Abstract
Anaplastic thyroid carcinomas (ATC) are rare, but represent the most lethal malignancy of the thyroid. Selective molecular markers and drivers distinguishing ATC from other thyroid carcinomas of follicular origin remain largely unknown, limiting advances in diagnosis and treatment. In a retrospective study, we analyzed gene expression in 36 ATC, 18 poorly differentiated, 132 papillary, and 55 follicular thyroid carcinoma, as well as 124 paired and unpaired normal thyroid tissues in three independent cohorts by RNA-sequencing and immunohistochemistry. RNA-sequencing data in the test cohort suggested selective ATC protein biomarkers. Evaluation of these revealed that ATCs are characterized by the de novo expression of various testis antigens, including melanoma-associated antigen A3 (MAGEA3), but most importantly the oncofetal IGF2 mRNA binding protein 1 (IGF2BP1). Shallow whole genome sequencing essentially excluded that IGF2BP1 upregulation results from gene copy number alterations. Immunohistochemical analyses in all three tumor cohorts confirmed the selective de novo expression of IGF2BP1 protein in ATC. In sum, 75% (27/36) of all tested ATC and 0.5% (1/204) of poorly and well-differentiated thyroid carcinoma tissue samples were positive for IGF2BP1 protein. This indicates that IGF2BP1 protein expression identifies ATC with a diagnostic odds ratio of 612 (95% CI: 74.6-5021). In addition, we found that MAGEA3 is exclusively, although less consistently upregulated in ATC, presenting with an odds ratio of 411 (95% CI: 23.8-7098.7). Importantly, we provide confirmatory evidence that IGF2BP1 and MAGEA3 expression distinguishes ATC from poorly differentiated thyroid carcinoma. IGF2BP1 furthermore identified ATC foci within low-grade follicular thyroid carcinoma. In conclusion, IGF2BP1 represents the most promising single-gene marker available for ATC, followed by MAGEA3, improving on current techniques. Robust markers are essential to help distinguish this high-grade malignancy from other thyroid carcinomas, to guide surgical decision making, therapy and post-resection/therapy monitoring strategies.