/FLUS) or IV (Follicular/Hurthle Neoplasm). Molecular testing is a common way to further risk stratify these nodules for surveillance versus surgical resection. BRAF-like mutations have a high probability of malignancy while RAS-like mutations have a variable risk of malignancy. This study examines the performance of RAS mutations in identifying malignancy or NIFTP in a real-world clinical setting. Methods: 2,372 sequential thyroid FNBs were performed in our center from July 2014 to June 2020. Of these, 367 were subjected to molecular testing with either Thyroseq V2 or V3. 55 RAS-mutated BCIII or IV cases were identified and retrospectively evaluated. NIFTP was considered malignant in the calculations. ATA and ACR ultrasound classification as well as Doppler grade of the study nodule were assessed. Results: 8 cases were excluded due to lack of follow-up or incomplete data. 40 underwent surgical resection based on the cytology and molecular results. 7 did not undergo surgery based on patient preference or comorbidities and had clinical follow up with stable ultrasound for at least 6 months. Surgical pathology results: 67.5% benign, 7.5% NIFTP and 25% malignant.All 4 of the Thyroseq “currently negative” RAS nodules were benign. Of the 36 “positive” cases, 6 with additional mutations had a 66.7% ROM and the 30 RAS-only cases had a 30% ROM. Of the 9 RAS-only cases that were not benign, 3 were NIFTP, 5 FVPTC (2 unencapsulated and 3 encapsulated) and 1 was angio-invasive PTC with extensive extrathyroidal extension (this case had high-risk pre-op imaging features). None had lymph node involvement.Higher allelic frequency, abnormal gene expression (GE) and copy number alterations (CNA) all increased the risk of malignancy/NIFTP; ROM was 50% when both GE and CNA were positive. Nodules that were ATA/ACR moderate and high suspicion had an approximately 40% and 67% ROM respectively with ATA and ACR performing similarly. High intra-nodular vascularity conferred a 46% ROM. Conclusions: RAS mutations represent the most frequent abnormality found in oncogene testing however isolated RAS mutations have a low risk of malignancy and when malignant are generally non-aggressive. When both GE and CNA are positive, the risk increases. US features, including Doppler also contribute to risk assessment. Combining all these features is recommended when counseling patients on active surveillance versus surgical resection of RAS-mutated thyroid nodules.