Expression of deiodinases-3 in HCC as predictor of poor tumor differentiation

Background/Aim The complex signalling pathways of thyroid hormones (TH) has been recently involved in the liver diseases pathogenesis. Particularly, it has been proven that the expression of deiodinases type-1 and 3 (D1-3) changes during liver injury, but their role is still poorly understood in hepatocarcinogenesis process. We aimed to evaluate the role of TH signalling pathway in liver carcinogenesis. Methods In this prospective study, 49 patients underwent liver surgery for HCC in cirrhosis (31 cases) or for other non-neoplastic indication in non-cirrhotic context (18 controls). We evaluated genes and protein expression of the main TH metabolism factors (D1, Monocarboxylate transporter-MCT8, Thyroid receptors-TR alpha and beta, Kruppel-like factor-KLF-9 and TH-responsive SPOT-14 gene), with RT-PCR and Western blot analysis in HCC, cirrhotic tissue and healthy liver samples. Results RT-PCR analysis showed a progressive statistically significant decrease of D1 (p=0.003), MCT-8 (p=0.001), and TRβ (p=0.04) mRNA expression from healthy liver to HCC. The expression of KLF9, involved in cell differentiation and proliferation, decreased accordingly (p=0.01). Similarly, the expression of SPOT14 decreased in relation to reduced TH activation (p=0.009). Western Blot analysis showed a decreased expression of D1 protein in all cirrhotic samples (p=0.01), while D3 increased in 50% of HCC (p=0.02). Among HCC patients, D3 expression was associated with poor tumor differentiation compared to D3 negative HCC subjects (p=0.007). Furthermore, a shorter Progression Free Survival was observed in D3 positive patients (19.2 months vs 47.2 months), even if not statistically significant Conclusions These preliminary data showed that D3 expression could define a more severe phenotype of HCC with poor tumor differentiation. Moreover, the signalling pathway of TH is dysregulated in HCC, with intrahepatic hypothyroidism. However, these results need to be confirmed in large cohorts