The genomic and evolutionary landscapes of anaplastic thyroid carcinoma

Authors

Zeng PYF

Paulson SD

Lee SY

Patel N

Sunderland RX

Castillo-Suyo MA

Titmuss M

Yap TN

Lindsay J

Barrett JW

Boutros PC

Nichols AC

Doi

10.1016/j.celrep.2024.113826

PMID: 38412093 · DOI: 10.1016/j.celrep.2024.113826 · Journal: Cell Reports (2024)

TL;DR

The Global Anaplastic Thyroid Cancer Initiative (GATCI), a 15-site consortium, sequenced tumor DNA from 329 thyroid cancer regions, including 213 from patients with primary anaplastic thyroid carcinomas (ATC). Using whole-exome sequencing, whole-genome sequencing, SNP arrays, and RNA-seq, the study defined the mutational landscape and molecular subtypes of ATC, demonstrating that ATC has a higher mutation burden than differentiated thyroid cancers but fewer mutations than most other adult cancer types. Multi-region whole-genome sequencing of 9 patients with co-occurring differentiated (DTC) and anaplastic components unambiguously demonstrated that ATC and DTC share a common clonal origin and diverge through acquisition of characteristic driver mutations, particularly in TP53, PIK3CA, and CDKN2A.

Cohort & data

  • 329 thyroid cancer regions from 292 patients across 15 institutions (PMID:38412093)
  • 213 regions from ATC patients (179 primary ATC, 1 metastatic ATC, 34 co-occurring DTC within ATCs) and 115 papillary thyroid cancer (PTC) regions
  • 141 ATCs characterized by WES (n=132) or WGS (n=9); 28 co-DTCs; 114 PTCs; 13 cell lines
  • Copy-number arrays on 110 ATCs, 22 co-DTCs, 112 PTCs, 1 metastasis, and 13 cell lines
  • RNA-seq on 24 primary ATCs and 13 cell lines, integrated with TCGA PTC data
  • Dataset: thyroid_gatci_2024
  • Methods: whole-exome-seq, whole-genome-seq, rna-seq, SNP microarrays

Key findings

  • ATCs harbor 3.8 +/- 1.2 SNVs/Mb and 120 +/- 44 CNAs per tumor (mean +/- 99% CI), more than PTCs but fewer than most adult cancer types (PMID:38412093)
  • 42 genes were recurrently mutated in ATC via SeqSig analysis (FDR < 0.05); top five by FDR: TP53, NRAS, BRAF, PIK3CA, and USH2A
  • Five distinct CNA subtypes (A-E) identified by consensus clustering, each with characteristic genomic and clinical features
  • Loss of CDKN2A in 42% of ATCs; loss of BRCA2 in 33.6% of ATCs; loss of RB1 also recurrent
  • BRAF V600E was the only driver more common in DTC (50.9% of PTCs) than ATC (21.3%); half of BRAF V600E variants in co-occurring cases were detected solely in the DTC component
  • TP53 mutations increased from 0.9% in PTCs to 21.4% in co-DTCs to 36.8% in ATCs
  • TERT promoter mutations correlated with TP53 mutations (FDR < 0.01)
  • BRAF V600E and RAS mutations were mutually exclusive; BRAF V600E and PIK3CA mutations co-occurred (FDR = 0.034)
  • Multi-region WGS of 9 patients confirmed that ATC and co-occurring DTC share a common clonal origin, with the common ancestor harboring ~95% of CNAs but only 19.1% +/- 7.9% of SNVs
  • ATC tumors lacking SNV drivers frequently harbored deletions in tumor suppressors CDKN2A, RB1, CDK7, or BRCA2 (10/21 such tumors)
  • COSMIC mutational signatures 1, 5, 6, and 13 (AID/APOBEC) were active in ATC
  • Germline variants in cancer predisposition genes detected: RECQL4 (5% of ATCs), BRCA2 (n=3), FANCF (n=3)
  • Patients with lower mutation rate (>10 SNVs/Mb) had significantly better survival (HR = 0.51, 95% CI 0.33-0.77, p = 0.002)
  • BRCA2 deletion was associated with better survival (HR = 0.48, 95% CI 0.29-0.80, p = 0.005)

Genes & alterations

  • TP53: Non-functional/partially functional non-synonymous SNVs in 36.8% of ATCs; associated with elevated mRNA abundance (p = 0.0067); frequency increases from PTC to co-DTC to ATC
  • BRAF: V600E in 21.3% of ATCs vs. 50.9% of PTCs; clonal/early subclonal timing; mutually exclusive with RAS
  • NRAS: Recurrently mutated in ATC; no frequency difference between PTC and ATC
  • PIK3CA: Co-occurs with BRAF V600E (FDR = 0.034); preferentially mutated in ATCs and co-DTCs
  • CDKN2A: Deleted in 42% of ATCs, recurrent in co-DTCs, rare in PTCs (~5%)
  • BRCA2: Deleted in 33.6% of ATCs, 13.6% of co-DTCs, 4.5% of PTCs; deletion associated with better survival
  • BRCA1: Recurrent somatic and germline alterations at both SNV and CNA level
  • RB1: Recurrent deletions in ATCs
  • ATM: Preferentially mutated in ATCs and co-DTCs; recurrent at both SNV and CNA level
  • ATR: Recurrently mutated in ATC
  • NF1: Focal CNA loss on chromosome 17 in ATC
  • EIF1AX: Mutations in 5 tumors including p.A113X splice-site mutation; tends to co-occur with RAS
  • TERT: Promoter mutations correlated with TP53 mutations
  • USH2A: Recurrently mutated (top 5 SeqSig FDR); previously described in ATC
  • MET, NOTCH1, FAT1, SPEN, TET1: Recurrent SNVs identified by SeqSig analysis
  • CDK7: Focal CNA loss on chromosome 5 in ATC
  • FLT3, FGF9: Focal CNA loss on chromosome 13 in ATC
  • HMCN1, HSPG2: Cell-adhesion genes with frequent alterations; implicated in tumor cell migration and invasion
  • VGLL3: 429 kb gain on chromosome 3p in 18% of ATCs
  • LRP1: Recurrent mutations previously described in ATC

Clinical implications

  • Treatment with radiotherapy (FDR = 1.7 x 10^-5) or surgery (FDR = 0.0089) was associated with improved overall survival; surgery and patient age strongly stratified ATC survival outcomes
  • Tumors with fewer CNAs more often presented with distant metastases (Wilcoxon rank-sum, FDR = 0.034)
  • CNA subtypes show distinctive clinical features: subtype A enriched for older patients with better survival; subtype D enriched for male patients with metastatic disease
  • Recurrent alterations in BRCA1, BRCA2, and ATM at both SNV and CNA level rationalize investigation of PARP inhibitors in ATC
  • Most samples were collected before FDA approval of dabrafenib and trametinib for ATC, limiting survival analysis in the targeted therapy era
  • The shared clonal origin of ATC and DTC suggests early molecular surveillance could detect progression risk before anaplastic transformation

Limitations & open questions

  • WES/WGS sequencing depth may have resulted in lower mutation rates than reported in other studies
  • Heterogeneity in platforms used across the 15-site consortium
  • Incomplete clinical data for all patients
  • Limited sample size for RNA-seq characterization (n=24 primary ATCs)
  • Most samples collected prior to widespread use of targeted therapy, potentially affecting survival outcomes
  • It is unclear whether the co-evolution model holds for ATCs arising without co-occurring DTCs
  • Functional studies are needed to determine which novel mutations drive dedifferentiation
  • Larger cohorts required to understand germline predisposition to ATC
  • Alternative driver mechanisms (e.g., fusions) in tumors lacking both SNV and CNA drivers remain unexplored

Citations from this paper used in the wiki

  • “We sequenced tumor DNA from 329 regions of thyroid cancer, including 213 from patients with primary anaplastic thyroid carcinomas.” (PMID:38412093)
  • “In this large genome-wide ATC cohort (n = 141, both WES and WGS), we identified 3.8 +/- 1.2 SNVs/Mb of DNA sequenced.” (PMID:38412093)
  • “The top five genes with the smallest SeqSig FDR were TP53, NRAS, BRAF, PIK3CA, and USH2A.” (PMID:38412093)
  • “Loss of CDKN2A was widespread (42% of ATCs), as was loss of BRCA2 (33.6% of ATCs).” (PMID:38412093)
  • “TP53 mutation frequency varied from 0.9% of PTCs to 21.4% of co-occurring DTCs and 36.8% of ATCs.” (PMID:38412093)
  • “Every single case shared two features. First, co-DTC and co-occurring ATC shared a common clonal origin, with a common ancestor splitting to form separate lineages.” (PMID:38412093)
  • “The common ancestor harbored ~95% of CNAs, but only 19.1% +/- 7.9% of SNVs.” (PMID:38412093)
  • “Recurrent alterations in BRCA1, BRCA2, and ATM at both the SNV and the CNA level rationalize the investigations of PARP inhibitors.” (PMID:38412093)

This page was processed by crosslinker on 2026-05-04.