Oregon Health and Science University
Introduction: Evaluation of thyroid nodules with indeterminate cytology is limited by current methods. We studied the tumor immune microenvironment (TiM) of thyroid cancer using a multiplex immunohistochemistry (mIHC) technique enabling assessment of leukocyte complexity and effector status. Based on previous results, we hypothesized that evaluation of the TiM using mIHC will identify immune biomarkers to improve the risk stratification of indeterminate nodules.
Methods: The TiM of adult thyroidectomy specimens containing benign and malignant neoplasms classified as indeterminate on FNA were evaluated using mIHC. Tissues were interrogated with 21 markers to identify immune cell phenotype and effector status. Results are presented as cell density and percentage of the total immune cell population.
Results: Unsupervised cluster analysis revealed 3 immune signatures: 1 enriched for malignancy, 1 enriched for benign adenomas, and 1 enriched for benign and malignant Hurthle cell neoplasms. Malignant nodules had a higher density of lymphocytes compared to benign nodules (p=0.0003). Evaluation of the myeloid and lymphoid compartments revealed a hyper-inflamed immune signature in malignant samples within both compartments (lymphoid p=0.0008, myeloid p=0.008). Within malignant samples, the inflammatory signature was skewed toward the lymphoid compartment. Deeper analysis of Hurthle cell neoplasms revealed a higher density of lymphocytes in Hurthle cell carcinoma (HCC) (p=0.0178) due to differences in the lymphoid compartment (p=0.0159), specifically the cytotoxic killer T cell population (CD8, p=0.0494). Analysis of CD8 T cell functional subsets revealed lower expression of Tbet in HCC (p=0.0009), denoting a lower level of activation of anti-tumor Th1 T cells in malignant samples.
Conclusions: Our study reveals immune biomarker signature differences between benign and malignant thyroid neoplasms, with results most marked in Hurthle cell neoplasms. Our results support the potential application of mIHC to risk stratification of Hurthle cell lesions in FNA samples obtained from indeterminate nodules.