TKI-258 | Peg300

Randomized, Open-Label Phase 2 Study Comparing Frontline Dovitinib Versus Sorafenib in Patients With Advanced Hepatocellular Carcinoma

Ann-Lii Cheng,1 Sumitra Thongprasert,2 Ho Yeong Lim,3 Wattana Sukeepaisarnjaroen,4 Tsai-Shen Yang,5 Cheng-Chung Wu,6 Yee Chao,7 Stephen L. Chan,8 Masatoshi Kudo,9 Masafumi Ikeda,10 Yoon-Koo Kang,11 Hongming Pan,12 Kazushi Numata,13 Guohong Han,14 Binaifer Balsara,15 Yong Zhang,15 Ana-Marie Rodriguez,16 Yi Zhang,15 Yongyu Wang,15 and Ronnie T. P. Poon17

Abstract

Angiogenesis inhibition by the vascular endothelial growth factor receptor (VEGFR) and platelet-derived growth factor receptor (PDGFR) inhibitor sorafenib provides survival beneﬁt in hepatocellular carcinoma (HCC); however, angiogenic escape from sorafenib may occur due to angiogenesis-associated ﬁbroblast growth factor receptor (FGFR) pathway activa- tion. In addition to VEGFR and PDGFR, dovitinib inhibits FGFR. Frontline oral dovitinib (500 mg/day, 5 days on, 2 days off; n 5 82) versus sorafenib (400 mg twice daily; n 5 83) was evaluated in an open-label, randomized phase 2 study of Asian-Paciﬁc patients with advanced HCC. The primary and key secondary endpoints were overall survival (OS) and time to tumor progression (TTP) as determined by a local investigator, respectively. Patients included in the study were ineligible for surgical and/or locoregional therapies or had disease progression after receiving these therapies. The median OS (95% conﬁdence interval [CI]) was 8.0 (6.6-9.1) months for dovitinib and 8.4 (5.4-11.3) months for sorafenib. The median TTP (95% CI) per investigator assessment was 4.1 (2.8-4.2) months and 4.1 (2.8-4.3) months for dovitinib and sorafenib, respectively. Common any-cause adverse events included diarrhea (62%), decreased appetite (43%), nausea (41%), vomiting (41%), fatigue (35%), rash (34%), and pyrexia (30%) for dovitinib and palmar-plantar erythrodysesthesia syndrome (66%) and decreased appetite (31%) for sorafenib. Subgroup analysis revealed a signiﬁcantly higher median OS for patients in the dovitinib arm who had baseline plasma soluble VEGFR1 (sVEGFR1) and hepatocyte growth factor (HGF) below median levels versus at or above the median levels (median OS [95% CI]: sVEGFR1, 11.2 [9.0-13.8] and 5.7 [4.3-7.0] months, respectively [P 5 .0002]; HGF, 11.2 [8.9-13.8] and 5.9 [5.0-7.6] months, respectively [P 5 0.0002]). Conclusion: Dovitinib was well tolerated, but activity was not greater than sorafenib as a frontline systemic ther- apy for HCC. Based on these data, no subsequent phase 3 study has been planned. (HEPATOLOGY 2016; 00:000–000)

Introduction

Overexpression of ﬁbroblast growth factor receptors (FGFRs) FGFR1, FGFR2, FGFR3, or FGFR4 and corresponding FGF ligands (FGF2, FGF8, FGF17, or FGF18) has been observed in human hepatocellular carcinoma (HCC) tumors.(1-3) HCC accounts for approximately 80% of primary liver cancer cases, the majority of which are diagnosed at an advanced stage of disease and are not candidates for sur- gical interventions.(4,5) FGF2, a potent angiogenic fac- tor in HCC, has been shown to augment vascular endothelial growth factor (VEGF)-mediated HCC development and angiogenesis, and perhaps may evade resistance to VEGFR modulating agents.(6-8)
Sorafenib (Nexavar; Whippany, NJ), a multikinase inhibitor of vascular endothelial growth factor receptor (VEGFR) and platelet-derived growth factor receptor (PDGFR), was the ﬁrst effective antiangiogenic ther- apy for advanced HCC and remains the only approved treatment for this disease.(9,10) Although sorafenib was shown to improve overall survival (OS) and radiologi- cal time to tumor progression (TTP) in Asian-Paciﬁc patients with advanced HCC (median OS, 6.5 months; median TTP, 2.8 months), better systemic therapy remains an unmet need for patients with HCC in the Asia-Paciﬁc region.
Dovitinib is a potent inhibitor of FGFRs, VEGFRs, and PDGFRb, with antitumor activity mediated by a dual mechanism of action, including antiproliferative and antiangiogenic effects.(11,12) Preliminary efﬁcacy for dovitinib has been reported in patients with metastatic renal cell carcinoma, metastatic melanoma, breast cancer, multiple myeloma, and acute myeloid leukemia.(12-15) In phase 1 studies in solid tumors, the maximum toler- ated dose was determined to be 500 mg/day on a 5 days on, 2 days off schedule.(12,16) Dovitinib activity has been evaluated in multiple preclinical xenograft models in HCC. In the sorafenib-sensitive PLC5 HCC model, dovitinib inhibited tumor growth in a dose-dependent manner.(17) Furthermore, in patient-derived HCC xeno- graft models, dovitinib demonstrated antitumor activity superior to that of sorafenib and antiangiogenic effects that correlated with FGFR, PDGFRb, and VEGFR2 signaling pathway activation.(18,19) These data support an investigation of dovitinib in patients with HCC. Here, we present the efﬁcacy and safety results of a phase 2 study of frontline dovitinib versus sorafenib in patients with advanced HCC.

Materials and Methods

STUDY DESIGN AND TREATMENT

This phase 2, open-label, multicenter, randomized study conducted in the Asia-Paciﬁc region evaluated the efﬁcacy and safety of dovitinib compared with sora- fenib in patients with advanced HCC. The protocol and all amendments were reviewed by the Independent Ethics Committee or Institutional Review Board for each study site. The study was conducted according to the ethical principles of the Declaration of Helsinki.
Patients were stratiﬁed according to Eastern Coop- erative Oncology Group performance status (ECOG PS; 0 versus 1) and were randomized 1:1 to receive oral dovitinib at 500 mg/day on a 5 days on, 2 days off schedule or sorafenib at the standard dose, 400 mg continuously twice daily, until disease progression, unacceptable toxicity, death, or discontinuation for any reason. Treatment crossover was not planned; however, some patients on the dovitinib arm were given sorafenib as subsequent therapy in countries where sorafenib was clinically available, based on the investigator’s judgment. Dose adjustments and interruptions (up to 3 weeks) were allowed for patients unable to tolerate dovitinib or sorafenib study doses. Dovitinib dose reduction for tox- icity could not be re-escalated; however, when deemed appropriate by the treating physician, treatment after interruption was resumed on the next day of planned dosing, with every effort made to return to the original 5 days on, 2 days off dosing schedule as soon as possible. Local prescribing information was used to guide sorafe- nib dose and/or schedule adjustments.

STUDY POPULATION

Patients aged 18 years with an ECOG PS of 1 and advanced stage B or C HCC according to the American Association for the Study of Liver Diseases guidelines(20) and the Barcelona Clinic Liver Cancer (BCLC) staging classiﬁcation(21) were eligible for this study. Patients were required to have at least one lesion as assessed by computed tomography or magnetic reso- nance imaging scans per the Response Evaluation Cri- teria in Solid Tumors (RECIST) version 1.1. Patients included in the study were either not eligible for surgical and/or locoregional therapies or had disease progression after receiving these therapies. All patients were required to have adequate bone marrow, liver, and renal function, and a current cirrhotic status of Child–Pugh Class A (5-6 points) with no encephalopathy. Patients were excluded from the study if they had received any sys- temic HCC therapy or sorafenib-based locoregional therapy, received a liver transplant or were awaiting immediate transplantation, or were currently receiving full-dose anticoagulation treatment with therapeutic doses of warfarin or antiplatelet therapy. Patients with clinically signiﬁcant third space ﬂuid accumulation (i.e., ascites requiring tapping despite use of diuretics, or pleural effusion that either required tapping or is associ- ated with shortness of breath), or impaired cardiac func- tion or clinically signiﬁcant cardiac diseases were also excluded. Patients were required to stop treatment with any locoregional therapies, radiotherapy (except pallia- tive radiotherapy for bone lesions, within 2 weeks), and major surgery within 4 weeks before study entry. Patients were permitted to receive prophylactic or anti- viral treatment as needed, per institutional guidelines. All patients provided written informed consent to par- ticipate in the study.

EFFICACY ASSESSMENTS

Tumor response was evaluated locally at investigator sites and centrally by an independent radiologist accord- ing to RECIST version 1.1. Criteria for disease progres- sion were also based on RECIST version 1.1. All target and nontarget lesions were assessed by chest, abdomen, and pelvis computed tomography or magnetic resonance imaging scans at baseline and every 6 weeks after the start of dovitinib or sorafenib treatment until radiologi- cal progression (see Supporting Information).

SAFETY ASSESSMENTS

Adverse events (AEs) occurring on or after the ﬁrst dose day through 30 days after the end of treatment were recorded. National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03 was used for grading. Hematology, blood chemistry, thyroid and cardiac function, and vital signs were also monitored throughout the study.

PHARMACOKINETICS AND BIOMARKER ASSESSMENTS

Blood samples for full pharmacokinetics (PK) analysis were collected from patients receiving dovitinib on day 1 of week 1 (postdose 1, 3, 6, and 24 hours) and on day 5 of week 2 (predose), week 4 (predose and postdose 1, 3, 6, and 24 hours), and week 6 (predose). Full PK blood sampling was used to estimate dovitinib PK parameters in patients with advanced HCC. A minimum of 18 patients who received dovitinib were enrolled for full PK analysis; the remaining patients in the dovitinib arm par- ticipated in sparse PK blood sampling, in which only the postdose 1-hour sample was collected on postdose col- lection days (day 1 of week 1 and day 5 of week 4).
For plasma pharmacodynamics analysis, blood sam- ples were collected at baseline and predose on day 1 of week 1; day 5 of week 2, 4, and 6; day 1 (63 days) of week 13; every 12 weeks thereafter; and at the end of treatment. Circulating growth factors, including hepa- tocyte growth factor (HGF), and soluble receptors, including soluble VEGFR1 (sVEGFR1), were eval- uated as core pharmacodynamics biomarkers for FGFR and VEGFR and were measured by way of enzyme-linked immunosorbent or multiplex assays.

STATISTICAL ANALYSIS

The primary endpoint of the study was OS, deﬁned as the time from date of randomization to date of death from any cause. If survival status for a patient was unknown at the time of data cutoff, the OS was censored at the last date of contact. Patients who discontinued dovitinib or sorafenib treatment were followed for survival every 6 weeks. Final OS analysis was performed after 130 deaths were observed.
The key secondary endpoint was radiologic TTP according to the assessment of a local investigator, which was deﬁned as the time from the date of ran- domization to the date of ﬁrst documented radiological disease progression. Patients who did not have a dis- ease progression event were censored on the date of last adequate tumor assessment before the date of data analysis cutoff, start of antineoplastic therapy, or death. Death due to progression without documented radio- logical disease progression did not represent a disease progression event. Additional secondary endpoints were local investigator-assessed disease control rate (sum of patients with best overall response of complete response, partial response, or stable disease), time to deﬁnitive ECOG PS deterioration by 1 point (time from the date of randomization to either the date of deﬁnitive deterioration of the ECOG PS by 1cate- gory of the score from baseline or death, whichever came ﬁrst), and safety.
The study population used for efﬁcacy analyses included all patients who had been randomized. The safety and exploratory PK and biomarker (dovitinib arm only) populations for analysis comprised all patients who received at least one dose of study medi- cation. Descriptive statistics were used to summarize patient demographics, baseline disease characteristics, AEs, PK parameters, and biomarker data, only allow- ing for an exploratory comparison between the two arms. A Cox proportional hazard model stratiﬁed by stratiﬁcation factor (ECOG PS [0 versus 1]) was used to estimate the hazard ratio and its 95% conﬁdence interval (CI). OS was analyzed using the Kaplan– Meier method, and the median OS along with 95% CIs were determined by treatment group.

Results

PATIENT DEMOGRAPHICS AND DISPOSITION

A total of 165 patients were randomized 1:1 to dovi- tinib (n 5 82) or sorafenib (n 5 83), stratiﬁed by ECOG PS (Fig. 1). Patient demographics were well balanced between the treatment arms (Table 1). All patients (100%) were Asian-Paciﬁc, with a median age of 56 years (range, 27-83 years), and a majority were men (85%) with an ECOG PS of 0 at baseline (64%). Most patients had BCLC stage C (98%) and Child– Pugh class A (99%) HCC with unknown histological grade (70%). Many patients received prior antineoplas- tic therapy (56%), including local HCC therapies (56%), surgery (36%), or radiotherapy (7%). Of the local HCC therapies, 35% of patients received anti- neoplastic medication and 12% received two or more regimens.
All patients discontinued study treatment (Fig. 1), most frequently due to progressive disease (dovitinib, 52%; sorafenib, 73%) or an AE (dovitinib, 29%; sora- fenib, 14%). In the dovitinib arm, three patients did not receive the study drug due to AEs (pulmonary embolism [n 5 1] and decreased platelets [n 5 1]) or change in Child-Pugh score from 6 to 7 at baseline (n 5 1) that occurred after study randomization but prior to receiving the ﬁrst dose. In patients who received at least one dose of the study drug, the median duration of exposure was 2.5 (range, 0.0-11.7) months in the dovitinib arm and 3.2 (range, 0.1-23.5) months in the sorafenib arm. A majority of patients (dovitinib, 72%; sorafenib, 63%) required dose adjustment or interruption for AEs, most commonly including increased bilirubin (14%), aspartate aminotransferase (13%), or alanine aminotransferase (11%), fatigue (11%), and diarrhea (10%) in the dovitinib arm, and palmar-plantar erythrodysesthesia syndrome (PPES; 30%) and increased aspartate aminotransferase (11%) in the sorafenib arm.

EFFICACY

A total of 136 OS events were observed (dovitinib, n 5 69; sorafenib, n 5 67), with a median follow-up of 113.9 weeks (26.2 months). The median OS (95% CI) was 34.6 (28.6-39.4) weeks (8.0 [6.6-9.1] months) and 36.7 (23.3-49.3) weeks (8.4 [5.4-11.3] months) for dovitinib and sorafenib, respectively, with a hazard ratio (95% CI) of 1.27 (0.90-1.79). Kaplan–Meier curves of the two treatment arms cross between 30 and 36 weeks (6.9 and 8.3 months), with a separation in curves in favor of dovitinib before crossing and in favor of sorafenib after crossing (Fig. 2A). The drop in the Kaplan–Meier plot for the dovitinib arm between 24 and 42 weeks was not due to toxicity; patients who died within 24-42 weeks (5.5-9.7 months) from ran- domization lived for 6.9-37.1 weeks (1.6-8.5 months) after discontinuing dovitinib (Supporting Table 1).
Deﬁnitive deterioration of ECOG PS was observed in 48% of patients in both treatment arms. The median time to deﬁnitive deterioration of ECOG PS (95% CI) was 22.3 (12.6-34.0) weeks (5.1 [2.9-7.8] months) and 21.3 (13.6-not estimable) weeks (4.9 [3.1-not estima- ble] months) for patients treated with dovitinib and sor- afenib, respectively (Supporting Table 2).

SAFETY

All patients experienced at least one AE regardless of study relationship during the study (Table 3). In the dovitinib arm, the most common AEs of any grade, regardless of cause, were diarrhea (62%), decreased appetite (43%), nausea (41%), vomiting (41%), fatigue (35%), rash (34%), and pyrexia (30%). PPES (66%), diarrhea (42%), and decreased appetite (31%) were common in the sorafenib arm. In the dovitinib arm, the most common grade 3/4 AEs, regardless of cause, were increased aspartate aminotransferase (20%), increased alanine aminotransferase (17%), fatigue (14%), hypertension (13%), diarrhea (11%), increased blood bilirubin (11%), and decreased neutrophil count (10%). Common grade 3 and 4 AEs for the sorafenib arm were increased aspartate aminotransferase (24%), PPES (16%), and hypertension (11%).
Serious AEs were experienced by 51% patients in the dovitinib arm and 41% patients in the sorafenib arm (Supporting Table 3), most commonly pyrexia (13% versus 6%). Other serious AEs occurring in 4% of patients were decreased appetite, hepatic encephal- opathy, fatigue, and increased blood bilirubin in the dovitinib arm and gastrointestinal hemorrhage in the sorafenib arm. All other serious AEs occurred in 2% of patients.
During the study, including up to 30 days after the end of treatment, a total of 18 patients died, ﬁve (6%) in the dovitinib arm and 13 (16%) in the sorafenib arm (Supporting Table 4). The most common cause of death was disease progression (dovitinib, n 5 4 [5%]; sorafenib, n 5 12 [14%]). Other causes of death were coronary artery disease (dovitinib only, n 5 1) and cer- ebral hemorrhage (sorafenib only, n 5 1).

DOVITINIB PK AND BIOMARKERS

PK analysis of patients with varying degrees of hepatic function and impairment who received doviti- nib revealed that exposure was comparable between patients with mild hepatic function impairment and normal hepatic function (Table 4). Subgroup analysis revealed that higher median OS was achieved by patients in both the dovitinib arm and the sorafenib arm who had baseline plasma sVEGFR1 and HGF below the median levels compared with patients who had baseline plasma sVEGFR1 and HGF at or above the median levels (Fig. 3); however, statistical signiﬁcance for this association was achieved only with dovitinib (sVEGFR1, P 5 0.0002; HGF, P 5 0.0002). The prognostic effect of baseline HGF and sVEGFR1 was not apparent compared with TTP as determined by the local investigator.

Discussion

Currently available treatment options for patients with HCC include the VEGFR and PDGFR inhibi- tor sorafenib, which has been shown to delay HCC progression through antiangiogenic effects.(9,10) How- ever, clinical beneﬁts observed with sorafenib are usu- ally limited, as angiogenic escape from sorafenib may occur due to FGFR pathway activation.(22) In addition to VEGFR and PDGFR, dovitinib inhibits FGFR(23) and has been hypothesized to provide more effective and sustainable antitumor activity in patients with advanced HCC. However, in this randomized phase 2 study, dovitinib activity was not greater than that of sorafenib as frontline therapy in Asian-Paciﬁc patients with advanced HCC.
In this study, the median OS was similar for doviti- nib and sorafenib (34.6 versus 36.7 weeks [8.0 versus 8.4 months]). Similarly, the median TTP as deter- mined by the local investigator did not differ with dovitinib and sorafenib treatment in this study (17.6 versus 17.9 weeks [4.0 versus 4.1 months]). These results are similar to those of studies evaluating other tyrosine kinase inhibitors (TKIs) versus sorafenib, although differences in toxicity and OS have been observed (24).
It is interesting to note that the OS and TTP results in this study are higher than those reported for sorafe- nib in the phase 3 Asia-Paciﬁc HCC trial (median OS, 6.5 months; median TTP, 2.8 months).(9) Although patient demographics and disease character- istics were similar, overall baseline ECOG PS was more favorable for patients in this study (dovitinib: 0, 63%; 1, 37%; 2, 0%; missing, 0%; sorafenib: 0, 64%; 1, 35%; 2, 0%; missing, 1%) than in the phase 3 Asia-
Paciﬁc HCC trial (sorafenib: 0, 25%; 1, 69%; 2, 5%), which may have contributed to the higher activity observed for sorafenib in this study. Similar results were noted in Asian subpopulations in studies of other TKIs versus sorafenib, in which the ECOG PS of 0 ( 50%-65%) and median OS (8.5-8.9 months) were comparable to those reported here.(25-27) The effect of ECOG PS on median OS was revealed in some of these TKI versus sorafenib studies, because patients (regardless of region) with ECOG PS 0 had a higher median OS than those with ECOG PS 1.(25,26) For example, in a phase 3 study evaluating linifanib versus sorafenib in HCC, the median OS for ECOG PS 0 compared with ECOG PS 1 was 10.2 versus 8.5 months for sorafenib and 10.2 versus 7.2 months for linifanib(26); in a phase 3 study of brivanib versus sora- fenib, the median OS for ECOG PS 0 compared with ECOG PS 1 was 12.8 versus 6.5 months for sorafenib and 11.6 versus 6.6 months for brivanib.(25) Likewise, ECOG performance status was identiﬁed as a prog- nostic indicator for OS in the SHARP study,(10,28) where risk of death was reduced in patients with ECOG PS 0 (hazard ratio, 0.68 [95% CI, 0.50-0.95]) compared with ECOG PS 1 (hazard ratio, 0.71 [95% CI, 0.52-0.96]).(10) Therefore, the differences in results between this study and the phase 3 Asia-Paciﬁc HCC trial(9) may be reﬂective of the difference in patient ECOG PS. However, it should also be noted that in this and more recently reported TKI studies, patients may have tolerated sorafenib better than those in the phase 3 Asia-Paciﬁc HCC trial, because sorafe- nib AE management has improved greatly in recent years.(29)
The data presented here support the observation from randomized phase 3 studies of sorafenib in patients with HCC, in which median OS and TTP with sorafenib were lower in an Asian-Paciﬁc popula- tion compared with a population that was primarily from Europe and North America (OS, 6.5 versus 10.7 months; TTP, 2.8 versus 5.5 months).(9,10) This result may potentially be attributed to differences in symptomatic disease and extrahepatic metastases at presentation, etiology (e.g., hepatitis B or C infection, alcohol), and varying regional treatment prac- tices.(9,10,30-32) For example, prevalence of hepatitis B infection in this study (dovitinib, 72%; sorafenib 64%) was more similar to that observed in the Asian subpo- pulation ( 65%) than in the non-Asian subpopulation ( 20%) in a recently reported study comparing suniti- nib versus sorafenib, which may explain why the median OS for sorafenib in this study (8.4 months) was more similar to that observed in the Asian subpo- pulation (8.8 months) than the non-Asian subpopula- tion (15.1 months). This is consistent with the association of hepatitis B infection with poor prognosis in patients with advanced HCC, as well as additional subgroup analyses in recently reported TKI versus sor- afenib studies demonstrating improved median OS in patients without hepatitis B.(25-27,30) However, although these data indicate that population variations need to be considered when comparing activity between TKI trials, it is important to note that etiology was not found to be a signiﬁcant predictor of OS in the SHARP trial or the Asia-Paciﬁc study of sorafe- nib.(9,10) Instead, recent studies have identiﬁed tumor stage, Child-Pugh class, and as signiﬁcant predictors of OS.(33,34)
The results of the AE analysis in this study were consistent with the known safety proﬁle of doviti- nib(35-38) and did not identify any new safety concerns associated with the use of dovitinib in patients with advanced HCC. The most common AEs experienced by patients on dovitinib and sorafenib were diarrhea (62% versus 42%), decreased appetite (43% versus 31%), nausea (41% versus 19%), vomiting (41% versus 12%), fatigue (35% versus 16%), rash (34% versus 22%), pyrexia (30% versus 28%), and PPES (14% ver- sus 66%), respectively. Overall, treatment with doviti- nib was shorter and interruptions due to AEs were more frequent, potentially indicating poorer tolerance to this agent; however, this could be attributed partly to the open label design of the study.
Dovitinib exposure in patients with mild hepatic function impairment was comparable to exposure in patients with normal hepatic function. Association of median OS with sVEGFR1 and HGF baseline plasma levels achieved statistical signiﬁcance for dovitinib.
Pattern of progression has recently been noted as a key parameter in HCC. For example, a recent analysis of patients with advanced HCC treated with sorafenib showed that the emergence of new extrahepatic metas- tases was an independent predictor of poor progno- sis.(39) Although patterns of progression (e.g., intrahepatic versus extrahepatic) and their association with postprogression survival were not assessed in this study, these factors may be worth exploring in future studies evaluating dovitinib in HCC.
In conclusion, though generally well tolerated, dovi- tinib did not appear to have improved activity over sor- afenib in patients with advanced HCC in this study, and OS analyses did not demonstrate any beneﬁt. Based on the TKI-258 data presented, there are no plans for a subsequent phase 3 study.

of human hepatoma cells by an autocrine mechanism. J Hepatol 1997;27:677-687.
4) Madkhali AA, Fadel ZT, Aljiffry MM, Hassanain MM. Surgical treatment for hepatocellular carcinoma. Saudi J Gastroenterol 2015;21:11-17.
5) Thomas MB, Zhu AX. Hepatocellular carcinoma: the need for progress. J Clin Oncol 2005;23:2892-2899.
6) Ribatti D, Vacca A, Nico B, Sansonno D, Dammacco F. Angio- genesis and anti-angiogenesis in hepatocellular carcinoma. Cancer Treat Rev 2006;32:437-444.
7) Yoshiji H, Kuriyama S, Yoshii J, Ikenaka Y, Noguchi R, Hicklin DJ, et al. Synergistic effect of basic ﬁbroblast growth factor and vascular endothelial growth factor in murine hepatocellular carci- noma. HEPATOLOGY 2002;35:834-842.
8) Wang L, Park H, Chhim S, Ding Y, Jiang W, Queen C, et al. A novel monoclonal antibody to ﬁbroblast growth factor 2 effec- tively inhibits growth of hepatocellular carcinoma xenografts. Mol Cancer Ther 2012;11:864-872.
9) Cheng AL, Kang YK, Chen Z, Tsao CJ, Qin S, Kim JS, et al. Efﬁcacy and safety of sorafenib in patients in the Asia-Paciﬁc region with advanced hepatocellular carcinoma: a phase III rand- omised, double-blind, placebo-controlled trial. Lancet Oncol 2009;10:25-34.
10) Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 2008;359:378-390.
11) Lopes de Menezes DE, Peng J, Garrett EN, Louie SG, Lee SH, Wiesmann M, et al. CHIR-258: a potent inhibitor of FLT3 kinase in experimental tumor xenograft models of human acute myelogenous leukemia. Clin Cancer Res 2005;11:5281- 5291.
12) Angevin E, Lopez-Martin JA, Lin CC, Gschwend JE, Harzstark A, Castellano D, et al. Phase I study of dovitinib (TKI258), an oral FGFR, VEGFR, and PDGFR inhibitor, in advanced or metastatic renal cell carcinoma. Clin Cancer Res 2013;19:1257-1268.
13) Sarker D, Molife R, Evans TR, Hardie M, Marriott C, Butzberger-Zimmerli P, et al. A phase I pharmacokinetic and pharmacodynamic study of TKI258, an oral, multitargeted recep- tor tyrosine kinase inhibitor in patients with advanced solid tumors. Clin Cancer Res 2008;14:2075-2081.
14) Andr´e F, Bachelot T, Campone M, Dalenc F, Perez-Garcia JM, Hurvitz SA, et al. Targeting FGFR with dovitinib (TKI258): preclinical and clinical data in breast cancer. Clin Cancer Res 2013;19:3693-3702.
15) Takiuchi H, Gotoh M, Yoshida M, Kii T, Yamashita K, Sunakawa Y, et al. A phase I study of dovitinib (TKI258) in Jap- anese patients with advanced solid tumors. ASCO Annual Meet- ing Abstracts 2012; Abstract 3088.
16) Kim KB, Chesney J, Robinson D, Gardner H, Shi MM, Kirkwood JM. Phase I/II and pharmacodynamic study of doviti- nib (TKI258)-an inhibitor of ﬁbroblast growth factor receptors and VEGF receptors-in patients with advanced melanoma. Clin Cancer Res 2011;17:7451-7461.
17) Tai WT, Cheng AL, Shiau CW, Liu CY, Ko CH, Lin MW, et al. Dovitinib induces apoptosis and overcomes sorafenib resistance in hepatocellular carcinoma through SHP-1- mediated inhibition of STAT3. Mol Cancer Ther 2012;11: 452-463.
18) Huynh HT, Chow P, Soo CK, Shi M. Dovitinib (TKI258), a multikinase inhibitor of FGFR, PDGFR, and VEGFR tyrosine