Thyroid cancer (TC), particularly anaplastic thyroid cancer (ATC), is the most common endocrine malignancy, with metastasis driving cancer-related deaths. CD44v6, a marker for metastatic TC cells, is currently targeted in CAR-T immunotherapy. however, residual tumors may survive due to a CD44v6- subpopulation that can reprogram into CD44v6+ cells, causing resistance and relapse. The identification of reprogramming-regulating mechanisms could help develop new therapeutic strategies.

We used established thyroid tumorigenesis model with human embryonic stem cells differentiated into thyroid progenitor cells (TPCs), focusing on double-mutant BRAF/TP53 and NRAS/TP53 TPCs. First, proliferation and invasion in vitro assay were performed and CD44v6-regulated pathways were analyzed by immunoblotting. Immunofluorescence evaluated CD44v6 expression in human samples. In vivo, metastatic ability was monitored by injecting CD44v6+ cells into immunocompromised mice. Barcode-mediated clonal tracking monitored transitions from CD44v6- to CD44v6+ . Also, CD44v6 expression changes were monitored in CD44v6---sorted cell by FACS. We evaluated CD44v6 CAR-T cells anti tumor activity, transducing isolated T-lymphocytes with lentivirus expressing human CD44v6-CAR and co-cultured them with TC cells in vitro at different effector-to-target (E:T) ratios. Cytotoxicity was assessed evaluate live TC cells after 48h using the Cell Titer-Glo Kit. In vivo, CAR-T efficiency was validated by orthotopic injection.

The highly expression of CD44v6 in double-mutant TPCs correlates with aggressive phenotype, as its signaling drives AKT phosphorylation, promoting proliferation and invasion, while its downregulation impairs these processes. Immunofluorescence further shows that CD44v6+ cells localize to metastatic sites, where they express nuclear β-catenin, reinforcing their metastatic potential. Also, only CD44v6+ cells show metastatic potential, generating primary tumors and lung metastasis in vivo. CD44v6- subclones are shown to be reprogrammable, switching to CD44v6+ upon specific stimuli as sorting-mediated perturbation with enhanced invasiveness and migration capabilities. CAR-T therapy eliminated CD44v6+ cells with 40% of BRAF/TP53 and 20% of NRAS/TP53 TPCs surviving in vitro. In vivo, CAR-T treatment reduced tumor size, metastases, and increased CAR-T cell infiltration. We plan to monitor CD44v6- resistant clones using barcode technology, isolating reprogrammed CD44v6+ cells for DNA and RNA analysis via NGS.

Our findings highlight CD44v6’s key role in TC progression, suggesting that CAR-T therapy could be effective. However, reprogramming of CD44v6- cells challenges treatment efficacy. Identifying the molecular mechanisms driving this switch could open avenues for combination therapies.