EGFR-mutated lung cancer accounts for a significant proportion of lung cancer cases worldwide. For these cases, osimertinib, a third-generation EGFR tyrosine kinase inhibitor, is extensively used as a first-line or second-line treatment. However, lung cancer cells acquire resistance to osimertinib in 1 to 2 years. Thus, a thorough clarification of resistance mechanisms to osimertinib is highly anticipated. Recent next-generation sequencing (NGS) of lung cancer samples identified several genetically defined resistance mechanisms to osimertinib, such as EGFR C797S or MET amplification. However, nongenetically defined mechanisms are not well evaluated. For a thorough clarification of osimertinib resistance, both genetic and nongenetic mechanisms are essential. By using our comprehensive protein phosphorylation array, we detected IGF1R bypass pathway activation after EGFR abolishment. Both of our established lung cancer cells and patient-derived lung cancer cells demonstrated IGF2 autocrine-mediated IGF1R pathway activation as a mechanism of osimertinib resistance. Notably, this resistance mechanism was not detected by a previously performed NGS, highlighting the essential roles of living cancer cells for a thorough clarification of resistance mechanisms. Interestingly, the immunohistochemical analysis confirmed the increased IGF2 expression in lung cancer patients who were treated with osimertinib and met the established clinical definition of acquired resistance. The findings highlight the crucial roles of cell-autonomous ligand expression in osimertinib resistance. Here, we report for the first time the IGF2 autocrine-mediated IGF1R activation as a nongenetic mechanism of osimertinib resistance in lung cancer at a clinically relevant level.

Implications:

Using comprehensive protein phosphorylation array and patient-derived lung cancer cells, we found that IGF2 autocrine-mediated IGF1R pathway activation is a clinically relevant and common mechanism of acquired resistance to osimertinib.

In non–small cell lung cancer (NSCLC), somatic mutations in the tyrosine kinase domain of EGFR have been identified in a significant subgroup of patients (1–3). In general, somatic mutations in the EGFR tyrosine kinase domain activate EGFR by promoting its active conformation (4–7). EGFR-mutated lung cancer cells are dependent on activated EGFR signaling for their survival and proliferation (8, 9). Multiple EGFR tyrosine kinase inhibitors have been developed (first generation: gefitinib and erlotinib; second generation: afatinib and dacomitinib; and third generation: osimertinib and rociletinib) to target EGFR signals in NSCLC. Multiple clinical trials involving EGFR-TKIs have demonstrated significant improvement in the prognosis of lung cancer patients with EGFR mutations (10–13).

The third-generation EGFR-TKI osimertinib was first approved for clinical use in the treatment of EGFR exon 20 T790M-positive NSCLC patients resistant to first- or second-generation EGFR-TKIs (14). Encouraging survival data and high tolerability have recently allowed the approval to be extended to first-line regimens in many countries (15). Currently, a significant number of EGFR mutation–positive lung cancer patients are treated with osimertinib as first- or second-line regimen. However, lung cancer cells acquire resistance to osimertinib in approximately 1 to 2 years. The median progression-free survival of osimertinib for the first- and second-line settings is 18.9 and 10.1 months, respectively (14, 15). Therefore, uncovering the mechanisms of osimertinib resistance is highly anticipated.

Next-generation sequencing (NGS) for tumor or serum samples from lung cancer patients on acquiring resistance to osimertinib clarified several genetically defined resistance mechanisms, including EGFR C797S, HER2, or MET amplification (16). However, nongenetically defined resistance mechanisms are not well clarified. For a thorough clarification of osimertinib resistance, both genetically and nongenetically defined mechanisms are essential.

In this study, by using our comprehensive protein phosphorylation array and patient-derived living cancer cells, we detected a nongenetically defined mechanism: IGF2 autocrine–mediated IGF1R bypass pathway activation as a novel and clinically relevant mechanism of osimertinib resistance. Additionally, we confirmed the increased expression of IGF2 in lung cancer patients who acquired resistance to osimertinib.

Here, we propose a clinically relevant mechanism of osimertinib resistance and a preclinical rationale for targeting the IGF2-IGF1R pathway in NSCLCs that have acquired resistance to this drug.

Cell lines

The PC9 [EGFR exon 19 deletion (delE746-A750)] and PC9ER [EGFR exon 19 deletion (delE746–A750) + T790M] cells were kindly gifted by S. Kobayashi (Beth Israel Deaconess Medical Center). PC9ER cells became resistant to erlotinib after chronic exposure to this molecule and acquired EGFR T790M mutation. H1975 [EGFR L858R + T790M] was purchased from the American Type Culture Collection (#CRL-5908DQ). Gefitinib-resistant PC9 (PC9GR), rociletinib-resistant PC9 (PC9COR), osimertinib-resistant PC9 (PC9AZDR), rociletinib-resistant H1975 (H1975COR), and osimertinib-resistant H1975 (H1975AZDR) were established as in previous studies (17, 18). All cell lines were cultured in RPMI-1640 growth medium supplemented with 10% fetal bovine serum (FBS) at 37°C in a humidified 5% CO2 incubator.

Establishment of patient-derived cancer cells from clinical specimens and collection of pleural effusion supernatants

Patient-derived cancer cells were established according to a previous report (19). Collected patient-derived fluid samples were filtered using a 40-μm cell strainer and centrifuged (500 × g, 5 minutes, 20°C). The pellets were suspended in RPMI-1640 growth medium supplemented with 10% FBS and transferred into standard cell culture dishes. Supernatants were stored at −80°C until use. Clinical information of the established cell lines is provided in Supplementary Table S1. All patients signed informed consent to participate in a protocol approved by the Keio University School of Medicine Review Board, giving permission for research to be performed on their samples. Cell lines were sequenced to confirm the presence of EGFR mutations identified through clinical testing of biopsy specimens from the same patient.

Reagents

Gefitinib (G-4408), erlotinib (E-4997), and rociletinib (R-3692) were purchased from LC Laboratories. AZD4547 (S2801), galunisertib (S2230), nintedanib (S1010), linsitinib (S1091), osimertinib (S7297), and crizotinib (S1068) were purchased from Selleck Chemicals. Xentuzumab was kindly provided by Boehringer Ingelheim. Recombinant human IGF2 (100-12) was purchased from PeproTech. Total EGFR antibody (#4267), phospho-IGF1R (Tyr1135/1136)/IR (Tyr1150/1151) antibody (#3024) (WB), total AKT antibody (#9272), phospho-AKT (Ser473/D9E) antibody (#4060), total p44/42 MAPK antibody (#9102), and phospho-p44/42 MAPK (Thr202/Tyr204) antibody (#9191) were purchased from Cell Signaling Technology. Phospho-EGFR (Tyr1068) antibody (#44-788G), phospho-IR/IGF1R (Tyr1162/Tyr1163) antibody (#44-804G) (IHC), and IGF2 antibody (#MA5-1709; WB; IHC) were purchased from Thermo Fisher Scientific. Actin antibody (#A5441) was purchased from Sigma-Aldrich.

Array design for phosphoproteome analysis

We developed an original protein array focused on phosphorylation in signal transduction pathways with reference to previous work (20). Furthermore, we also developed a computational system for the measured data on the designed array. The designed array, therefore, enabled us to estimate active pathways as well as the degree of phosphorylation of proteins (substrates) in the lysate. Further details of the protein array are described in the Supplementary Methods.

Cell proliferation assay

The MTS assay was performed using CellTiter 96 Aqueous One Solution Assay (Promega) following the manufacturer's protocol. Cells (2 × 103) were seeded in 96-well plates and treated with the relevant agents. Control cells were treated with the same concentration of the vehicle, DMSO. Absorbance was measured at 72 or 120 hours after treatment. Synergy was evaluated using the Bliss independence model previously reported (21–23).

Phospho-receptor tyrosine kinase (phospho-RTK) array

The human phospho-RTK array kit was purchased from R&D Systems and used following the manufacturer's instruction, with 150 μg of protein for each experiment. Signal intensity was calculated using LumiVision Analyzer software.

Western blot analysis

Cells were lysed in Cell Lysis Buffer (Cell Signaling Technology). Proteins were quantified using BCA protein assay (Thermo Fisher Scientific), and equal amounts of protein (by weight) were separated by SDS-PAGE and transferred to polyvinylidene fluoride membranes. Next, the membranes were incubated overnight with primary antibodies at 4°C, followed by incubation with secondary antibodies for 1 hour. Immunoreactive proteins were visualized with LumiGLO reagent and peroxide (Cell Signaling Technology) and exposed to X-ray films.

Apoptosis assay

Cells were seeded in 6-well plates (50,000/well) and treated with various drugs individually or in combination for 72 hours. Control cells were treated with DMSO. The apoptotic status was analyzed using the TACS Annexin V–FITC Apoptosis Detection Kit (R&D Systems) following the manufacturer's instructions. The proportion of apoptotic cells was evaluated by flow-cytometric analysis using the BD FACSCalibur system (Becton Dickinson).

Quantitative RT-PCR (qRT-PCR) and quantitative PCR

Total RNA was isolated from cultured cells using an RNeasy Mini Kit (Qiagen), and genomic DNA was isolated using a DNeasy Blood and Tissue kit (Qiagen). RNA was subjected to reverse transcription using the High-Capacity RNA-to-cDNA Kit (Thermo Fisher Scientific) following the manufacturer's protocol. Quantitative RT-PCR was performed using fluorescent SYBR Green and Applied Biosystems 7500 Fast (Thermo Fisher Scientific). Human GAPDH was used to normalize input cDNA. The LINE1 repetitive element was used as a reference for MET and IGF2 copy-number analysis. Primer sequences are provided in Supplementary Tables S2 and S3.

ELISA for IGF2

The amount of IGF2 in the pleural effusion was assayed using Human IGF2 Quantikine ELISA kit (R&D Systems) based on the manufacturer's instructions.

siRNA-mediated gene knockdown

siRNAs targeting IGF2 (s7214 and s7216), IGF1R (s7211 and s7212), and negative control (4390843) were purchased from Thermo Fisher Scientific. Cells (2 × 105) were transfected with 100 nmol/L of the respective siRNAs using siLentFect transfection reagent (Bio-Rad) following the manufacturer's instructions. Forty-eight hours after transfection, cells were plated and incubated for the appropriate time periods prior to the performance of the different types of experiments (cell proliferation assay or qRT-PCR). We used qRT-PCR to confirm the efficiency of gene knockdown.

Targeted sequencing

Tumor genomic DNA was extracted from patient-derived resistant cell lines using DNeasy Blood and Tissue kit (Qiagen). Library preparation for each sample was performed using Ion AmpliSeq Library kit 2.0 (Thermo Fisher Scientific) according to the manufacturer's instructions. Library quantification was performed by the Ion Library TaqMan Quantitation kit (Thermo Fisher Scientific), and each library was amplified and enriched by the OneTouch system using Ion PI Hi-Q OT2 200 kit (Thermo Fisher Scientific) following the manufacturer's instructions. Libraries were sequenced with Ion AmpliSeq Comprehensive Cancer Panel (Thermo Fisher Scientific), which could target 409 genes with ∼16,000 amplicons. Sequencing read mapping and variant calling were performed with Ion Torrent Suite v5.0.4 (Ion Torrent PGM server; Thermo Fisher Scientific). Sequence reads were aligned to the human reference genome UCSC hg19. The presence or absence of hotspot genes, such as mutations registered in the Catalogue of Somatic Mutations in Cancer, was identified.

Immunohistochemical analysis

All specimens were acquired from patients under the guidance of the Keio University School of Medicine. Written informed consent was obtained in all cases, and the protocols were approved by the review board. All ethical regulations were followed. Tissues were fixed in formalin and embedded in paraffin. Tissue sections of patient-derived xenograft and patients' samples were cut into 4-μm sections. The paraffin sections were deparaffinized in xylene and rehydrated in multiple concentrations of alcohol. Antigens were unmasked by boiling in citrate buffer (pH 6) for 10 minutes. Sections were then sequentially incubated with 3% H2O2 for 10 minutes, blocking buffer for 60 minutes, and primary antibodies overnight at 4°C. Both IGF2 and pIGF1R antibodies were diluted 200-fold. The reaction was visualized using the ABC kit (Vector Laboratories), and the color was developed using the DAB Color kit (Vector Laboratories). The slides were counterstained with hematoxylin. Expression levels of IGF2 were quantified in the cytoplasm by a lung cancer pathologist (Dr. Hayashi) using a four-value intensity score (0, 1+, 2+, and 3+) and the percentage (0%–100%) of the stained extension. A final score was obtained by multiplying the intensity score by the extension values (range, 0–300). TBS-T without any antibody was used as a negative control for immunostaining.

Mouse xenograft model

All animal experiments were approved by the Laboratory Animal Center, Keio University School of Medicine (approval no. 12115). Twenty-four female NOD/SCID mice (6 weeks old, approximately 20 g/mouse) were purchased from Charles River. KOLK43 cells, suspended in RPMI-1640 growth medium (1 × 106 cells/50 μL), were mixed with 50 μL of Matrigel (Corning) and implanted subcutaneously into the mouse. To estimate the tumor size, the length (L) and width (W) of each tumor were measured using calipers, and the tumor volume (TV) was calculated as TV = (L × W2). Once average TV reached approximately 150 mm3, the mice were randomized to receive the vehicle alone, osimertinib (5 mg/kg; 5 days per week, orally), linsitinib (30 mg/kg; 5 days per week, orally), or a combination of both (6 mice per group). Each drug was prepared in the following solvent: osimertinib (1% DMSO and 30% PEG 300) and linsitinib (30% PEG 400, 0.5% tween 80, and 5% propylene glycol). During the drug treatment, animals showed no sign of toxicity, such as body weight loss (>10%), decreased food intake, or diarrhea. Animals were humanely exsanguinated, and tumor tissues were harvested.

Statistical analysis

Statistical analysis was performed using GraphPad Prism software, version 7.0 (GraphPad Software). Two-sided Student t test was used for comparison between two groups. Values of P < 0.05 were regarded as statistically significant.

IGF1R bypass pathway activation as a compensatory response to osimertinib

To elucidate the dynamic regulation of multiple pathways in response to EGFR inhibition, we developed a comprehensive protein phosphorylation array, in which the phosphorylation intensity of 1,205 proteins can be measured (Fig. 1A). By combining mathematical computation approaches, this array could also estimate the activity of 307 signaling pathways defined in the Reactome and Kyoto Encyclopedia of Genes and Genomes databases. Using this array, we performed analysis for EGFR-dependent PC9 cells, a well-known lung cancer cell line harboring an EGFR exon 19 deletion, treated with an EGFR-TKI, erlotinib. Of note, erlotinib was selected as an EGFR-TKI in this experiment due to its high selectivity for EGFR (24).

Figure 1.

IGF2 overexpression–mediated IGF1R activation contributes to resistance development in PC9AZDR cells. A, Schematic of the analysis workflow. PC9 cells were treated with 0.1 μmol/L erlotinib for 0, 1, 4, and 24 hours before the extraction of the cell lysates. Data for 0 hours of erlotinib exposure were used for comparison. Phosphorylation degrees of selected proteins were determined by the protein array. The activity of the 377 pathways was estimated by a two-computational method, so-called network screening and pathway analysis. B, MTS cell proliferation assays following treatment with the indicated concentration of osimertinib with or without linsitinib in PC9AZDR cells. The assay was performed after 72 hours of incubation. Error bars, SD. *, P < 0.05; **, P < 0.01. C, Flow-cytometric data for PC9AZDR cells treated with DMSO, osimertinib, linsitinib, and osimertinib/linsitinib in combination for 48 hours. The numbers (%) indicate the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. D, Immunoblotting of EGFR and IGF1R/IR signaling pathway components. PC9, PC9AZDR, and PC9GR cells was treated with indicated drug combinations for 4 hours before protein extraction. E, Heat map for relative gene expression in PC9AZDR cells compared with that in PC9 cells. F, MTS cell proliferation assay in PC9AZDR cells following treatment with xentuzumab 100 μg/mL. The assay was performed after 72 hours of the incubation period. Error bars, SD. **, P < 0.01. G, Immunoblotting of indicated proteins in PC9, PC9AZDR, and PC9GR cells treated with DMSO, osimertinib 1 μmol/L, xentuzumab 100 μg/mL, or combination of both for 4 hours.

Figure 1.

IGF2 overexpression–mediated IGF1R activation contributes to resistance development in PC9AZDR cells. A, Schematic of the analysis workflow. PC9 cells were treated with 0.1 μmol/L erlotinib for 0, 1, 4, and 24 hours before the extraction of the cell lysates. Data for 0 hours of erlotinib exposure were used for comparison. Phosphorylation degrees of selected proteins were determined by the protein array. The activity of the 377 pathways was estimated by a two-computational method, so-called network screening and pathway analysis. B, MTS cell proliferation assays following treatment with the indicated concentration of osimertinib with or without linsitinib in PC9AZDR cells. The assay was performed after 72 hours of incubation. Error bars, SD. *, P < 0.05; **, P < 0.01. C, Flow-cytometric data for PC9AZDR cells treated with DMSO, osimertinib, linsitinib, and osimertinib/linsitinib in combination for 48 hours. The numbers (%) indicate the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. D, Immunoblotting of EGFR and IGF1R/IR signaling pathway components. PC9, PC9AZDR, and PC9GR cells was treated with indicated drug combinations for 4 hours before protein extraction. E, Heat map for relative gene expression in PC9AZDR cells compared with that in PC9 cells. F, MTS cell proliferation assay in PC9AZDR cells following treatment with xentuzumab 100 μg/mL. The assay was performed after 72 hours of the incubation period. Error bars, SD. **, P < 0.01. G, Immunoblotting of indicated proteins in PC9, PC9AZDR, and PC9GR cells treated with DMSO, osimertinib 1 μmol/L, xentuzumab 100 μg/mL, or combination of both for 4 hours.

Close modal

Interestingly, visualized active pathways estimated by a computational method called “network screening” revealed transient downregulation of multiple pathways within 4 hours after EGFR-TKI treatment and reactivation within 24 hours (Supplementary Fig. S1A). To narrow down the specific pathways that are strongly induced after EGFR inhibition, we performed an additional calculation called the “pathway analysis” and identified 6 pathways. A list of the detected pathways and their calculated value scores are indicated in Supplementary Table S4. Considering this result, we hypothesized that pathways upregulated in 24 hours after EGFR-TKI treatment are more likely to contribute to survival after the abolishment of EGFR signals. Therefore, we especially focused on the following four pathways: (i) VEGFR2-mediated vascular permeability B, (ii) GAB1 signalosome A, (iii) phosphorylation and activation of VAV1, and (iv) p-Y-IRS1 p-Y-IRS2 bind PI3K. To examine whether the activation of these pathways contributed to the resistance to EGFR-TKI in PC9 cells, an MTS cell proliferation assay using various RTK inhibitors was performed (Supplementary Fig. S1B). As expected, RTK inhibitors, which suppress the detected pathways, inhibited the proliferation of erlotinib treated PC9 cells, indicating that the “p-Y-IRS1 p-IRS2 bind PI3K” and “VEGFR2-mediated vascular permeability B” pathways contribute to the resistance to EGFR-TKI in PC9 cells.

Next, we hypothesized that these compensatory pathways may contribute to the acquired resistance in EGFR-TKI–resistant cells. Therefore, we evaluated the activation levels of these four pathways in EGFR-TKI–resistant cells, which we previously established after chronic exposure to EGFR-TKIs, including erlotinib-resistant PC9 (PC9ER), gefitinib-resistant PC9 (PC9GR), and osimertinib-resistant PC9 (PC9AZDR) cells (17, 18). Interestingly, the array detected the high activity of the “p-Y-IRS1 p-IRS2 bind PI3K” pathway, which is involved in IGF1R signaling in PC9AZDR but not in PC9ER and PC9GR cells (Table 1).

Table 1.

Activity of the detected four pathways in EGFR-TKI–resistant PC9 cells.

PathwayPC9ERPC9GRPC9AZDR
VEGFR2-mediated vascular permeability B 0.47 0.77 0.43 
GAB1 signalosome A 0.56 0.35 0.38 
Phosphorylation and activation of VAV1 0.74 0.92 0.99 
p-Y-IRS1 p-Y-IRS2 bind PI3K 0.32 0.55 0.06 
PathwayPC9ERPC9GRPC9AZDR
VEGFR2-mediated vascular permeability B 0.47 0.77 0.43 
GAB1 signalosome A 0.56 0.35 0.38 
Phosphorylation and activation of VAV1 0.74 0.92 0.99 
p-Y-IRS1 p-Y-IRS2 bind PI3K 0.32 0.55 0.06 

Note: Pathway activity was estimated by “network screening.” The thresholds of the probabilities were set to 0.3 in “network screening.”

IGF1R pathway activation confers resistance to osimertinib in PC9AZDR cells

To evaluate the functional contribution of the IGF1R pathway in PC9AZDR cells, we performed MTS cell proliferation assay with a combination of an EGFR inhibitor and linsitinib. As predicted by the protein array, linsitinib demonstrated significant restoration of the sensitivity to osimertinib in PC9AZDR cells (Fig. 1B). Synergy was also observed (Supplementary Table S5). None of the other resistant cell lines showed combined effectiveness with linsitinib, including our previously established PC9COR, H1975AZDR, and H1975COR cells (Supplementary Fig. S1C). To further confirm that linsitinib restores sensitivity to osimertinib, we performed apoptosis assay by flow cytometry or downstream pathway evaluation by immunoblotting using PC9AZDR cells. Induction of apoptosis by linsitinib in combination with osimertinib was observed in PC9AZDR cells (Fig. 1C). Osimertinib/linsitinib cotreatment suppressed the pathway activation downstream of AKT to a greater extent than either drug alone (Fig. 1D). Consistent with the previous study, immunoblotting revealed that IGF1R bypass signaling mainly activated the PI3K/AKT pathway in the EGFR-TKI–resistant cells (25, 26). The above findings indicated that IGF1R bypass signaling is one of the mechanisms of acquired resistance to osimertinib in PC9AZDR cells.

The currently known resistance mechanisms to third-generation EGFR-TKIs include EGFR C797S mutation (27); increased activation of the MAPK pathway by mutated KRAS, MEK, or BRAF (28, 29); and activation of bypass pathways such as the MET or HER2 pathway (30–32). In a previous study (18), whole-exome sequencing of PC9AZDR cells indicated the KRAS G13D mutation as a potential resistance mechanism for this cell line. Referring to this result and the sustained pERK in the immunoblotting assays, we evaluated the effect of trametinib, a MEK inhibitor, in PC9AZDR cells (Supplementary Fig. S1D). Interestingly, trametinib also showed a combined effect with osimertinib and linsitinib, indicating that MAPK activation could also contribute to osimertinib resistance.

IGF2 overexpression induces IGF1R activation, a mechanism of acquired resistance to osimertinib in lung cancer

In order to identify the mechanism of IGF1R activation in PC9AZDR cells, we evaluated the expression profiles of IGF-related molecules by quantitative RT-PCR (qRT-PCR). Previous studies have suggested that resistance to EGFR-TKIs is mediated by the activation of the IGF1R pathway due to the reduction of IGFBP3 expression via its promoter methylation (25, 26, 33). Interestingly, our qRT-PCR analysis did not show downregulation of IGFBP3 but revealed an increased expression of IGF2 in PC9AZDR cells compared with that in PC9 and other resistant cells (Fig. 1E; Supplementary Fig. S2A). This result prompted us to determine whether the inhibition of the ligand itself would affect the proliferation in PC9AZDR cells. For this, we performed an MTS assay in the presence of xentuzumab, an IGF1/IGF2-neutralizing antibody. As expected, xentuzumab attenuated cell proliferation and induced apoptosis in PC9AZDR cells (Fig. 1F; Supplementary Fig. S2B). Additionally, xentuzumab and osimertinib cotreatment attenuated p-AKT in PC9AZDR cells (Fig. 1G), indicating that intracellular AKT signaling was maintained by IGF2/IGF1R signaling after EGFR inhibition. Next, we performed experiments using recombinant human IGF2 to confirm the effect on parental PC9 cells. IGF2 addition attenuated the sensitivity to osimertinib in PC9 cells (Supplementary Fig. S2C). Moreover, IGF2 addition suppressed the apoptosis induced by osimertinib in PC9 cells, and upregulation of p-IGF1R/IR followed by the activation of AKT was observed by immunoblotting (Supplementary Fig. S2D and S2E). These data indicated that PC9AZDR cells were dependent on IGF2–IGF1R pathway activation. Here, we propose a novel acquired resistance mechanism, namely, IGF2 overexpression–mediated IGF1R activation, in lung cancer; however, the clinical relevance of this mechanism should be evaluated further.

IGF1R bypass pathway contributes to acquired resistance in osimertinib-resistant patient-derived cancer cells (PDC)

To clarify the acquired resistance mechanisms to TKIs at a clinically relevant level, PDCs were derived from NSCLC patients who were treated with EGFR-TKIs and after that experienced disease progression. Four of 10 (40%) EGFR-TKI–resistant PDCs were successfully established using pleural effusion or ascites from individual patients. The characteristics of the PDCs are summarized in Supplementary Table S1. We confirmed that KOLK17 and KOLK37 were resistant to first- and second-generation EGFR-TKIs owing to their well-reported MET amplification (Supplementary Fig. S3A–S3E), indicating the clinical relevance of our PDC models.

KOLK43 was derived from a patient who was clinically resistant to first- and third-generation EGFR-TKIs (Fig. 2A). First, its resistance to the first- and third-generation EGFR-TKI, erlotinib, and osimertinib was confirmed by the MTS assay (Fig. 2B; Supplementary Fig. S4A). Target sequencing of KOLK43 cells detected EGFR exon 21 L858R but not EGFR exon 20 T790M (Supplementary Table S6). It was previously reported that EGFR exon 20 T790M occasionally disappears in third-generation EGFR-TKI–resistant cells (34), consistent with our result. Additionally, none of the gene mutations that have been previously reported to contribute to resistance to third-generation EGFR-TKIs was detected by targeted sequencing. In order to investigate the bypass RTK signaling pathway activation in KOLK43 cells, phospho-RTK array screening was performed. Interestingly, phosphorylation of IGF1R was observed compared with that of the other RTKs (Fig. 2C). Linsitinib restored the sensitivity to osimertinib and displayed synergism (Fig. 2D; Supplementary Table S7). None of the other PDCs showed combined effectiveness with linsitinib (Supplementary Fig. S4B). Again, as shown in PC9AZDR cells, the IGF1R pathway contributed to the acquired resistance to osimertinib in KOLK43 cells. Induction of apoptosis by linsitinib in combination with osimertinib was observed in KOLK43 cells (Fig. 2E). Osimertinib/linsitinib cotreatment suppressed the pathway activation downstream of AKT to a greater extent than either drug alone in KOLK43 cells (Fig. 2F). These data indicated that the activation of IGF1R signaling contributed to the development of osimertinib resistance in KOLK43 cells.

Figure 2.

IGF1R bypass pathway activation contributes to osimertinib resistance in PDCs. A, Representative image of KOLK43 cells. Scale bars, 100 μm. B, MTS cell proliferation assays in PC9 and KOLK43 cells treated with increasing concentrations of osimertinib for 72 hours. Error bars, SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001. C, Results of human phospho-RTK array experiments in KOLK43 cells. The spots for EGFR, IR, and IGF1R signaling are indicated with arrows. The spots are shown in duplicate. D, MTS cell proliferation assays following treatment with the indicated concentration of osimertinib with or without linsitinib in KOLK43 cells. The assay was performed after a 120-hour incubation period. Error bars, SD. *, P < 0.05; **, P < 0.01. E, Flow-cytometric data for KOLK43 cells. KOLK43 cells were treated with osimertinib, linsitinib, or combination of both for 48 hours. The number (%) indicates the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. F, Immunoblotting of the indicated proteins in PC9, KOLK43, and KOLK37 cells. PC9, KOLK43, and KOLK37 were treated with DMSO, osimertinib, linsitinib, or combination of both for 4 hours.

Figure 2.

IGF1R bypass pathway activation contributes to osimertinib resistance in PDCs. A, Representative image of KOLK43 cells. Scale bars, 100 μm. B, MTS cell proliferation assays in PC9 and KOLK43 cells treated with increasing concentrations of osimertinib for 72 hours. Error bars, SD. *, P < 0.05; **, P < 0.01; ***, P < 0.001. C, Results of human phospho-RTK array experiments in KOLK43 cells. The spots for EGFR, IR, and IGF1R signaling are indicated with arrows. The spots are shown in duplicate. D, MTS cell proliferation assays following treatment with the indicated concentration of osimertinib with or without linsitinib in KOLK43 cells. The assay was performed after a 120-hour incubation period. Error bars, SD. *, P < 0.05; **, P < 0.01. E, Flow-cytometric data for KOLK43 cells. KOLK43 cells were treated with osimertinib, linsitinib, or combination of both for 48 hours. The number (%) indicates the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. F, Immunoblotting of the indicated proteins in PC9, KOLK43, and KOLK37 cells. PC9, KOLK43, and KOLK37 were treated with DMSO, osimertinib, linsitinib, or combination of both for 4 hours.

Close modal

Of note, consistent with PC9AZDR cells, the inhibitory effects of osimertinib/linsitinib cotreatment were observed on IGF1R and AKT phosphorylation, but not on ERK signaling. Once again, we performed additional MTS proliferation assays evaluating the effect of trametinib in KOLK43 cells (Supplementary Fig. S4C) and detected the combined effect with this drug. As in PC9AZDR cells, the activation of MAPK signaling may have also promoted resistance to osimertinib. These findings suggest the therapeutic efficacy of MAPK pathway targeting in conjunction with EGFR/IGF1R pathway inhibition.

IGF2 overexpression–mediated IGF1R activation contributes to acquired resistance in KOLK43 cells

Considering the result of PC9AZDR cells, we evaluated the expression profiles of IGF-related molecules in KOLK43 cells, in which IGF1R pathway activation contributed to the acquired resistance to osimertinib. Consistent with the results of PC9AZDR cells, KOLK43 did not exhibit downregulation of IGFBP3. However, it showed a significantly increased expression of IGF2 and IGFBP5 compared with PC9 cells (Fig. 3A). IGFBPs secreted by cancer cells are known to either enhance or inhibit growth by modulating IGF1 or IGF2 activity (35). In KOLK43 cells, due to restoration of sensitivity to osimertinib as a result of IGFBP5 addition (Supplementary Fig. S4D), IGFBP5 presumably suppressed the IGF2–IGF1R activation by binding to the ligand. As mentioned above, next-generation sequencing-based targeted sequencing did not identify any IGF1R pathway–related mutation (Supplementary Table S6). Thus, IGF2 overexpression was suspected of causing IGF1R activation in KOLK43 cells in addition to PC9AZDR. First, we confirmed that there was no upregulation of IGF2 in other PDCs (Supplementary Fig. S4E). We have also verified that IGF2 and pIGF1R were not upregulated in the patient samples before EGFR-TKI treatment (Supplementary Fig. S4F). DNA quantification by quantitative PCR revealed that the DNA copy number of IGF2 did not increase in KOLK43 cells (Fig. 3B), suggesting that the elevated expression of IGF2 was not due to gene amplification. IGF2 overexpression was validated at the protein level by immunoblotting. Interestingly, compared with the mature form of IGF2 (7.5 kDa), the unprocessed forms of IGF2 such as pro-IGF2 (10–18 kDa) were mostly expressed (Fig. 3C). Pro-IGF2 mainly exists as a binary complex that can leave the circulation, allowing it to exhibit higher bioavailability than the mature form and resulting in favorable conditions for the proliferation of cancer cells (36). Xentuzumab restored the sensitivity to osimertinib in KOLK43 cells (Fig. 3D). Analysis of apoptosis by flow cytometry also revealed an increase in the number of Annexin V–positive cells in the xentuzumab/osimertinib combination treatment group (Fig. 3E). Cotreatment also suppressed p-AKT pathway activation, as shown by immunoblotting (Fig. 3F). Additionally, knockdown of IGF2 and IGF1R by siRNA suppressed the proliferation in KOLK43 cells (Supplementary Fig. S4G and S4H). These data indicate the functional roles of IGF2 and IGF1R in KOLK43 cells.

Figure 3.

IGF2 overexpression induces IGF1R activation and contributes to osimertinib resistance in KOLK43 cells. A, Heat map for relative gene expression in KOLK43 and PC9 cells. B, Relative gene copy number of IGF2 normalized by LINE1 in PC9 and KOLK43 cells. Error bars, SD. C, Immunoblotting of the indicated proteins in PC9 and KOLK43 cells. D, MTS cell proliferation assays following treatment with the indicated osimertinib concentrations with or without xentuzumab 100 μg/mL in KOLK43 cells. The assay was performed after 120 hours of the incubation period. Error bars, SD. **, P < 0.01. E, Flow-cytometric data for KOLK43 cells treated with DMSO, osimertinib 1 μmol/L, xentuzumab 100 μg/mL, or the combination of both for 48 hours. The number (%) indicates the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. F, Immunoblotting of the indicated proteins in PC9, KOLK43, and KOLK37 cells treated with DMSO, osimertinib 1 μmol/L, xentuzumab 100 μg/mL, or combination of both for 4 hours.

Figure 3.

IGF2 overexpression induces IGF1R activation and contributes to osimertinib resistance in KOLK43 cells. A, Heat map for relative gene expression in KOLK43 and PC9 cells. B, Relative gene copy number of IGF2 normalized by LINE1 in PC9 and KOLK43 cells. Error bars, SD. C, Immunoblotting of the indicated proteins in PC9 and KOLK43 cells. D, MTS cell proliferation assays following treatment with the indicated osimertinib concentrations with or without xentuzumab 100 μg/mL in KOLK43 cells. The assay was performed after 120 hours of the incubation period. Error bars, SD. **, P < 0.01. E, Flow-cytometric data for KOLK43 cells treated with DMSO, osimertinib 1 μmol/L, xentuzumab 100 μg/mL, or the combination of both for 48 hours. The number (%) indicates the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. F, Immunoblotting of the indicated proteins in PC9, KOLK43, and KOLK37 cells treated with DMSO, osimertinib 1 μmol/L, xentuzumab 100 μg/mL, or combination of both for 4 hours.

Close modal

To evaluate the dynamics of IGF2 expression, KOLK43 cells were cultured in various conditioned media. Interestingly, IGF2 expression and the phosphorylation level of IGF1R/IR in KOLK43 cells elevated in a low percentage of FBS and osimertinib treatment (Supplementary Fig. S5A). Additionally, we found that KOLK43 cells cultured in 10% FBS medium lost its IGF1R phosphorylation along with the reduction of IGF2 during cell passages (Supplementary Fig. S5B and S5C). These findings indicate that IGF2 expression is influenced by culture conditions.

IGF2 is secreted into the microenvironment and activates IGF1R in KOLK43 cells

Interestingly, pleural effusion supernatants from which KOLK43 was established exhibited higher quantities of IGF2 than that observed in the other lung cancer patients (Fig. 4A). This indicates that IGF2 was once secreted into the microenvironment and activated IGF1R. MTS cell proliferation assay revealed that PC9 cells became resistant to osimertinib in conditioned medium with 25% pleural effusion supernatants from KOLK43 (PC9 + PE; Fig. 4B). RTK array analysis was performed to evaluate whether any bypass pathway activation occurred in PC9 + PE cells. As expected, IGF1R phosphorylation was upregulated in these cells (Supplementary Fig. S5D). Based on this result, MTS cell proliferation assay using linsitinib was performed, and restoration of sensitivity to osimertinib in PC9 + PE cells was detected (Fig. 4C). Furthermore, linsitinib and osimertinib cotreatment induced apoptosis in an apoptosis assay (Fig. 4D). Cotreatment also suppressed p-AKT pathway activation, as observed by immunoblotting (Fig. 4E).

Figure 4.

EGFR-TKI–sensitive lung cancer cells become resistant to osimertinib in conditioned medium with pleural effusion supernatants of KOLK43. A, Quantification of IGF2 using pleural effusion supernatants from multiple lung cancer patients. **, P < 0.01. B, MTS cell proliferation assay in PC9 cells following treatment with the indicated osimertinib concentrations and the addition of pleural effusion supernatants of KOLK43. The assay was performed after a 72-hour incubation period. Error bars, SD. C, MTS cell proliferation assay in PC9 cells incubated with 25% pleural effusion supernatants of KOLK43 following treatment with the indicated concentrations of osimertinib with or without linsitinib. The assay was performed after 72 hours of incubation. Error bars, SD. **, P < 0.01; ***, P < 0.001. D, Flow-cytometric data for PC9 cells incubated with 25% pleural effusion supernatants of KOLK43 treated with DMSO, osimertinib, linsitinib, or combination of both for 48 hours. The number (%) indicates the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. E, Immunoblotting of indicated proteins in PC9 cells with or without the addition of 25% pleural effusion supernatants of KOLK43. Cells were treated with DMSO, osimertinib, linsitinib, or combination of both for 4 hours.

Figure 4.

EGFR-TKI–sensitive lung cancer cells become resistant to osimertinib in conditioned medium with pleural effusion supernatants of KOLK43. A, Quantification of IGF2 using pleural effusion supernatants from multiple lung cancer patients. **, P < 0.01. B, MTS cell proliferation assay in PC9 cells following treatment with the indicated osimertinib concentrations and the addition of pleural effusion supernatants of KOLK43. The assay was performed after a 72-hour incubation period. Error bars, SD. C, MTS cell proliferation assay in PC9 cells incubated with 25% pleural effusion supernatants of KOLK43 following treatment with the indicated concentrations of osimertinib with or without linsitinib. The assay was performed after 72 hours of incubation. Error bars, SD. **, P < 0.01; ***, P < 0.001. D, Flow-cytometric data for PC9 cells incubated with 25% pleural effusion supernatants of KOLK43 treated with DMSO, osimertinib, linsitinib, or combination of both for 48 hours. The number (%) indicates the proportion of Annexin V–FITC- and/or propidium iodide–stained cells. E, Immunoblotting of indicated proteins in PC9 cells with or without the addition of 25% pleural effusion supernatants of KOLK43. Cells were treated with DMSO, osimertinib, linsitinib, or combination of both for 4 hours.

Close modal

Overcoming IGF2–IGF1R–mediated acquired resistance via osimertinib and linsitinib combination treatment in vivo

To further evaluate the efficacy of the combination treatment using linsitinib and osimertinib, KOLK43 tumor xenografts were established in SCID mice. IGF2 expression in KOLK43 xenograft tumors was evaluated by IHC staining and immunoblotting (Fig. 5A; Supplementary Fig. S6A). Among them, the IHC staining provided additional information that IGF2 was mainly expressed by the cancer cells themselves, indicating that IGF2 activates IGF1R in an autocrine manner. During drug treatment, mice bodyweights and tumor volumes were monitored (Fig. 5B; Supplementary Fig. S6B). Notably, tumor volumes increased in the control and single-treatment groups. However, osimertinib/linsitinib cotreatment demonstrated significant inhibition of tumor growth. Collectively, these data support the possibility that combination therapy may be a practical approach to overcome IGF2–IGF1R–mediated acquired resistance.

Figure 5.

Clinical potential of combination treatment in overcoming IGF2 overexpression–mediated acquired resistance. A, Representative IHC of IGF2 in KOLK43 xenograft tumors. Scale bars, 100 μm. B, KOLK43 derived tumor-bearing mice were randomized into control, osimertinib, linsitinib, or osimertinib/linsitinib combination treatment groups. Tumor size was measured to calculate the TV. Values indicate average TV ± SEM in each group. ***, P < 0.001 for the combination of osimertinib/linsitinib versus linsitinib alone. C, IGF2 IHC scores of patients' tumor tissue obtained from the first biopsy (A), second biopsy (B), and third biopsy (C). D, Representative IHC images of IGF2 in first, second, and third biopsies. IGF2 IHC scores are shown in Supplementary Table S8. Scale bars, 100 μm.

Figure 5.

Clinical potential of combination treatment in overcoming IGF2 overexpression–mediated acquired resistance. A, Representative IHC of IGF2 in KOLK43 xenograft tumors. Scale bars, 100 μm. B, KOLK43 derived tumor-bearing mice were randomized into control, osimertinib, linsitinib, or osimertinib/linsitinib combination treatment groups. Tumor size was measured to calculate the TV. Values indicate average TV ± SEM in each group. ***, P < 0.001 for the combination of osimertinib/linsitinib versus linsitinib alone. C, IGF2 IHC scores of patients' tumor tissue obtained from the first biopsy (A), second biopsy (B), and third biopsy (C). D, Representative IHC images of IGF2 in first, second, and third biopsies. IGF2 IHC scores are shown in Supplementary Table S8. Scale bars, 100 μm.

Close modal

Increased IGF2 expression in human NSCLC resistant to EGFR-TKI

To determine the clinical relevance of the above findings and evaluate the frequency of IGF2–IGF1R pathway activation in human EGFR-TKI–treated lung cancer samples, we examined the expression level of IGF2 by IHC. The third biopsy tumor samples, matched with the first and/or second biopsy, were collected from 6 NSCLC patients. All patients were treated with gefitinib, erlotinib, or afatinib and met the clinical definition of acquired resistance due to EGFR exon 20 T790M mutation. Further information on the clinical backgrounds of the 6 patients is shown in Supplementary Table S8. As a result, 4 of 6 (66.7 %) patients eventually showed elevation of IGF2 levels after osimertinib resistance (Fig. 5C). Representative IHC images of IGF2 in first, second, and third biopsy samples are shown in Fig. 5D. These data indicate the possibility that IGF2–IGF1R–mediated acquired resistance is a clinically relevant mechanism of acquired resistance to osimertinib.

Lung cancers positive for EGFR gene mutations rely heavily on EGFR signaling for the abnormal proliferation of cancer cells, and EGFR inhibition is expected to eliminate these cells (8, 9). However, in the clinical setting, a fraction of the cancer cell population is known to either intrinsically possess or gradually acquire resistance to EGFR inhibition (37–39). As reported in previous studies (40, 41), in the early phase of EGFR-TKI exposure, complex alterations of signal transduction pathways induce prosurvival signals. Some of these prosurvival signals do not depend on EGFR kinase activity (40, 42). Using our protein array, we successfully detected the activation of the IGF1R pathway in the early phase of EGFR-TKI exposure. These data highlight the critical roles played by the IGF1R pathway in response to EGFR inhibition.

In this study, we acquired PDCs from multiple patients with NSCLC who were treated with EGFR-TKIs and experienced disease progression. The success rate for PDCs was as high as 40%. Using experimentally established cell lines and PDCs, we discovered IGF2 overexpression–mediated IGF1R activation as a novel and common resistance mechanism in lung cancer (Fig. 6). Notably, in both cases, this resistant mechanism was not detected by the NGS that was previously performed.

Figure 6.

Schematic representation of the resistant mechanism induced by IGF2 overexpression. Proposed model of IGF2 overexpression inducing an IGF1R bypass and contributing to the acquired resistance to osimertinib.

Figure 6.

Schematic representation of the resistant mechanism induced by IGF2 overexpression. Proposed model of IGF2 overexpression inducing an IGF1R bypass and contributing to the acquired resistance to osimertinib.

Close modal

NGS of clinical samples from patients enables the detection of genetic alterations that contribute to drug resistance, such as EGFR T790M and EGFR C797S mutations (27, 43). However, it does not always offer an appropriate treatment strategy for patients with acquired resistance to EGFR-TKIs (44). In contrast, the presence of living cancer cells allows researchers to biologically evaluate the roles of mutations or bypass pathways. For a thorough clarification of resistance mechanisms, living cancer cells, in addition to genomic information, are essential.

IGF1R bypass activation contributing to EGFR-TKI resistance has already been reported, but all the previous studies indicated the loss of IGFBP3 as the mechanism underlying increasing IGF1R activation (25, 33). To the best of our knowledge, this is the first study to elucidate the acquired resistance mechanism involving IGF2-mediated IGF1R activation in lung cancer. Interestingly, increased expression of IGF2 was observed in a proportion of lung cancer patients, indicating that IGF2-mediated IGF1R pathway activation is a clinically relevant mechanism of acquired resistance to osimertinib.

Additionally, we demonstrated the efficacy of the IGF1/2-neutralizing antibody xentuzumab, which is currently under evaluation in clinical trials for breast cancer treatment. The advantage of this antibody over that of IGF1R inhibitors is that it inhibits IGF1/2 signaling through the IGF1R, IR-A, IR-B, and hybrid receptors, without affecting insulin signaling through the insulin receptor (45). Compared with IGF1R inhibitors, xentuzumab showed fewer side effects, such as hyperglycemia, nausea, vomiting, fatigue, and anorexia (46).

In conclusion, using established lung cancer cell lines and PDCs, we confirmed that IGF2-mediated IGF1R bypass pathway activation contributes to the acquired resistance to osimertinib. We also demonstrated that in addition to the genomic information, living cancer cells are essential in order to elucidate the resistant mechanism and to determine the optimal treatment for individual patients. The findings suggest a clinically relevant mechanism of resistance to osimertinib and may help develop an effective treatment for EGFR-mutated lung cancer.

K. Soejima reports receiving commercial research grants from Nippon Boehringer Ingelheim, AstraZeneca K.K., and Taiho Pharmaceutical, has received honoraria from the speakers bureau of AstraZeneca K.K., Chugai Pharmaceutical, Ono Pharmaceutical, Bristol-Myers Squibb Japan, MSD Oncology, Lily Japan, and Novartis Pharma K.K. No potential conflicts of interest were disclosed by the other authors.

Conception and design: T. Manabe, H. Yasuda, H. Terai, S. Ikemura, I. Kawada, K. Soejima

Development of methodology: T. Manabe, H. Yasuda, H. Terai, H. Kagiwada, J. Hamamoto, K. Horimoto

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): T. Manabe, H. Terai, H. Kagiwada, J. Hamamoto, T. Ebisudani, K. Masuzawa, I. Kawada

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T. Manabe, H. Yasuda, H. Terai, H. Kagiwada, J. Hamamoto, T. Ebisudani, K. Kobayashi, K. Masuzawa, Y. Hayashi, K. Fukui, K. Horimoto

Writing, review, and/or revision of the manuscript: T. Manabe, H. Yasuda, H. Terai, H. Kagiwada, K. Kobayashi, S. Ikemura, I. Kawada, K. Horimoto, K. Soejima

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): H. Yasuda, H. Terai, J. Hamamoto

Study supervision: H. Yasuda, H. Terai, S. Ikemura, I. Kawada, K. Fukunaga, K. Soejima

This work was supported in part by the Japan Society for the Promotion of Science to T. Manabe (grants #17K16059 and #19J12401), H. Terai (grant #18K08184), S. Ikemura (grant #18K15251), and H. Yasuda (grant #17K09667). This work was also supported in part by Takeda Science Foundation to H. Yasuda and H. Terai and by the Development of Diagnostic Technology for Detection of miRNA in Body Fluids grant from the Japan Agency for Medical Research and Development to K. Horimoto (grant #18ae0101016s0105). We thank Mrs. Chinatsu Yonekawa for her excellent technical assistance.

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.

1.
Giordano
TJ
. 
The cancer genome atlas research network: a sight to behold
.
Endocr Pathol
2014
;
25
:
362
5
.
2.
Kosaka
T
,
Yatabe
Y
,
Endoh
H
,
Kuwano
H
,
Takahashi
T
,
Mitsudomi
T
. 
Mutations of the epidermal growth factor receptor gene in lung cancer: biological and clinical implications
.
Cancer Res
2004
;
64
:
8919
23
.
3.
Sharma
SV
,
Bell
DW
,
Settleman
J
,
Haber
DA
. 
Epidermal growth factor receptor mutations in lung cancer
.
Nat Rev Cancer
2007
;
7
:
169
81
.
4.
Jura
N
,
Zhang
X
,
Endres
NF
,
Seeliger
MA
,
Schindler
T
,
Kuriyan
J
. 
Catalytic control in the EGF receptor and its connection to general kinase regulatory mechanisms
.
Mol Cell
2011
;
42
:
9
22
.
5.
Yasuda
H
,
Park
E
,
Yun
CH
,
Sng
NJ
,
Lucena-Araujo
AR
,
Yeo
WL
, et al
Structural, biochemical, and clinical characterization of epidermal growth factor receptor (EGFR) exon 20 insertion mutations in lung cancer
.
Sci Transl Med
2013
;
5
:
216ra177
.
6.
Yun
CH
,
Boggon
TJ
,
Li
Y
,
Woo
MS
,
Greulich
H
,
Meyerson
M
, et al
Structures of lung cancer-derived EGFR mutants and inhibitor complexes: mechanism of activation and insights into differential inhibitor sensitivity
.
Cancer Cell
2007
;
11
:
217
27
.
7.
Zhang
X
,
Gureasko
J
,
Shen
K
,
Cole
PA
,
Kuriyan
J
. 
An allosteric mechanism for activation of the kinase domain of epidermal growth factor receptor
.
Cell
2006
;
125
:
1137
49
.
8.
Lynch
TJ
,
Bell
DW
,
Sordella
R
,
Gurubhagavatula
S
,
Okimoto
RA
,
Brannigan
BW
, et al
Activating mutations in the epidermal growth factor receptor underlying responsiveness of non-small-cell lung cancer to gefitinib
.
N Engl J Med
2004
;
350
:
2129
39
.
9.
Paez
JG
,
Janne
PA
,
Lee
JC
,
Tracy
S
,
Greulich
H
,
Gabriel
S
, et al
EGFR mutations in lung cancer: correlation with clinical response to gefitinib therapy
.
Science
2004
;
304
:
1497
500
.
10.
Mok
TS
,
Wu
YL
,
Thongprasert
S
,
Yang
CH
,
Chu
DT
,
Saijo
N
, et al
Gefitinib or carboplatin-paclitaxel in pulmonary adenocarcinoma
.
N Engl J Med
2009
;
361
:
947
57
.
11.
Sequist
LV
,
Yang
JC
,
Yamamoto
N
,
O'Byrne
K
,
Hirsh
V
,
Mok
T
, et al
Phase III study of afatinib or cisplatin plus pemetrexed in patients with metastatic lung adenocarcinoma with EGFR mutations
.
J Clin Oncol
2013
;
31
:
3327
34
.
12.
Shepherd
FA
,
Rodrigues Pereira
J
,
Ciuleanu
T
,
Tan
EH
,
Hirsh
V
,
Thongprasert
S
, et al
Erlotinib in previously treated non-small-cell lung cancer
.
N Engl J Med
2005
;
353
:
123
32
.
13.
Zhou
C
,
Wu
YL
,
Chen
G
,
Feng
J
,
Liu
XQ
,
Wang
C
, et al
Erlotinib versus chemotherapy as first-line treatment for patients with advanced EGFR mutation-positive non-small-cell lung cancer (OPTIMAL, CTONG-0802): a multicentre, open-label, randomised, phase 3 study
.
Lancet Oncol
2011
;
12
:
735
42
.
14.
Mok
TS
,
Wu
YL
,
Ahn
MJ
,
Garassino
MC
,
Kim
HR
,
Ramalingam
SS
, et al
Osimertinib or platinum-pemetrexed in EGFR T790M-positive lung cancer
.
N Engl J Med
2017
;
376
:
629
40
.
15.
Soria
JC
,
Ohe
Y
,
Vansteenkiste
J
,
Reungwetwattana
T
,
Chewaskulyong
B
,
Lee
KH
, et al
Osimertinib in untreated EGFR-mutated advanced non-small-cell lung cancer
.
N Engl J Med
2018
;
378
:
113
25
.
16.
Yang
Z
,
Yang
N
,
Ou
Q
,
Xiang
Y
,
Jiang
T
,
Wu
X
, et al
Investigating novel resistance mechanisms to third-generation EGFR tyrosine kinase inhibitor osimertinib in non-small cell lung cancer patients
.
Clin Cancer Res
2018
;
24
:
3097
107
.
17.
Terai
H
,
Soejima
K
,
Yasuda
H
,
Nakayama
S
,
Hamamoto
J
,
Arai
D
, et al
Activation of the FGF2-FGFR1 autocrine pathway: a novel mechanism of acquired resistance to gefitinib in NSCLC
.
Mol Cancer Res
2013
;
11
:
759
67
.
18.
Nukaga
S
,
Yasuda
H
,
Tsuchihara
K
,
Hamamoto
J
,
Masuzawa
K
,
Kawada
I
, et al
Amplification of EGFR wild-type alleles in non-small cell lung cancer cells confers acquired resistance to mutation-selective EGFR tyrosine kinase inhibitors
.
Cancer Res
2017
;
77
:
2078
89
.
19.
Crystal
AS
,
Shaw
AT
,
Sequist
LV
,
Friboulet
L
,
Niederst
MJ
,
Lockerman
EL
, et al
Patient-derived models of acquired resistance can identify effective drug combinations for cancer
.
Science
2014
;
346
:
1480
6
.
20.
Takeda
H
,
Kawamura
Y
,
Miura
A
,
Mori
M
,
Wakamatsu
A
,
Yamamoto
J
, et al
Comparative analysis of human SRC-family kinase substrate specificity in vitro
.
J Proteome Res
2010
;
9
:
5982
93
.
21.
Bliss
CI
. 
The toxicity of poisons applied jointly
.
Ann App Biol
1939
;
26
:
585
615
.
22.
Greco
WR
,
Bravo
G
,
Parsons
JC
. 
The search for synergy: a critical review from a response surface perspective
.
Pharmacol Rev
1995
;
47
:
331
85
.
23.
Geary
N
. 
Understanding synergy
.
Am J Physiol Endocrinol Metab
2013
;
304
:
E237
53
.
24.
Moyer
JD
,
Barbacci
EG
,
Iwata
KK
,
Arnold
L
,
Boman
B
,
Cunningham
A
, et al
Induction of apoptosis and cell cycle arrest by CP-358,774, an inhibitor of epidermal growth factor receptor tyrosine kinase
.
Cancer Res
1997
;
57
:
4838
48
.
25.
Cortot
AB
,
Repellin
CE
,
Shimamura
T
,
Capelletti
M
,
Zejnullahu
K
,
Ercan
D
, et al
Resistance to irreversible EGF receptor tyrosine kinase inhibitors through a multistep mechanism involving the IGF1R pathway
.
Cancer Res
2013
;
73
:
834
43
.
26.
Guix
M
,
Faber
AC
,
Wang
SE
,
Olivares
MG
,
Song
Y
,
Qu
S
, et al
Acquired resistance to EGFR tyrosine kinase inhibitors in cancer cells is mediated by loss of IGF-binding proteins
.
J Clin Invest
2008
;
118
:
2609
19
.
27.
Thress
KS
,
Paweletz
CP
,
Felip
E
,
Cho
BC
,
Stetson
D
,
Dougherty
B
, et al
Acquired EGFR C797S mutation mediates resistance to AZD9291 in non-small cell lung cancer harboring EGFR T790M
.
Nat Med
2015
;
21
:
560
2
.
28.
Eberlein
CA
,
Stetson
D
,
Markovets
AA
,
Al-Kadhimi
KJ
,
Lai
Z
,
Fisher
PR
, et al
Acquired resistance to the mutant-selective EGFR inhibitor AZD9291 is associated with increased dependence on RAS signaling in preclinical models
.
Cancer Res
2015
;
75
:
2489
500
.
29.
Oxnard
GR
,
Hu
Y
,
Mileham
KF
,
Husain
H
,
Costa
DB
,
Tracy
P
, et al
Assessment of resistance mechanisms and clinical implications in patients with EGFR T790M-positive lung cancer and acquired resistance to osimertinib
.
JAMA Oncol
2018
;
4
:
1527
34
.
30.
Ortiz-Cuaran
S
,
Scheffler
M
,
Plenker
D
,
Dahmen
L
,
Scheel
AH
,
Fernandez-Cuesta
L
, et al
Heterogeneous mechanisms of primary and acquired resistance to third-generation EGFR inhibitors
.
Clin Cancer Res
2016
;
22
:
4837
47
.
31.
Mizuuchi
H
,
Suda
K
,
Murakami
I
,
Sakai
K
,
Sato
K
,
Kobayashi
Y
, et al
Oncogene swap as a novel mechanism of acquired resistance to epidermal growth factor receptor-tyrosine kinase inhibitor in lung cancer
.
Cancer Sci
2016
;
107
:
461
8
.
32.
Lin
CC
,
Shih
JY
,
Yu
CJ
,
Ho
CC
,
Liao
WY
,
Lee
JH
, et al
Outcomes in patients with non-small-cell lung cancer and acquired Thr790Met mutation treated with osimertinib: a genomic study
.
Lancet Respir Med
2018
;
6
:
107
16
.
33.
Park
JH
,
Choi
YJ
,
Kim
SY
,
Lee
JE
,
Sung
KJ
,
Park
S
, et al
Activation of the IGF1R pathway potentially mediates acquired resistance to mutant-selective 3rd-generation EGF receptor tyrosine kinase inhibitors in advanced non-small cell lung cancer
.
Oncotarget
2016
;
7
:
22005
15
.
34.
Piotrowska
Z
,
Niederst
MJ
,
Karlovich
CA
,
Wakelee
HA
,
Neal
JW
,
Mino-Kenudson
M
, et al
Heterogeneity underlies the emergence of EGFRT790 wild-type clones following treatment of T790M-positive cancers with a third-generation EGFR inhibitor
.
Cancer Discov
2015
;
5
:
713
22
.
35.
Livingstone
C
. 
IGF2 and cancer
.
Endocr Relat Cancer
2013
;
20
:
R321
39
.
36.
Unger
C
,
Kramer
N
,
Unterleuthner
D
,
Scherzer
M
,
Burian
A
,
Rudisch
A
, et al
Stromal-derived IGF2 promotes colon cancer progression via paracrine and autocrine mechanisms
.
Oncogene
2017
;
36
:
5341
55
.
37.
Maemondo
M
,
Inoue
A
,
Kobayashi
K
,
Sugawara
S
,
Oizumi
S
,
Isobe
H
, et al
Gefitinib or chemotherapy for non-small-cell lung cancer with mutated EGFR
.
N Engl J Med
2010
;
362
:
2380
8
.
38.
Miller
VA
,
Hirsh
V
,
Cadranel
J
,
Chen
YM
,
Park
K
,
Kim
SW
, et al
Afatinib versus placebo for patients with advanced, metastatic non-small-cell lung cancer after failure of erlotinib, gefitinib, or both, and one or two lines of chemotherapy (LUX-Lung 1): a phase 2b/3 randomised trial
.
Lancet Oncol
2012
;
13
:
528
38
.
39.
Rosell
R
,
Carcereny
E
,
Gervais
R
,
Vergnenegre
A
,
Massuti
B
,
Felip
E
, et al
Erlotinib versus standard chemotherapy as first-line treatment for European patients with advanced EGFR mutation-positive non-small-cell lung cancer (EURTAC): a multicentre, open-label, randomised phase 3 trial
.
Lancet Oncol
2012
;
13
:
239
46
.
40.
Rusan
M
,
Li
K
,
Li
Y
,
Christensen
CL
,
Abraham
BJ
,
Kwiatkowski
N
, et al
Suppression of adaptive responses to targeted cancer therapy by transcriptional repression
.
Cancer Discov
2018
;
8
:
59
73
.
41.
Sharma
SV
,
Lee
DY
,
Li
B
,
Quinlan
MP
,
Takahashi
F
,
Maheswaran
S
, et al
A chromatin-mediated reversible drug-tolerant state in cancer cell subpopulations
.
Cell
2010
;
141
:
69
80
.
42.
Liu
Q
,
Yu
S
,
Zhao
W
,
Qin
S
,
Chu
Q
,
Wu
K
. 
EGFR-TKIs resistance via EGFR-independent signaling pathways
.
Mol Cancer
2018
;
17
:
53
.
43.
Kobayashi
S
,
Boggon
TJ
,
Dayaram
T
,
Janne
PA
,
Kocher
O
,
Meyerson
M
, et al
EGFR mutation and resistance of non-small-cell lung cancer to gefitinib
.
N Engl J Med
2005
;
352
:
786
92
.
44.
Skoulidis
F
,
Papadimitrakopoulou
VA
. 
Targeting the gatekeeper: osimertinib in EGFR T790M mutation-positive non-small cell lung cancer
.
Clin Cancer Res
2017
;
23
:
618
22
.
45.
Allison
M
. 
Clinical setbacks reduce IGF-1 inhibitors to cocktail mixers
.
Nat Biotechnol
2012
;
30
:
906
7
.
46.
Friedbichler
K
,
Hofmann
MH
,
Kroez
M
,
Ostermann
E
,
Lamche
HR
,
Koessl
C
, et al
Pharmacodynamic and antineoplastic activity of BI 836845, a fully human IGF ligand-neutralizing antibody, and mechanistic rationale for combination with rapamycin
.
Mol Cancer Ther
2014
;
13
:
399
409
.

Supplementary data