According to different mechanisms and structures, HGF/c-Met axis inhibitors are categorized into anti-HGF and anti-c-Met antibodies as well as small molecule c-Met kinase inhibitors. Based on chemical structures and different docking sites with kinases, there are three types of small-molecule c-Met kinase inhibitors, including selective, non-selective, and special structure c-Met inhibitors (Parikh and Ghate, 2018). So far, most of the HGF/c-Met inhibitors in digestive system cancer have been assessed in preclinical studies (Table 1) and phase I/II/III clinical trials (Table 2).

Primary liver cancer is one of the most common malignant tumors in clinics. The condition ranks sixth and second based on incidence and mortality, respectively (Siegel and Miller, 2019). About 75–85% of primary liver cancer cases are HCC (Siegel and Miller, 2019). Sorafenib [a small-molecule multi-kinase inhibitor that targets vascular endothelial growth factor receptors (VEGFRs) rapidly accelerated fibrosarcoma (RAF) and other protein kinases] has been regarded as the standard of treatment for patients with advanced HCC since 2007 (Marisi et al., 2018). However, survival among patients even after treatment with sorafenib is still low, necessitating urgent development of new and effective treatment methods. Studies have demonstrated the role of HGF/c-Met axis in cell survival, cancer proliferation, and metastasis in HCC cases (Bahrami et al., 2017; García-Vilas and Medina, 2018; Ogunwobi et al., 2019) making anti-HGF/c-Met a promising therapeutic target for HCC treatment.

Tivantinib, a staurosporine derivative, is a non-ATP competitive selective small-molecule inhibitor and acts mainly as an anti c-Met inhibitor (Munshi et al., 2010). Previous studies that have targeted the molecule have, in turn, reported encouraging results in several phase I/II clinical trials. For instance, a randomized, controlled, phase II clinical trial evaluated the efficacy of tivantinib in patients with advanced HCC and Child–Pugh A cirrhosis with the results indicating that patients treated with the drug had a longer median disease progression time relative to the placebo (Santoro et al., 2013). Among patients with high c-Met expression, a more significant (2.7 months vs. 1.4 months; P = 0.03) curative effect was recorded (Santoro et al., 2013). However, some serious complications recorded in the tivantinib group included anemia and neutropenia, especially when a high dosage was administered. Unfortunately, in a phase III trial of METIV-HCC, tivantinib did not improve overall patient survival compared to the placebo (Rimassa et al., 2018). Nevertheless, there is evidence that a high expression of c-Met in HCC is related to a poor prognosis (Rimassa et al., 2016). Since the METIV-HCC trial selected patients with high c-Met levels, this does not necessarily mean that tivantinib has no role in targeted treatment in HCC patients. Therefore, determination of an accurate dosage coupled with guided patient stratification is key to the use of tivantinib as a target drug for HCC. Some scholars have pointed out that tivantinib cannot inhibit the c-Met tyrosine autophosphorylation, and its biological activity is mainly antimitotic. It kills any kind of cells regardless of the expression of c-Met protein, and the cytotoxicity of tivantinib is related to inhibit microtubule assembly (Basilico et al., 2013). This novel viewpoint may explain the results of phase III clinical trials.

Cabozantinib is a novel small-molecule multi-target tyrosine kinase inhibitor of c-Met, VEGFRs, RET, KIT, and AXL (Yakes et al., 2011). In mice xenotransplantation models, cabozantinib treatment has been shown to inhibit the growth and metastasis of HCC (Xiang et al., 2014). In a phase II placebo-controlled randomized discontinuation study, cabozantinib suppressed tumor growth, disease stability and reduced alpha fetoprotein (AFP) in HCC (Kelley et al., 2017). Recently, a phase III study assessment of the effect of cabozantinib on advanced HCC patients who had been previously treated showed a longer total survival time taken in the cabozantinib group compared to the controls (10.2 vs. 8 months; p = 0.005) (Abou-Alfa et al., 2018). The progression-free survival (PFS) period (5.2 vs. 1.9 months; p < 0.001) and objective response rate (ORR) (4 vs. < 1%; p = 0.009) also improved significantly. The main adverse events were erythrodysesthesia, hypertension, increased aspartate aminotransferase level, fatigue, and diarrhea.

Foretinib is a multi-target inhibitor for c-Met, AXL, RON, VEGFR2, and TIE-2 (Giordano and Columbano, 2014). A phase I/II clinical study evaluating the safety, pharmacodynamics, pharmacokinetics, and activity of foretinib in a late first-line trial reported 30 mg as the maximum dosage for foretinib in phase I. A total of 39 patients received foretinib treatment in phase II (38 patients were evaluable for efficacy), with a median overall survival time (OS) of 15.7 months and a median progression time of 4.2 months reported (Yau et al., 2017). The most common adverse reactions including hypertension (43.6%), loss of appetite (28.2%), ascites (25.6%), and fever (25.6%) were also recorded, although these were not considered serious. Overall, foretinib showed satisfactory antitumor activity in Asia in patients with advanced HCC.

Tepotinib is an effective and highly selective c-Met inhibitor, and its c-Met selectivity exceeds that of most kinase inhibitors with a cell-based 50% inhibitory concentration (IC50) of 1.7 nM (Falchook et al., 2020). In contrast, highly selective c-Met inhibitors have little effect on other targets and are expected to produce less toxicity. Therefore, a sufficient dose can guarantee effective c-Met inhibition. Tepotinib was first approved for the treatment of patients with advanced NSCLC in 2020 in Japan (Markham, 2020). A phase II clinical study on c-Met-positive advanced HCC showed that tepotinib had a longer progression time and PFS time than sorafenib (NCT01988493). The time to progression was 2.9 months in the tepotinib group and 1.4 months in the sorafenib group. The median PFS of the tepotinib group and sorafenib group was 2.8 and 1.4 months, respectively, and the median overall survival was 9.3 and 8.6 months, respectively. Another phase Ib/II study evaluated the efficacy and safety of tepotinib in patients with c-Met-positive HCC who have failed sorafenib treatment, and the PFS rate of 12 weeks was 63.3% (NCT02115373).

Capmatinib (INC280) is a highly selective c-Met inhibitor who has an IC50 of 0.13 nM for c-Met (Bouattour et al., 2018). A phase II study indicated that capmatinib is beneficial to tumor suppression in patients with c-Met-high HCC (Qin et al., 2019). The overall response rate was 30.0% and the disease control rate (DCR) was 50.0% in the c-Met high expression group. It even includes a long-term complete response of more than 600 days and two partial responses. Another phase Ib/II clinical trial is studying the safety and antitumor activity of capmatinib combined with PDR001 [a programmed death 1 (PD-1) inhibitor] in patients with HCC (NCT02795429). The clinical trial is anticipated to be completed at the end of 2020.

Several other similar c-Met inhibitors, such as golvatinib (E7050) and SU11274 have also been reported to have antitumor activity against HCC (Nakagawa et al., 2010; Inagaki et al., 2011). However, studies in relation to their effects in early clinical stages as well as their efficacy and safety are yet to be carried out. In view of the limited therapeutic effect of targeted drugs, a combination of therapies may be a more effective strategy for HCC treatment.

Pancreatic cancer (PC) is a highly malignant disease with the worst prognosis due to its strong invasion and metastasis ability. The disease’s 5-year survival rate is approximately 9%, the lowest of any cancer (Siegel and Miller, 2019). Pancreatic cancer patients with high c-Met and HGF expression levels tend to show poor prognosis and low survival rates (Yan et al., 2014). Studies have shown that c-Met expression levels in pancreatic tumors are related to tumor grades (Gardian et al., 2012). Similarly, it has been demonstrated that c-Met expression increased in pancreatic ductal adenocarcinoma, suggesting that c-Met might be a molecular marker for predicting prognosis in patients with pancreatic cancer (Zhu et al., 2011).

Crizotinib is an ATP competitive multi-target protein kinase inhibitor that targets c-Met, ALK, and ROS1 (Takiguchi et al., 2017). In an orthotopic mouse model experiment, crizotinib was found to increase the concentration of gemcitabine in blood and tissues, reduce sizes of tumors, inhibit peritoneal dissemination of highly Met-expressing pancreatic cancer, and prolong survival (Avan et al., 2013). Similarly, an in vitro experiment showed that crizotinib can be used to treat peritoneal spread of pancreatic cancer based on inhibition of cell proliferation and metastasis mediated by phosphorylation of the c-Met pathway (Takiguchi et al., 2017). However, it is not clear whether the above effects were mediated by the c-Met signaling pathway.

Reports have also indicated that long-term treatments involving cabozantinib induce less resistance and can improve the efficacy of gemcitabine and can overcome gemcitabine resistance in pancreatic cancer (Hage et al., 2013). The effects are achieved by suppressing total c-MET, controlling downstream phosphorylation of c-MET, and decreasing expression of transcription factor SRY-related HMG-box 2 (SOX2) (Hage et al., 2013). An experiment conducted to determine maximum tolerable concentrations of cabozantinib and gemcitabine in 10 advanced pancreatic cancer patients showed that more than 25% of the patients in all concentration groups had dose-limiting toxicity (≥grade 3 ALT or AST elevations or ≥grade 3 thrombocytopenia) (Zhen et al., 2016). Despite the small sample size used in the study, a certain level of significance for future research of cabozantinib can be still be derived from the findings.

Gastric cancer is the fifth most common form of cancer worldwide and the third leading cause of cancer-related deaths (Padmanabhan et al., 2017). Gastric cancer patients who have a high expression of c-Met show poor prognosis compared to those with c-Met negative tumors (Fuse et al., 2016). Numerous preclinical and clinical studies have generated data on monoclonal antibodies against HGF/c-Met axis, which provide possible targets for development of gastric cancer treatments.

Rilotumumab (AMG102), a monoclonal antibody of the human IgG2 targeting HGF, can block binding of HGF to c-Met. A randomized phase II clinical trial showed that this antibody in combination with epirubicin, cisplatin, and capecitabine (ECX) could prolong PFS in advanced gastric cancer patients relative to the control group, especially in patients with high c-Met expression (Iveson et al., 2014). Based on these results, a further study on the effects of rilotumumab combined with ECX in treatment of gastric cancer was conducted using two phase III trials (RILOMET-1 and 2) (Doi et al., 2015; Catenacci et al., 2017). Unfortunately, both trials were terminated because RILOMET-1 results showed an increase in the number of deaths due to complications in patients treated with rilotumumab compared to the placebo (Catenacci et al., 2017).

Onartuzumab is a recombinant humanized anti-c-Met monoclonal antibody, which is produced in Escherichia coli (Merchant et al., 2013). Evaluation of the safety and efficacy of onartuzumab combined with mFOLFOX6 in patients with HER2-negative gastric cancer showed that adding onartuzumab to mFOLFOX6 failed to improve efficacy of c-Met immunohistochemically positive population or even the whole population (Shah et al., 2016). Similarly, a phase III clinical trial revealed no significant improvements in clinical benefits in first-line chemotherapy for HER2-negative and c-Met-positive advanced gastroesophageal adenocarcinoma (GEC). Another small molecule, AMG 337, which is a selective inhibitor targeting c-Met has also been reported (Hughes et al., 2016). A phase II clinical study assessing 45 patients with gastric/gastroesophageal junction/esophageal tumor and AMG 337 monotherapy patients diagnosed with c-Met-amplified tumors showed an ORR of 18%, indicating that AMG 337 has certain antitumor activities (Van Cutsem et al., 2019).

Emibetuzumab (LY2875358) is a humanized immunoglobulin G4 monoclonal anti-Met antibody. The drug inhibits the activation of HGF/c-Met pathway by preventing HGF from binding to its receptor, c-Met, and by degrading c-Met (Liu et al., 2014). A phase II study evaluated the safety and efficacy of emibetuzumab in patients with advanced gastric cancer. The 8-week PFS rate is 47%, and the DCR is 40%. Although the sample size is small, monotherapy of emibetuzumab was well tolerated and showed certain anti-tumor activity (Sakai et al., 2017). A phase Ib/II study evaluated the efficacy of emibetuzumab combined with ramucirumab (a monoclonal anti-VEGFR-2 antibody) in 97 patients with solid tumors (including 16 gastric or gastroesophageal junction adenocarcinoma, 45 HCC). The results showed that the combination therapy had more significant antitumor activity.

Telisotuzumab (ABT-700) is a novel anti-c-Met antibody that binds c-Met with high affinity and inhibits c-Met signaling. Unlike most other c-Met antibodies, it destroys the productive dimerization and activation induced by HGF or c-Met on the cell surface independent of ligand (Wang et al., 2016). A phase I study evaluated the safety and efficacy of telisotuzumab in patients with advanced solid tumors. However, significant clinical antitumor activity was only observed in patients with MET-amplified gastroesophageal cancer (Strickler et al., 2020). Considering that primary MET genomic amplification a low-frequency event in most tumors, the researchers developed an antibody–drug conjugate, telisotuzumab–vedotin, also known as ABBV-399, which is composed of telisotuzumab and cytotoxic monomethyl auristatin E (MMAE) via a valine–citrulline linker (Wang et al., 2017). Telisotuzumab–vedotin does not require the tumors to be addicted to the oncogene. It delivers cytotoxin MMAE directly to c-Met-positive tumor cells and has demonstrated antitumor activity in tumors without increased copy number of the MET gene (Strickler et al., 2018). A phase I clinical trial is studying the safety and preliminary efficacy of telisotuzumab–vedotin in patients with advanced solid tumors (NCT02099058). The clinical trial is anticipated to be completed at the beginning of 2022.

Savolitinib is a novel and highly selective c-Met inhibitor, which has shown encouraging results in the phase III randomized clinical trial of papillary renal cell carcinoma (PRCC) (Choueiri et al., 2020). In a preclinical study, savolitinib showed tolerable side effects and significant antitumor effect in c-Met-amplified mice (Gavine et al., 2015). Recently, a multiple-arm clinical trial evaluated the efficacy of targeted therapy in 772 patients with metastatic gastric cancer (Lee et al., 2019). The results showed that there was the highest response rate in the fourth arm (c-Met amplification–savolitinib monotherapy). One patient developed malignant ascites after failure of capecitabine/oxaliplatin treatment. Following savolitinib treatment, the tumor was significantly reduced and the patient achieved curative resection.

Tivantinib, foretinib, and other c-Met tyrosine kinase inhibitors (TKIs), have previously resulted in no clear antitumor activity in gastric cancer patients (Shah et al., 2013; Kang et al., 2014). Although results from several recent clinical trials do not show promise as anti-c-Met drugs, a combined therapeutic strategy for multiple downstream signal transduction could have clinical potential, especially in c-Met-positive patients.

Colorectal cancer is the third most common cancer in the world, although a decline in CRC mortality has been observed in the recent past (Siegel and Miller, 2019). Various studies have reported that overexpression or amplification of c-Met in early (I and II) and late CRC (III and IV) is closely related to its invasion and distant metastasis (Zeng et al., 2008; Garouniatis et al., 2013; Lee et al., 2018). For this reason, studies have hypothesized that inhibition of c-Met activation may decrease tumor activity in CRC.

Ficlatuzumab (AV-299) is a humanized, high-affinity monoclonal antibody that can block HGF/c-Met binding and mediate downstream phosphorylation of the signaling pathway (Mok et al., 2016). A study assessing tolerability and safety of ficlatuzumab in advanced CRC patients demonstrated that the antibody has the ability to regulate HGF/c-Met pathway and downstream signaling (Tabernero et al., 2014).

The therapeutic effect of rilotumumab in CRC has also been evaluated in a randomized phase Ib/II research where panitumumab (a monoclonal antibody targeting EGFR) was combined with rilotumumab or placebo in patients with kirsten rat sarcoma (KRAS) wild-type metastatic CRC who met the prespecified criterion for improvement in ORR (Van Cutsem et al., 2014). Although the combined inhibition of HGF/c-Met and EGFR showed some encouraging results, no further research findings have been reported with regard to rilotumumab and panitumumab development in CRC.

Another new compound, SU11274, that targets the c-Met ATP-binding site has also been reported and implicated in blocking of HGF-dependent c-Met activation (Gao W. et al., 2013). In vivo experiments showed that daily administration of SU11274 in mice resulted in inhibition of tumor growth in xenografts. It was also found that SU11274 could significantly inhibit survival and growth of colon cancer cells (Gao S. H. et al., 2013). However, a study of melanoma in vitro showed that the off-target action of SU11274 resulted in the increase of tumorigenic potential (Kucerova et al., 2016). Whether SU11274 can be used as a candidate drug to inhibit c-Met necessitates further clinical trials.

Analysis of clinical effects of tivantinib, following its gradual introduction into CRC, reveals negative results. For instance, a randomized controlled phase I/II clinical trial of tivantinib combined with irinotecan and cetuximab showed no significant improvement of PFS or OS in metastatic CRC patients with wild-type KRAS (Eng et al., 2016). Similarly, a new multicenter phase II study suggested that a combination of tivantinib and cetuximab did not achieve the expected efficacy in CRC patients who had high c-Met expression and acquired resistance to anti-EGFRs (Rimassa et al., 2019). However, a study indicated that inhibition of the HGF/c-MET pathway can improve the sensitivity of CRC to EGFR inhibitors, indicating that combination therapy could still be a future research direction (Liska et al., 2011).

Currently, relatively few reports about HGF inhibitors in CRC exist. In general, though the research on anti-HGF/c-Met axis for CRC is still in its infancy, they have great potential in the future treatment.

In addition to digestive system cancers, the study of c-Met inhibitors is also very active for other cancer. A large number of clinical trials have been carried out on melanoma, breast cancer, PRCC, NSCLC, and medullary thyroid cancer (MTC), providing a lot of reliable experience for the treatment of digestive system cancers (Table 3).

A phase II clinical trial evaluated the efficacy of cabozantinib compared with traditional chemotherapy in the treatment of uveal melanoma (Luke et al., 2020). The result showed that cabozantinib did not demonstrate improvement in PFS (cabozantinib vs. temozolomide or dacarbazine: 60 vs. 59 days). In another phase II study, cabozantinib showed bone-centric activity in patients of breast cancer with bone metastases (Xu et al., 2020). Savolitinib has shown encouraging results in the phase III randomized clinical trial of PRCC (Choueiri et al., 2020). In this clinical trial, 60 patients were randomly divided into group savolitinib (n = 33) and group sunitinib (n = 27). The primary end point was PFS; group savolitinib and group sunitinib were 7.0 and 5.8 months, respectively. Tivantinib and tepotinib have demonstrated significant antitumor activity in NSCLC. In a phase III study of tivantinib in patients with NSCLC, tivantinib combined with erlotinib led to a significant prolongation of PFS (13.0 vs. 7.5 months) and OS (25.5 vs. 20.3 months) compared to erlotinib monotherapy (Scagliotti et al., 2018). A phase II study evaluated the efficacy of tepotinib in patients with advanced NSCLC with METex14 alterations (Paik et al., 2020). The response rate of the liquid-biopsy group (n = 66) and the tissue-biopsy group (n = 60) were 48% and 50%, respectively. Median duration of response was 11.1 months. In a phase III clinical trial, 330 patients with progressive metastatic MTC were randomly assigned to the cabozantinib group (n = 220) or the placebo group (n = 110) (Elisei et al., 2013). Median duration of PFS was 11.2 and 4.0 months in cabozantinib and placebo, respectively. ORR was 28% in the cabozantinib group and 0% in the placebo group. The therapeutic effect of cabozantinib on MTC was statistically significant.