(A305) Diagnostic Performance of Afirma Genomic Sequencing Classifier in Underrepresented Minority Patients With Indeterminate Thyroid Nodules

Introduction: The Afirma Genome Sequence Classifier (GSC) uses next-generation sequencing to stratify the risk of malignancy in thyroid nodules. This study specifically correlates the malignancy risk of thyroid nodules of Bethesda categories III and IV based on Afirma GSC to the gold standard of surgical pathology.

Methods: All patients who had GSC orders sent to Afirma at a single academic medical center between 2017 and 2024 were included in this study. Demographics, relevant clinical history, Afirma GSC results, and surgical pathology reports were obtained via retrospective chart review. Univariate analysis was conducted to elucidate demographic and clinical characteristics of this population as well as to generate confusion matrices of Afirma GSC test performance.

Results: 176 patients were included in this cohort with a median age of diagnosis of 50 years, 149 (84.7%) females, 86 (48.9%) Black, and 67 (38.1%) Hispanic/ Latino patients. Median nodule size was 2.5cm, 192 were Bethesda III, and 11 were Bethesda IV. There were a total of 203 thyroid nodules that were tested with GSC, of which 141 (69.5%) were reported as benign. In this overall patient cohort, 58 patients (33.0%) underwent surgery, revealing 19 malignancies. The observed values of sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were 90.5%, 84.6%, 43.2%, and 98.6%, and for specimens with dominant hürthle cell cytology (n=29) were 100.0%, 80.0%, 44.4%, and 100.0% respectively. In patients with suspicious GSC but no identifiable mutation, the PPV was 23.9%.

Conclusions: The Afirma GSC data reported by our institution suggests that GSC can aid in the decision to defer surgery in those with indeterminate thyroid nodules that have been reported to have low risk of malignancy. However, it should be noted that this data showed a notably lower specificity and positive predictive value, indicating a greater proportion of false positives compared to the 2018 Afirma GSC validation study.