Abstract
Purpose
Recent evidence has highlighted the role of hepatocyte growth factor (HGF) as a putative biomarker to predict EGFR inhibitor resistance. This study investigated the impact of plasma HGF levels on EGFR inhibition and the counter effect of MET inhibition in KRAS, NRAS, and BRAF (RAS/RAF) wild-type colorectal cancers (CRCs).
Methods
Plasma HGF levels were analyzed with clinical outcomes of patients with metastatic CRC (mCRC) receiving palliative first-line chemotherapy. Then, in vitro experiments were conducted to validate the clinical findings and to establish pre-clinical evidence of MET inhibition by capmatinib.
Results
A total of 80 patients were included: cetuximab + FOLFIRI (n = 35) and bevacizumab + FOLFIRI (n = 45). Both progression-free survival (PFS) and overall survival (OS) were significantly lesser in the high vs low HGF group: median 11.8 vs. 24.7 months, respectively, for PFS (p = 0.009), and median 21.1 months vs. not reached, respectively, for OS (p = 0.018). The difference was significantly evident in the cetuximab group. In five RAS/RAF wild-type CRC cells, the addition of HGF activated ERK1/2 and AKT via MET phosphorylation, resulting in cetuximab resistance in vitro. In cetuximab-sensitive Caco-2 and SNU-C4 cells, capmatinib overcame cetuximab resistance in the presence of HGF by attenuating HGF-induced MET signaling activation.
Conclusion
Patients with mCRC receiving cetuximab + FOLFIRI who presented with high plasma HGF levels had significantly worse PFS and OS. Cetuximab resistance induced by HGF was mediated by AKT and ERK activation and overcome by MET inhibition in vitro.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00432-021-03872-4.
Keywords: FOLFIRI, Cetuximab, Capmatinib, Hepatocyte growth factor, Colorectal cancer
Introduction
A combination of doublet chemotherapy and a targeted agent that blocks epidermal growth factor receptor (EGFR) or vascular endothelial growth factor (VEGF) is the standard initial treatment regimen in patients with metastatic colorectal cancer (Benson et al. 2021). Nevertheless, most patients experience disease progression and, eventually, death. Therefore, there have been several trials to understand and overcome the resistance mechanisms of chemotherapy (Van der Jeught et al. 2018; Xie et al. 2020).
For EGFR inhibitors, cetuximab or panitumumab, the MET signaling pathway has been suggested as a mechanism for primary and secondary resistance (Bardelli et al. 2013; Kishiki et al. 2014; Liska et al. 2011). In colorectal cancer, compared with other cancers including lung adenocarcinoma, de novo MET amplification and mutations are generally rare and approximately 2‒4% (Liu et al. 2018; Raskob et al. 2018). Hepatocyte growth factor (HGF), a ligand for MET, was shown to activate downstream signaling pathway molecules including AKT and MAPK and was, thus, associated with poor prognosis in various cancers (Parizadeh et al. 2019; Raskob et al. 2018; Xing et al. 2016). In addition, plasma HGF has been shown to be linked with cetuximab resistance in several retrospective studies of metastatic colorectal cancer (Takahashi et al. 2014; Yonesaka et al. 2015). Thus, it is speculated that HGF is one of the causes of resistance of colorectal cancer. However, unfortunately, previous studies lacked validation cohorts and did not successfully overcome the HGF-mediated resistance.
Recent data suggest that MET inhibitors could overcome the resistance to EGFR inhibitors in EGFR-mutant non-small cell lung cancer (Wang et al. 2019). Therefore, we hypothesized that MET inhibition could overcome the cetuximab resistance induced by high HGF levels in KRAS, NRAS, and BRAF wild-type colorectal cancers. This study aimed to investigate the clinical implications of baseline plasma HGF levels on EGFR inhibition in patients with KRAS, NRAS, and BRAF wild-type metastatic colorectal cancers. In addition, an in vitro study was conducted to elucidate the biological mechanism and therapeutic potential of an MET inhibitor, capmatinib, in the presence or absence of HGF in KRAS, NRAS, and BRAF wild-type colorectal cancers.
Materials and methods
Clinical study
Patients
This study was conducted using consecutively collected blood samples for two studies: a pharmacogenomic study of patients with advanced solid tumors receiving palliative chemotherapy (IRB registration No. B-1603/340-305) and a biomarker study of patients with metastatic colorectal cancer (IRB registration No. B-1211/180-007). From March 2015 to September 2019, a total of 80 patients were consecutively enrolled at Seoul National University Bundang Hospital, Seongnam, Korea. Of the 80 patients, 35 were treated with cetuximab and 45 patients were treated with bevacizumab in combination with FOLFIRI chemotherapy. The FOLFIRI regimen consisted of an intravenous infusion of irinotecan, 180 mg/m2 on day 1, followed by leucovorin, 400 mg/m2 infusion and 5-fluorouracil 400 mg/m2 bolus, then 5-fluorouracil 2,400 mg/m2 infusion over 46 h every 2 weeks. The blood samples were obtained before the chemotherapy after obtaining informed consent, and additional retrospective medical record review was conducted after approval by the institutional review board of Seoul National University Bundang Hospital (IRB registration No. B-2102-667-301). The present study was performed in accordance with the ethical principles in the Declaration of Helsinki and its later revision in 2013.
Sample collection and measurement of plasma HGF level
Following collection of peripheral blood from patients, plasma was obtained by centrifugation at 3000 × g for 10 min. Plasma HGF levels were measured using human HGF immunoassay kits (DHG00B, R&D Systems, Minneapolis, MN, USA) according to the manufacturer’s protocol.
In vitro study
Cell culture
The Korean Cell Line Bank, Seoul, Korea provided the human colorectal cancer cell lines, Caco-2, SNU-C4, SNU-C1, COLO 320DM, and KM12C, used in this study. RPMI 1640 medium (LM011-51, WELGENE, Gyeongsan, Korea) containing 10% fetal bovine serum (FBS) (26140079, Gibco, Waltham, MA, USA) and 1% penicillin–streptomycin (LS202-02, WELGENE) was used as culture medium for Caco-2, SNU-C4, SNU-C1, and COLO320DM cells. Meanwhile, KM12C cells were cultured with MEM medium (11095080, Gibco) supplemented with 10% FBS and 1% penicillin–streptomycin. Cells were maintained in a 37℃ humidified CO2 incubator and subcultured every 3 to 4 days.
Cytotoxicity assay
Before the assay, the cells were counted using a Countess™ II FL Automated Cell Counter (AMQAF1000, Thermo Scientific, Waltham, MA, USA) to quantify the number of viable cells. Cells were seeded into 384-well plates (781080, Greiner Bio-One, Kremsmünster, Austria) at a density of 1000 cells/well a day prior to treatment with the drug and recombinant human HGF. After 5 days of treatment, cell viability was measured using a CellTiter-Glo luminescent cell viability assay solution (G7573, Promega, Madison, WI, USA).
Western blot analysis
Prior to Western blot analysis, cells were cleaned with cold Dulbecco’s phosphate-buffered saline (PBS) (LB 001-02, WELGENE) followed by lysis for 1 h on ice in radio immunoprecipitation assay (RIPA) buffer (BRA0500, Biomax, Seoul, Korea) supplemented with protease and phosphatase inhibitors. Next, the mixture was centrifuged for 20 min at 13,000 rpm to collect the whole protein lysate. Protein quantification was conducted using a bicinchoninic acid (BCA) protein assay kit (23227, Thermo Scientific). The expression level of the designated targets was examined using 15 μg of protein. An 8% sodium dodecyl sulfate (SDS)–polyacrylamide gel was used to separate the protein which was then transferred onto a PVDF membrane (IPVH00010, Millipore, MA, USA). The membrane was probed with primary antibodies overnight at 4 °C followed by 1 h probing with HRP-conjugated secondary antibodies at room temperature (RT). The expression level of the protein was measured using a ChemiDoc Touch Imaging System (BioRad, Hercules, CA, USA).
Antibodies and therapeutics
Phospho-MET Y1234/1235 (#3077), MET (#8198), phospho-EGFR Y1068 (#3777), EGFR (#2646), phospho-AKT S473 (#4058), AKT (#4685), phospho-ERK1/2 T202/Y204 (#4376), ERK1/2 (#4695), phospho-STAT3 Y705 (#9145), STAT3 (#9139), and vinculin (#13901) were purchased from Cell Signaling Technology (Danvers, MA, USA). HRP-conjugated anti-rabbit (#111-035-003) and mouse (#115-035-003) secondary antibodies were obtained from Jackson ImmunoResearch (West Grove, PA, US). Cetuximab (A2000) and capmatinib (S2788) were purchased from Selleckchem (Houston, TX, USA). HGF (100-39H) was purchased from PeproTech (Rocky Hill, CT, USA).
Colony-forming assay
Based on the appropriate cell density (1500 cells/well for Caco-2 and 1000 cells/well for SNU-C4), cells were seeded to a 6-well plate (30006, SPL Life Science, Pocheon, Korea) a day prior to treatment with cetuximab, capmatinib, and HGF. Cetuximab (15 μg/mL for Caco-2 and 35 μg/mL for SNU-C4), capmatinib 10 nM, and HGF 40 ng/mL were added to the cells twice a week for 2 weeks. Subsequently, Coomassie brilliant blue R-250 solution (1610436, BioRad) was used to stain the cells for 2 h at room temperature. The representative colony images were taken using the ChemiDoc Touch Imaging System (BioRad). Colony de-staining was performed using a 1% SDS solution to analyze the relative colony area which was determined by the optical density value at 595 nm using a Synergy H1 microplate reader (BioTek, Winooski, VT, USA).
Additional measurements and definition of variables
As a part of clinical practice, KRAS, NRAS, and BRAF mutation statuses were evaluated by polymerase chain reaction (PCR) or next-generation sequencing, and microsatellite instability was determined according to the revised Bethesda guidelines (Eisenhauer et al. 2009; Umar et al. 2004). Performance status (PS) of the patients were evaluated using the Eastern Cooperative Oncology Group (ECOG) scale, and disease status was evaluated using computed tomography (CT) or magnetic resonance imaging (MRI) according to routine clinical practice. Treatment response was measured according to the response evaluation criteria in solid tumor (RECIST) version 1.1 (Eisenhauer et al. 2009).
The overall clinical response rate (ORR) was defined as the proportion of patients who had a partial or complete response after palliative treatment. The progression-free survival (PFS) was determined from the start of the palliative first-line chemotherapy to the date of documented disease progression, relapse, or death of any cause. Overall survival (OS) was measured from the start day of the first-line chemotherapy to the date of death of any cause.
Statistical analysis
Maxstat R package (a maximal Chi-square method) was used in R 4.0.3 (R Development Core Team, Vienna, Austria) to define the optimal cutoff level for plasma HGF levels. The cutoff value of plasma HGF was calculated using PFS of the entire patient population. The survival outcomes were calculated with the Kaplan–Meier method and compared with the log-rank test.
In vitro experiments were conducted at least in triplicate. The differences between groups were evaluated by Student’s t test and two-way ANOVA with Bonferroni post hoc test using GraphPad Prism 5 software (San Diego, CA, USA). Error bars represent mean ± standard deviation. A p value < 0.05 was considered statistically significant (*p < 0.05, **p < 0.01, and *** p < 0.001).
Results
Choice of cetuximab or bevacizumab on palliative first-line chemotherapy outcome
Eighty patients were consecutively enrolled from the 2 biomarker studies (Table 1). Among them, 35 patients received cetuximab plus FOLFIRI, and 45 patients received bevacizumab plus FOLFIRI as their palliative first-line chemotherapy. Age, sex, and PS were relatively well balanced between the two groups. The proportion of left-sided cancer was 60.0% (21 patients) in the cetuximab group and 37.8% (17 patients) in the bevacizumab group. KRAS or NRAS mutations were found in 39 patients (86.6%) in the bevacizumab group, while none (0%) was found in patients in the cetuximab group (p < 0.001).
Table 1.
Baseline patient characteristics
| Cetuximab (n = 35) | Bevacizumab (n = 45) | p value | |
|---|---|---|---|
| Age at diagnosis | 66 (24–85) | 60 (39–83) | 0.396 |
| Sex | 0.383 | ||
| Male | 20 (57.1%) | 30 (66.7%) | |
| Female | 15 (42.9%) | 15 (33.3%) | |
| ECOG PS | 0.459 | ||
| 0–1 | 33 (94.3%) | 40 (88.9%) | |
| 2 | 2 (5.7%) | 5 (11.1%) | |
| Primary site | |||
| Right-sided | 5 (14.3%) | 15 (33.3%) | 0.350 |
| Left-sided | 21 (60.0%) | 17 (37.8%) | |
| Rectum | 9 (25.7%) | 13 (28.9%) | |
| Pathology | 0.674 | ||
| Adenocarcinoma, well differentiated | 2 (5.7%) | 2 (4.4%) | |
| Adenocarcinoma, moderately differentiated | 29 (82.9%) | 36 (80.0%) | |
| Adenocarcinoma, poorly differentiated | 3 (8.6%) | 6 (13.3%) | |
| Adenocarcinoma, NOS | 1 (2.9%) | 1 (2.2%) | |
| RAS mutation | < 0.001 | ||
| KRAS | 0 (0.0%) | 33 (73.3%) | |
| Codon 12/13/61 | 21(63.6%) / 9(27.3%) / 3(9.1%) | ||
| NRAS | 0 (0.0%) | 6 (13.3%) | |
| Codon 12/13/61 | 3(50.0%)/1(16.7%)/2(33.3%) | ||
| RAS WT | 35 (100.0%) | 6 (13.3%) | |
| BRAF mutation | 1.000 | ||
| BRAF MT | 2 (5.9%) | 2 (4.4%) | |
| BRAF WT | 32 (94.1%) | 43 (95.6%) | |
| Not evaluated | 1 (2.9%) | 0 (0.0%) | |
| MSI | 0.519 | ||
| MSS | 30 (85.7%) | 40 (88.9%) | |
| MSI-L | 4 (11.4%) | 5 (11.1%) | |
| MSI-H | 1 (2.9%) | 0 (0.0%) | |
| CEA (ng/mL) | 23.1 (1.4–15,480.0) | 27.3 (1.0–15,000.0) | 0.890 |
| CA19-9 (U/mL) | 72.5 (2.0–16,900.0) | 68.0 (0.0–20,000.0) | 0.939 |
ECOG PS Eastern Cooperative Oncology Group performance status; NOS not otherwise specified MSI microsatellite instability; MSS microsatellite stable; MSI-L microsatellite instability-low; MSI-H microsatellite instability-high; CEA carcinoembryonic antigen; CA19-9 carbohydrate antigen
Treatment response of first-line chemotherapy was evaluable in 34 patients (97.1%) in the cetuximab group and 43 patients (95.6%) in the bevacizumab group. The ORR was 60.0% in the cetuximab group and 37.8% in the bevacizumab group (p = 0.162). During a median follow-up period of 29.3 months (range, 1.2–71.3 months), the median PFS was 23.1 months in the cetuximab group (95% CI, 15.8–30.4 months) and 20.9 months in the bevacizumab group (95% CI, 14.3–27.6 months) (p = 0.427). The median OS was not significantly different in the two groups: 36.3 months (95% CI, 28.1–44.5 months) in the cetuximab group and 35.2 months (95% CI, 26.7–43.7 months) in the bevacizumab group (p = 0.481).
High plasma HGF levels are associated with lower PFS and OS
The median plasma HGF level was 374.75 pg/mL (range, 9.43–30,644.44 pg/mL), and the cutoff value of HGF level was 12.70 pg/mL, obtained by the maximal Chi-square method. Of 80 patients, 72 patients were categorized as the high HGF group. The median baseline HGF level was significantly higher among patients with RAS mutant tumors (p = 0.027), and accordingly, among those with bevacizumab (p = 0.034) (Supplementary Table 1). Otherwise, there was no significant association between baseline HGF levels and key clinicopathologic features. The ORR was not different between the high and low HGF groups (p = 0.691) (Table 2). However, the PFS was significantly lesser in the high HGF group vs. the low HGF group: median 11.8 months (95% CI, 10.5–13.1 months) vs. median 24.7 months (95% CI, 23.4–25.9 months; p = 0.009). OS was also significantly lower in the high HGF group (median 21.1 months; 95% CI, 17.9–24.3 months) than in the low HGF group (not reached [NR]; p = 0.018) (Fig. 1A–B).
Table 2.
Treatment response
| Total patient cohort | HGF high (n = 72) | HGF low (n = 8) | p value |
|---|---|---|---|
| PR | 34 (49.3%) | 4 (50.0%) | 0.691 |
| SD | 29 (42.0%) | 4 (50.0%) | |
| PD | 6 (8.7%) | 0 (0.0%) | |
| NA* | 3 | 0 |
| Cetuximab cohort | HGF high (n = 28) | HGF low (n = 7) | p value |
|---|---|---|---|
| PR | 17 (63.0%) | 4 (57.1%) | 0.851 |
| SD | 7 (25.9%) | 3 (42.9%) | |
| PD | 3 (11.1%) | 0 (0.0%) | |
| NA* | 1 | 0 |
PR partial response; SD stable disease; PD progressive disease; NA not assessed
*These patients were not evaluable as they were lost to follow-up before the first disease evaluation
Fig. 1.
Progression-free survival (PFS) and overall survival (OS) according to plasma HGF levels. A PFS according to plasma HGF levels in the entire patient cohort. B OS according to plasma HGF levels in the entire patient cohort C PFS according to plasma HGF levels in the cetuximab group. D OS according to plasma HGF levels in the cetuximab group
This trend was similar in the cetuximab group (n = 35). The ORR was not significantly different according to plasma HGF levels with the same cut-off value (p = 0.851). However, the PFS was significantly lesser in patients with high HGF levels (median 10.9 months; 95% CI, 8.6–13.2 months) than in those with low HGF levels (NR; p = 0.003). The OS was also significantly lower in patients with high HGF levels: median 19.0 months (95% CI, 2.9–35.1 months) than in those with low HGF levels (NR; p = 0.035) (Fig. 1C–D).
In the bevacizumab group, only one patient was in the low HGF group (n = 8), and thus, further statistical analysis could not be performed.
HGF activates MET signaling pathways and induces cetuximab resistance
The cell viability assay was employed to differentiate cetuximab-sensitive cell lines among human colorectal cancer (CRC) cells with wild-type KRAS, NRAS, and BRAF genes. Caco-2 and SNU-C4 cells were relatively sensitive to cetuximab compared with SNU-C1, COLO 320DM, and KM12C cells (Fig. 2A). The half-maximal inhibitory concentration (IC50) values could be calculated in Caco-2 (76.3 μg/mL) and SNU-C4 (479.5 μg/mL) cells. Therefore, Caco-2 and SNU-C4 cells were selected for further experiments.
Fig. 2.
Effect of HGF on the expression levels of signaling molecules and the anti-proliferative activity of cetuximab. (A) Cetuximab cytotoxicity assay in various RAS/RAF wild-type colorectal cancer cell lines Cetuximab was treated for 5 days, and cell viability was measured using CellTiter-Glo. B Western blot analysis of signaling molecules in the presence or absence of HGF. Cells were serum-starved for 1 day and then stimulated with 40 ng/mL HGF for 10 min. HGF activated MET and its downstream signaling molecules, AKT and ERK. C Cetuximab cytotoxicity assay in the presence or absence of HGF. Serially diluted cetuximab and 40 ng/mL of HGF were added for 5 days. After 5 days, cell viability was quantified by CellTiter-Glo
The Western blot analysis was performed to evaluate the effect of HGF on the phosphorylation of MET and it is downstream signaling molecules in KRAS, NRAS, and BRAF wild-type CRC cells (Fig. 2B). The phosphorylation of MET Y1234/1235 and its downstream p-ERK1/2 T202/Y204 were increased by HGF in all cell lines. In addition, HGF increased the phosphorylation of AKT S473 in three of the five cell lines: Caco-2, SNU-C4, and COLO 320DM. In the presence of HGF, increased phosphorylation of STAT3 was not evident in the five cell lines. These data suggest that HGF activates MET and its downstream molecules, ERK1/2 and AKT.
Next, the effect of HGF on cetuximab sensitivity in Caco-2 and SNU-C4 cells was evaluated (Fig. 2C). In the presence of HGF, cell proliferation was significantly increased in Caco-2 and SNU-C4 cells, and HGF did not directly influence cetuximab sensitivity. As a result, cetuximab alone could not retard the cancer cell proliferation effectively, when compared with HGF-free condition. Therefore, we hypothesized that MET inhibition may overcome cetuximab resistance in CRC cells under high HGF condition.
Capmatinib overcomes cetuximab resistance in the presence of HGF
Caco-2 and SNU-C4 cells were treated with capmatinib, an MET inhibitor, to inhibit the effect of HGF in the presence or absence of cetuximab treatment using colony-forming assays (Fig. 3A, B). The two cell lines did not harbor MET alteration including single nucleotide variants, small indels, and copy number variants according to Cancer Cell Line Encyclopedia (Ghandi et al. 2019). Without HGF, cetuximab treatment resulted in significant anti-proliferative effects, while capmatinib showed no additive effect in both cell lines. In contrast, in the presence of HGF, the anti-proliferative effect of cetuximab was significantly reduced. To attenuate the effect of HGF, capmatinib was added to cetuximab. As a result, adding capmatinib to cetuximab showed significantly increased anti-tumor effects compared with cetuximab alone under high HGF condition.
Fig. 3.
The effect of capmatinib in combination with cetuximab, in the absence or presence of HGF. A A colony-forming assay for evaluation of the long-term effect of HGF, cetuximab, and capmatinib on cell viability. HGF, cetuximab, and capmatinib were added twice a week. After 2 weeks, colonies were stained, and representative images were captured by ChemiDoc Touch. B Colony quantification by optical density at 595 nm. C Western blot analysis of EGFR and MET down-stream signaling molecules to evaluate the combined effects of cetuximab and capmatinib on HGF-stimulated cells. Prior to addition of drugs and HGF, the cells were serum-starved for 1 day. The cells were stimulated with 40 ng/mL of HGF for 10 min, and then 1 μM of capmatinib and cetuximab (75 μg/mL for Caco-2 and 200 μg/mL for SNU-C4) for 1 h
To evaluate the mechanism of interaction of cetuximab and capmatinib in Caco-2 and SNU-C4 cells, we analyzed the EGFR and MET signaling pathway by Western blot analysis (Fig. 3C). In the absence of HGF, cetuximab treatment showed inhibitory effect on the phosphorylation of EGFR, AKT, and ERK signaling pathways. Capmatinib did not significantly alter EGFR downstream signaling pathways in both cell lines.
In contrast, in the presence of HGF, the inhibitory effect of cetuximab on phosphorylation of AKT and ERK1/2 was decreased. Notably, the combination of cetuximab and capmatinib abrogated the effect of HGF on EGFR and MET downstream signaling pathways.
Discussion
High plasma HGF levels were associated with poor PFS and OS in patients with colorectal cancer receiving palliative first-line FOLFIRI backbone chemotherapy. This difference was particularly evident in patients receiving cetuximab plus FOLFIRI. Our clinical findings suggest that the resistance mechanism resulting from high HGF levels may be specifically associated with cetuximab-based chemotherapy. In the cetuximab-sensitive, MET wild-type, colorectal cancer cell lines, the addition of HGF resulted in cetuximab resistance through phosphorylation of MET and its downstream pathways. This study verified the in vitro mechanism of resistance to anti-EGFR treatment in the presence of HGF. Furthermore, we explored the possibility that MET inhibition could be an effective therapeutic strategy in KRAS, NRAS, and BRAF wild-type colorectal cancers characterized by high HGF levels.
The concept that HGF inhibits the anti-proliferative effect of EGFR inhibitors through MET activation has been suggested in metastatic colorectal cancer (Liska et al. 2011; Luraghi et al. 2014). Subsequent studies have attempted to elucidate the relationship between HGF levels and treatment outcomes in patients. Yonesaka et al. analyzed the relationship between circulating plasma HGF levels and clinical outcomes in 51 patients with KRAS wild-type metastatic colorectal cancer and reported low disease control rate and lesser survival outcome in the high HGF group (Takahashi et al. 2014; Yonesaka et al. 2015). Similarly, another Japanese group suggested that, in 103 patients receiving anti-EGFR treatment, high plasma HGF levels were associated with anti-EGFR treatment resistance (Takahashi et al. 2014). Since the finding was validated in three independent patient cohorts including ours, it would be less likely that our results are false positive although our study did not have a validation cohort. However, a major limitation of the previous studies was that they did not comprise a control group that did not receive anti-EGFR treatment. In addition, since there were no PFS2 analyses after anti-EGFR treatment, it was not known whether the resistance mechanism by HGF was specific to the anti-EGFR treatment. In contrast, our study showed that there was no significant difference in PFS2 according to baseline plasma HGF levels, suggesting that the effect of HGF is specific to anti-EGFR treatment.
Furthermore, several studies have also been conducted on colon cancer tissue samples. In 2007, Kammula et al. suggested that mRNA co-expression of HGF and c-MET oncogene was associated with a metastatic phenotype and poor survival outcome in metastatic colorectal cancer (Kammula et al. 2007). However, tissue mRNA expression analysis has some obvious limitations. When using FFPE samples, mRNA analysis results may not be reproducible and inaccurate in repeated exams. In addition, protein levels may not always be correlated with mRNA expression levels when protein half-life is different. Seneviratne et al. showed that, using immunohistochemistry and Western blot analysis, HGF activation induced by genomic instability can result in resistance to necroptosis in colorectal cancer samples (Seneviratne et al. 2015). Compared with immunohistochemistry, plasma protein quantification has some advantages, especially in terms of cost and feasibility.
Liska et al. experimentally demonstrated that EGFR/MET activation through HGF enhances proliferation of colorectal cancer cells and inhibits cetuximab-induced G1 arrest and apoptosis (Liska et al. 2011). However, this study was limited to pre-clinical data using colorectal cancer cell lines, and thus, it was still unclear whether high HGF levels negatively influenced patients who receive palliative anti-EGFR chemotherapy. Moreover, it should be noted that the MET kinase inhibitor in this study, PHA-665752, had a rather high IC50 value of 9 nM for MET kinase activity. Moreover, PHA-665752 has been less investigated in humans relative to more recently developed MET inhibitors. In our study, capmatinib is a more potent MET inhibitor with a low IC50 value of 0.13 nM for kinase activity.
Recently, several therapeutic agents targeting HGF or MET signaling pathways have been studied in colorectal cancer (Hu et al. 2016; Lee et al. 2018; Tabernero et al. 2014; Zhi et al. 2018). Additional HGF-MET targeted agents are under clinical investigation (Xie et al. 2020). Moreover, based on the recent GEOMETRY mono-1 trial, capmatinib is a standard targeted agent for patients with MET exon 14-mutated or MET-amplified metastatic non-small cell lung cancer (Wolf et al. 2020). Therefore, our proof-of-concept study may provide a basis for clinical trials to add capmatinib to the palliative anti-EGFR treatment of patients with high plasma HGF levels.
One of the limitations of our study is the lack of a validation cohort. Nevertheless, based on the results of the two previous studies, our study confirmed again that plasma HGF is a predictive biomarker for anti-EGFR treatment outcome in these patients (Takahashi et al. 2014; Yonesaka et al. 2015). In addition, as the treatment regimens were targeted agents in combination with cytotoxic drugs, it was difficult to accurately evaluate the effect of HGF on the efficacy of targeted agent alone. Moreover, our study of capmatinib was limited to pre-clinical experiments and could not evaluate the clinical benefit of administration of MET inhibitors to these patients. This should be investigated in further clinical trials.
In conclusion, high plasma HGF levels were significantly associated with worse PFS and OS in patients with metastatic colorectal cancer receiving cetuximab plus FOLFIRI chemotherapy. The HGF induced cetuximab resistance by activation of AKT and ERK. The MET inhibitor, capmatinib, significantly increased the anti-tumor effects of cetuximab in the presence of HGF in vitro.
Supplementary Information
Below is the link to the electronic supplementary material.
Acknowledgements
This research was funded by Seoul National University Bundang Hospital Research Fund (No. 14-2015-024) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MIST) (No. 2018R1D1A1A02086240). The funders had no role in the study design, data collection and analysis, the decision to publish, or the preparation of the manuscript.
Author contributions
SAK, HJP, and J-WK drafted the article; Prof. J-WK and KWL revised it critically for important intellectual content; SAK, HJP, KJK, J-WK, JHS, KJS, JYL, SHK, J-OL, JWK, YJK, JHK, S-MB, JSL, and K-WL contributed to acquisition of data, analysis and interpretation of data and approved the final version to be published.
Funding
This research was funded by Seoul National University Bundang Hospital Research Fund (No. 14–2015-024) and the National Research Foundation of Korea (NRF) grant funded by the Korea government (MIST) (No. 2018R1D1A1A02086240).
Code availability
Not applicable.
Declarations
Conflict of interest
The authors declare that they have no conflict of interest.
Ethic approval
The blood samples were obtained before the chemotherapy after obtaining informed consent, and this study was conducted after the approval by the institutional review board of Seoul National University Bundang Hospital (IRB registration No. B-2102-667-301). The study was conducted in accordance with the precepts established by the Helsinki Declaration.
Consent to publication
All the authors consented to the publication of this article.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Contributor Information
Kui-Jin Kim, Email: kjkim@snubh.org.
Ji-Won Kim, Email: jiwonkim@snubh.org.
References
- Bardelli A, Corso S, Bertotti A, Hobor S, Valtorta E, Siravegna G, Siena S (2013) Amplification of the MET receptor drives resistance to anti-EGFR therapies in colorectal cancer. Cancer Discov 3(6):658–673. 10.1158/2159-8290.CD-12-0558 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Benson AB, Venook AP, Al-Hawary MM, Arain MA, Chen YJ, Ciombor KK, Gurski LA (2021) Colon cancer, version 22021, NCCN clinical practice guidelines in oncology. J Natl Compr Canc Netw 19(3):329–359. 10.6004/jnccn.2021.0012 [DOI] [PubMed] [Google Scholar]
- Eisenhauer EA, Therasse P, Bogaerts J, Schwartz LH, Sargent D, Ford R, Verweij J (2009) New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer 45(2):228–247. 10.1016/j.ejca.2008.10.026 [DOI] [PubMed] [Google Scholar]
- Ghandi M, Huang FW, Jane-Valbuena J, Kryukov GV, Lo CC, McDonald ER 3rd, Sellers WR (2019) Next-generation characterization of the cancer cell line encyclopedia. Nature 569(7757):503–508. 10.1038/s41586-019-1186-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hu HJ, Lin XL, Liu MH, Fan XJ, Zou WW (2016) Curcumin mediates reversion of HGF-induced epithelial-mesenchymal transition via inhibition of c-Met expression in DU145 cells. Oncol Lett 11(2):1499–1505. 10.3892/ol.2015.4063 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kammula US, Kuntz EJ, Francone TD, Zeng Z, Shia J, Landmann RG, Weiser MR (2007) Molecular co-expression of the c-Met oncogene and hepatocyte growth factor in primary colon cancer predicts tumor stage and clinical outcome. Cancer Lett 248(2):219–228. 10.1016/j.canlet.2006.07.007 [DOI] [PubMed] [Google Scholar]
- Kishiki T, Ohnishi H, Masaki T, Ohtsuka K, Ohkura Y, Furuse J, Sugiyama M (2014) Overexpression of MET is a new predictive marker for anti-EGFR therapy in metastatic colorectal cancer with wild-type KRAS. Cancer Chemother Pharmacol 73(4):749–757. 10.1007/s00280-014-2401-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee J, Kim ST, Park S, Lee S, Park SH, Park JO, Park YS (2018) Phase I trial of Anti-MET monoclonal antibody in MET-overexpressed refractory cancer. Clin Colorectal Cancer 17(2):140–146. 10.1016/j.clcc.2018.01.005 [DOI] [PubMed] [Google Scholar]
- Liska D, Chen CT, Bachleitner-Hofmann T, Christensen JG, Weiser MR (2011) HGF rescues colorectal cancer cells from EGFR inhibition via MET activation. Clin Cancer Res 17(3):472–482. 10.1158/1078-0432.CCR-10-0568 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Liu Y, Sethi NS, Hinoue T, Schneider BG, Cherniack AD, Sanchez-Vega F, Laird PW (2018) Comparative molecular analysis of gastrointestinal adenocarcinomas. Cancer Cell 33(4):721–735728. 10.1016/j.ccell.2018.03.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Luraghi P, Reato G, Cipriano E, Sassi F, Orzan F, Bigatto V, Boccaccio C (2014) MET signaling in colon cancer stem-like cells blunts the therapeutic response to EGFR inhibitors. Cancer Res 74(6):1857–1869. 10.1158/0008-5472.CAN-13-2340-T [DOI] [PubMed] [Google Scholar]
- Parizadeh SM, Jafarzadeh-Esfehani R, Fazilat-Panah D, Hassanian SM, Shahidsales S, Khazaei M, Avan A (2019) The potential therapeutic and prognostic impacts of the c-MET/HGF signaling pathway in colorectal cancer. IUBMB Life 71(7):802–811. 10.1002/iub.2063 [DOI] [PubMed] [Google Scholar]
- Raskob GE, van Es N, Verhamme P, Carrier M, Di Nisio M, Garcia D, Hokusai VTECI (2018) Edoxaban for the treatment of cancer-associated venous thromboembolism. N Engl J Med 378(7):615–624. 10.1056/NEJMoa1711948 [DOI] [PubMed] [Google Scholar]
- Seneviratne D, Ma J, Tan X, Kwon YK, Muhammad E, Melhem M, Zarnegar R (2015) Genomic instability causes HGF gene activation in colon cancer cells, promoting their resistance to necroptosis. Gastroenterology 148(1):181–191117. 10.1053/j.gastro.2014.09.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tabernero J, Elez ME, Herranz M, Rico I, Prudkin L, Andreu J, Baselga J (2014) A pharmacodynamic/pharmacokinetic study of ficlatuzumab in patients with advanced solid tumors and liver metastases. Clin Cancer Res 20(10):2793–2804. 10.1158/1078-0432.CCR-13-1837 [DOI] [PubMed] [Google Scholar]
- Takahashi N, Yamada Y, Furuta K, Honma Y, Iwasa S, Takashima A, Shimada Y (2014) Serum levels of hepatocyte growth factor and epiregulin are associated with the prognosis on anti-EGFR antibody treatment in KRAS wild-type metastatic colorectal cancer. Br J Cancer 110(11):2716–2727. 10.1038/bjc.2014.230 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Umar A, Boland CR, Terdiman JP, Syngal S, de la Chapelle A, Ruschoff J, Srivastava S (2004) Revised Bethesda guidelines for hereditary nonpolyposis colorectal cancer (Lynch syndrome) and microsatellite instability. J Natl Cancer Inst 96(4):261–268. 10.1093/jnci/djh034 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Van der Jeught K, Xu HC, Li YJ, Lu XB, Ji G (2018) Drug resistance and new therapies in colorectal cancer. World J Gastroenterol 24(34):3834–3848. 10.3748/wjg.v24.i34.3834 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wang Q, Yang S, Wang K, Sun SY (2019) MET inhibitors for targeted therapy of EGFR TKI-resistant lung cancer. J Hematol Oncol 12(1):63. 10.1186/s13045-019-0759-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wolf J, Seto T, Han JY, Reguart N, Garon EB, Groen HJM, Investigators G (2020) Capmatinib in MET Exon 14-mutated or MET-amplified non-small-cell lung cancer. N Engl J Med 383(10):944–957. 10.1056/NEJMoa2002787 [DOI] [PubMed] [Google Scholar]
- Xie YH, Chen YX, Fang JY (2020) Comprehensive review of targeted therapy for colorectal cancer. Signal Transduct Target Ther 5(1):22. 10.1038/s41392-020-0116-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Xing F, Liu Y, Sharma S, Wu K, Chan MD, Lo HW, Watabe K (2016) Activation of the c-Met pathway mobilizes an inflammatory network in the brain microenvironment to promote brain metastasis of breast cancer. Cancer Res 76(17):4970–4980. 10.1158/0008-5472.CAN-15-3541 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yonesaka K, Satoh T, Ueda S, Yoshida T, Takeda M, Shimizu T, Nakagawa K (2015) Circulating hepatocyte growth factor is correlated with resistance to cetuximab in metastatic colorectal cancer. Anticancer Res 35(3):1683–1689 [PubMed] [Google Scholar]
- Zhi J, Li Z, Lv J, Feng B, Yang D, Xue L, Jia Y (2018) Effects of PHA-665752 and vemurafenib combination treatment on in vitro and murine xenograft growth of human colorectal cancer cells with BRAF(V600E) mutations. Oncol Lett 15(3):3904–3910. 10.3892/ol.2018.7770 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Not applicable.