Third-generation EGFR-TKIs such as osimertinib (AZD9291) and CO-1686 have shown promising activity in treatment-resistant EGFR mutation-positive NSCLCs containing the T790M mutation [20, 21]. However, acquisition of an EGFR C797S mutation has been identified as a novel mechanism of resistance to osimertinib (AZD9291) [22].

An additional 10–20 % of tumors from refractory EGFR-mutant NSCLC patients undergo MET gene amplification, which causes HER3-dependent activation of the signaling cascade downstream of EGFR despite its inhibition by TKIs [15]. Recently, MET inhibitors have been administered to NSCLC patients who are naïve or resistant to EGFR-TKIs [23]. This phase II study showed that PFS was extended in the group treated with erlotinib plus the MET inhibitor tivantinib than in the group treated with erlotinib alone, especially among patients harboring KRAS mutations [23]. Yano et al. also showed that HGF, a ligand of the MET oncoprotein, induces gefitinib resistance in EGFR-mutant lung adenocarcinoma cells by restoring PI3K/AKT signaling via phosphorylation of MET, but not EGFR or ErbB3 [16]. Inhibition of HGF-MET signaling may be a useful strategy to overcome resistance to EGFR-TKIs depending on the status of HGF-MET signaling.

Morphological transformation to SCLC from NSCLC represents one of the mechanisms of acquired resistance to EGFR-TKIs. Sequist et al. reported the morphological transformation of five drug-resistant NSCLC tumors (14 % of a total of 37 patients) containing EGFR mutations into SCLC [18]. The existing EGFR mutation was maintained during SCLC transformation in all cases [18]. These transformed SCLC tumors were sensitive to standard SCLC chemotherapy.

EMT is a progressive biological phenomenon that includes loss of epithelial cell adhesion and induction of a mesenchymal phenotype. Several studies have demonstrated that EMT is associated with reduced drug sensitivity and acquisition of resistance to EGFR-TKIs in NSCLC, whereas retention of an epithelial phenotype ensured a good response to EGFR-TKIs [24, 25]. Recent studies revealed that overexpression of AXL led to resistance to EGFR-TKI in NSCLC cells undergoing EMT and that AXL was a potential therapeutic target in patients with acquired resistance to EGFR-TKIs [19]. These reports suggested that EMT might be a mechanism of resistance to EGFR-TKI. However, the molecular mechanisms of the development of EMT-related resistance to EGFR-TKI are still not fully understood. Therapeutic strategies aiming to prevent EMT with a view to restoring sensitivity to EGFR-TKIs remain to be investigated.

Cancer stem cells (CSCs) are characterized by the capacity for pluripotency and self-renewal and are thought to represent a renewable source of cancer cells. The significance of CSC-like properties to the mechanism of resistance to EGFR-TKIs in NSCLCs has recently been investigated [26, 27]. Sharma et al. identified a drug-tolerant cancer cell subpopulation (DTP) that maintains viability under conditions where the vast majority of the cell population is rapidly killed following treatment with gefitinib [26]. The putative CSC marker CD133 was shown to be overexpressed in these DTPs, suggesting a CSC-like phonotype. Shien et al. established 13 gefitinib-resistant, EGFR-mutant NSCLC cell lines. Four of the latter lines showed an EMT phenotype and CSC-like properties, accompanied by overexpression of the CSC markers ALDHA1, ABCG2, and CD44 [27]. These findings may provide clues to overcoming resistance to EGFR-TKIs.

Loss of PTEN expression has been associated with decreased sensitivity to EGFR-TKIs owing to activation of PI3K-AKT signaling, impairment of ligand-induced ubiquitination, and degradation of activated EGFR in EGFR-mutant cells [17].

A recent study showed that knockdown of FAS and NF-κB enhanced cell death induced by the EGFR-TKI erlotinib in EGFR-mutant lung cancer cells [28]. Increased expression of the NF-κB inhibitor IκB predicted an improved response and PFS in EGFR-mutant NSCLC patients who received EGFR-TKI therapy [28]. These findings suggest that simultaneous inhibition of EGFR and NF-κB may be useful for the treatment of EGFR-mutant NSCLC.

CRKL overexpression induced resistance to EGFR-TKI mediated by ERK and AKT signaling in EGFR-mutant NSCLC cells [29]. CRKL amplification was found in a lung adenocarcinoma treated with an EGFR inhibitor [29]. These results suggest that CRKL is a therapeutic target for a subset of EGFR-mutant NSCLCs that harbor CRKL amplifications.

MicroRNAs (miRNAs) can function as either tumor suppressors or oncogenes and are used as diagnostic, prognostic, and therapeutic biomarkers in lung cancer. Four miRNAs (miR-30b, miR-30c, miR-221, and miR-222) have been shown to play important roles in gefitinib-induced apoptosis and EMT in NSCLC cells in vitro [30]. Several reports demonstrated that EGFR-activated miR-21 is a potential therapeutic target in tumors with mutations in EGFR [31, 32]. Members of the miR-200 family, targeting the E-cadherin suppressors ZEB1 and ZEB2, have been recognized as key suppressors of EMT associated with the resistance to EGFR-TKIs [27, 32]. These findings suggest that miRNAs may be promising therapeutic targets in EGFR-mutant NSCLC patients.

Recent studies have identified molecules associated with intrinsic resistance to EGFR-TKIs. BIM (BCL2L11) is a member of the Bcl-2 family encoding a proapoptotic protein. Upregulation of BIM is required for apoptosis induction by EGFR-TKIs in EGFR-mutant NSCLCs. Notably, a BIM deletion polymorphism occurs naturally in 13 % of East Asian individuals. This polymorphism can mediate intrinsic resistance to and reduced responses to EGFR-TKIs in EGFR-mutant NSCLC patients [33]. PFS in response to first-line EGFR-TKIs was significantly shorter in patients with the BIM deletion polymorphism than in those patients with wild-type BIM (8.6 and 4.6 months, respectively [p = 0.004]) [34]. An HDAC inhibitor could restore BIM function and resistance to EGFR-TKI [35]. Treatment with an HDAC inhibitor combined with an EGFR-TKI could represent an attractive strategy to treat EGFR-mutant NSCLC patients harboring a BIM deletion polymorphism.

Yamaguchi et al. demonstrated that thyroid transcription factor-1 (TTF-1)-induced ROR1 was required to sustain EGFR survival signaling as an “Achilles heel” in lung adenocarcinoma. ROR1 could activate kinase-dependent c-Src as well as kinase-independent EGFR-ErbB3, ErbB3 phosphorylation, and PIK3 signaling [36]. Inhibition of ROR1 expression could also restore the sensitivity to the EGFR-TKI. ROR1 expression was evaluated in erlotinib-pretreated tumor samples from 45 EGFR-mutant patients in the EURTAC trial to assess its potential as a predictive biomarker of PFS and overall survival (OS) [6, 37]. The PFS of patients with elevated ROR1 expression was significantly shorter than in those patients with low/intermediate ROR1 expression (11.8 months and 5.8 months, respectively) [37].

Noro et al. recently reported that MET gene amplification could predict short PFS and overall survival (OS) after gefitinib treatment in lung adenocarcinoma (LADC) harboring EGFR mutations [38]. MET-FISH-positive LADC patients defined by MET amplification and gene copy number gains (CNGs) exhibited significantly shorter PFS and OS than patients who were MET-FISH negative [38].

These findings suggest that BIM, ROR1, and MET gene status are involved in intrinsic resistance to EGFR-TKIs and may thus be predictive biomarkers for selecting patients who would benefit from EGFR-TKI therapy.

Understanding mechanisms of intrinsic and acquired resistance to EGFR-TKI, followed by the development of drugs targeted to molecules that can overcome this resistance, could serve as an important advance for targeting EGFR which is activated in NSCLC. Third-generation EGFR-TKIs such as osimertinib (AZD9291) and CO-1686 showed promising activity in EGFR mutation-positive NSCLCs harboring the T790M mutation [20, 21]. However, an EGFR C797S mutation arose as a novel mechanism of resistance to osimertinib (AZD9291) [22]. EGFR-TKIs combined with platinum-doublet chemotherapy showed prolonged PFS in EGFR-mutant NSCLCs in several clinical trials [39, 40]. A recent study demonstrated that the combination of erlotinib plus bevacizumab could be an effective first-line regimen to treat EGFR mutation-positive NSCLC patients [41]. PFS was significantly prolonged in patients receiving erlotinib plus bevacizumab compared to those that received erlotinib alone (16.0 months and 9.7 months, respectively). There are various mechanisms of acquired resistance to EGFR-TKIs. Future studies should clarify whether there exist other as yet unidentified mechanisms associated with acquired resistance to EGFR-TKI. In addition, treatments of individual patients should be based on assessment of the details of particular mechanism of resistance observed in the patient using re-biopsy or liquid biopsy samples.

NSCLC patients harboring an ALK fusion gene have shown a strong response to ALK-TKIs such as crizotinib, alectinib, and ceritinib [8–10]. Based on the clinical data, first-line treatment with crizotinib has become the standard therapy for NSCLC patients with ALK fusion gene. However, acquired resistance to ALK-TKI remains virtually inevitable.