Novel genomic model spots genes linked to more aggressive anaplastic thyroid cancer

A new genomic predictive model has identified mutated genes that are associated with the most aggressive anaplastic thyroid cancer (ATC), according to a study.

A team of investigators conducted this retrospective cohort study using a multi-institutional, international genomic database. They included 204 ATC samples from the cBioPortal database, which were divided into training (80 percent) and validation cohorts (20 percent).

Multivariate analysis was performed to identify prognostic genes, which were used to build a point-based risk model. The investigators also carried out KEG pathway enrichment and BRAF subanalyses.

Overall, 14 genes were found to have a significant association with an increased risk of ATC, namely TET1, MAPK12, ATP10A, PIK3CA, MUC4, PNPLA2, PLD4, EGLN2, BSN, FLNC, RADIL, ZMYND8, FRAS1, and RECQL4. Using the maximally selected rank statistic, the investigators determined more aggressive (n=37) and less aggressive tumour cohorts (n=128), yielding a point threshold of 0.27.

The risk model demonstrated an acceptable predictive performance, with a C-index of 0.74. Kaplan-Meier analysis further revealed that the 1-year survival differed for more aggressive compared with less aggressive tumour patients (0 percent vs 32 percent).

In the validation cohort, survival remained significantly different between the more aggressive and less aggressive cohorts, as well as on BRAF subanalysis. Each risk cohort then underwent KEGG pathway enrichment analysis, which revealed a significantly increased enrichment across several pathways for those with more aggressive tumours.

This genomic predictive model may thus be “used in preoperative risk assessment when evaluating patients for surgery for curative intent,” the investigators said.