Mutant-Selective Irreversible EGFR Inhibitor, Naquotinib, Inhibits Tumor Growth in NSCLC Models with EGFR-Activating Mutations, T790M Mutation, and AXL Overexpression

EGFR mutations are detected in approximately 10% of non–small cell lung cancer (NSCLC) in Caucasian patients and approximately 40% of NSCLC in East Asian patients (1–3). EGFR mutations lead to constitutive activation of EGFR signaling, including the MAPK/ERK and PI3K/AKT pathways (4, 5) and oncogenic transformation, such as increased malignant cell survival, proliferation, invasion, metastatic spread, and tumor angiogenesis in NSCLC (6, 7). The most common EGFR mutations are deletion in exon 19 (del ex19) and leucine-to-arginine substitution at amino acid position 858 (L858R) in exon 21 mutations, which together account for approximately 90% of activating EGFR mutations in NSCLC (8, 9).

The first-generation reversible EGFR tyrosine kinase inhibitors (TKI), gefitinib and erlotinib, and second-generation covalent EGFR-TKIs, afatinib and dacomitinib, dramatically improve progression-free survival in patients with advanced NSCLC. However, almost all patients eventually develop resistance to these drugs after a median of 10–13 months (10–13). The EGFR T790M mutation is the most predominant mechanism for acquired resistance, detected in approximately 50%–60% of patients with clinical resistance to these EGFR-TKIs (14–17).

Several third-generation irreversible EGFR-TKIs, including WZ4002, osimertinib, rociletinib, and olmutinib, have been developed to target activating EGFR mutations, as well as T790M mutation while sparing wild-type (WT) EGFR because second-generation EGFR-TKIs cause dose-limiting epithelial toxicities due to their activity against WT EGFR and the normal physiologic role of this kinase in skin and gastrointestinal tissues (18–23). Although osimertinib has been approved for second-line treatment in patients with NSCLC with EGFR T790M mutation, this therapeutic option is limited.

Aside from the EGFR T790M resistance mutation, activation of alternative bypassing pathways, such as the MET, HER2, or AXL signaling pathways, has also been indicated as a resistance mechanism to EGFR-TKIs (15–17, 24–29). AXL overexpression and AXL-mediated resistance have been detected in many human cancers, including breast cancer, pancreatic cancer, prostate cancer, and NSCLC (28–33). Increased AXL expression in tumors often results from pharmacologic selective pressure to multiple chemotherapies and targeted therapies (34). Inhibition of activated EGFR may trigger the tyrosine kinase switch to transactivation of AXL and its downstream pathways to maintain tumor growth, given that they share the same downstream pathways such as the MAPK/ERK and PI3K/AKT pathways (29, 35, 36). AXL may mediate acquired resistance to TKIs in an epithelial-to-mesenchymal transition setting in EGFR-mutant NSCLC (29, 32, 33). These findings suggest that AXL is a potential therapeutic target in patients with NSCLC with acquired resistance to EGFR-TKIs.

Here, we describe the preclinical in vitro and in vivo characteristics of naquotinib, which has a selective inhibitory effect on EGFR-activating mutations and T790M resistance mutation over WT. In addition, we show the inhibitory effect of naquotinib on the AXL signaling pathway, an alternative bypassing pathway in the development of resistance to EGFR-TKIs.

Naquotinib and erlotinib were prepared at Astellas Pharma Inc. as described in PCT Patent Application WO 2016/121777 and WO2001/34574, respectively. Afatinib and osimertinib were synthesized according to the PCT Patent Application WO 2002/50043 and WO 2013/014448, respectively.

NCI-H1975, HCC827, NCI-H292, and NCI-H1666 were obtained from ATCC in 2012, 2012, 2008, and 2008, respectively. PC-9 was obtained from Immuno-Biological Laboratories in 2011. II-18, A431, and Ba/F3 were obtained from RIKEN BRC Cell Bank in 2013, 2013, and 2008, respectively. Cells were cultured in RPMI1640 medium supplemented with 10% heat-inactivated FBS at 37°C in 5% CO₂ atmosphere. The cell lines used in this study were not authenticated in our laboratory but were purchased from the providers of authenticated cell lines and stored at early passages in a central cell bank at Astellas Pharma Inc., with Mycoplasma testing performed using PCR. The experiments were conducted using low-passage cultures of these stocks.

Cell viability was assessed using the CellTiter-Glo Luminescent Cell Viability Assay (Promega). A detailed protocol is provided in the Supplementary Materials and Methods.

Protein was extracted using Cell Lysis Buffer (Cell Signaling Technology) or RIPA Buffer (Thermo Fisher Scientific) supplemented with a Phosphatase Inhibitor Cocktail (Thermo Fisher Scientific) and Protease Inhibitor Cocktail (Sigma-Aldrich or Nacalai Tesque). Protein concentrations of the lysates were determined using a BCA protein assay reagent kit or Pierce 660 nm Protein Assay Kit (Thermo Fisher Scientific). Equal amounts of total protein were resolved by SDS-PAGE and transferred to a polyvinylidene fluoride membrane. After blocking at room temperature with Blocking One (Nacalai Tesque), each membrane was incubated overnight at 4°C with the primary antibodies. After washing with TBS with Tween 20 (TBST), membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody for 1 hour at room temperature. Proteins of interest were visualized by enhanced chemiluminescence using ECL-Prime (GE Healthcare) and detected using ImageQuant LAS 4000 (GE Healthcare). Antibodies for Western blotting are described in the Supplementary Materials and Methods.

Plasma and tumor concentrations of naquotinib were determined by LC-MS/MS (Shimadzu 20AD LC System 2, Shimadzu Corporation and Triple Quad 5500, AB Sciex LLC). Naquotinib and deuterated naquotinib, used as an internal standard (IS), were added to 25 μL of mouse plasma and tumor homogenate containing 0.025% 2,2-dichlorovinyl dimethyl phosphate (Wako Pure Chemical Industries, Ltd.) with 300 μL of 0.5 mol/L NaHCO₃ solution (Kokusan Chemical Co., Ltd.) and 3 mL of tert-butyl methyl ether (Kanto Chemical Co., Inc.), and the mixture was shaken and centrifuged. The organic layer was collected and evaporated to dryness under a stream of nitrogen gas at about 40°C. The residue was dissolved in 400 μL of 50 mmol/L NH₄HCO₃ buffer (Kanto Chemical Co., Inc.)/acetonitrile (Kokusan Chemical Co., Ltd.; 45:55, v/v), and 3 μL of the resulting solution was injected into the LC-MS/MS. The analytic column was a CAPCELL PAK C18 MGII (3.0 mm inner diameter × 75 mm, particle size 3 μm, Shiseido Co., Ltd.) and the mobile phase comprised of 50 mmol/L NH₄HCO₃ buffer (45%) and acetonitrile (55%). The parent and product ions, m/z 563 and m/z 323 for naquotinib, and m/z 571 and m/z 323 for the IS, respectively, were monitored using positive multiple reaction monitoring. The IS method with peak area ratio was used to determine levels of naquotinib.

All animal experimental procedures were approved by the Institutional Animal Care and Use Committee of Astellas Pharma Inc. Furthermore, Astellas Pharma Inc., Tsukuba Research Center has been awarded accreditation status by the Association for Assessment and Accreditation of Laboratory Animal Care International. HCC827, NCI-H1975, A431, PC-9vec, and PC-9AXL cells were inoculated subcutaneously into the flank of male CAnN.Cg-Foxn1nu/CrlCrlj mice. A xenograft study using LU1868 cells, a human NSCLC patient-derived xenograft (PDX) model with EGFR L858R/T790M, was conducted at Crown Bioscience Inc. LU1868 cells were inoculated subcutaneously into the flank of female BALB/c nude mice.

Mice were randomized and administered vehicle, naquotinib, erlotinib, or osimertinib at indicated doses (for details, see the figures for each experiment). Body weight and tumor diameter were measured twice a week, and tumor volume was determined by calculating the volume of an ellipsoid using the formula: length × width² × 0.5. All values are expressed as mean ± SEM.

Tissue sections (4–5 μm thick) were deparaffinized and rehydrated. For antigen retrieval, sections were heated in 10 mmol/L sodium citrate buffer, pH 6.0 at 95°C for 10 minutes. Slides were washed with TBST three times and immersed in 3% H₂O₂ for 10 minutes at room temperature to quench endogenous peroxidases. Sections were treated with 5% goat serum used as a blocking agent for 1 hour at room temperature and then incubated with phosphorylated ERK (pERK) 1/2 rabbit mAb (Cell Signaling Technology) diluted with SignalStain Antibody Diluent (Cell Signaling Technology). Sections were washed with TBST three times for 5 minutes and treated with SignalStain Boost IHC Detection Reagent (HRP, Rabbit, Cell Signaling Technology) for 30 minutes at room temperature, before reacting with 3,3′-diaminobenzidine at room temperature. Sections were counterstained with hematoxylin.

Data were statistically analyzed using Dunnett multiple comparison test or Student t test with SAS Software (SAS Institute) or GraphPad Prism (GraphPad Software). The IC₅₀ value of each experiment was calculated using Sigmoid-Emax model nonlinear regression analysis and was expressed as the geometric mean of three independent experiments.

Naquotinib is a pyrazine carboxamide–based compound with a reactive acrylamide moiety. It is structurally different from other third-generation EGFR-TKIs such as osimertinib, rociletinib, and olmutinib, which have pyrimidine-based chemical structures (refs. 37–39; Supplementary Fig. S1). In vitro biochemical enzymatic assays revealed that naquotinib inhibited EGFR del ex19, L858R, del ex19/T790M, and L858R/T790M with IC₅₀ values of 5.5, 4.6, 0.26, and 0.41 nmol/L, respectively, and an IC₅₀ value of 13 nmol/L against WT EGFR (Supplementary Table S1). In vitro cell proliferation assays in human cancer cell lines harboring mutant EGFR or WT EGFR revealed that naquotinib inhibited the growth of NCI-H1975 (L858R/T790M), HCC827 (del ex19), PC-9 (del ex19), II-18 (L858R), A431 (WT), NCI-H292 (WT), and NCI-H1666 (WT) cells with IC₅₀ values of 26, 7.3, 6.9, 43, 600, 260, and 230 nmol/L, respectively (Table 1). We further evaluated the inhibitory effect of naquotinib on EGFR and its downstream signals, ERK and AKT, by Western blotting. Naquotinib dose-dependently suppressed the phosphorylation of EGFR, ERK, and AKT in HCC827 and NCI-H1975 cells at concentrations that showed antiproliferative effects in these cell lines. In contrast, the inhibitory effect of naquotinib on these molecules in A431 cells was weaker (Fig. 1A). In addition, we compared the inhibitory effect of naquotinib on EGFR activating mutations and T790M resistance mutation under the same genetic background using Ba/F3 transfectants. Naquotinib showed comparable inhibitory effects on the growth of Ba/F3 cells expressing EGFR del ex19, L858R, del ex19/T790M, and L858R/T790M with IC₅₀ values of 4.1, 9.4, 2.1, and 4.2 nmol/L, respectively (Supplementary Table S2).

Naquotinib covalently binds to the Cys-797 residue of EGFR L858R/T790M via its acrylamide moiety (ref. 42; Supplementary Fig. S2). We therefore investigated whether the covalent binding of naquotinib to mutant EGFR results in the prolonged inhibition of EGFR phosphorylation in NCI-H1975 cells by Western blotting. Naquotinib continuously inhibited the phosphorylation of EGFR for at least 24 hours after removal of the compound (Fig. 1B).

The antitumor activity of naquotinib was evaluated in murine xenograft models using HCC827, NCI-H1975, and A431 cells and in a PDX model using LU1868 cells. In HCC827 and NCI-H1975 xenograft models, once-daily oral administration of naquotinib inhibited tumor growth with tumor regression at 10, 30, and 100 mg/kg (Fig. 2A and B) without affecting body weight (Supplementary Fig. S3). In contrast, naquotinib did not significantly inhibit tumor growth at 10 and 30 mg/kg in an A431 xenograft model, but inhibited tumor growth at 100 mg/kg (Fig. 2C). These results indicate that naquotinib is selective for mutant EGFR in cell line xenograft models. In the PDX model using LU1868 (L858R/T790M) cells, erlotinib at 50 mg/kg did not inhibit tumor growth, which is in agreement with the resistance shown in clinical settings. In contrast, naquotinib significantly inhibited tumor growth at 10, 30, and 100 mg/kg, which is comparable with its effects in cell line xenograft models (Fig. 2B and D).

As described above, robust tumor regression was observed after 14 days of treatment with naquotinib in several xenograft models with EGFR mutations. We next investigated the duration of the antitumor efficacy of naquotinib in a mutant EGFR model. In the NCI-H1975 xenograft model, while naquotinib at 10 mg/kg inhibited tumor growth for around 30 days, tumor regrowth was observed despite treatment. Naquotinib at 30 and 100 mg/kg induced tumor regression and the antitumor effect was sustained during the treatment period of 90 days with no sign of recurrence (Fig. 2E).

Naquotinib showed almost dose-proportional pharmacokinetics in both plasma and tumors after a single oral dose of the compound in the NCI-H1975 xenograft model. Interestingly, naquotinib showed a higher concentration and a longer elimination half-life in tumors than plasma (Fig. 3A and B). Evaluation of the phosphorylation of EGFR in tumors under the same conditions confirmed that the inhibitory effect of naquotinib on EGFR phosphorylation in tumors also showed dose proportionality, and revealed that inhibition continued for at least 24 hours after dosing at 100 mg/kg (Fig. 3C). These data may suggest that, in addition to the irreversible covalent binding property of naquotinib, the retention of naquotinib in tumors also contributes to its prolonged EGFR inhibition in tumors.

To investigate the inhibitory effect of naquotinib, erlotinib, and afatinib on WT EGFR in skin, we conducted IHC staining for pERK in nude mice. Repeated dosing with erlotinib at 100 mg/kg and afatinib at 25 mg/kg eliminated the pERK signal in skin compared with nontreated controls. In contrast, naquotinib did not change the staining pattern of pERK even at 100 mg/kg (Fig. 4).

In vitro biochemical enzymatic assays revealed that naquotinib inhibited AXL with IC₅₀ values of 6.2 nmol/L at 50 μmol/L ATP and 37 nmol/L at 1 mmol/L ATP (Supplementary Table S3). These results indicate that naquotinib inhibited AXL kinase activity in an ATP-competitive manner. To evaluate the inhibitory effect of naquotinib on AXL overexpression, we constructed a retrovirus-based expression vector to upregulate AXL expression in PC-9 cells (PC-9AXL) and a vector control (PC-9vec), and investigated the pharmacologic effects of EGFR-TKIs on these cells. In the cell proliferation assays, PC-9AXL cells showed resistance to erlotinib and osimertinib, whereas PC-9vec cells did not; neither compound showed cell growth inhibition in PC-9AXL cells at up to 1 μmol/L, although both inhibited cell growth in PC-9vec cells. In contrast, naquotinib showed a comparable inhibitory effect on PC-9AXL and PC-9vec cells with IC₅₀ values of 18 and 9.6 nmol/L, respectively (Fig. 5A and B). In Western blotting, all tested compounds inhibited the phosphorylation of EGFR in PC-9AXL and PC-9vec cells, but only naquotinib inhibited the phosphorylation of AXL, AKT, and ERK in PC-9AXL cells (Fig. 5C and D). In xenograft models, osimertinib showed partial tumor growth inhibition in PC-9AXL cells and tumor growth inhibition with tumor regression in PC-9vec cells. In contrast, naquotinib showed complete tumor growth inhibition in PC-9AXL cells and tumor growth inhibition with tumor regression in PC-9vec cells (Fig. 5E and F).

In this study, we characterized the efficacy of naquotinib, a pyrazine carboxamide–based compound, as a third-generation EGFR-TKI. In in vitro experiments, naquotinib inhibited cell growth and EGFR signaling in NSCLC cell lines harboring mutant EGFR with or without T790M resistance mutation (Table 1). In xenograft models, naquotinib showed robust inhibition of tumor growth in NSCLC cell lines with EGFR mutations (Fig. 2A and B) and in a PDX model, which is thought to maintain intratumoral heterogeneity, indicating clinical efficacy in more than just monoclonal cell lines (Fig. 2D). Our findings also suggest that naquotinib shows mutant selectivity and spares WT EGFR. Naquotinib inhibited cell growth and EGFR signaling in NSCLC cell lines harboring mutant EGFR with or without T790M more potently than those harboring WT EGFR in vitro (Fig. 1A; Table 1; Supplementary Table S2). While we did not examine Ba/F3 cells overexpressing WT EGFR, our collaborator previously demonstrated that naquotinib showed very low activity against these cells (40). IHC evaluation of EGFR signaling molecules in murine skin showed that erlotinib and afatinib inhibited the phosphorylation of ERK, which corresponds to the prevalence of severe skin rashes in clinical settings (11, 22). Naquotinib showed comparable mutant selectivity, sparing WT EGFR, to that of erlotinib in vitro (Table 1), and showed some inhibition of tumor growth in the A431 xenograft model (Fig. 2C). In contrast, naquotinib did not inhibit the phosphorylation of ERK in skin at 100 mg/kg (Fig. 4), the same dose that induced tumor regression in xenograft models with EGFR mutations (Fig. 2A, B, and D). Pharmacokinetic studies in the NCI-H1975 xenograft model showed that naquotinib had a higher concentration and a longer half-life in tumors than in plasma (Fig. 3A and B). This distribution may explain why naquotinib did not inhibit EGFR downstream signaling in skin. Considering that second-generation EGFR-TKIs are limited because of toxicity that likely arises from WT EGFR inhibition, the WT-sparing selectivity of naquotinib suggests that it may exert antitumor activity in patients with NSCLC harboring EGFR T790M resistance mutation without such limitations. This is in agreement with recent reports from naquotinib phase I–II clinical trials, in which skin toxicities were rarely observed (41).

We also examined the inhibitory effect of naquotinib on AXL overexpression, an alternative bypassing pathway involved in acquired resistance. We generated PC-9AXL cells stably expressing AXL. Studies have demonstrated that erlotinib, gefitinib, afatinib, rociletinib, and osimertinib do not inhibit any AXL kinase activity (37, 38, 42, 43). Consistent with these reports, PC-9AXL cells were resistant to erlotinib and osimertinib. In contrast, naquotinib inhibited AXL kinase activity in an ATP-competitive manner (Supplementary Table S3). Furthermore, naquotinib inhibited the phosphorylation of AXL, ERK, and AKT, and decreased cell proliferation more potently than erlotinib and osimertinib. In addition, in xenograft models, naquotinib showed complete tumor growth inhibition both in PC-9AXL and PC-9vec cells, whereas osimertinib showed only partial growth inhibition in PC-9AXL cells (Fig. 5). Furthermore, naquotinib inhibited the phosphorylation of AXL in PC-9AXL cells more potently than the phosphorylation of EGFR in A431 cells (Figs. 1A and 5C), and decreased the cell proliferation of PC-9AXL cells more potently than WT EGFR–expressing cells in vitro (Fig. 5A and Table 1). In xenograft models, naquotinib showed complete tumor growth inhibition in PC-9AXL cells and partial tumor growth inhibition in A431 (WT) cells (Figs. 2C and 5E). Interestingly, AXL expression has not only been observed in tumors from EGFR-TKI–resistant patients but also in those from a certain rare subgroup of patients with treatment-naïve NSCLC (28, 29, 44, 45). In addition, patients with NSCLC with EGFR L858R+T790M mutations and high AXL mRNA expression show resistance to osimertinib (46). Further investigation, such as in drug exposure–derived resistant cells with high AXL expression, is expected to reveal how AXL activation may cooperate with other genomic or genetic alterations to induce acquired resistance to EGFR-TKIs.

In conclusion, naquotinib, a mutant-selective irreversible EGFR inhibitor, shows antitumor activity against NSCLC tumors that harbor EGFR-activating mutations, as well as EGFR T790M resistance mutation and that overexpress AXL.

No potential conflicts of interest were disclosed.

Conception and design: H. Sakagami, N. Kaneko, S. Konagai, Y. Yamanaka, M. Mori, M. Takeuchi, H. Koshio, M. Hirano, S. Kuromitsu

Development of methodology: H. Sakagami, S. Konagai, M. Yuri, Y. Yamanaka

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): H. Tanaka, N. Kaneko, S. Konagai, H. Yamamoto, T. Matsuya, M. Yuri, Y. Yamanaka, H. Koshio

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): H. Tanaka, H. Sakagami, N. Kaneko, S. Konagai, M. Yuri, Y. Yamanaka, M. Mori, H. Koshio, M. Hirano

Writing, review, and/or revision of the manuscript: H. Tanaka, H. Sakagami, N. Kaneko, S. Konagai, T. Matsuya, M. Yuri, Y. Yamanaka, M. Mori, H. Koshio, M. Hirano, S. Kuromitsu

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): H. Tanaka, H. Sakagami

Study supervision: H. Sakagami, M. Mori, M. Takeuchi, M. Hirano, S. Kuromitsu

This research was funded by Astellas Pharma Inc. The authors are grateful to Masafumi Kudo, Ryuichi Takezawa, Yukihiro Takemoto, Nobuaki Shindo, and Tomohiro Eguchi for their helpful discussions and work, and thank Regina Switzer, Choice Healthcare Solutions, for logistical support during the development of this article, which was funded by the study sponsor.

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.