E7090, a Novel Selective Inhibitor of Fibroblast Growth Factor Receptors, Displays Potent Antitumor Activity and Prolongs Survival in Preclinical Models

The FGF signaling pathway comprises 18 ligands and 4 FGFR subtypes, FGFR1, -2, -3, and -4, which are receptor-type tyrosine kinases. Upon ligand binding, FGFRs activate an array of downstream signaling pathways, such as the MAPK and the PI3K/Akt pathways (1, 2).

In cancer, FGFRs with genetic abnormalities (gene fusion, mutation, or amplification) activate their signaling pathways, which contribute to proliferation, survival, and migration of cancer cells, tumor angiogenesis, and drug resistance (2). These abnormalities have been reported to be involved in cancer types including lung cancer (3–6), breast cancer (3, 7), multiple myeloma (1), endometrial cancer (8), gastric cancer (9), glioblastoma (10), cholangiocarcinoma (11), and urothelial cancer (3, 12, 13). Therefore, FGFRs are considered promising targets for cancer therapy, and FGFR inhibitors may serve as therapeutic agents.

The most clinically advanced FGFR inhibitors are multitarget kinase inhibitors such as ponatinib, nintedanib, dovitinib, and lucitanib. These inhibitors suppress multiple tyrosine kinases, but FGFRs are not inhibited as strongly as other kinases such as VEGFRs, platelet-derived growth factor receptor (PDGFR), or ABL1 (14). Therefore, these inhibitors seem to be unlikely to achieve the required drug exposure for sufficient inhibition of the FGFR signaling pathway in tumors in a clinical setting.

Recently, the field of FGFR targeting has progressed exponentially, and several FGFR-selective inhibitors, such as AZD4547 (15), NVP-BGJ398 (16), and JNJ-42756493 (17), are in clinical trials, including some phase II studies, targeting patients who have genetic alterations of FGFRs (14). Although these inhibitors produced partial responses in some cancer patients harboring FGFR gene abnormalities in phase I studies, all of them are still under clinical development. Especially, the contribution of these inhibitors to prolongation of survival time remains to be clarified. Thus, there is a need to develop novel compounds that show potent FGFR inhibitory activity, suitable physicochemical property, pharmacokinetics profile, and the ability to prolong survival.

Here, we report the preclinical pharmacologic analysis of a novel selective FGFR inhibitor, E7090, an orally available tyrosine kinase inhibitor with a new type of binding mode to FGFR1. The analysis included target binding, antitumor activity in several subcutaneous human cancer cell line xenograft models, and survival benefit in a preclinical lung metastasis model. E7090 is currently being investigated in a phase I study (NCT 02275910).

E7090, E7090 succinate, AZD4547, ponatinib, and PD173074 were synthesized at Eisai Co. Ltd., in accordance with previously reported procedures (18, 19) and are described in patent publications WO 2014129477 (US 20140235614), WO 2016027781, WO 2008075068, and WO 2007075869. For in vitro studies, all compounds were prepared as a 20 mmol/L stock solution in DMSO and diluted in the relevant assay media. For in vivo studies, E7090 was formulated in a 3.5% (v/v) DMSO/6.5% (v/v) Tween 80/5% (w/v) glucose solution, and the administration volume (0.2 mL/10 g body weight) was calculated from the body weight before administration. E7090 succinate was dissolved in distilled water and the administration volume (0.1 mL/10 g body weight) was calculated from the body weight before administration.

Cell lines were obtained from the ATCC (from 1993 to 2013), Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH (in 2009 and 2010), European Collection of Authenticated Cell Cultures (from 2008 to 2012), Immuno-Biological Laboratories Co., Ltd. (from 1990 to 2005), and Japanese Collection of Research Bioresources Cell Bank (in 2009 and in 2013). All cell lines were authenticated by the cell banks with short tandem repeat (STR) analysis from 2013 to 2016.

Antibodies against phospho-FGFR (Tyr653/654), FGFR1, phospho-FRS2α (Tyr436), phospho-ERK1/2 (Thr202/Tyr204), ERK1/2, phospho-AKT (Ser473), and AKT were obtained from Cell Signaling Technology. An antibody against FGFR2 was obtained from Santa Cruz Biotechnology. An antibody against β-actin was obtained from Sigma-Aldrich. Peroxidase-conjugated secondary antibodies were obtained from Cell Signaling Technology.

Inhibitory activities of E7090 against 93 purified recombinant protein kinases (including tyrosine kinases and serine threonine kinases) were examined by using an Off-Chip Mobility Shift Assay from Carna Biosciences, Inc. Briefly, a DMSO solution of E7090 was mixed with enzyme, substrate, ATP, and metals such as magnesium, calcium, or manganese under appropriate buffer conditions for each protein kinases. The readout value of the reaction control (complete reaction mixture) was set as 0% inhibition, and the readout value of the background (in the absence of enzyme) was set as 100% inhibition; the percent inhibition of each test solution was then calculated. IC₅₀ values (the half maximal inhibitory concentration) were calculated from plots of concentration versus percent inhibition by using Microsoft Excel or GraphPad Prism.

Experiments to determine the binding affinities and kinetic rate constants of interactions between compounds and FGFR1 were performed as described previously (20, 21). Briefly, FGFR1 (final concentration 6.7 nmol/L) was preincubated with the reporter probe at a concentration equal to its binding affinity (K_d) in a reaction buffer consisting of 20 mmol/L 3-morpholinopropanesulfonic acid (MOPS; pH 7.0), 1 mmol/L dithiothreitol (DTT), and 0.01% Tween20. The final reaction volume was 10 μL in black NBS 384-well polypropylene plates. After transfer of serially diluted compounds, probe displacement was monitored for 60 minutes. K_d values were calculated by using the Cheng–Prusoff equation from the IC₅₀ values obtained from the percentage displacement values at the last time point measured. Association rate constants of the inhibitors were calculated from the decay rate of probe displacement. Dissociation rate constants were determined as the product of K_d × association rate constant.

Cells (500–5,000) were seeded in a 96-well cell culture plate in RPMI1640 (WAKO) containing 10% (v/v) FBS (Sigma-Aldrich). One day after cell seeding, various concentrations of compounds were added, and the cells were incubated at 37°C for 3 to 14 days. Then Cell Counting Kit-8 (Dojindo Laboratories) was added to each well, followed by incubation for the appropriate time. The optical density (OD) of each well was measured at 450 nm (reference wavelength: 650 nm) with a SpectraMax 250 (Molecular Devices) or EnVision (Perkin Elmer) microplate reader.

SNU-16 cells were treated with the indicated concentrations of E7090 succinate in RPMI1640 containing 10% FBS for 4 hours and lysed with RIPA buffer (Sigma-Aldrich) containing protease and phosphatase inhibitors. The protein expression of the indicated proteins was detected by Western blot analysis. The blots were developed with ECL Prime Western Blotting detection reagents (GE Healthcare), and chemiluminescence was detected with an Image Analyzer LAS-4000 (Fujifilm). The chemiluminescence intensity of each band was measured using Multi Gauge software (version 3.1, Fujifilm).

The percentage of phosphorylation was determined by the following formula:

where T is the chemiluminescence intensity of p-FGFRs in the drug-treated sample lane; Bt is the chemiluminescence intensity of background in the drug-treated sample lane; C is the chemiluminescence intensity of p-FGFRs in the control sample lane; and Bc is the chemiluminescence intensity of background in the control sample lane.

siRNAs targeting mouse Fgfr1 (FGFR1 Stealth Select RNAi MSS204294), mouse Fgfr2 (FGFR2 Stealth Select RNAi MSS204298), mouse Fgfr3 (FGFR3 Stealth Select RNAi MSS204301) and mouse Fgfr4 (FGFR4 Stealth Select RNAi MSS236498), Stealth RNAi Negative Control med GC (as a control for siFGFR1, -2, -3), and Stealth RNAi Negative Control low GC (as a control for siFGFR4) were purchased from Thermo Scientific. Transfection complexes were prepared in antibiotic-free and serum-free medium by mixing 1 μmol/L of siRNA solution (final concentration 3 nmol/L) and DharmaFECT4 transfection reagents (Thermo Scientific, final concentration 0.1 μL/well) in accordance with themanufacturer's recommended protocol. Then, 4T1 cells (500 cells/well) were incubated with transfection complexes for 72 hours, and the effect of cell growth was detected using Cell-Counting Kit-8. The OD of each well was measured at 450 nm (reference wavelength: 650 nm) with EnVision (Perkin Elmer) microplate reader. The differences in OD450 to OD650 nm between the negative control-treated and siFgfr1-4 groups were analyzed by one-way ANOVA followed by the Dunnett multiple comparison test (more than three groups; the negative control vs. siFgfr1, siFgfr2, and siFgfr3) or t test (two groups; the negative control vs. siFgfr4). A value of P < 0.05 (two-sided) was considered statistically significant. Statistical analyses were performed using GraphPad Prism.

Mice were maintained under specific pathogen-free conditions and housed in barrier facilities on a 12-hour light/dark cycle, with food and water ad libitum. Animal experiments were conducted in accordance with the Institutional Animal Care and Use Committee guidelines of Eisai Co., Ltd.

Cultured human cancer cells were prepared in Hank balanced salt solution (Thermo Fisher Scientific Inc.) and mixed with an equal volume of BD Matrigel (BD Biosciences) to yield a suspension of 3.6 × 10⁷ to 9.0 × 10⁷ cells/mL. A 0.1 mL aliquot of the cell suspension was transplanted subcutaneously into the right flank region of female nude mice (7 weeks old, CLEA Japan or Japan SLC). When tumor volumes reached around 100–300 mm³, mice were selected on the basis of their tumor volumes and general condition, and were randomly divided into groups according to their tumor volumes (n = 4–6 per group). The oral administration of E7090 or E7090 succinate (6.25, 12.5, 25, or 50 mg/kg) or vehicle (control) was started on day 1, and the administration continued once daily for 11 to 14 days. The tumor was measured in two dimensions, and the volume was calculated using the following formula: tumor volume (mm³) = 1–2 length (mm) × [width (mm)]².

The differences in tumor volume between the vehicle-treated and E7090 or E7090 succinate–treated groups were analyzed by one-way ANOVA followed by the Dunnett multiple comparison test. A value of P < 0.05 (two-sided) was considered statistically significant. Statistical analyses were performed using GraphPad Prism.

SNU-16 tumors and plasma samples were collected at the indicated times after single oral administration. Nude mice bearing SNU-16 tumors were prepared as described above. Mice were given vehicle (n = 5) or indicated doses of E7090 (n = 3 per group) orally, and blood samples and tumors were collected at the indicated times.

Plasma concentrations of E7090 were determined by LC-MS/MS. In short, plasma samples and standards were deproteinized with methanol–acetonitrile [3:7 (v/v)] containing an internal standard and centrifuged. The supernatant was then filtered and analyzed by LC/MS-MS.

For signal inhibition analysis, tumors were homogenized and lysed with Cell Lysis Buffer (Cell Signaling Technology) containing protease and phosphatase inhibitors. The protein expression of the indicated proteins was detected and the band intensity was measured as described above. The ratio of phosphorylated FGFRs to total FGFR2 was calculated for each sample. Then the percentage of phosphorylation was determined by the following formula:

where T′ is the ratio of p-FGFRs/FGFR2 for the drug-treated sample lane, and C′ is the ratio of p-FGFRs/FGFR2 for the control sample lane.

Nude mice bearing SNU-16 tumors were prepared as described above and administered vehicle (n = 5) or the indicated doses of E7090 (n = 3 per group) orally. Twenty-four hours later, plasma was collected, and the amount of plasma FGF23 was measured by using an FGF23 ELISA Kit (Kainos Laboratories) in accordance with the manufacturer's protocol.

The differences in the amount of FGF23 between the vehicle-treated and E7090-treated groups were analyzed by one-way ANOVA followed by the Dunnett multiple comparison test. A value of P < 0.05 (two-sided) was considered statistically significant. Statistical analyses were performed using GraphPad Prism.

The 4T1 lung metastasis mouse model was established as previously reported (22) with some modification. Briefly, 5 × 10⁴ 4T1 cells were injected into the tail veins of female Balb/c mice (8 weeks old, Charles River). Eleven days after the cell injection, the mice were randomly divided into groups (n = 8 per group) and treated with vehicle or various concentrations of E7090 (orally, once daily). Mice were sacrificed according to the guideline for animal experiment of Eisai Co., Ltd.

Survival times between the E7090-treated group and the vehicle-treated group were compared with the log-rank (Mantel–Cox) test. Values of P < 0.01 were considered statistically significant. Statistical analyses were performed using GraphPad Prism.

E7090 (5-({2-[({4-[1-(2-Hydroxyethyl)piperidin-4-yl]phenyl}carbonyl)amino]pyridin-4-yl}oxy)-6-(2-methoxyethoxy)-N-methyl-1H-indole-1-carboxamide) was synthesized as an orally available and selective inhibitor of FGFR1, -2, and -3. The structure of E7090 succinate is shown in Fig. 1A. We used E7090 or E7090 succinate in each study. To evaluate the selectivity of E7090 against FGFR1, -2, and -3, we performed a kinase inhibition assay against 93 kinases, including wild-type and mutated members of the FGFR family, other receptor tyrosine kinases, and serine/threonine kinases (Fig. 1B; Supplementary Table S1). The IC₅₀ values of E7090 on the enzymatic activity of FGFR1, -2, -3, and -4 were 0.71, 0.50, 1.2, and 120 nmol/L, respectively. The IC₅₀ values of E7090 against mutated forms of FGFR3 (K650E and K650M) were 3.1 and 16 nmol/L, respectively. In addition, E7090 inhibited only three additional tyrosine kinases (RET, DDR2, and FLT1) with IC₅₀ values lower than 10 nmol/L among the 93 kinases. A kinome map is shown in Supplementary Fig. S1. These results indicate that E7090 is a selective inhibitor against FGFR1, -2, and -3 tyrosine kinases.

To gain insight into the mode of interaction between E7090 succinate and FGFR1, we analyzed the kinetics of the interaction by using the Proteros reporter displacement assay (23). We also determined the kinetic parameters for the interactions of FGFR1 with AZD4547 and ponatinib, which, from their cocrystal structures, are representative type I and type II FGFR1 inhibitors, respectively (24). The k_on and k_off values of E7090 succinate were 3.4 × 10⁵ s⁻¹ M⁻¹ and 8.6 × 10⁻⁴ s⁻¹, resulting in a K_d value of 2.5 nmol/L (Table 1). The kinetic parameters for the interactions of AZD4547 and ponatinib with FGFR1 were consistent with the reported features of the interaction modes of these inhibitors (24). Ponatinib had relatively slow association and dissociation kinetics, with a residence time of 57 minutes. In contrast, AZD4547 had fast association and dissociation kinetics, with a residence time of 7 minutes. E7090 succinate had interaction kinetics with FGFR1 kinases intermediate between those of the two representative inhibitors, and the residence time of E7090 succinate was 19 minutes. These data suggest that the interaction of E7090 with FGFR1 possesses unique kinetics among selective FGFR inhibitors, similar to that of type V inhibitors (23).

To evaluate the inhibitory activity of E7090 on the cellular FGFR signaling pathway, SNU-16 cells were treated with the indicated concentrations of E7090 succinate, and cell lysates were prepared for Western blotting analysis. SNU-16 is a human gastric cancer cell line which contains high copy numbers of FGFR2 and high levels of FGFR2 protein expression and tyrosine phosphorylation (25, 26). E7090 succinate inhibited FGFR phosphorylation with an IC₅₀ value of 1.2 nmol/L (Fig. 2A). E7090 treatment also inhibited the phosphorylation of FRS2α, ERK1/2, and AKT, molecules downstream of FGFRs, in a dose-dependent manner (Supplementary Fig. S2). E7090 succinate inhibited SNU-16 cell proliferation with an IC₅₀ value of 5.7 nmol/L (Fig. 2B), in a good agreement with the in vitro inhibition of FGFR phosphorylation.

E7090 was tested on a panel of 39 human cancer cell lines in cell proliferation assays (Supplementary Table S2). The IC₅₀ values observed ranged from 2 nmol/L to greater than 10,000 nmol/L (Fig. 2C). Thirteen cell lines were highly sensitive to E7090, with IC₅₀ values less than 100 nmol/L. In this group of cell lines, 12 of the 13 cell lines harbored an FGFR abnormality, including FGFR1 and/or FGFR2 amplification, FGFR1 fusion, FGFR2 mutation, FGFR3 mutation, or FGFR3 fusion. These data indicate that E7090 has selective antiproliferative activity against cancer cell lines harboring FGFR genetic abnormalities.

Next, the in vivo characteristics of E7090 were investigated in a SNU-16 subcutaneous xenograft model. Nude mice bearing SNU-16 xenografts were orally treated with various concentrations of E7090 succinate once daily for 14 days. Administration of 6.25 to 50 mg/kg E7090 succinate significantly inhibited tumor growth (Fig. 3A) without severe loss of body weight (Fig. 3B). To confirm the selective antitumor activity of E7090 against tumors harboring FGFR gene abnormalities in vivo, we performed additional in vivo experiments using several human cancer cell lines (Table 2). E7090 also had dose-dependent antitumor activity against other xenograft tumors harboring FGFR abnormalities such as FGFRs amplifications or fusions (NCI-H1581, DMS114, RT112/84, or MFM223), whereas even a dose of 50 mg/kg E7090 lacked antitumor activity against xenograft tumors with no FGFR abnormalities (MCF-7, HCC1806, HCC1427, or MDA-MB-468). These data indicate that E7090 has selective antitumor activity against xenograft tumors harboring FGFR genetic abnormalities such as amplifications or fusions.

We next performed a pharmacodynamic analysis of the level of FGFR phosphorylation in tumors in nude mice bearing SNU-16 tumors. FGFR phosphorylation was inhibited by E7090 in a dose-dependent and time-dependent manner as assessed in a Western blotting analysis. E7090 at doses of 6.25 to 50 mg/kg reduced phosphorylation of FGFR to less than 10% of control levels at 4 hours after oral administration (Fig. 3C and Supplementary Fig. S3), after which the level of FGFR phosphorylation recovered to the control level by 24 hours. In the 50 mg/kg dosing group, maximal inhibition of FGFR phosphorylation lasted for 12 hours after drug administration. We also examined the effect of E7090 on plasma FGF23 level. FGF23 belongs to a subgroup of FGF ligands that function as endocrine factors, and elevation of plasma FGF23 has been demonstrated to be a surrogate pharmacodynamic biomarker of FGFR inhibition in both nonclinical and clinical studies (27, 28). Dose-dependent elevation of plasma FGF23 was observed at 24 hours after drug administration (Fig. 3D).

In addition, a pharmacokinetic analysis of E7090 in nude mice bearing SNU-16 tumors was conducted. E7090 showed dose-dependent increase of maximum concentration (C_max) and AUC (C_max values of 9.31, 23.8, 94.2, 179.7, and 415 ng/mL and AUC values of 27.5, 84.4, 228.5, 481.6, and 1172.6 ng· h/mL for 3.13, 6.25, 12.5, 25, and 50 mg/kg, respectively; Fig. 3E).

To evaluate the effect of E7090 on survival, we established a lung metastasis model using 4T1, a mouse breast cancer cell line, in Balb/c mice as a clinically relevant model. E7090 inhibited tumor growth of 4T1 with an IC₅₀ value of 22 nmol/L in vitro (Fig. 4A). A siRNA knockdown experiment revealed that each Fgfr2 siRNA or Fgfr3 siRNA decreased cell growth of 4T1 cell, meaning that 4T1 cell growth is dependent on both FGFR2 and the FGFR3 signaling pathway (Supplementary Fig. S4). We also confirmed the dependency of 4T1 cell growth on the FGFR signaling pathway by performing a cell proliferation assay using PD173074 (Fig. 4A), another FGFR1, -2, and -3 inhibitor that has previously shown selective inhibitory activity against FGFR kinase in preclinical experiments (29). We next performed a challenge experiment (Fig. 4B). Cultured 4T1 cells were intravenously injected into mice, and 11 days after cell inoculation, engraftment of 4T1 cells was observed in the lung. On the day engraftment was observed, administration of E7090 was started. In this model, mice became moribund within 20 days, suffering from cancer-induced cachexia and pulmonary dysfunction. Daily administration of E7090 significantly prolonged the survival of mice in the 12.5 to 50 mg/kg dosing groups (Fig. 4C; Table 3).

We have developed E7090, an orally available selective inhibitor of the FGFR1, -2, and -3 tyrosine kinases with a unique kinetic profile among FGFR inhibitors. E7090 inhibited SNU-16 cell proliferation as expected from its in vitro inhibition of FGFR phosphorylation. E7090 shows selective antiproliferative activity against cancer cell lines harboring FGFR abnormalities and showed antitumor activity in mouse subcutaneous xenograft models using cell lines with dysregulated FGFR signaling. Furthermore, E7090 administration significantly prolonged the lives of mice with lung metastases. Thus, E7090 is a selective FGFR inhibitor that shows promising antitumor activities with wide therapeutic windows in preclinical models of cancers harboring FGFR abnormalities.

Some multikinase inhibitors, including several FGFR-selective inhibitors and subtype-specific inhibitors, have been developed and are in clinical studies. Although some cancer patients harboring FGFR abnormalities have had partial responses in these clinical studies, none of them have exhibited any defined proof of concept in clinical studies. We present the first evidence that the interaction of E7090 with FGFR1 has unique kinetics that is different from those of other selective FGFR inhibitors. More than 80% of approved kinase inhibitors possess the characteristics of either type I inhibitors or type II inhibitors (30). Selective FGFR inhibitors such as AZD4547 have been reported to possess the characteristics of type I inhibitors. Generally, type I inhibitors have more rapid association and dissociation kinetics, whereas type II inhibitors such as ponatinib have slow binding kinetics, leading to a prolonged residence time (31, 32). E7090 succinate associated more rapidly with FGFR1 than did ponatinib and E7090 dissociated more slowly and had a relatively longer resident time, than did AZD4547, which is a representative type I inhibitor. Our analysis revealed that E7090 possesses kinetic properties more similar to the type V inhibitors, such as lenvatinib, a VEGFR and multiple receptor tyrosine kinase inhibitor, which is the only type V kinase inhibitor approved for clinical use at this time (30). Further studies, including crystallization and costructural determination of the FGFR1–E7090 complexes, are required to fully understand the details of E7090 binding to FGFRs at the amino acid level. In addition, the relationship between the unique binding kinetics of E7090 and its efficacy should be explored.

E7090 had selective antiproliferative activity against cancer cell lines harboring FGFR abnormalities, but not all cell lines with these abnormalities were sensitive to E7090. One of the reasons is the protein expression level of FGFRs. In fact, some lung and breast cancer cell lines reported to have high copy numbers of FGFR1 (33–35) were not sensitive to E7090, and we confirmed by Western blotting analysis that these cell lines in fact expressed low levels of FGFR1 (Supplementary Fig. S5 and data not shown; ref. 36). Consistent with our results, J82, a human bladder cancer cell line harboring an FGFR3 K652E mutation, has been reported to be insensitive to FGFR inhibitors due to low protein expression (37). Another reason may be activation of other signaling pathways, which reduce activity of the FGFR inhibitor. Amplification of PDGFRA was also found in NCI-H1703 (38). KRAS and RTK alterations occurred in a near mutually exclusive pattern, but PIK3CA, PTEN alterations, and FGFR alterations overlapped in patients with lung squamous carcinoma (6) and endometrial cancer (8). In squamous cell lung cancer, MYC was coexpressed in about 40% of FGFR1-amplified tumors, and this coexpression was associated with sensitivity to FGFR inhibitors (39). These results suggest that in addition to the presence of FGFR abnormalities, the expression level of FGFRs or the status of other molecules should be investigated to select patients suitable for treatment with E7090.

E7090 showed antitumor activity in SNU-16 subcutaneous xenograft models in a dose-dependent manner without inducing severe weight loss. The C_max value at a dose of 50 mg/kg, which is maximum tolerated dose (MTD) of E7090 in mice, is 415 ng/mL. The unbound C_max value is calculated to be 12.5 ng/mL with unbound fraction in plasma (fu, 0.03), which is equivalent to 21.2 nmol/L. This result supports that there were few off-target effects even at the C_max of the MTD dose. We also confirmed the mRNA expression level of FGFR2 was more than 500-fold higher than those of FLT-1, RET, DDR2, FLT-4, PDGFRA, and KDR, which E7090 showed inhibitory activity with IC₅₀ values of lower than 20 nmol/L in cell-free kinase assay, in SNU-16 cells by using data extracted from the Cancer Cell Line Encyclopedia (CCLE) database (ref. 40; data not shown). Combined with these data, we propose that antitumor activity of E7090 in SNU-16 sc xenograft model is based on inhibition of FGFR signaling and the possibility that inhibition of other targets involved in its activity in this model is low. Administration of E7090 produced dose-dependent increases of C_max and AUC and sustained inhibition of FGFR phosphorylation in SNU-16 tumors, suggesting that the level of phosphorylated FGFRs would be an appropriate pharmacodynamic marker of target inhibition by E7090. In addition, E7090 administration also elevated plasma FGF23 levels in a dose-dependent manner 24 hours after administration, suggesting that plasma FGF23 is a good candidate biomarker for noninvasive estimation of target inhibition in humans. The dose-dependent inhibition of tumor growth, FGFR phosphorylation, and FGF23 elevation were consistent with the plasma exposure. Yamazaki and colleagues, reported an excellent example of pharmacokinetic/pharmacodynamic (PK/PD) modeling using crizotinib, an orally available, ATP-competitive dual inhibitor for anaplastic lymphoma kinase (ALK) and the hepatocyte growth factor receptor MET (also named cMet or HGFR). They characterized the PK/PD relationships among crizotinib systemic concentration, ALK or MET inhibition, and tumor growth inhibition in human tumor xenograft models in a quantitative manner and have produced clinically applicable results (41). Therefore, PK/PD modeling would be thought to be a useful approach linking drug exposure to pharmacologic responses as a function of time, providing a quantitative assessment of in vivo drug potency with mechanistic insight of drug action. We will investigate the relationship between the pharmacokinetic data and pharmacodynamic changes observed in this study, such as phosphorylation of FGFR in tumors and increased plasma FGF23 levels, in preclinical models as a method to predict human dosing in near future.

In the setting of a clinical study, the survival benefit to patients is one of the most important endpoints for a new cancer therapeutic agent. Here, we used a mouse lung metastasis model to evaluate the effect of E7090 on survival. E7090 significantly prolonged the lives of mice in a dose-dependent manner. In this model, drug administration was started when 4T1 tumors grew in the lung 11 days after tumor cells were injected into the tail veins. This means that E7090 could be effective for metastases as well as primary lesions. However, even in the 50 mg/kg dosing group the mice became moribund and were euthanized. Further analysis of the residual tumors in the lung may help us to understand the mechanism of resistance and allow us to explore suitable drugs for use in combination with E7090.

In summary, we propose that E7090 could be a novel therapeutic agent for cancer patients with abnormalities in the FGF/FGFR signaling pathway. On the basis of the preclinical results reported here, E7090 is currently being investigated in a phase I clinical trial (NCT 02275910).

No potential conflicts of interest were disclosed by the other authors.

Conception and design: S. Watanabe Miyano, Y. Yamamoto, Y. Miyajima, J. Matsui, A. Tsuruoka

Development of methodology: Y. Yamamoto

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Watanabe Miyano, Y. Yamamoto, Y. Miyajima, M. Mikamoto, T. Nakagawa, H. Kuramochi, N.H. Sugi, K. Okamoto, Y. Minoshima

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Watanabe Miyano, Y. Yamamoto, Y. Miyajima, M. Mikamoto

Writing, review, and/or revision of the manuscript: S. Watanabe Miyano, Y. Yamamoto, K. Kodama, Y. Miyajima, S. Funasaka, J. Matsui, A. Tsuruoka

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Yamamoto

Study supervision: S. Funasaka, T. Matsushima, J. Matsui, M. Iwata, T. Uenaka, A. Tsuruoka

Other (design and synthesis of E7090): S. Funasaka, S. Nagao, Y. Nakatani, Y. Karoji, I. Ohashi, Y. Yamane, T. Okada

We thank Midori Kawamata, Shizuka Taniguchi, Ayaka Ishii, Haruka Suzuki, Tomoya Shiina, Yusuke Niwa, and Syugo Hasuike for technical support. We would like to thank Norimasa Miyamoto and Kohei Sawada for suggestions on manuscript preparation.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.