Involvement of KEAP1/NRF2 pathway in non-BRAF mutated squamous cell carcinoma of the thyroid

Abstract

Squamous cell carcinoma (SCC) of the thyroid is a rare tumor that is classified as an anaplastic thyroid cancer (ATC) due to its similar unresponsiveness to chemoradiotherapy and an outstandingly poor prognosis. Due to its rarity, current knowledge about this tumor is mostly based on single-case reports. The tumor-cell-origin and molecular pathogenesis remain unclear, although the presence of BRAF mutations in some cases suggest it may evolve from papillary thyroid carcinoma (PTC). Here we provide direct evidence of derivation of SCC of the thyroid from PTC, based on a unique combination of likely pathogenic mutations in KEAP1, STK11 (LKB1), and RB1 found in both tumor components, along with loss of one copy of chromosome 11 and additional somatic mutations in the SCC tumor. Transdifferentiation from PTC to SCC was also evident by immunohistochemistry. Out of eight attempted patient-derived xenografts (PDX) from advanced thyroid cancers, only one derived from thyroid SCC successfully engrafted in immunodeficient NOG mice. Untreated PDXs showed high Ki67 indices but did not reproduce the conspicuous stromal invasion of CDH1^low/SNAI2⁺/CDH2⁺ cells that characterized the primary tumor. Based on the mutation profile (NFE2L2, PIK3CA, CDKN2A, and TP53), experiments were designed to evaluate targeted drug therapy using third-passage PDX transplants. The combination of TRK and PI3K inhibitors, cabozantinib and GDC-0326, additively reduced PDX growth by nearly 90%. Remarkably, CB-839 (telaglenastat), a glutaminase inhibitor targeting metabolic rewiring downstream of NRF2 activation, was equally effective. Both combined treatment with cabozantinib + GDC-0326 and CB-839 monotherapy diminished the expression of NQO1, an NRF2 transcriptional target, in tumor cells. Glutaminase inhibition further promoted squamous differentiation in engrafted tumors. Both investigated SCC tumors were negative for BRAFV600E or any other common driver mutation of thyroid cancer. Collectively, these findings indicate that aberrant activation of the KEAP1/NRF2 pathway due to somatic mutations is a previously unrecognized feature of thyroid SCC and suggest that glutaminase inhibition may serve as a potential therapeutic option for this subgroup of ATC patients. © 2025 The Author(s). The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.