Introduction
Hepatocellular carcinoma (HCC) accounts for about 90% of liver cancers. Most patients with HCC are diagnosed at an advanced tumor stage when treatment options are very limited. Until 2018, when lenvatinib was approved, sorafenib was considered the gold standard in the first-line setting for the treatment of advanced HCC.1 However, choosing between these two agents remains challenging due to their low impact on survival and their similar and well-tolerated safety profiles.2 Recently, the atezolizumab-bevacizumab combination has emerged as the first-line systemic treatment, but sorafenib is still a relevant choice for refractory patients and those ineligible for immunotherapy.3 Sorafenib exerts its action through the inhibition of tumor cell proliferation and angiogenesis via the targeting of several oncogenic signaling pathways involving serine/threonine and tyrosine kinases (RAF1, BRAF, VEGFR 1, 2, 3, PDGFR, KIT, FLT3, FGFR1, and RET).4,5 However, sorafenib resistance remains a major challenge for improving the effectiveness of HCC treatment. The underlying mechanisms for inter-individual variability in response to therapy have not been fully elucidated, and no validated markers have been found that are capable of predicting clinical outcomes or sorafenib tolerability.6,7 Thus, the identification of suitable biomarkers for patient stratification for sorafenib response in HCC may potentially help physicians in guiding the selection of tailored treatments.
HCC is a hypervascular tumor in which angiogenesis plays an important role for tumor growth and progression. Among others, VEGF/VEGFR, angiopoietin (ANGPT), endothelial nitric oxide synthase (eNOS or NOS3), and hypoxia-inducible factor-1α (HIF-1α) signaling play an important role in regulating tumor angiogenesis.8 Single nucleotide polymorphisms (SNPs) in angiogenesis-related genes have been reported to influence outcomes in HCC patients treated with sorafenib.9-11
In the present study, we selected 5 SNPs in these angiogenesis-related genes for the genotyping of 34 HCC patients, of which 9 showed response (responders) to therapy and 25 no response (non-responders). Additionally, in a subgroup of HCC patients, we evaluated 1,931 SNPs and 5 copy number variations in 231 genes involved in drug absorption, distribution, metabolism, and excretion (ADME) using the DMET Plus microarray assay for the identification of new potential predictive biomarkers of response and outcome.12-14 Through a model learning process, we proceeded to apply rules to classify all patients in terms of the detected SNPs and genotypes and according to sorafenib response in order to identify a predictive genetic signature which could allow the stratification of non-responder/responder patients to sorafenib for tailored prescriptions. Furthermore, the correlation between angiogenesis- and ADME-related genes was confirmed by a cumulative genetic risk score (GRS) and by network and pathway enrichment analysis, which demonstrated the association of 8/12 identified genes placed in topological key points of the interaction networks involved in several key common biological pathways correlated to HCC and sorafenib. Our findings should be considered as a “proof of concept” to be further validated in follow-up studies for the stratification of HCC patients towards the improvement of therapeutic choices.