MET-Targeting Anticancer Drugs—De Novo Design and Identification by Drug Repurposing

Abstract

The Met protein is a cell surface receptor tyrosine kinase predominantly expressed in epithelial cells. Aberrant regulation of MET is manifested by numerous mechanisms including amplification, mutations, deletion, fusion of the MET proto-oncogene, and protein overexpression. They represent the common causes of drug resistance to conventional and targeted chemotherapy in numerous cancer types. There is also accumulating evidence that MET/HGF signaling drives an immunosuppressive tumor microenvironment and dampens the efficacy of cancer immunotherapy. Substantial research effort has been invested in designing Met-targeting drugs with different mechanisms of action. In this review, we summarized the current preclinical and clinical research about the development of Met-targeting drugs for cancer therapeutics. Early attempts to evaluate Met-targeted therapies in clinical trials without selecting the appropriate patient population did not produce satisfactory outcomes. In the era of personalized medicine, cancer patients harboring MET exon 14 alterations or MET amplification have been found to respond well to Met-inhibitor therapy. The application of Met inhibitors to overcome drug resistance in cancer patients is discussed in this paper. Given that kinases play critical roles in cancer development, numerous kinase-mediated signaling pathways are attractive targets for cancer therapy. Existing kinase inhibitors have also been repurposed to new kinase targets or new indications in cancer. On the other hand, non-oncology drugs have also been repurposed for treating cancer through kinase inhibition as one of their reported anticancer mechanisms.

1. Introduction

The MET proto-oncogene is located on chromosome 7q21-31 and encodes a receptor tyrosine kinase Met (also known as c-Met) for the hepatocyte growth factor (HGF) [1,2,3]. HGF is a cytokine produced by stromal cells and fibroblasts to stimulate the migration of epithelial cells. The binding of HGF to Met triggers the dimerization of the receptor and autophosphorylation of the Met kinase domain, which subsequently activates downstream signaling pathways to promote cell survival and proliferation [4,5]. The HGF/Met axis also plays a critical role in regulating embryonic development in normal physiology [6]. In addition, it also controls the epithelial-to-mesenchymal (EMT) transition process and promotes the differentiation of multipotent cells to numerous other cell types [7]. Furthermore, the HGF/Met pathway is also involved in tissue regeneration [8].

The oncogenic role of Met was first reported in a human osteosarcoma cell line carrying a TPR (translocated promoter region)-MET genomic rearrangement after exposure to the carcinogen N-methyl-N′-nitro-N-nitrosoguanidine [9]. Hyperactivation of the HGF/Met signaling pathway is frequently observed in numerous cancer cell types and it is generally associated with dismal patient survival [10]. Aberrant activation of MET is usually caused by mutations, genomic rearrangement, amplification, and protein overexpression, which promotes cancer development by activating the downstream PI3K/AKT, Ras/MAPK, JAK/STAT, SRC, and Wnt/β-catenin signaling pathways [11,12,13,14].

With the high propensity of MET dysregulation in numerous cancers, Met is considered an attractive therapeutic target for cancer therapy [15,16]. However, early attempts to target Met for cancer therapy in the clinical setting were not satisfactory. To date, only a few Met inhibitors have been clinically approved. The major hurdle is the weak Met-inhibitory activity of the drug candidates and a lack of an appropriate patient selection for the clinical trials [17,18,19].

MET-exon-14-skipping mutations represent a valuable predictive biomarker to identify non-small-cell lung cancer (NSCLC) patients who are likely to respond well to Met tyrosine kinase inhibitors (TKIs) [20]. MET exon 14 contains the amino acid residue Y1003, which is critical for the binding of the Met protein to an E3 ubiquitin ligase CBL. MET-exon-14-skipping mutations produce a shorter exon-14-spliced Met protein, thereby impairing CBL-mediated Met protein degradation and subsequently leading to the constitutional activation of MET signaling. Moreover, Met TKIs were also shown to display promising efficacy in NSCLC patients harboring other MET abnormalities including amplification or gene fusions. Following this successful clinical observation, numerous newer Met TKIs are currently under clinical investigation.

This review aims to provide an updated summary of the design and preclinical and clinical research of Met-targeting drugs for cancer treatment. Unlike early clinical trials investigating Met-targeted therapy in unselected patient cohorts, more recent clinical studies focusing on MET exon 14 alterations and MET amplification have produced encouraging treatment responses to Met-inhibitor therapy. The novel strategies of using Met inhibitors to overcome drug resistance and potentiate immunotherapy in cancer patients are discussed. Drug repurposing refers to the application of clinically approved drugs with characterized pharmacological properties and known adverse effect profiles to new indications. Given that kinases play critical roles in cancer development, numerous kinase-mediated signaling pathways are attractive targets for cancer therapy, and existing kinase inhibitors have also been repurposed to new kinase targets or new indications in cancer. On the other hand, non-oncology drugs have been repurposed for treating cancer through kinase inhibition as one of their reported anticancer mechanisms.

2. Recent Development of Met-Targeting Drug Candidates for Cancer Therapy

Currently, there are four major strategies to inhibit the MET signaling pathway for cancer therapy: (i) preventing the extracellular binding of Met and HGF by a neutralizing antibody; (ii) blocking the dimerization of Met; (iii) preventing the phosphorylation of the tyrosine kinase domain of Met with a small molecule inhibitor; and (iv) blocking other signaling molecules downstream of MET. Their site of action in the HGF/MET pathway is illustrated in Figure 1. The updated status about development of the first three strategies (i–iii) is summarized below. Strategy (iv) involves the inhibition of very diverse signaling pathways and is out of the scope of the current review. Readers may refer to other recent excellent reviews on the topic for more detailed descriptions [21].

2.1. Antibody-Based Inhibitors of the HGF-MET Axis

Given that the interaction of HGF and Met is a prerequisite for the activation of the MET signaling pathway, both proteins are promising molecular targets for therapeutic intervention. Therapeutic agents designed to target HGF or Met have been evaluated alone as a monotherapy or in combination with other targeted cancer therapies. They are in various stages of preclinical and clinical development. HGF-neutralizing antibodies work by binding to the fully processed HGF molecules and preventing them from interacting with the Met receptor. On the other hand, anti-Met monoclonal antibodies (mAb) work by competing with HGF, and they do not trigger Met receptor dimerization for signal transduction. Decoys of Met have been designed to interact with intact Met or HGF, or both, to disrupt the dimerization of the Met receptor. To date, no antibody-based biotherapeutics targeting Met/HGF have been clinically approved. The latest research and development of these antibody-based therapeutic agents (including classical therapeutic monoclonal antibodies, bispecific antibodies, and antibody–drug conjugates (ADC)) is summarized in the following section.

2.1.1. Anti-HGF mAbs

HGF, also known as scatter factor (SF), is a paracrine multifunctional cytokine secreted by mesenchymal cells that regulates cell growth, motility, and morphogenesis. It is secreted as an inactive polypeptide, which is cleaved by serine proteases into an alpha chain (69 kDa) and a beta chain (34 kDa) during activation. A disulfide bond formation between the alpha and beta chains generates the active heterodimeric molecule. Upon the binding of HGF to its receptor, Met, the downstream intracellular signal transduction pathways are activated to regulate cell proliferation, migration, invasion, angiogenesis, and apoptosis. Elevated tumoral and plasma levels of HGF have been reported in patients with breast cancer, glioma, multiple myeloma, and sarcomas [22,23]. HGF-targeting antibodies bind to extracellular HGF to prevent its interaction with Met and the subsequent activation of the HGF/Met pathway.

Rilotumumab (AMG-102) is an mAb that was developed against the full-length HGF protein. It is the first HGF inhibitor to reach phase 3 clinical testing. It binds to the HGF β-chain and inhibits HGF–Met interaction. In a phase 1 trial (NCT01791374), rilotumumab was well-tolerated in patients with solid tumors, with only low-grade adverse events being observed (fatigue, constipation, anorexia, nausea, and vomiting) [24]. In a phase 2 trial (NCT00719550) studying the combination of rilotumumab with epirubicin, cisplatin, and capecitabine, overall survival (OS) and progression-free survival (PFS) were extended in Met-positive patients with gastric cancer or gastroesophageal junction adenocarcinoma [25]. However, two phase 3 clinical trials (RILOMET-1 and RILOMET-2) were halted because of a lack of efficacy or patient deaths [26]. In the RILOMET-1 trial, a statistically higher number of patient deaths were observed in the rilotumumab arm (128 deaths) than in the placebo arm (107 deaths). Moreover, a shorter median OS was also noted in the patient group receiving rilotumumab than the placebo (9.6 months versus 11.5 months, hazard ratio: 1.37, p = 0.016). However, further subgroup analysis also did not reveal any survival benefit achieved by rilotumumab in patients with a higher Met expression [26].

Ficlatuzumab (AV-299) is another humanized HGF-targeting mAb under clinical investigation. Results from a phase 2 study in lung cancer patients harboring EGFR mutations demonstrated that the VeriStrat-poor patient subgroup may gain survival benefits from the inclusion of ficlatuzumab in the treatment [27]. The VeriStrat test is a serum-based assay that identifies responders according to a serum mass spectrometry proteomic signature. The intensity and pattern of eight mass spectral features were captured by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry and compared with a control. The test examined pretreatment serum samples from the patients and assign them to a “VeriStrat Good” or “VeriStrat Poor” status as an indicator of treatment response. However, in the phase 2 FOCAL trial (NCT02318368) evaluating a first-line treatment with a ficlatuzumab–erlotinib combination versus erlotinib monotherapy in VeriStrat-poor patients with EGFR-mutated NSCLC, the study was prematurely terminated due to a high discontinuation rate in the drug combination group [28].

MP0250 is a designed-ankyrin-repeat-protein (DARPin)-based drug candidate that interacts with both HGF and vascular endothelial growth factor (VEGF) [29]. It also has an additional domain to bind to human serum albumin (HSA), which enhances its plasma half-life and promotes tumor penetration. DARPins are antibody-mimetic-engineered proteins. Compared to antibodies, DARPins are smaller in size, and they generally exhibit favorable pharmacokinetic profiles and facilitate both a high-affinity binding and efficacy. MP0250 comprises one anti-VEGF-A, one anti-HGF, and two anti-human serum albumin (HSA) DARPin^® domains within a single polypeptide chain. The HSA-binding DARPin domain serves as a carrier to extend the half-life and leads to a more favorable pharmacokinetic profile. In preclinical studies, MP0250 was shown to bind to and inhibit both HGF and VEGF [29]. When used alone or in combination with bortezomib, MP0250 inhibited myeloma cell migration, invasion, and bone destruction in an orthotopic mouse model of multiple myeloma [30]. Currently, MP0250 is under clinical investigation in a multicenter phase 2 trial (NCT03136653) in combination with bortezomib and dexamethasone for patients with refractory and relapsed multiple myeloma.

2.1.2. Anti-Met mAbs

Onartuzumab (MetMAb) is an mAb raised against Met, which interferes with the binding of the HGF α-chain to its Met-ligand-binding domain [31]. Results from a phase 2 trial (NCT00854308) showed a superior clinical efficacy from the onartuzumab–erlotinib combination compared with erlotinib alone in lung cancer patients with a tumor expressing a high Met protein level [32,33]. However, subsequent testing in another phase 3 OAM4971g trial (NCT01456325, METLung) was discontinued due to shorter patient survival in the onartuzumab–erlotinib combination group [19,34].

Emibetuzumab (LY2875358) is a humanized bivalent IgG4 monoclonal antibody targeting the Met protein. It blocks HGF binding to Met and therefore the downstream HGF-induced signaling pathways [35]. Unlike onartuzumab, emibetuzumab also induces the internalization and degradation of the Met receptor protein, thus enabling it to exert an anticancer effect in both HGF-dependent and -independent (including MET-amplified) cancer [35]. In a recent phase 2 trial (NCT0190652) comparing an emibetuzumab–erlotinib combination with emibetuzumab monotherapy in stage IV NSCLC patients with acquired resistance to erlotinib, the disease control rate (DCR) and PFS were higher for the drug combination (50%/3.3 months) than for emibetuzumab alone (26%/1.6 months) [36]. Similar results were observed in patients with a high tumoral Met protein expression (i.e., >60% of cells expressing Met at >2 + IHC staining intensity).

ARGX-111 was developed by screening from a panel of anti-Met mAbs generated by the SIMPLE antibody platform [37]. It is a fully human mAb. Its Fc region was afucosylated (where the oligosaccharides in the Fc region of the antibody do not have any fucose sugar unit) for increased tissue penetration and enhanced antibody-dependent cell-mediated cytotoxicity (ADCC) [37]. In cancer-cell-line studies, ARGX-111 was shown to inhibit both HGF-dependent and -independent MET activation [37]. It exhibited potent cytotoxicity through ADCC against human cancer cell lines and patient-derived leukemia cells expressing variable Met levels. In orthotopic tumor models, ARGX-111 was found to remarkably reduce the number of circulating tumor cells and suppress distant metastasis [37]. In a phase 1 clinical trial (NCT02055066) for patients with hematological and solid tumors, ARGX-111 displayed favorable safety profiles [38]. One patient (1/16) with metastatic, MET-amplified gastric cancer refractory to platinum- and taxane-based chemotherapy maintained a stable disease for 6 months following treatment with ARGX-111.

SAIT301 is a Met-specific mAb that promotes leucine-rich repeats (LRR) and immunoglobulin-like domain-containing protein 1 (LRIG1)-mediated Met degradation in a Cbl-independent manner [39]. This humanized anti-Met antibody was originally derived from the mouse antibody F46. It binds to the Sema domain of Met with a higher affinity than HGF, subsequently promoting the cellular internalization and degradation of the receptor in both an HGF-dependent and -independent manner [40]. Moreover, SAIT301 also competes with HGF for binding to Met, thus preventing the downstream signaling pathways and leading to reduced cell proliferation and diminished cancer invasiveness in an HGF-dependent manner [40]. Findings from a phase 1 trial (NCT02296879) in Met-overexpressed colorectal cancer patients showed that SAIT301 was well-tolerated [41]. Partial response and stable disease were achieved by SAIT301 in the cohort of 16 metastatic colorectal cancer patients [41].

Telisotuzumab (ABT-700 or h224G11) is a humanized bivalent Met-specific mAb that inhibits Met dimerization and activation [42]. It gave rise to long-term tumor regression in tumor xenograft models harboring Met overexpression, amplification, and autocrine HGF stimulation [42]. When combined with other chemotherapeutic drugs (including gemcitabine and docetaxel), telisotuzumab produced a synergistic anticancer effect and a more prolonged duration of a tumor-suppressive effect [42]. Recent data from a phase 1 dose-escalation and -expansion trial (NCT01472016) showed that telisotuzumab was well-tolerated in patients bearing MET-amplified solid tumors [43]. Due to the promising anti-Met and anti-tumor activities of telisotuzumab, an antibody–drug conjugate (ADC) called telisotuzumab vedotin (ABBV-399) is being developed, which will be described in a later section on ADC [44].

Recently, an interesting strategy of employing two anti-Met mAbs to target non-overlapping Met epitopes called Sym015 has been reported [45]. Sym015 binds to human and monkey Met with a high affinity and promotes rapid Met cellular internalization and degradation, which is accompanied by dampened MET signaling in both HGF-dependent and -independent manners. In various cancer cell lines with or without MET overexpression or amplification, Sym015 was shown to inhibit proliferation, migration, invasion, and prominent cell death [45,46]. Moreover, Sym015 was also found to induce complement-dependent cytotoxicity (CDC) and ADCC to promote cellular lysis [45,46]. Importantly, xenograft tumors generated by cancer cells that are resistant to emibetuzumab were reported to be sensitive to Sym015 [46], thus suggesting its utility in treating tumors that have acquired resistance after previous anti-Met antibody-based therapy. Currently, Sym015 is being investigated in a phase 1 trial for patients with advanced solid tumors (NCT02648724).

Employing an antibody-decoy strategy, a hybrid molecule capable of simultaneously inhibiting HGF and Met has been developed [47]. The hybrid molecule was constructed by the “head-to-tail” or “tail-to-head” fusion of a single-chain Fab derived from a Met mAb (MvDN30) with a “decoy” (i.e., recombinant engineered Met extracellular domain; decoyMET). The anti-Met mAb facilitated the removal of Met from the cell surface, whereas the decoy provided the HGF-sequestering ability (Figure 2). To prevent antibody/decoy interaction and subsequent neutralization, a specific amino acid residue in the extracellular domain of Met (lysine 842) critical for MvDN30 binding was engineered to K842E in the recombinant decoyMET [47]. Using an orthotopic pancreatic tumor model in SCID mice, the MvDN30–decoyMET hybrid molecule was shown to sequester Met and HGF simultaneously, thereby inhibiting HGF-induced Met downstream signaling to inhibit cancer growth and metastatic spread [47].

2.1.3. Bispecific Antibodies Simultaneously Targeting Met and Other Signaling Proteins

Simultaneously Targeting Met and EGFR

The crosstalk between Met and other oncogenic signaling pathways to propel cancer development has been well established [48]. The aberrant regulation of MET also contributes to drug resistance acquired by cancer cells following EGFR-targeted therapy [49]. Therefore, the simultaneous inhibition of both MET and EGFR signaling cascades represents a promising approach to enhance the anticancer effects of Met- and EGFR-targeting therapies. To date, five bispecific antibodies targeting both Met and EGFR (MET-HER1, amivantamab, LY3164530, ME22S, and B10v5x225-H/M) have been reported. The proof-of-concept evaluation of this approach was first established in the investigation of Met-HER1 in 2013 [50]. Currently, amivantamab has advanced to phase 2 clinical trials, whereas LY3164530 was discontinued due to toxicities and limited clinical efficacy [51].

Simultaneously Targeting Met and VEGFR-2

VEGFR-2 is an important molecular target for anti-angiogenesis therapy in cancer treatment [52]. Bevacizumab is a humanized anti-VEGF monoclonal antibody clinically approved for cancer therapy by inhibiting the VEGF signaling pathway. However, it is well known that increased Met expression represents a major resistance mechanism contributing to bevacizumab treatment [53]. To this end, the development of bispecific monoclonal antibody targeting both Met and VEGFR2 to overcome bevacizumab has been reported [54]. The amino acid sequences from onartuzumab (an anti-Met mAb) and tanibirumab (an anti-VEGFR2 mAb) were used to construct one such bispecific Met- and VEGFR2-targeting mAb (called “bsVEMET”) using an engineered heterodimeric fragment crystallizable region (Fc) platform [54]. In human vascular endothelial cells, HUVEC, overexpressing both Met and VEGFR2, bsVEMET was confirmed to bind to the two target antigens simultaneously with high affinities [54]. Moreover, bsVEMET was also found to inhibit the HGF/VEGF-induced phosphorylation of Met and VEGF2 and their downstream signaling molecules. Furthermore, in a gastric cancer MKN45 tumor xenograft model in vivo, the antitumor effect of bsVEMET was stronger than that derived from the individual use of either onartuzumab or tanibirumab and was also slightly more effective than that mediated by the combination of the two mAbs [54].

Simultaneously Targeting Met and PD-1

Bispecific mAbs simultaneously inhibiting MET signaling and blocking PD-1 (anti-programmed cell death receptor) for the restoration of T-cell-mediated anticancer response have also been developed. PD-1 is an inhibitory checkpoint protein expressed on immune cells (such as T cells, B cells, and natural killer cells), which normally suppress a host’s immunity from attacking other cells in the body. Upon the binding of PD-1 to its major ligand, PD-L1, expressing on tumor cells, the T-cell-mediated cancer-killing effect is suppressed. In recent years, various mAbs targeting either PD-1 or PD-L1 (such as nivolumab, pembrolizumab, and ipilimumab) have been approved for cancer immunotherapy. Numerous strategies have also been developed to potentiate the clinical outcome of immunotherapeutic drugs. To this end, bispecific mAbs (including Met-PD1 BsAb, DVD-Ig, and IgG-ScFv) targeting both Met and PD-1 have been designed and evaluated in preclinical studies [55,56,57]. In gastric and NSCLC cell lines, these bispecific mAbs were shown to interact with Met and inhibit HGF-dependent cancer proliferation and migration and induce apoptosis. Using a co-culture system consisting of cancer cells and peripheral blood T cells, Met-PD-1 was shown to partially restore the ability of T cells to produce interleukin 2 and other cytokines, but the effect was only very modest [55]. In a gastric MKN45 tumor xenograft model, the three bispecific mAbs were found to moderately delay tumor growth [55,56,57]. Moreover, a concurrent administration of the bispecific mAb and peripheral blood leucocytes (PBMC) was only found to slightly enhance the overall tumor-growth-suppressive effect. Therefore, targeting PD-1 by using bispecific mAbs for restoring cytotoxic T cell activity may not produce a dramatic antitumor effect in an experimental setting.

2.1.4. Antibody–Drug Conjugates (ADC) Targeting Met

ADCs are highly specific biopharmaceutical drugs composed of an antibody linked, via a chemical linker, to a cytotoxic compound for targeted cancer therapy. A few anti-Met ADCs have been designed, and they are currently in clinical investigation.

Telisotuzumab vedotin (ABBV-399) was designed by conjugating telisotuzumab (ABT-700, anti-Met mAb) with monomethyl auristatin E (MMAE, a cytotoxic drug) via a cleavable dipeptide linker [44]. It was shown to produce a potent anticancer effect against Met-overexpressing cancer cell lines with either Met protein overexpression or MET gene amplification. The specificity of ABBV-399 against Met-overexpressing cancer cells stems from the fact that a threshold level of cell surface Met expression (≥100,000 Met molecules per cell) is needed for the cells to experience the cytotoxic effect from the ADC [44]. Thus, normal cells (including fibroblasts, endothelial cells, and non-cancerous epithelial cells) that have a low cell surface expression of Met protein are insensitive to ABBV-399 [44]. In various Met-overexpressing xenograft tumor models in vivo, ABBV-399 was shown to produce remarkable and durable tumor regression, and it also produced synergistic antitumor effects with other chemotherapeutic drugs. Results from a recent phase 1 trial for patients with advanced solid cancers (NCT02099058) showed that ABBV-399 was well-tolerated, and it produced partial response (three out of sixteen patients) in a sixteen patient Met-expressing NSCLC cohort [58].

SHR-A1403 is another ADC formed by conjugating the humanized anti-Met monoclonal antibody HTI-1066 with auristatin analog SHR152852 via a non-cleavable linker [59]. In animal studies, SHR-A1403 was shown to be stable, and free toxin was not detectable in the serum [60]. It was shown to significantly retard xenograft tumor growth in various animal models from different cancer types with Met overexpression [59,60]. SHR-A1403 is currently being evaluated in a phase 1 clinical trial (NCT03856541).

TR1801-ADC is considered a “third generation” ADC generated by the site-specific conjugation of the humanized monoclonal antibody hD12 to tesirine via a PBD toxin linker [61]. Among all the anti-Met ADCs tested, TR1801-ADC was shown to be the most cytotoxic in cancer cell lines expressing variable Met levels [62]. Importantly, TR1801-ADC was also found to display a potent (at a single dose of 1 mg/kg) and long-lasting (up to 4 weeks) tumor-suppressive effect in a CRC-patient-derived tumor xenograft model [61]. Currently, TR1801-ADC is being investigated in a phase 1 trial (NCT03859752) in adult patients with Met-overexpressing solid tumors.

B10v5x225-H-vc-MMAE and B10v5x225-M-vc-MMAE are two recently reported dual-acting ADCs that target both Met and EGFR [63]. These bispecific antibodies were designed by connecting optimized protein sequences specific to the EGFR epitope derived from the EGFR mAb (cetuximab C225, with either a high or moderate binding affinity (therefore termed 255-H and 255-M, respectively)) and another one specific to the Met SEMA domain (termed B10v5, derived from a phage-displayed anti-Met mAb B10) [63]. On the other hand, monomethyl auristatin E (MMAE) is a commonly used payload in ADC design. Results from cell line studies showed that both B10v5x225-M and B10v5x225-H could inhibit the ligand-induced receptor activation of Met and EGFR, thereby blocking MET signaling and EGFR phosphorylation [63]. The two bispecific ADCs were shown to display potent anticancer effects (with IC₅₀ ranging from 0.4 to 1 nM) in a panel of cancer cell lines expressing various levels of Met and EGFR [63].

2.2. Inhibiting Met Dimerization

The dimerization of the Met receptor protein is known to be an essential signal transduction mechanism regulating the HGF/Met pathway. DNA-based aptamers have been designed to bind specifically to the extracellular domain of the Met receptor protein and compete with HGF, thereby blocking Met dimerization [64,65]. Two representative aptamer designs are described below. CLN0003_SL1 (abbreviated as SL1) is a 50-mer Met-binding DNA aptamer, which was shown to work as a Met antagonist by competing with HGF for Met binding [66,67]. SL1 exhibited high specificities and affinities to recombinant and cellular-expressed Met. It inhibited HGF-dependent Met activation, the downstream signaling pathway, and cancer cell migration. In a more recent study, Aida et al. reported a photocleavable molecular glue, ^PCGlue-NBD, that could inhibit receptor dimerization and subsequent activation [68]. The ^PCGlue-NBD molecule is a dendrimer with nine guanidinium ion (Gu⁺) pendants, which form a multivalent salt bridge with the oxyanionic groups on the target protein of interest. Moreover, there are photocleavable linkages on the glue molecule. When it is subjected to UV irradiation, protein-bound ^PCGlue-NBD is degraded and subsequently released from the protein. Using the HGF-Met axis as a proof-of-concept demonstration, the interaction between HGF and Met could be blocked when HGF was bound by ^PCGlue-NBD. However, upon exposure to UV light, the high-affinity interaction between HGF and Met was restored to allow Met dimerization and activation of the downstream signaling pathways [65].

2.3. Small-Molecule Met Tyrosine Kinase Inhibitors (TKIs)

The first crystal structure of Met kinase domain bound with the microbial alkaloid K-252a was resolved in 2002 [69]. The availability of more crystal structures of Met kinase bound to various small molecule inhibitors has facilitated the rational design of Met TKIs capable of occupying the ATP pocket of the enzyme [70]. Underiner et al. compiled an excellent review about the detailed structure–activity relationship evaluation of the pioneer Met TKIs [71].

Numerous Met TKIs have been designed and developed in recent years [72]. They can be classified as type Ia, Ib, II, and III according to their binding mode to the Met tyrosine kinase domain [73]. The stage of their clinical investigation and approval status is summarized in

Table 1. Representative Met TKIs approved or under clinical investigation for treating cancers with MET dysregulation.

Type	Name	Molecular Target(s) *	Approval Status (Indications)	Representative Clinical Trials
Ia	Crizotinib (PF-02341066)	ROS1 > Met > ALK	Approved (indicated for advanced NSCLC with ALK or ROS1; breakthrough therapy for advanced NSCLC with MET exon 14 skipping as second-line therapy)	Phase 1 (NCT00585195)—first-in-class drug candidate demonstrating clinical efficacy in NSCLC patients bearing MET-exon-14-skipping mutations Phase 2 (NCT02465060) Phase 2 (NCT02499614) Phase 2 (NCT02034981)
Ib	Capmatinib (INCB28060)	Met	Approved (indicated for advanced NSCLC with MET exon 14 skipping)	Phase 2 (NCT02414139, GEOMETRY mono-1)—pivotal trial demonstrating substantial antitumor efficacy of capmatinib in NSCLC patients with MET-exon-14-skipping mutations Phase Ib/II (NCT01610336)—combination with geftinib Phase Ib/II (NCT02468661)—combination with erlotinib
	Tepotinib (EMD1214063)	Met	Approved (breakthrough therapy for advanced NSCLC with MET exon 14 skipping as first-line therapy)	Phase 2 (NCT02864992, VISION)—pivotal trial demonstrating favorable ORR and mPFS from tepotinib in NSCLC patients with MET-exon-14-skipping mutations Phase Ib/II (NCT01982955, INSIGHT)—combination with gefitinib Phase II (NCT03940703, INSIGHT 2)—combination with osimertinib
	Savolitinib (AZD6094, volitinib)	Met	Conditionally approved in China for advanced NSCLC with MET exon 14 skipping mutations	Phase 2 (NCT02897479)—pivotal trial demonstrating favorable clinical outcome and safety profile from single-agent savolitinib in Chinese patients with PSC, brain metastasis, and NSCLC patients with MET-exon-14-skipping mutations. Savolitinib was conditionally approved in China. Phase Ib (NCT02143466, TATTON), Phase 2 (NCT03778229; SAVANNAH) and Phase 2 (NCT03944772; ORCHARD)—demonstrating clinical efficacy of savolitinib–osimertinib combination in acquired resistance setting in advanced NSCLC with MET alterations.
	APL-101 (Bozitinib)	Met	Under clinical investigation	Phase 1 (NCT03175224)
	SAR125844	Met	Under clinical investigation	Phase 1/2 (NCT01391533) Phase 2 (NCT02435121)
II	Cabozantinib (XL-184, BMS-907351)	VEGFR2 > Met > Ret > Kit > Flt-1/2/3/4 > AXL > Tie2	Approved (indicated for renal cell carcinoma and advanced metastatic medullary thyroid carcinoma)	Phase 3 (NCT01865747)—cabozantinib improved PFS compared to everolimus in RCC patients who progressed after VEGFR-targeted therapy Phase 2 (NCT01639508)—cabozantinib showing favorable clinical efficacy in patients with RET-rearranged lung cancer
	Merestinib (LY2801653)	DDR1 > Met ~ AXL > MKNK1/2 > FLT3 > DDR2 > MERTK > MST1R > ROS1	Under clinical investigation	Phase 1 (NCT03027284) Phase 2 (NCT02711553) Phase 2 (NCT02920996)
	Glesatinib (MGCD265)	Met > RON > VEGFR1/2 /3 > Tie-2	Under clinical investigation	Phase 1 (NCT00697632) Phase 2 (NCT02544633)
	Sitravatinib (MGCD516)	TAM receptors (Axl, Mer) > VEGFR2 > KIT > Met	Under clinical investigation	Phase 2 (NCT03606174)—combination with PD-1 checkpoint inhibitor Phase 3 (NCT03906071)—combination with PD-1 checkpoint inhibitor in metastatic NSCLC
	Altiratinib (DCC-2701)	Met > Tie2 > VEGFR2	Under clinical investigation	Phase 1 (NCT02228811)
	Foretinib (GSK1363089)	Met > AXL > RON > VEGFRs	Product development terminated by sponsor company in 2015	Phase 2 (NCT00726323) Phase 2 (NCT02034097)—product development terminated by sponsor
III	Tivantinib (ARQ197)	Met > RON	Under clinical investigation	Phase 2 (NCT00988741) Phase 2 (NCT01892527)—combination with cetuximab in resistant MET high subjects Phase 2 (NCT01519414)

Abbreviations: mPFS, median progression-free survival; NSCLC, non-small-cell lung cancer; ORR, objective response rate; RCC, renal cell carcinoma. * The relative rank of target inhibition is shown according to Ki value in cell-free or cell-based assays.

. There is a conserved region called the Asp-Phe-Gly (DFG) motif at the N terminus of the activation loop of protein kinases including Met [74]. Upon conformational change, the Met tyrosine kinase is switched from a catalytically active state (“DFG-in” conformation) to an inactive state (“DFG-out” conformation) [75]. Both type I and II Met TKIs are ATP-competitive inhibitors. More specifically, type I inhibitors bind to the U-shaped ATP-binding pocket of the active “DFG-in” state through the M1160 residue by hydrogen bonding [70]. They also form π-π stacking interactions with the Y1230 residue within the activation loop and link with the hinge [70]. This makes type I Met TKIs more selective than other types of Met TKIs. Type I Met TKIs can be further classified into type Ia and Ib according to their binding to the solvent front glycine residue (G1163) on the Met receptor. Type II Met TKIs are also ATP-competitive inhibitors. They bind to the hydrophobic pocket of the inactive “DFG-out” state [76], which exerts an enzyme-inhibitory effect by locking the molecule in the inactive state [77]. On the other hand, type III Met TKIs are allosteric inhibitors that do not compete with ATP, and they were designed to interact with the inactive Met conformation [78].

2.3.1. Type Ia Met TKIs

Crizotinib (PF-02341066) is an ATP-competitive multitargeted TKI inhibiting Met/ALK/ROS1 [79]. A detailed structure-based drug design strategy and lead compound optimization was reported by Cui et al. [79]. The structural backbone of 3-substituted indolin-2-ones (known to be a potent class of kinase inhibitors, as exemplified by sunitinib) was used in the search for lead compounds of novel Met TKIs. The cocrystal of PHA-665752 (analog of sunitinib) and Met complex revealed a novel binding model of Met, which was subsequently used to guide both the lead candidate prioritization and further structural optimization. A novel series of 2-amino-5-aryl-3-benzloxy-2-aminopyridine analogs was generated, which were shown to interact with the ATP-binding pocket of Met with a better ligand efficiency than PHA-665752. Further structural modification of the lead series gave rise to crizotinib as a potent Met and ALK inhibitor [79]. In the Profile 1001 trial (NCT00585195), crizotinib was the first-in-class drug candidate shown to exhibit clinical efficacy in NSCLC patients bearing MET-exon-14-skipping mutations [80]. Among the eighteen evaluable patients in the study, there were eight (44%) partial responses and nine (50%) stable diseases. Importantly, none of the subjects exhibited progressive disease. The drug-related adverse events were mostly grade 1 or 2. No grade 4 adverse event was registered. After crizotinib therapy, the objective response rate (ORR) was 32%, and the median progression-free survival (mPFS) was 7.3 months [81]. Following the Profile 1001 trial, crizotinib was granted a breakthrough therapy designation by the US Food and Drug Administration (FDA) for the management of metastatic NSCLC patients bearing MET exon 14 alterations who progressed after receiving platinum-based chemotherapy. Moreover, crizotinib is also the first small-molecule TKI approved by FDA for the treatment of ALK-mutated NSCLC [82]. Furthermore, crizotinib is currently also recommended by the National Comprehensive Cancer Network (NCCN) for MET-mutated NSCLC [83].

2.3.2. Type Ib Met TKIs

Capmatinib and tepotinib are the first two types Ib Met TKIs approved by FDA specifically for the treatment of metastatic NSCLC harboring MET-exon-14-skipping mutations [84]. Capmatinib (INCB28060) and tepotinib (EMD1214063) are both orally administered, selective, and highly potent Met TKIs with more than a 10,000-fold selectivity over other major human kinases [85,86]. The approvals were based on the multi-cohort phase II GEOMETRY mono-1 (capmatinib) [87] and VISION (tepotinib) studies [88]. In the GEOMETRY study (NCT02414139), capmatinib displayed a substantial antitumor efficacy in advanced NSCLC patients with MET-exon-14-skipping mutations, particularly in those not treated previously [87]. Interestingly, the clinical efficacy of capmatinib in MET-amplified NSCLC patients was higher in tumors with a higher MET copy number than those with a lower MET copy number [87]. In the VISION trial (NCT02864992), NSCLC patients harboring MET-exon-14-skipping mutations were recruited according to their liquid biopsy (DNA-based) or tissue biopsy (RNA-based) [88]. The ORR and mPFS of all recruited patients were 46% and 11.1 months, respectively. The most common adverse effects of tepotinib were an increase in blood creatinine and peripheral oedema, which were all manageable. Due to its favorable clinical outcome, tepotinib was subsequently approved in Japan for NSCLC patients with MET-exon-14-skipping mutations.

There are a few more type Ib Met TKIs currently under clinical investigation, (savolitinib, APL-101, and SAR125844). Savolitinib (also known as AZD6094 and volitinib) is a highly selective Met TKI that is conditionally approved in China for advanced NSCLC with MET-exon-14-skipping mutations [89]. In a pivotal phase 2 trial (NCT02897479), savolitinib as a monotherapy demonstrated notable clinical responses and a well-tolerated safety profile in Chinese patients with pulmonary sarcomatoid carcinoma (PSC), brain metastasis, and other NSCLC subtypes positive for MET exon 14 mutations [90]. The ORRs of the PSC group and the NSCLC subtypes were 55% and 50%, respectively [90]. Moreover, the combination of savolitinib and osimertinib has been evaluated in NSCLC patients refractory to first- to third-generation EGFR TKIs, and it exhibited MET amplification or Met immunohistochemistry (IHC) 3+ in a phase Ib TATTON study (NCT02143466). The drug combination was found to achieve similar ORR and mPFS results in NSCLC patients with or without EGFR T790M mutation (ORR of 67% and mPFS of 11.0 months versus ORR of 65% and mPFS of 9.0 months, respectively). More recently, preliminary results from ongoing phase 2 clinical studies (SAVANNAH study (NCT03778229) and ORCHARD study (NCT03944772)) also confirmed the efficacy and safety of a savoltinib–osimertinib combination in an acquired resistance setting in advanced NSCLC and its subtypes with MET alterations [91,92]. A few phase 3 trials (NCT04923945 and NCT05261399) are ongoing that are investigating the clinical efficacy of savolitinib for NSCLC patients harboring MET exon 14 mutations or NSCLC patients whose disease progressed on osimertinib.

SAR125844 is a highly selective type Ib Met TKI that is administered intravenously [93] that is being investigated in a phase 1/2 clinical study (NCT01391533). Preliminary findings revealed that SAR125844 gave rise to a partial response (PR) and stable disease (SD) in five and seventeen (out of twenty-two) NSCLC patients with MET amplification, whereas no response was noted in patients with high-p-Met tumors [94].

PLB-1001 (also known as APL-101) is another potent and ATP-competitive type of Ib Met TKI. In various preclinical models, it displayed remarkable anticancer potency by selectively inhibiting MET-altered cancer cells [95]. In a phase 1 clinical trial (NCT03175224) enrolling MET-altered chemo-resistant glioma patients, PLB-1001 was well-tolerated, and it produced partial response in at least two advanced patients [96].

2.3.3. Type II Met TKIs

Numerous studies have reported that certain mutations near the active site of Met could lead to resistance to type I Met TKIs [97]. Based on clinical experience with type I Met TKIs, the response rates ranged from 32 to 55% and the median PFS achieved were about 5–12 months [81,98,99,100,101]. Importantly, type I Met TKIs are limited by the inevitable emergence of acquired drug resistance. To this end, type II Met TKIs do not rely on A-loop interactions for binding to the kinase, but they recognize the inactivated conformation of the Met activation loop (DFG-out). Therefore, type II Met TKIs are generally considered more effective in killing cancer cells that possess the type I Met inhibitor resistance-causing mutations because their binding interactions extend outside the Met active site [97,102,103,104,105,106,107]. On the other hand, due to the binding mode of type II Met TKIs with the targeted Met kinase, they are considerably less selective than type I Met TKIs and therefore they are associated with an increased risk of toxicity [108].

Cabozantinib (XL-184 or BMS-907351) is a type II Met inhibitor that was approved by the FDA for the treatment of metastatic medullary thyroid cancer (November 2012), renal cell carcinoma (April 2016), and hepatocellular carcinoma (January 2019). It inhibits vascular endothelial growth factor receptor 2, Met, FMS-like tyrosine kinase 3, and Kit (stem cell factor receptor). Cabozantinib and a few other type II Met TKIs were developed with a 6,7-disubstituted-4-(2-fluorophenoxy) quinoline backbone (Figure 3). Analysis of the structure–activity relationships of quinoline-based Met inhibitors revealed that the 6,7-disubstituted-4-phenoxyquinoline backbone (moiety A) and an aryl fragment (moiety B) formed hydrogen bonds and van der Waals interactions with Met kinase [109,110,111]. Moreover, moiety B fitted into the hydrophobic pocket. These 6,7-disubstituted-4-phenoxyquinoline derivatives display two common structural features (5-atom regulation and containing both hydrogen-bond donor and acceptor) within the linkers between the two aromatic moieties (A and B) (Figure 3) [112,113,114]. Further structural modification with a suitable linker may be pursued to develop new quinoline-based type II Met inhibitors.

Other type II Met inhibitors in clinical development include merestinib, glesatinib, sitravatinib, and altiratinib. Merestinib (LY2801653) is a multi-kinase inhibitor that is effective against MET, MST1R, FLT3, AXL, MERTK, TEK, ROS1, NTRK1/2/3, and DDR1/2 [115,116]. It is currently under investigation in an ongoing phase 2 study (NCT02920996) for advanced NSCLC patients bearing MET exon 14 mutations or other advanced cancer patients harboring an NTRK1/2/3 rearrangement. In a few case reports, merestinib was shown to exhibit a promising antitumor activity in NSCLC patients with the MET-exon-14-skipping mutation who did not respond to crizotinib or capmatinib [117]. Glesatinib (MGCD265) is another multi-kinase inhibitor targeting various oncogenes such as MET, AXL, VEGFR1/2/3, RON, and TIE-2 [118,119]. In a recently published phase 1 clinical trial (NCT00697632), glesatinib was well-tolerated [120]. Antitumor activity was observed following glesatinib therapy alone, with an ORR of 25.9% and 30.0% in cancer patients with MET/AXL mutations and MET-activating mutations, respectively. The findings from this study have led to the initiation of another phase 2 study (NCT02544633) to investigate the clinical efficacy of glesatinib in NSCLC patients stratified by different types of MET alterations. On the other hand, glesatinib was also reported to exhibit efficacy in patients with MET-exon-14-skipping mutations and acquired resistance to crizotinib [117,118].

Sitravatinib (MGCD516) is a broad-spectrum TKI targeting TAM receptors (TYRO3, AXL, and MERTK), VEGFR2, c-Kit, and Met. These receptors regulate several immune-suppressive cell types in the tumor microenvironment (including M2-polarized macrophages, MDSCs, and T regulatory cells), which play a critical role in mediating resistance to the immune checkpoint inhibitors in cancer immunotherapy [121]. Therefore, the combination of sitravatinib and checkpoint inhibitors to augment antitumor efficacy has been actively pursued in recent clinical trials [122,123]. Importantly, biomarker analyses in some of these trials supported an immunostimulatory mechanism of action [122].

Altiratinib (DCC-2701), a novel multi-kinase inhibitor, was designed to inhibit not only the mechanism of tumor initiation and progression but also the drug resistance mechanisms within the tumor microenvironment through the balanced inhibition of MET, TIE2, and VEGFR2 kinases [124]. Its binding to the switch control pocket of all three kinases was optimized, thus inducing type II inactive conformations [124]. Altiratinib was found to inhibit both the wild-type and mutated forms of Met in vitro and in vivo. By inhibiting the three oncogenic kinases (MET, TIE2, and VEGFR2), altiratinib was shown to inhibit the three major vascularization and resistance signaling pathways (HGF, ANG, and VEGF), thereby blocking tumor invasion and metastasis in preclinical tumor models [124,125]. However, the phase 1 trial (NCT02228811) evaluating the safety of altiratinib was terminated by the sponsor company in 2016.

Foretinib (GSK1363089) is another oral multi-kinase inhibitor known to target Met, RON, AXL, and VEGFRs [126] that has also been extensively investigated in clinical trials. While foretinib was originally designed as a Met TKI, it was later identified as a potent ROS1 inhibitor in an unbiased high-throughput kinase inhibitor screening assay [127]. Oncogenic ROS1 signaling is activated by interchromosal translation or intrachromosomal deletion that results in N-terminal ROS1 fusion genes that has been reported in a subset of patients with glioblastoma, NSCLC, and cholangiocarcinoma [128]. In clinical practice, the two types Ib Met TKIs (capmatinib and tepotinib) are indicated for the first-line treatment of NSCLC bearing the MET-exon-14-skipping mutation (METex14). However, the emergence of acquired resistance to capmatinib and tepotinib is almost inevitable, which is contributed to mainly by the D1228X(N/H/Y/E) or Y1230X(C/H/N/S) secondary MET mutations [129]. A recent preclinical study was conducted to investigate whether six type II Met TKIs were effective against both D1228X and Y1230X Met after the failure of capmetainib/tepotinib in NSCLC patients with METex14 [130]. Importantly, only foretinib was found to exhibit a potent inhibitory activity against both D1228X and Y1230X secondary MET mutations in vitro and in vivo [130]. Therefore, foretinib may be suitable for the second-line treatment of NSCLC harboring METex14 after campatinib/tepotinib failure due to secondary D1228 or Y1230 mutations. About 10 phase 1/2 trials have been conducted for foretinib. However, single-agent foretinib did not demonstrate notable clinical efficacy in unselected patients in most studies [131]. In 2015, GlaxoSmithKline decided to terminate the product development of foretinib. Despite the discontinuation of the development of foretinib, the TKI is still commonly used as a control Met-targeting candidate during the investigation of a newer generation of Met TKIs because of its unique structure and binding behavior.

2.3.4. Type III Met TKIs

Tivantinib (formerly known as ARQ197) represents a type III Met TKI that has reached the most advanced stage of clinical investigation. It interacts with the inactive non-phosphorylated configuration of Met and inhibits the autophosphorylation of the kinase [78]. It is highly selective for Met (10- to 100-times more selective for Met than 229 other kinases tested) with an inhibitory constant (Ki) of 355 nM [132]. In preclinical studies, tivantinib exhibited broad-spectrum anticancer effects in various cancer types, including lung cancer, melanoma, breast cancer, colon cancer, ovarian cancer, gastric cancer, and hepatocellular carcinoma [133]. It is noteworthy that tivantinib can also bind directly to microtubules and disrupt microtubule function to induce mitotic catastrophe and apoptosis [134,135]. As tivantinib could inhibit cancer cell growth regardless of the cellular activation status of Met, microtubule inhibition has been proposed to be the key mechanism mediating the tivantinib-associated anticancer effect in hepatocellular carcinoma [136]. The clinical efficacy of tivantinib has been evaluated in a few phase 2 and 3 trials for patients with advanced Met-positive hepatocellular carcinoma. No substantial clinical benefit was observed. Only in a phase 2 study (NCT00988741) was tivantinib shown to produce a significantly longer PFS than a placebo (1.6 months versus 1.4 months; p = 0.04) [137]. However, there was no significant difference in the median PFS (tivantinib 1.5 months, placebo 1.4 months; p = 0.06) and OS (tivantinib 6.6 months, placebo 6.2 months; p = 0.63) between the tivantinib and placebo groups [137].

2.3.5. Novel Met TKIs with Distinct Binding Mode

Collie et al. recently screened a DNA-encoded chemical library against the isolated kinase domains of the wild-type and D1228V acquired resistance mutant forms of Met [138]. One drug candidate (compound 1) that was capable of inhibiting both forms of the Met kinase was selected for a more detailed investigation. Using X-ray crystallography, compound 1 was shown to bind to the D1228V Met kinase in an unprecedented manner [138]. In the compound 1-D1228V Met cocrystal structure (crystal structure was deposited in the Protein Data Bank with accession code 8ANS), the indazole group of compound 1 binds to the hinge region of the kinase (i.e., the ATP-binding site), and the drug molecule extends into the back pocket towards the αC helix [138]. While this binding drug conformation is similar to a type II Met inhibitor, the conformation of the drug-bound Met kinase is highly unusual. Specifically, the conserved DFG motif adopts an “out” conformation, but the vacated DFG pocket is not occupied by the compound. Instead, it is occupied by the rearranged αC helix of the kinase molecule. With the DGF motif in the “out” conformation, the A-loop is largely disordered. Importantly, V1228 (the mutated residue known to render Met kinase resistant to type I Met inhibitors) was shown to be around 15 Å from compound 1 in the cocrystal, obviously not playing any role in the compound interaction. Importantly, the compound was also shown to be highly specific to Met kinase in a recombinant kinase profiling assay. Moreover, the new compound was also shown to inhibit the dimerization of the Met receptor in a cell-based assay using a time-resolved FRET assay [138]. Collectively, these findings suggest a novel mode of Met inhibition by the new compound. The new compound may be used to simultaneously target the wild-type and drug-resistant D1228V mutant form of Met without inducing toxicity due to non-specific inhibition of the whole kinome.

2.3.6. Met Inhibitors Derived from Natural Sources

While most of the Met inhibitors reported originated from a de novo drug design and structural optimization, a few natural compounds were also found to exhibit a Met-inhibitory effect. Dictamnine is a naturally occurring small-molecule furoquinoine alkaloid isolated from the root bark of Dictamnus dasycarpus Turcz. Its Met-inhibitory activity has been recently reported [139]. The binding mode of dictamnine with the crystal structure of Met protein (4IWD) was predicted by Autodock [140]. It was later verified by a cellular thermal-shift assay (CETSA) and a drug-affinity-responsive target stability (DARTS) assay [139]. Consistent with its attenuation of the PI3K/AKT/mTOR pathway, dictamnine was shown to exhibit a synergistic anticancer effect with gefitinib and osimertinib in EGFR-TKI-resistant lung cancer cell lines [139]. Withaferin A (a steroidal lactone derived from Withania somnifera) and carnosol (a naturally occurring phenolic diterpene found in rosemary) were reported to target pancreatic cancer stem cells as novel Met inhibitors [141].

2.3.7. Met-Targeting PROTACs

Proteolysis-Targeting chimeras (PROTACs) are chimeric bifunctional molecules that target a protein of interest for ubiquitination and degradation [142]. A typical PROTAC molecule consists of a ligand capable of binding to an E3 ligase that is connected via a linker to another ligand capable of binding to the target protein (Figure 4). Therefore, PROTACs promote the ubiquitination and subsequent proteasome-mediated degradation of the protein target. A salient feature of PROTACs is that they not only bind to but also eliminate those protein targets even without a distinct functionality (i.e., “undruggable” proteins and non-enzymatic proteins). While Met TKIs have been developed for nearly 20 years, their clinical efficacies and progress are modest. It has been suggested that a kinase-independent function of Met may drive cancer growth and metastasis. Therefore, the degradation of the Met protein may be more advantageous over the inhibition of the Met kinase activity for cancer therapy.

Crew’s research group developed a Met-targeting PROTAC based on a type II inhibitor, foretinib [143,144,145]. VHL and CRBN are two popular E3 ligases recruited by PROTACs to induce protein ubiquitination and degradation. Met-targeting PROTACs utilizing either VHL or CRBN E3 ligase were shown to induce the rapid degradation of Met protein in a concentration- and time-dependent manner. It is noteworthy that exon-14-deleted Met lacks the juxta membrane domain recruitment site (Y1003) for its endogenous E3 ligase. To this end, foretinib-based VHL PROTAC was shown to induce the degradation of exon-14-deleted Met.

More recently, Sachkova et al. also reported the design, synthesis, and in vitro evaluation of some other cabozantinib-based PROTACs to target Met for cancer treatment [146]. As described above, cabozantinib is a clinically approved type II Met inhibitor for the treatment of medullary thyroid cancer, advanced renal cell carcinoma, and hepatocellular carcinoma. Among the cabozantinib-based PROTAC molecules tested, two molecules bearing a VHL-ligand as the E3-ligase-binding moiety and a 10- to 12-atom glycol-based linker were found to be the most effective in degrading Met protein and eliciting anticancer activity in Met-overexpressing breast cancer cells in the nanomolar range [146].

3. Repurposing of Non-Oncology Drug as Met Inhibitors

Drug repurposing refers to the application of clinically approved drugs with known safety profiles and defined pharmacokinetic properties for new indications. It has emerged as an attractive approach for the search of effective and durable cancer treatment. Compared with the de novo development of novel drug candidates, drug repurposing represents a time-saving and cost-efficient method that can dramatically reduce the risk of drug development. Numerous systematic methods utilizing multi-omics analyses, molecular docking, artificial intelligence, and machine learning techniques have been used to facilitate the identification of repurposed drugs for cancer therapy [147]. A few recent attempts to identify repurposed drugs involving Met inhibition are described below.

3.1. High-Content-Analysis (HCA)-Based Screening for Met Inhibitors

HCA is widely used in biological research to identify small molecules, peptides, or RNAi that could alter cell phenotypes with the simultaneous readout of several parameters [148]. Oh et al. recently reported an HCA-based novel therapeutics screening method for Met-addicted glioblastoma [149]. Tumor cells isolated from 12 patients with glioblastoma were cultured ex vivo and subjected to high-content screening. Multiple cellular parameters, including Met protein immunofluorescence, cell viability, apoptosis, cell motility, and migration, were assessed. Intriguingly, the tumor cells derived from one glioblastoma patient (PDC6) exhibited a distinctively high Met level, and they were highly susceptible to the anticancer effect of Met inhibitors [149]. Subsequent genetic, immunoblot, and drug sensitivity characterizations of PDC6 cells confirmed that the specific glioblastoma patient had Met overexpression, thus supporting the reliability of the screening platform. This method was subsequently expanded for use as a drug-repurposing screen. The patient-derived tumor cells were treated with 60 clinically approved drug candidates. The concentration–response results of the mean phosphorylated Met intensity and relative AUC value were analyzed. As the positive controls, all the Met-targeting drugs (including crizotinib, cabozantinib, foretinib, capamatinib, and SAIT301) were shown to significantly decrease the phosphorylated Met (p-Met) protein level (fluorescent image intensity). Intriguingly, the specific CDK4/6 inhibitor (abemaciclib), but not palbociclib and ribociclib, was found to drastically suppress the p-Met intensity. Consistent with this novel finding, subsequent large-scale drug sensitivity screening in 59 cancer cell lines (classified as sensitive or resistant to crizotinib and cabozantinib) and 125 glioblastoma-patient-derived cancer cells confirmed that abemaciclib response correlates well with the response to known Met inhibitors [149].

3.2. In Silico Structure-Based Repurposing Screening for Met Inhibitors

Cutinho et al. recently reported a structure-based and in silico pharmacophore-modeling approach to identify possible Met-targeting drug candidates for repurposing [150]. A structure-based pharmacophore model was built using the optimized crystallographic structure of Met protein (PDB ID: 3LQ8). The pharmacophore model was then screened with publicly available databases of natural compounds and the FDA approved drug database. Drug molecules sharing similar pharmacophoric features as the control Met TKIs (cabozantinib and foretinib) were then subjected to molecular docking for a detailed examination of binding interactions and conformations at the enzyme binding site. Two clinically approved drugs (bicalutamide—anti-androgenic drug; diphenidol hydrochloride—anti-emetic drug) were identified as putative Met kinase inhibitors. The capacity of the drug candidates to form hydrogen bonds with MET1160 appears to be essential for them to potentially function as Met kinase inhibitors [150]. The highest-ranking hits were also inspected by using various in silico techniques, including SwissADME and pkCSM, to check their bioavailability and pharmacokinetic parameters, respectively. Further experimental validation is pending.

3.3. Kinobeads Technology for Kinase Drug Repurposing

Klaeger et al. recently analyzed the cellular molecular targets of 243 clinically approved kinase drug candidates using chemical proteomics [151]. The kinobeads technology was employed, where the tested kinase drug candidates were incubated with cellular extracts containing endogenous full-length proteins harboring various posttranslational modifications and in the presence of regulatory proteins and metabolites [152]. The assay was conducted in a competition-binding format. The promiscuous kinase inhibitors were immobilized on a bead to capture the cellular kinases in the lysates. The extent of the binding was quantified by mass spectrometry after pulldown assays. The kinobead assay has an additional advantage in that some non-kinase proteins can also be identified as novel targets of clinical kinase inhibitors. Using this kinobead assay, Klaeger et al. showed that the Met TKI cabozantinib could also potently inhibit the tyrosine kinase fusion product FLT3-ITD [151]. Importantly, cell lines bearing the FLT3-ITD rearrangement but not the wild-type AML cell lines were found to be sensitive to cabozantinib treatment. Cabozantinib was also shown to remarkably inhibit the phosphorylation of the FLT3 downstream target STAT5 and exhibited an antitumor efficacy in a tumor xenograft model in vivo [151]. It is noteworthy that clinical trials are underway to evaluate the safety and efficacy of cabozantinib in cancer patients harboring the FLT3/ITD rearrangement (NCT04116541, NCT03425201, and NCT01961765).

4. Combination of Met Inhibitors with Other Cancer Treatment Modalities to Overcome Drug Resistance

4.1. Use of Met TKIs to Overcome Drug Resistance to EGFR-Targeted TKIs

MET dysregulation is a well-known mechanism that contributes to drug resistance to targeted cancer therapies [153]. Met hyperactivation was reported to account for about 22% of acquired resistance cases following first-generation EGFR-targeted therapy [154,155]. While EGFR-TKIs often induce on-target drug resistance mechanisms (i.e., the EGFR T790M mutation), MET gene amplification represents the most frequent cause of bypass pathway activation as an acquired resistance mechanism to EGFR-TKIs. MET amplification was reported in up to 50% of cases after the failure of second-line osimertinib therapy [156] or in 7–15% of cases after an insufficient clinical response from first line osimertinib treatment [157]. MET amplification also represents an important mechanism of intrinsic drug resistance to osimertinib [158]. MET amplification/hyperactivation leads to the persistent activation of signaling pathways downstream of EGFR (including MAPK, STAT, and PI3K-AKT signaling). Importantly, MET amplification was detected in resistant cancer cells with or without other concomitant resistance mechanisms, including the loss of the T790M mutation in patients refractory to previous EGFR-TKI therapy. In addition, MET exon 14 mutations are also known to be the key resistance mechanism to EGFR-TKI therapy [159].

As osimertinib is currently the standard treatment for NSCLC in the first-line setting, extensive research has been conducted to find out an effective strategy for resistance circumvention. A few recent preclinical studies have reported the use of Met inhibitors (crizotinib being the most-studied) to overcome osimertinib resistance in EGFR mutant NSCLC cell lines harboring MET gene amplification [160,161]. The combination of crizotinib and osimertinib has also been investigated in clinical trials in NSCLC patients with acquired resistance to osimertinib and MET amplification [162,163]. The interim results from a multicenter, open-label, phase 1b TATTON trial also revealed an encouraging antitumor efficacy from the combination of osimertinib and savolitinib (a type Ib Met TKI) in advanced NSCLC patients with MET-amplified tumors who had disease progression on a previous EGFR TKI [164]. Following the favorable clinical outcome from the TATTON trial, a phase 2 SAVANNAH trial (NCT03778229) was initiated to evaluate the efficacy of osimertinib–savolitinib in similar patient cohorts with prior osimertinib exposure. One of the cohorts of the multi-arm clinical trial ORCHARD (NCT03944772) was also set out to evaluate the osimertinib–savolitinib combination. More recently, a combination of tepotinib (another type Ib Met TKI) and osimertinib is currently under evaluation in the open-label, multicenter phase 2 INSIGHT 2 trial (NCT03940703) in patients with metastatic EGFR-mutant NSCLC and acquired resistance to first-line osimertinib and MET amplification [165].

In another phase 1 (CHRYSALIS) trial, the combination of lazertinib (a third-generation EGFR TKI) and amivantamab (a bi-specific antibody targeting both Met and EGFR) was investigated in NSCLC patients who were chemotherapy-naïve but refractory to osimertinib treatment [166,167]. While patients with a Met- or EGFR/Met-based mechanism of resistance exhibited an ORR of 50%, the response rate was increased to 90% in a patient subgroup (10 patients) with high EGFR/Met expression as observed by immunohistochemistry. Another ongoing CHRYSALIS-2 trial (NCT04077463) has included a patient cohort to validate this exciting finding.

4.2. Use of Met Inhibitors to Potentiate Antitumor Response to Cancer Immunotherapy

Programmed cell death receptor-1 (PD-1) is an inhibitory checkpoint protein expressed on immune cells, including activated T cells, B cells, and natural killer cells. Upon the binding of PD-1 to its major ligand, PD-L1, expressed in tumor cells, the T-cell-mediated cancer-killing effect by the host’s immunity is suppressed. Blockade of the PD-1/PD-L1 interaction represents an effective immunotherapeutic strategy for cancer treatment. However, the response rate to the PD-1/PD-L1 inhibitor is limited. Many patients do not respond, whereas others cannot achieve a durable clinical response. Extensive research has been conducted to search for novel strategies that can enhance anticancer immunity.

To this end, Met signaling is known to induce PD-L1 expression in cancer cells [168]. While this implies a reduced responsiveness of Met-driven cancers to PD-1/PD-L1 immunotherapy, there are limited reports in the literature demonstrating the efficacy of single-agent immune checkpoint inhibitors in cancer patients with MET-amplified/hyperactive tumors. A recent retrospective study was conducted to investigate the clinical efficacy of a single-agent PD-1/PD-L1-directed immune checkpoint inhibitor in advanced NSCLC patients bearing at least one oncogenic driver alteration [169]. The ORR in the Met subgroup (16%) was considerably lower than in the other patient subgroups (KRAS: 26%; BRAF: 24%; ROS1: 17%) [169]. Therefore, patients carrying actionable tumor alterations are recommended to receive targeted therapy and chemotherapy before considering the monotherapy of anti-PD-1/PD-L1 immunotherapy [169].

It is noteworthy that NSCLC patients bearing EGFR or ALK oncogenic mutations tend to have minimal to no smoking history [170]. In contrast, NSCLC patients carrying Met-driven tumors consist of a relatively greater proportion of smokers [171]. Therefore, Met-driven NSCLC tumors are also likely to carry a higher tumor mutation burden [172]. Consistent with this observation, only a modest clinical benefit is usually observed in Met-dependent NSCLC patients [173]. Importantly, numerous preclinical studies have suggested that the concomitant inhibition of Met could potentiate the efficacy of PD-1/PD-L1 immune checkpoint blockade [174,175]. It has been shown that IFNγ could induce PD-L1 expression more readily in MET-amplified tumors [175]. To this end, the Met TKI JNJ-605 and a specific Met-blocking antibody were shown to significantly impair the induction of both PD-L1 and PD-L2 by IFNγ [175]. Intriguingly, IFN-stimulated PD-L1/PD-L2 expression was not reversed by Met inhibitors in cancer cells that were not amplified [175].

More recently, bispecific anti-Met/PD-1 antibodies (also called diabodies) have been developed to potentiate the antitumor efficacy of anti-PD blockade cancer immunotherapy [176]. On the one hand, these diabodies were demonstrated to suppress HGF-induced cancer proliferation, migration, and invasion by preventing the binding of HGF to Met [176]. On the other hand, they were also able to stimulate T cell activation by blocking the PD-1 pathways. Interestingly, the diabodies with a higher Met-binding affinity (diabody-mp) were shown to exhibit a greater antitumor effect than the corresponding diabodies with a lower Met-binding affinity (diabody-pm) [176].

In fact, the combination of PD blockade therapy and Met inhibitors has been actively pursued in numerous multicenter clinical trials (summarized in

Table 2. Recent representative clinical trials investigating the combination of PD-1/PD-L1 blockade therapy and Met inhibitors (clinicaltrials.gov, accessed on 1 May 2023).

Combination	Cancer Type	clinicatrials.gov Identifier (Phase)	Status
Cabozantinib (Met TKI) + nivolumab (anti-PD-1 mAb) versus Sunitinib (multitargeted TKI)	Previously untreated advanced RCC	NCT03141177 (Phase 3)	Completed; nivolumab + cabozantinib had significant benefits over sunitinib with respect to PFS and OS.
cabozantinib (Met TKI) + Nivolumab (anti-PD-1 mAb) with or without ipilimumab (anti-CTLA mAb)	Metastatic genitourinary tumors	NCT02496208 (Phase 1)	Active, not recruiting (last update posted 26 April 2023)
APL-101 (Met TKI) + genolimzumab/nivolumab (anti-PD-1 mAb)	Locally advanced or metastatic HCC or RCC	NCT03655613 (Phase 1/2)	Terminated (due to administrative reasons; status update on 6 May 2022)
Capmatinib (Met TKI) + pembrolizumab (anti-PD-1 mAb)	NSCLC with PD-L1 expression > 50% and no EGFR mutation or ALK rearrangement	NCT04139317 (Phase 2)	Terminated (due to toxicity in the drug combination arm; status update on 27 February 2023)
Cabozantinib (Met/VEGFR TKI) + nivolumab (anti-PD-1 mAb) + ipilimumab (anti-CTLA mAb)	Metastatic soft-tissue sarcoma	NCT04551430 (Phase 2)	Active, not recruiting (last update posted 31 March 2023)
Cabozantinib (Met/VEGFR TKI) + nivolumab (anti-PD-1 mAb)	Metastatic microsatellite-stable colorectal cancer	NCT04963283 (Phase 2)	Recruiting (last update posted 15 December 2022)
Cabozantinib (Met/VEGFR TKI) + nivolumab (anti-PD-1 mAb)	Advanced HCC who progressed upon first-line therapy	NCT05039736 (Phase 2)	Not yet recruiting (estimated start date 24 February 2023)
Cabozantinib (Met/VEGFR TKI) + ipilimumab (anti-CTLA-4 mAb) + nivolumab (anti-PD-1 mAb)	Refractory cutaneous melanoma	NCT05200143 (Phase 2)	Recruiting (last update posted 13 June 2022)

Abbreviations: HCC, hepatocellular carcinoma; NSCLC, non-small-cell lung cancer; OS, overall survival; PFS, progression-free survival; RCC, renal cell carcinoma; TKI, tyrosine kinase inhibitor.

). In 2021, the combination of nivolumab (an anti-PD-1 monoclonal antibody) and cabozantinib (type II Met TKI) was approved by the FDA as first-line treatment for patients with advanced renal cell carcinoma after a fast-track real-time oncology review. The antitumor efficacy of the combination was primarily substantiated by the encouraging results from the phase 3 international multicenter CheckMate 9ER trial (NCT03141177) [177,178].

4.3. Use of Met TKIs to Overcome Drug Resistance to Chemotherapy

Although immunotherapy and targeted drugs are revolutionizing the treatment paradigm for numerous cancer types, chemotherapy still retains a vital role in most cancer patients. Classical chemotherapy suppresses cancer proliferation and reduces the tumor burden by exerting cytotoxic effects. However, treatment relapse and chemotherapy resistance are almost inevitable. Aberrant MET regulation is frequently observed in various cancer types and is associated with poor prognosis. Constitutive MET signaling is known to mediate chemoresistance by protecting cancer cells from apoptosis, promoting tumor invasion, and facilitating the epithelial-to-mesenchymal transition. There are two excellent review articles published by Wood et al. and To et al. about the role of MET in chemoresistance and the application of Met inhibitors to overcome drug resistance [179,180].

In preclinical studies, the HGF expression level was found to be higher in chemoresistant cancer cells than their chemosensitive counterparts [181,182,183]. HGF overexpression was shown to activate MET signaling in an autocrine manner, which increased basal Met phosphorylation to promote cancer cell survival. In addition, HGF secreted from cancer-associated fibroblasts has been reported to protect breast, glioblastoma, lung, and ovarian cancer against classical chemotherapeutic-drug (camptothecin, cisplatin, and doxorubicin)-induced apoptosis [184]. Moreover, dysregulated HGF/Met signaling could also activate various downstream effectors including the cellular Src kinase (c-Src), phosphotidylinsitol-3-OH kinase (PI3K), serine/threonine protein kinase (Akt), mitogen-activated protein kinase (MAPK), and the Wnt/β-catenin pathway to promote cell proliferation, invasiveness, morphogenesis, and angiogenesis. The application of Met inhibitors to overcome chemoresistance has also been investigated in clinical studies.

Table 3. Representative clinical trials investigating the combination of Met inhibitors and classical chemotherapeutic drugs.

Combination	Cancer Type	ClinicaTrials.gov Identifier (Phase)	Key Findings/Current Trial Status
Tivantinib (Met inhibitor) + cetuximab (EGFR mAb) + irinotecan (topoisomerase I inhibitor)	CRC	NCT01075048 (Phase 2)	No significant difference in PFS between the two treatment groups with or without tivantinib. Patient subgroup analysis revealed a better outcome in patients with MET-high tumors, but subgroup patient size was too small to draw conclusions.
Onartuzumab (anti-Met mAb) + FOLFOX + bevacizumab (anti-VEGF mAb)	CRC	NCT01418222 (Phase 2)	No significant difference in PFS between the two treatment groups with or without onartuzumab. Met immunohistochemistry was not a predictive biomarker for treatment outcome.
Rilotumumab (anti-HGF mAb) + epirubicin + Cisplatin + capecitabine	Gastric cancer	NCT00719550 (Phase 2)	Modest improvement in PFS in rilotumumab combination over placebo (PFS: rilotumumab combination, 5.7 months vs cytotoxic drugs without rilotumumab, 4.2 months (HR, 0.60; p = 0.016)).
Rilotumumab (anti-HGF mAb) + mitoxantrone + prednisone	Castration-resistant prostate cancer who had received previous taxane chemotherapy	NCT00770848 (Phase 2)	No significant PFS difference between rilotumumab combination over cytotoxic drugs combination. Unfavorable OS associated with patients with high tumor Met expression regardless of treatment.
Crizotinib (Met/ALK inhibitor) + cyclophosphamide + topotecan	Refractory solid tumors or nnaplastic large-cell lymphoma	NCT01606878 (Phase 1)	Patients experienced increased toxicity in combination treatment that was not explained by the relative bioavailability or exposure.
Cabozantinib (Met/VEGFR inhibitor) + topotecan + cyclophosphamide	Refractory Ewing sarcoma or osteosarcoma	NCT04661852 (Phase 1)	Trial completed on 24 October 2022.
Cabozantinib (Met/VEGFR inhibitor) + cisplatin / doxorubicin / methotrexate	Newly diagnosed osteosarcoma	NCT05691478 (Phase 2/3)	Suspended (scheduled interim monitoring as of 26 April 2023).

Abbreviations: CRC, colorectal cancer; FOLFOX, chemotherapeutic regimen consists of leucovorin, fluorouracil, and oxaliplatin; HCC, hepatocellular carcinoma; PFS, progression-free survival; OS, overall survival.

summarizes the representative clinical studies investigating the potentiation of chemotherapy by Met inhibition.

5. Inherent and Acquired Resistance to Met TKIs

5.1. Inherent Resistance to Met TKIs

While cancer patients are selected to receive personalized therapy with Met TKIs using a relevant MET-activation biomarker such as the MET-exon-14-skipping mutation, the presence of other concurrent oncogenic driver mutations in the Ras-Raf-MAPK or PI3K/Akt pathways are known to reduce drug sensitivity to Met TKIs [185,186]. In fact, it has been reported that treatment-naïve NSCLC patients bearing the MET-exon-14-skipping mutation did not respond to Met TKIs if they also have other oncogenic RAS or PI3K mutations [186,187]. The hyperactive RAS or PI3K may be the driver oncogenes in the tumors of those patients.

5.2. Acquired Resistance to Met TKIs

5.2.1. On-Target Resistance Mechanisms

Similar to other oncogene-targeted therapies, cancer cells could acquire resistance to Met TKIs following drug treatment via either on-target or off-target mechanisms. Employing the Ba/F3 cell model expressing MET exon 14 alterations, Fujino et al. evaluated various specific secondary mutations as resistance mechanisms to different types of Met TKIs [188]. As expected, mutations of D1228 and Y1230 in the Met kinase activation loop (A-loop) were found to confer resistance to type I Met TKIs by disrupting the drug binding. On the other hand, mutations of L1195 and F1200 outside the A-loop could mediate resistance to type II Met TKIs. In addition, a few specific mutations including D1228A/Y were found to confer resistance to both type I and II Met TKIs in the Ba/F3 cell model [188]. Interestingly, the two sets of mutations (D1228/Y1230 and L1195/F1200) appear to be complementary to each other. Type II Met TKIs were shown to overcome acquired resistance due to the secondary mutations (D1228/Y1230) caused by type I Met TKIs and vice versa [188]. Indeed, this secondary mutation complementarity was supported by clinical findings that glesatinib (type II Met TKI) could overcome resistance due to Y1230H/S induced by prior crizotinib treatment (type Ia Met TKI) [118]. Similarly, merestinib (type II Met TKI) maintained antitumor activity against D1228N, which occurred after prior capmatinib treatment (type Ib Met TKI) [117].

In a clinical setting, Recondo et al. recently compared paired pre- and post-Met TKI tumor specimens from twenty NSCLC patients with MET-exon-14-skipping mutations [117]. The on-target (secondary mutations) and off-target (to be discussed below) resistance mechanisms were present in 35% and 45% of the patients, respectively [117]. Some patients acquired both on-and off-target resistance mechanisms [117]. Consistent with the aforementioned Ba/F3 cell line study, D1228/Y1230 secondary mutations were primarily induced after treatment with type I Met TKIs (crizotinib, capmatinib, and glesatinib). It is also noteworthy that multiple secondary mutations could occur simultaneously in one patient, thus pinpointing the heterogeneity in the Met TKI resistance mechanisms [117,118]. It may be difficult to treat patients according to the resistance mechanism(s) identified from a single biopsied tumor lesion because different secondary mutations may be operating in other tumor lesions. In other clinical studies, high-grade MET and HGF gene amplifications have also been reported as on-target mechanisms to mediate Met TKI resistance, respectively [117,187]. Most recently, switching between different types of Met TKIs (type I and II) was shown to provide repeated clinical responses in cases with second-site mutations in NSCLC patients bearing the MET-exon-14-skipping mutation [189].

5.2.2. Off-Target Resistance Mechanism

As with inherent resistance to Met TKIs, concomitant mutations in other parallels (EGFR, HER3, KIT) or downstream signaling pathways (Ras/Raf/MAPK and PI3K/AKT) were reported to contribute to acquired resistance to Met TKIs [185,187,190,191]. However, it remains elusive as to whether there were tumor clones bearing these co-driver mutations prior to Met TKI therapy or the tumor acquired these co-driver mutations during the therapy.

In a recent study evaluating the role of aberrant PI3K regulation in Met TKI resistance, NSCLC patients bearing PI3KCA mutations and PTEN loss were shown to be refractory to Met TKIs [186]. To this end, PI3K inhibition was found to restore Met TKI sensitivity in MET-exon-14-mutated cell lines with PI3K alteration, thus suggesting the ability of the respective drug combination to overcome resistance. MET-exon-14-mutated cell lines with Ras/MAPK pathway dysregulation (Kras overexpression or NF1 downregulation) were also found to be resistant to Met TKIs, which could be reversed by a combination of crizotinib and the MEK inhibitor trametinib [192].

6. Further Perspectives

MET dysregulation is a well-established oncogenic mechanism driving cancer development, and it also mediates drug resistance to classical chemotherapy, targeted therapy, and immunotherapy in the treatment of cancer. Accumulating preclinical evidence suggests that MET is a druggable target for the circumvention of drug resistance [153,179]. The combination of EGFR and Met inhibitors in clinical settings has been proposed to overcome drug resistance because MET amplification is common in EGFR-mutated NSCLC patients not responding to EGFR-targeted therapies [156,193,194]. However, early attempts in clinics to investigate Met-targeted therapies in unselected cancer patients generally produced disappointing results. In recent years, by stratifying cancer patients with specific MET exon 14 alterations and MET amplification, more promising treatment outcomes following treatment with Met inhibitors were observed. Therefore, the combination of Met inhibitors with other cancer treatment modalities for the circumvention of drug resistance in a personalized manner has been advocated for [147]. It follows that robust and standardized methods will be needed to evaluate the specific MET alterations in tumor specimens both at diagnosis and at relapse.

It is commonly believed that the failure of the early clinical trials investigating Met-targeting drugs could have resulted from lack of or inappropriate patient selection. In the early trials, patient selection was based on Met expression in tumor specimens. Although Met protein overexpression represents the most common manifestation of MET dysregulation in NSCLC, it was not proven to be an effective biomarker to predict clinical response to Met inhibitors in most cases [195]. The association between Met protein overexpression and its activation is not very clear [195]. Therefore, the measurement of Met protein phosphorylation (p-Met) at tyrosine 1234/1235 may be more appropriate in reflecting the status of MET activation. Indeed, only a fraction of Met-protein-overexpressing tumors actually expressed the activated p-Met form [196]. To our knowledge, all clinical trials employing immunohistochemical methods for the assessment of Met protein overexpression and patient stratification did not produce a favorable clinical outcome for the Met-targeting therapies under investigation [197]. Moreover, the detection of the phosphorylated form of the protein (p-Met) is technically demanding [198]. Met phosphorylation could be lost during specimen processing and during the immunohistochemical (IHC) staining protocol [198]. Another critical drawback of using IHC for identifying tumors with MET dysregulation was further demonstrated by a recent study conducted by the Lung Cancer Mutation Consortium [187]. In this study, the correlation between MET amplification/MET exon 14 alterations and Met protein IHC intensity was examined in tumor specimens from 181 patients with Met-IHC-positive NSCLC. Surprisingly, almost all Met-IHC-positive cases were revealed to be negative for MET amplification or MET exon 14 alterations [187].

Therefore, instead of relying on the measurement of Met protein expression, the identification of MET amplification or MET-exon-14-skipping mutations is currently recommended as the preferred patient stratification strategy for the investigation of MET-targeted therapies in NSCLC [20,171,199,200]. To this end, the cutoff value for defining clinically relevant MET amplification in tumor specimens remains controversial. It has been proposed that only the amplification of the MET locus represents a bona fide oncogenic event, which is defined by a high ratio of MET to centromere 7 [201]. Meanwhile, the detection of MET-exon-14-skipping alterations in circulating tumor DNA has been employed as a novel patient selection strategy [198,201]. A companion diagnostic test, called FoundationOne^® CDX, has been recently approved to select NSCLC patients with MET-exon-14-skipping mutations for capmatinib therapy [202]. Nevertheless, it is noteworthy that the prevalence of MET-exon-14-skipping mutations is low, with rates of up to 4% being observed in NSCLC [203]. Given that Met inhibitors are mostly multitargeted TKIs, the selection of a small subpopulation of patients may miss some potential responders [204].

MET dysregulation is known to mediate resistance to classical chemotherapy, targeted therapy, and immunotherapy. The combination of Met inhibitors and other cancer treatment modalities is considered a rational strategy to overcome drug resistance. Given the limited efficacy of PD-1/PD-L1 immunocheckpoint inhibitors in MET-dependent NSCLC, the use of immunotherapy alone should be avoided. However, the results from a recent phase 1/2 clinical trial (CheckMate 370) investigating a combination of nivolumab (PD-1 inhibitor) and crizotinib (Met/ALK inhibitor) raised toxicity concerns [32]. More investigation into the combination of Met inhibitors and immunotherapy is warranted.

Similar to other targeted therapeutic drugs, acquired resistance to Met TKIs is almost inevitable. The development of effective strategies to overcome Met TKI resistance is needed to fully unleash the utility of this unique drug class. In both preclinical and clinical studies, MET D1228 and Y1230 have been reported as hotspots for the secondary resistant mutations for type I Met TKIs in NSCLC harboring MET-exon-14-skipping mutation. Switching from the type I to type II Met TKI may represent an effective strategy to overcome the D1228/Y1230-mutation-mediated resistance. However, a recent clinical case report revealed that cabozantinib (type II Met TKI) could only eliminate the Y1230C mutant but left behind the D1228N/Y/H mutant alleles [205]. In another clinical study about the sequential treatment of type I and type II Met TKIs for MET-dependent NSCLC patients harboring D1228X and Y1230X, only the D1228X mutant was detected in the tumor biopsies at the time of disease progression [206]. Therefore, MET D1228X appears to be more resistant than Y1230X to the currently available type II Met TKIs. The development of novel Met TKIs could focus on drug candidates that are effective against the D1228X secondary resistance mutant.