T790M-Selective EGFR-TKI Combined with Dasatinib as an Optimal Strategy for Overcoming EGFR-TKI Resistance in T790M-Positive Non–Small Cell Lung Cancer

Treatment for advanced non–small cell lung cancer (NSCLC) depends on the molecular characteristics of the tumor (1). Mutations in the gene encoding EGFR are present in approximately 32% of Asians and approximately 7% of individuals of other ethnicities with NSCLC, with deletions in exon 19 and an L858R point mutation in exon 21 accounting for approximately 90% of genetic alterations detected at diagnosis (2). NSCLC tumors harboring EGFR mutations, including those mentioned above, are oncogene addicted and therefore sensitive to treatment with EGFR-tyrosine kinase inhibitors (TKI).

Despite initially responding to EGFR-TKIs, most patients acquire resistance to these agents within 1 to 2 years (3–5), which is associated with secondary mutations in EGFR; the most common of these is the substitution of methionine for threonine at position 790 (T790M), which is detected in approximately 50% of patients with acquired resistance to EGFR-TKIs (6–9). T790M-selective or third-generation EGFR-TKIs, including osimertinib (AZD9291) and ASP8273 (10), have been developed to overcome T790M-related resistance. Osimertinib was approved by the FDA for use in NSCLC patients harboring a T790M mutation whose disease progressed during treatment with other EGFR inhibitors, whereas ASP8273 is currently in clinical trials to evaluate the efficacy in patients with T790M-positive EGFR-mutated NSCLC.

Despite the improvement in progression-free survival (PFS) demonstrated by osimertinib compared with a combination of pemetrexed and platinum-based chemotherapy in patients with EGFR T790M, not all patients benefit from this treatment, with most developing resistance within approximately 10 months (11). Further study is needed to optimize the treatment for T790M-positive NSCLC and improve patient survival.

We previously reported that Src family kinases (SFK) act as codrivers of resistance along with T790M and that the Src inhibitor dasatinib enhances the antitumor activity of the pan-EGFR-TKI afatinib or the T790M-selective inhibitor WZ4002 (12). Given that WZ4002 is an agent only used in preclinical models, in the current study, we investigated the efficacy of clinically relevant T790M-selective EGFR-TKIs in combination with dasatinib in T790M-positive EGFR-mutated NSCLC.

The human PC9 NSCLC cell line was provided from the Tokyo Medical University (Tokyo, Japan) in 1997. The PC9GR cell line was previously generated in our institution (12). The human H1975 NSCLC cell line was obtained from the ATCC in 2009. Cells were maintained under a humidified atmosphere of 5% CO₂/95% air at 37°C in RPMI1640 medium containing 10% FBS. The cells were routinely tested for mycoplasma using MycoAlert (LT07; Lonza) and were negative. Erlotinib, dasatinib, and osimertinib were purchased from Chemietek. Bosutinib (SKI-606) and navitoclax (ABT-263; ref. 13) were purchased from Selleck Chemicals. ASP8273 was provided by Astellas Pharma thorough a material transfer agreement.

Cells were transferred to 96-well flat-bottomed plates and cultured overnight before exposure to various concentrations of erlotinib, ASP8273, osimertinib, and dasatinib in medium containing 5% FBS for 72 hours. Cell Counting Kit-8 solution (CK04; Dojindo) was added to each well, and cells were incubated for 3 hours at 37°C before measuring absorbance with a Multiskan Spectrum instrument (Thermo Fisher Scientific); values are expressed as a percentage of the absorbance of untreated cells. The combination index (CI) was calculated using CalcuSyn v.2.1 software (Biosoft); values <1, =1, and >1 indicated synergism, additive effect, and antagonism, respectively.

Cells were washed twice with ice-cold PBS and then lysed with 1× Cell Lysis Buffer (Cell Signaling Technology) composed of 20 mmol/L of Tris-HCl (pH 7.5), 150 mmol/L NaCl, 1 mmol/L EDTA (disodium salt), 1 mmol/L EGTA, 1% Triton X-100, 2.5 mmol/L sodium pyrophosphate, 1 mmol/L β-glycerophosphate, 1 mmol/L Na₃VO₄, 1 μg/mL leupeptin, and 1 mmol/L phenylmethylsulfonylfluoride. The protein concentration of lysates was determined with a Bicinchoninic Acid Assay Kit (Thermo Fisher Scientific), and equal amounts of protein were subjected to SDS-PAGE on a 7.5% gel for analysis of intracellular signaling, or on a 12% gel for analysis of cell apoptosis (Bio-Rad). Separated proteins were transferred to a nitrocellulose membrane that was incubated with Blocking One or (for phosphorylated proteins) Blocking One-P solution (both from Nakalai Tesque) for 20 minutes at room temperature before overnight incubation at 4°C with primary antibodies against EGFR (#4405), phosphorylated EGFR [phospho- (p-)Tyr1068; #2234], Src (#2108), p-Src (Tyr416; #2101), Akt (#9272), p-Akt (Ser473; #9271), Erk (#9102), cleaved PARP (#5625), Bcl-xL (#2762), and β-tubulin (#2128), which were obtained from Cell Signaling Technology. The antibody against p-Erk (Thr202/Tyr204; sc-16982) was from Santa Cruz Biotechnology, and the antibody against β-actin (#10021) was from Sigma-Aldrich. The membrane was washed with PBS containing 0.05% Tween 20 before incubation for 2 hours at room temperature with horseradish peroxidase–conjugated secondary antibodies (NA934; GE Healthcare). Immune complexes were detected with enhanced chemiluminescence reagent (RPN3244; GE Healthcare).

The binding of Annexin-V to cells was measured using the Annexin-V-FLUOS Staining Kit (#11858777001; Roche). Cells were exposed to Accumax (#AM-105; Innovative Cell Technologies), washed with PBS, and harvested by centrifugation at 200 × g for 5 minutes. Cell pellets were resuspended in 100 μL Annexin-V-FLUOS labeling solution, incubated for 10 to 15 minutes at 15°C to 25°C, and then analyzed for fluorescence using a BD FACSCanto II system and BD FACSDiva software (BD Biosciences).

Animal procedures were performed in accordance with the Recommendations for Handling of Laboratory Animals for Biomedical Research compiled by the Committee on Safety and Ethical Handling Regulations of Laboratory Animal Experiments, Kindai University (Osaka, Japan). The study protocol was reviewed and approved by the Animal Ethics Committee of Kindai University. H1975 cells (5 × 10⁶ per mouse) were subcutaneously injected into the flank of 7-week-old female athymic nude mice (BALB/cAJcl-nu/nu) obtained from CLEA Japan. The mice were divided into four treatment groups; treatments were initiated when tumors in each group achieved an average volume of 100 to 200 mm³. Mice were treated over 3 weeks by daily oral gavage of vehicle, osimertinib (1 mg/kg), dasatinib (50 mg/kg), or osimertinib (1 mg/kg) + dasatinib (50 mg/kg); a 0.5% (w/v) aqueous solution of hydroxypropyl methyl cellulose and 50% propylene glycol was used as vehicle for osimertinib and dasatinib, respectively; the vehicle in the control group consisted of 50 μL each of 0.5% hydroxypropyl methylcellulose and 50% propylene glycol. The control group consisted for 4 mice and treatment groups of 6 mice each. Tumor volume was determined from caliper measurements of tumor length (L) and width (W) according to the formula LW²/2. Tumor size and body weight were measured twice weekly. Mice were sacrificed at the end of the treatment period, and tumor tissue was flash frozen at −80°C for immunoblot analysis.

Quantitative data are presented as mean ± SE unless otherwise indicated. The significance of differences in the Annexin-V binding assay was evaluated with the Wilcoxon rank-sum test using GraphPad Prism v.7 software (GraphPad Inc.). The repeated-measures model generated with STATA14 (StataCorp) was used to evaluate differences between groups in the in vivo study. A P value <0.05 was considered statistically significant.

PC9 human NSCLC cells harbor an EGFR exon 19 deletion, which is sensitive to the first-generation EGFR-TKIs gefitinib and erlotinib. PC9GR cell line generated from PC9 cells harbors the T790M mutation and is resistant to gefitinib and erlotinib (12). H1975 is a de novo erlotinib-resistant cell line with L858R and T790M mutations. Consistent with previous reports, we confirmed that erlotinib decreased the viability of PC9 cells (Fig. 1A), whereas PC9GR and H1975 were resistant to erlotinib treatment (Fig. 1B and C). As expected, the T790M-selective EGFR-TKIs ASP8273 and osimertinib reduced PC9GR and H1975 cell viability (Fig. 1B, C, E and F). However, IC₅₀s for ASP8273 and osimertinib in these cells were approximately 2-fold higher than those in PC9 cells, suggesting that there is room for improvement in the treatment of T790M-positive NSCLC. We previously showed that dasatinib increased antitumor efficacy in combination with a pan-EGFR-TKI afatinib, or the T790M-selective EGFR-TKI WZ4002, which has only been used in preclinical models. Given these findings, we examined the efficacy of combined dasatinib and the novel T790M-selective EGFR-TKI ASP8273 or osimertinib in T790M-positive PC9GR or H1975 cells (Fig. 1B, C, E and F; ref. 12). In agreement with our previous study, we observed increased sensitivity to the combination of dasatinib and ASP8273 or osimertinib as compared with ASP8273 or osimertinib as single agents. In contrast, in T790M-negative PC9 cells, the antitumor effects of dasatinib were unaltered when administered in combination with ASP8273 or osimertinib.

To evaluate the combined effects of dasatinib and osimertinib, CI values were calculated from the results of the cell viability assay (Table 1). All CI values in both PC9GR and H1975 cells were <1, suggesting a synergistic effect between dasatinib and T790M-selective EGFR-TKIs. As the 1:1 ratio of osimertinib and dasatinib showed the lowest CI value among the various ratios tested, this was used in subsequent experiments.

To clarify the antitumor mechanism of dasatinib in combination with ASP8273 or osimertinib, we examined the effects of these combinations on signaling pathways in PC9GR and H1975 cells (Fig. 2A–D). In both cell lines, erlotinib had a partial effect on the activation of Akt and Erk, which act downstream of EGFR, whereas ASP8273 or osimertinib caused an apparent reduction in p-Akt and p-Erk levels. Src activity was inhibited by dasatinib alone but not by either ASP8273 or osimertinib alone. Simultaneous inhibition of p-Src, p-Akt, and p-Erk was only achieved using a combination of dasatinib and ASP8273 or osimertinib.

We investigated whether the combination of dasatinib and T790M-selective EGFR-TKIs could induce apoptosis in PC9GR and H1975 cells (Fig. 3A–D). The results of the Annexin-V binding assay revealed that the rate of apoptosis was lower in cells treated with erlotinib or dasatinib as compared with ASP8273 or osimertinib alone. Although the number of Annexin-V–positive apoptotic cells was higher upon combined treatment with dasatinib and erlotinib as compared with either agent alone, it was nonetheless lower than the rate of apoptosis of cells treated with ASP8273 or osimertinib alone. Dasatinib in combination with ASP8273 or osimertinib further increased the number of Annexin-V–positive apoptotic cells as compared with either one of the T790M-selective EGFR-TKIs (P < 0.05), indicating that Src antagonism in T790M-positive NSCLC cells treated with ASP8273 or osimertinib potently induces apoptosis. As dasatinib is a multikinase inhibitor that also inhibits Abl and c-Kit, we examined the effects of the Src inhibitor bosutinib in combination with osimertinib on apoptosis with the Annexin-V binding assay (Fig. 3E and F). As expected, bosutinib in combination with osimertinib increased the number of Annexin-V–positive apoptotic cells in PC9GR and H1975 cells, indicating that Src inhibition increases the antitumor activity of T790M-selective EGFR-TKI in T790M-positive cells.

Given that dasatinib combined with ASP8273 or osimertinib induced apoptosis of T790M-positive cells, we examined the expression of apoptosis-related proteins by immunoblotting (Fig. 4A and B). The level of cleaved PARP, a marker of apoptosis, was increased in PC9GR and H1975 cells treated with dasatinib combined with osimertinib. Single-agent treatment induced Bcl-xL degradation, whereas Bcl-xL inhibition was enhanced by combining the two agents. To investigate whether enhanced apoptosis was induced by inhibition of Bcl-xL, we evaluated the combination of navitoclax (ABT-263), an inhibitor of Bcl-xL and Bcl-2, and osimertinib in PC9GR and H1975 cells (Fig. 4C and D). The number of Annexin-V–positive apoptotic cells was exacerbated by the combination of osimertinib and navitoclax in T790M-positive cells. This result supports the idea that inhibition of Bcl-xL is the cause of enhancement of apoptosis in T790M-positive cells.

The combined effect of dasatinib and osimertinib in T790M-positive cells was evaluated in vivo using a H1975 xenograft model. Tumors treated with dasatinib (50 mg/kg) or osimertinib (1 mg/kg) alone inhibited tumor progression relative to the control group, whereas treatment with a combination of the two agents suppressed tumor growth (Fig. 5A). The combination therapy was well tolerated, as evidenced by the negligible reduction in body weight (<5% of the starting weight; Fig. 5B). We compared Bcl-xL and cleaved PARP levels in posttreatment tissue samples from xenograft-bearing mice by immunoblotting (Fig. 5C) and found that coadministration of dasatinib and osimertinib increased PARP cleavage by decreasing Bcl-xL expression, consistent with our in vitro findings. This suggests that dasatinib enhances osimertinib-induced growth inhibition and apoptosis in vivo.

The results of this study demonstrate that the antitumor activity of dasatinib in T790M-positive NSCLC cells is enhanced in vitro and in vivo when administered in combination with clinically relevant T790M-selective EGFR-TKIs. Dasatinib combined with T790M-selective EGFR-TKIs synergistically inhibited the growth of T790M-positive cells, with simultaneous inhibition of Src, Akt, and Erk; it also increased apoptosis, as determined by the Annexin-V binding assay. Dasatinib is a multikinase inhibitor; however, bosutinib, another Src inhibitor, also increased apoptosis in combination with osimertinib in T790M-positive cells. In addition, in our previous study, lentiviral transfection of PC9GR and H1975 cells with SFKs (SRC, FYN) harboring gatekeeper mutations caused resistance to dasatinib combined with afatinib, indicating that dasatinib acts as an Src inhibitor in T790M-positive cells (11). Thus, concurrent inhibition of Src and EGFR signaling is an effective strategy for overcoming T790M-associated EGFR-TKI resistance. Targeting Src to overcome drug resistance has been investigated as a potential treatment approach in several cancers (14, 15). Increased Src activation has been observed in trastuzumab-resistant breast cancer cells, and the Src inhibitor saracatinib combined with trastuzumab reduced trastuzumab resistance (14). Dasatinib was also shown to sensitize KRAS-mutant colorectal cancer to cetuximab via Src inhibition (15). Although the specific mechanism underlying Src inhibition is not fully understood, it is possible that Src has a universal role in mediating multiple resistance pathways, as SFKs are transducers of mitogenic signals originating from a number of receptor tyrosine kinases, such as EGFR, HER2, FGFR, platelet-derived growth factor receptor, colony-stimulating factor 1 receptor, and hepatocyte growth factor receptor (c-Met; refs. 16–21). Src is an upstream activator and downstream effector of EGFR and is phosphorylated in about one third of lung tumors (22, 23). In our previous study, we profiled PC9 and PC9GR cell lines by immunoaffinity purification of tyrosine-phosphorylated peptides and mass spectrometry–based identification/quantification and found that SFKs act as codrivers of resistance along with T790M (12). Simultaneously targeting both SFKs and EGFR T790M mutation with dasatinib combined with afatinib or WZ4002 had synergistic antitumor effects in T790M-positive cells. The synergism between dasatinib and ASP8273 or osimertinib observed in the current study is consistent with these previous findings. However, the mechanistic link between Src activity and T790M remains unclear and warrants further study, as there was no obvious difference in Src phosphorylation between T790M-positive and negative cells (12).

T790M-positive EGFR mutants are resistant to the proapoptotic effects of T790M-selective EGFR-TKIs; this can reportedly be overcome by navitoclax (24). Bcl-xL is an antiapoptotic Bcl-2 family member that is a major determinant of the apoptotic response to PI3K and MAPK kinase blockade (25, 26). In the current study, we showed that the combination of dasatinib and T790M-selective EGFR-TKI induced apoptosis and inhibited Bcl-xL expression in T790M-positive cells, which was confirmed in a xenograft model. This finding suggests that existence of Bcl-xL is associated with intrinsic apoptosis resistance in T790M-positive cells, which can be overcome by dasatinib in conjunction with T790M-selective EGFR-TKIs, which may function via Src inhibition (27–29). Taken together, our results suggest that inhibition of Akt and Erk signaling by T790M-selective EGFR-TKIs as well as Src signaling by dasatinib leads to the suppression of Bcl-xL and induction of apoptosis. The effect of the dasatinib/osimertinib combination has been reported by another study in which Cripto-1 overexpression was found to contribute to intrinsic resistance to EGFR-TKIs via Src activation (30). However, in this earlier report, H1975 cells harboring T790M showed low Cripto-1 expression. In the current study, we demonstrated for the first time the efficacy of dasatinib combined with osimertinib in T790M-positive NSCLC, which is the main cause (>50% of cases) of EGFR-TKI resistance in clinical practice; additionally, we found that Bcl-xL downregulation is a mechanism underlying the antitumor effects of this drug combination, which has not been previously reported.

Treatment of EGFR-mutated NSCLC patients with dasatinib has not thus far been successful. In a phase II study, single-agent dasatinib showed limited antitumor effects in advanced NSCLC patients with EGFR mutation who had developed resistance to gefitinib or erlotinib (31). Although dasatinib combined with the first-generation EGFR-TKI erlotinib showed tolerable response in 59% of patients in a phase I/II study (32), there have been no published clinical trials for dasatinib combined with T790M-selective EGFR-TKI. The current treatment recommended for T790M-positive NSCLC patients is T790M-selective EGFR-TKI (33). Given the limited PFS of single-agent T790M-selective EGFR-TKI in T790M-positive NSCLC patients, the synergistic effects of dasatinib combined with T790M-selective EGFR-TKI reported here suggest a promising and novel therapeutic strategy. A clinical trial is currently under way (NCT02954523) to evaluate the combination of osimertinib and dasatinib in patients with EGFR mutation–positive NSCLC, including those with tumors harboring the T790M mutation.

J. Tanizaki has received speakers bureau honoraria from Boehringer-Ingelheim Japan Inc. and Taiho Pharmaceutical Co. Ltd. H. Hayashi has received speakers bureau honoraria from AstraZeneca and Bristol-Myers Squibb and is a consultant/advisory board member for AstraZeneca. J. Tsurutani has received speakers bureau honoraria from Chugai, Eizai, Kyowa Hakko Kirin, Novartis, and Taiho and is a consultant/advisory board member for Asahi Kasei, Daiichi Sankyo, Eizai, and Novartis. K. Nakagawa reports receiving commercial research grants from and has received speakers bureau honoraria from AstraZeneca, Boehringer Ingelheim, Chugai Pharmaceutical, and Pfizer. No potential conflicts of interest were disclosed by the other authors.

Conception and design: S. Watanabe, T. Yoshida, H. Kawakami, K. Nakagawa

Development of methodology: S. Watanabe, H. Kawakami, J. Tanizaki, K. Nakagawa

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): S. Watanabe, H. Kawakami

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): S. Watanabe, T. Yoshida, H. Kawakami, J. Tanizaki, H. Hayashi, M. Takeda, J. Tsurutani, K. Nakagawa

Writing, review, and/or revision of the manuscript: S. Watanabe, T. Yoshida, H. Kawakami, J. Tsurutani, K. Nakagawa

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): H. Kawakami, N. Takegawa, J. Tanizaki, K. Yonesaka

Study supervision: T. Yoshida, H. Kawakami, J. Tanizaki, H. Hayashi, M. Takeda, J. Tsurutani, K. Nakagawa

We thank Michiko Kitano, Haruka Sakamoto, and Yume Shinkai for providing technical support.

This work was supported by a Japan Society for the Promotion of Science KAKENHI Grant-in-Aid for Scientific Research (grant no. JP16K18461) and the Osaka Medical Research Foundation for Intractable Diseases.

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.