Lenvatinib inhibits angiogenesis and tumor fibroblast growth factor signaling pathways in human hepatocellular carcinoma models

Abstract

Unresectable hepatocellular carcinoma (uHCC) is one of the most lethal and prevalent cancers worldwide, and current systemic therapeutic options for uHCC are limited. Lenvatinib, a multiple receptor tyrosine kinase inhibitor targeting vascular endothelial growth factor receptors (VEGFRs) and fibroblast growth factor receptors (FGFRs), recently demonstrated a treatment effect on overall survival by statistical confirmation of noninferiority to sorafenib in a phase 3 study of uHCC. Here, we investigated mechanisms underlying the antitumor activity of lenvatinib in preclinical HCC models. In vitro proliferation assay of nine human HCC cell lines showed that lenvatinib selectively inhibited proliferation of FGF signal-activated HCC cells including FGF19-expressing Hep3B2.1-7. Lenvatinib suppressed phosphorylation of FRS2, a substrate of FGFR1-4, in these cells in a concentration-dependent manner. Lenvatinib inhibited in vivo tumor growth in Hep3B2.1-7 and SNU-398 xenografts and decreased phosphorylation of FRS2 and Erk1/2 within the tumor tissues. Lenvatinib also exerted antitumor activity and potently reduced tumor microvessel density in PLC/PRF/5 xenograft model and two HCC patient-derived xenograft models. These results suggest that lenvatinib has antitumor activity consistently across diverse HCC models, and that targeting of tumor FGF signaling pathways and anti-angiogenic activity underlies its antitumor activity against HCC tumors.

Keywords: Lenvatinib; angiogenesis; fibroblast growth factor; hepatocellular carcinoma; vascular endothelial growth factor.

© 2018 The Authors. Cancer Medicine published by John Wiley & Sons Ltd.

Figures

Figure 1

In vitro antiproliferative activity of lenvatinib and sorafenib in nine human HCC cell lines. (A) Inhibition of HCC cell proliferation by lenvatinib and sorafenib. Data are presented as means of three independent experiments performed in triplicate. (B) IC 50 values for HCC cells, placed in the order of ascending values for lenvatinib. (C) Expression levels of FGF19, FGFR1–4, and β‐Klotho, as determined by Western blot analysis. Multiple bands represent different isoforms and/or post‐translational modifications of FGFRs and β‐Klotho.

Figure 2

Inhibitory effects of lenvatinib and sorafenib on FGF signaling pathways in human HCC cells. (A) Hep3B2.1‐7 cells were treated with lenvatinib, sorafenib, or vehicle (DMSO) for 1 h, followed by stimulation with FGF19 for 5 min. (B) HuH‐7 cells and (C) SNU‐398 cells were treated with each drug or vehicle for 1 h. Phosphorylated FRS2 (p‐FRS2), FRS2, and β‐actin were detected by Western blot analysis.

Figure 3

Effects of lenvatinib and sorafenib in human HCC xenograft models. Mice bearing xenograft tumors were orally administered lenvatinib (3–30 mg/kg), sorafenib (10, 30 mg/kg), or the corresponding vehicle for 7 (Hep3B2.1‐7) or 11 (SNU‐398) days. (A, B) Antitumor activity in the Hep3B2.1‐7 (A) and SNU‐398 (B) xenograft models. Data are means ± SEM (n = 8) *P < 0.05, ***P < 0.001, ****P < 0.0001 versus vehicle control. (C–F) Inhibitory activity against phosphorylation of FRS2 and Erk1/2 in the xenograft‐derived tumors. Tumors were collected 2 h after single treatment with lenvatinib (C, E) or sorafenib (D, F) at the indicated doses. Lysates of Hep3B2.1‐7 tumors (C, D) and SNU‐398 tumors (E, F) were then subjected to Western blot analysis.

Figure 4

Antitumor and anti‐angiogenic activities of lenvatinib and sorafenib in the PLC/PRF/5 xenograft model. Mice bearing xenograft tumors were orally administered lenvatinib (LEN; 3–30 mg/kg), sorafenib (SOR; 10, 30 mg/kg), or the corresponding vehicle (Veh) for 7 days. (A) Tumor growth curves of PLC/PRF/5 xenograft model. Data are means ± SEM (n = 6). (B, C) Formalin‐fixed paraffin‐embedded sections of tumors collected on Day 8 were stained with anti‐CD31 antibody to visualize microvessels in the tumors. (B) Representative images of the tumor microvessels. Bar = 100 μm. (C) Quantification of MVD. Data are presented as means + SEM (n = 6). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 versus vehicle control.

Figure 5

Effects of lenvatinib and sorafenib in the HCC PDX‐derived cell line (LIXC‐012) xenograft model. Mice bearing LIXC‐012 tumors were orally administered lenvatinib (3–30 mg/kg), sorafenib (10, 30 mg/kg), or the corresponding vehicle (Veh). Formalin‐fixed paraffin‐embedded sections of tumors collected the day after the last dosing were stained with anti‐Ki‐67 antibody to visualize proliferating tumor cells. (A) Tumor growth curves of the LIXC‐012 xenograft model. Data are means ± SEM (n = 8). *P < 0.05, ***P < 0.001 versus vehicle control on Days 1–11 (two‐way analysis of variance followed by Bonferroni's multiple comparison test). (B) Numbers of Ki‐67 positive cells (%). Data are means + SEM (n = 7 [Veh for lenvatinib] to 8). *P < 0.05, ****P < 0.0001 versus vehicle control.

Figure 6

Antitumor and anti‐angiogenic activities of lenvatinib and sorafenib in two HCC PDX models. Mice bearing tumors were orally administered lenvatinib (10, 30 mg/kg), sorafenib (30 mg/kg), or vehicle (Veh; 3 mmol/L HCl only) for 28 days. (A) Tumor growth curve of LI0050 model. In this study sorafenib was poorly tolerated, with the death of five mice (one on Day 10 and four on Day 19) and multiple dose suspensions, and 30 mg/kg lenvatinib treatment of one mouse was suspended for 13 days owing to transient BWL on Day 15. Data are means ± SEM; n = 15 (vehicle control and lenvatinib groups) or n = 10–15 (sorafenib group). NA, not applicable. (B) Tumor growth curve of LI0334 model. Data are means ± SEM (n = 15). NS, not significant. (C, D) MVD in the LI0050 (C) and LI0334 (D) models. Data are means + SEM. ***P < 0.001, ****P < 0.0001 versus vehicle control.