Antiapoptotic BCL-2 proteins determine sorafenib/regorafenib resistance and BH3-mimetic efficacy in hepatocellular carcinoma

Sorafenib, systemic treatment for advanced hepatocellular carcinoma (HCC), and regorafenib, novel second line treatment after sorafenib failure, have efficacy limited by evasive mechanisms of acquired-drug resistance. BCL-2 proteins participate in the response to tyrosine kinase inhibitors; however, their role in HCC therapy with sorafenib/regorafenib remains uncertain. BH3-mimetic ABT-263 (navitoclax) enhanced sorafenib activity, inducing cell death via a mitochondrial caspase-dependent mechanism, after BCL-xL/BCL-2 inhibition. Sorafenib-resistant hepatoma cells (HepG2R and Hep3BR) exhibited altered mRNA expression of BCL-2 and other anti-apoptotic family members, such as MCL-1, priming drug-resistant cancer cells to death by BH3-mimetics. ABT-263 restored sorafenib efficacy in sorafenib-resistant cell lines and HCC mouse models. Moreover, in mice xenografts from patient-derived BCLC9 cells, better tumor response to sorafenib was associated to higher changes in the BCL-2 mRNA pattern. HCC non-treated patients displayed altered BCL-2, MCL-1 and BCL-xL mRNA levels respect to adjacent non-tumoral biopsies and an increased BCL-2/MCL-1 ratio, predictive of navitoclax efficacy. Moreover, regorafenib administration also modified the BCL-2/MCL-1 ratio and navitoclax sensitized hepatoma cells to regorafenib by a mitochondrial caspase-dependent mechanism. In conclusion, sorafenib/regorafenib response is determined by BCL-2 proteins, while increased BCL-2/MCL-1 ratio in HCC sensitizes drug resistant-tumors against ABT-263 co-administration. Thus, changes in the BCL-2 profile, altered in HCC patients, could help to follow-up sorafenib efficacy, allowing patient selection for combined therapy with BH3-mimetics or early switch them to second line therapy.