BOSTON MEDICAL CENTER Brighton, Massachusetts, UNITED STATES
Background Genomic classifiers such as ThyroSeq guide diagnosis and therapy in differentiated thyroid cancer. We evaluated coverage-aware prevalence and co-mutation patterns of ThyroSeq-relevant alterations across papillary (PTC), follicular (FTC), and Hürthle/oncocytic (Hürthle) carcinomas using a large multi-institutional real-world clinicogenomic dataset. Methods The public AACR Project GENIE Cohort v18.0-public was queried in cBioPortal (accessed January 2026). We restricted to OncoTree-annotated PTC, FTC, and Hürthle carcinomas and retained one tumor per patient. Available data types in the filtered cohort were somatic mutations (SNVs/indels; n=2,219 samples), structural variants/fusions (n=1,891), and discrete copy-number alterations (n=1,342). The predefined ThyroSeq-relevant gene set included BRAF; HRAS/NRAS/KRAS; TERT promoter; EIF1AX; PTEN; PIK3CA; TP53; and fusions in RET, NTRK1/NTRK3, ALK, and PAX8–PPARG. Denominators were coverage-aware (only samples tested for each alteration and data type). Prevalence was summarized with Wilson 95% CIs. Pairwise co-occurrence/mutual exclusivity within each histology used Fisher’s exact tests with Benjamini–Hochberg FDR (10%). Sensitivity analyses restricted to genes covered across ≥90% of samples in all three histologies and stratified by sequencing panel/center. Results We identified PTC (n=1,867; 55.3% of thyroid cancers in GENIE), FTC (n=207; 6.1%), and Hürthle (n=145; 4.3%). PTC exhibited a BRAF^V600E-dominant architecture with comparatively low RAS alterations; BRAF and RAS were mutually exclusive after FDR correction, and TERT promoter mutations co-occurred with high-risk drivers. FTC showed enrichment of RAS (predominantly NRAS), infrequent BRAF, and a PAX8-PPARG fusion subset; RAS and BRAF were mutually exclusive, and RAS showed positive association with PAX8-PPARG. Hürthle tumors were BRAF-sparse with relatively higher RAS and a distinct secondary-event profile. All major signals persisted in intersection-gene analyses and after panel/center stratification. Fusion and CNA prevalences reflected data-type availability (SV in 1,891 tested samples; CNA in 1,342), underscoring the need for coverage-aware denominators.
Conclusions In a large real-world cohort, ThyroSeq-relevant alterations display clear, histology-specific architectures-BRAF-driven PTC, RAS-enriched FTC (with PAX8-PPARG), and BRAF-sparse, RAS-skewed Hürthle. Rigorous coverage-aware handling is essential in heterogeneous targeted-panel datasets. These findings support translational use of genomics to refine diagnostic classifiers and prioritize targeted therapy trials in differentiated thyroid cancer.
*Unless otherwise noted, all abstracts presented at ENDO must not be released to the press or the public until the date and time of presentation. For oral presentations, the abstracts are embargoed until the session begins. The Endocrine Society reserves the right to lift the embargo on specific abstracts that are selected for promotion prior to or during ENDO.*
.jpg)
.jpg)