Background
Thyroid nodule location is an independent risk factor for thyroid cancer with isthmus nodules having the highest risk. Among lobar nodules, those from the upper poles have been associated with increased risk for malignancy. In this study, we sought to interrogate the Afirma thyroid tumor database for molecular characteristics of upper pole thyroid nodules relative to other lobar locations.
Methods
This was a retrospective evaluation of samples for which Afirma Genomic Sequencing Classifier (GSC) was ordered as part of routine clinical care from April 2018 to December 2024. Inclusion criteria included samples that passed quality control, had (B)ethesda III-VI cytology, and had nodule location (lower, middle, upper) designated. Isthmus nodules and samples where only the laterality of a lobar nodule was noted were excluded. Clinical and molecular features were compared between upper pole nodules and lobar nodules from other locations.
Results
Of approximately 300,000 nodules, 197,783 met inclusion criteria of which 33,928 were from the upper poles. Mean age was 58 years, and the majority were female in both groups (P = NS). Upper pole nodules were smaller (2.19 vs 2.45 cm, P < .01) and more likely to have cytologic features suspicious for or diagnostic of malignancy (B V/VI—7.0% vs 4.9%, P < .01). Among B III/IV, upper pole nodules were more frequently classified as GSC-(S)uspicious (32.4% vs 31.5%, P < .01). BRAF p.V600E was more prevalent in upper pole nodules; this included 8.9% of B III/IV, GSC-S (vs 6.8%, P < .01) and 64.6% of B V/VI (vs 58.4%, P < .01). No other genomic alterations were statistically different by nodule location. Considering gene expression signatures that may relate to tumor biology and prognosis, there was no significant difference in thyroid differentiation score by nodule location. The BRAF-RAS score was more BRAF-like, whereas ERK and follicular mesenchymal transition signature scores were higher in upper pole nodules (P < .0005), though the absolute difference was small.
Discussion/Conclusion
In this large cohort of lobar thyroid nodules sent for Afirma GSC testing, upper pole nodules were smaller, though they had a higher proportion of B V/VI cytology, GSC-S, and BRAF p.V600E. Some gene expression signatures associated with aggressive disease were also higher in upper pole nodules. A study limitation is the lack of clinical outcomes data including histopathology of surgical specimens if thyroidectomy was pursued. However, these data suggest that upper pole nodules harbor more suspicious features than those arising from other lobar locations. Whether this information alters clinical practice requires further evaluation.