MET Expression and Amplification in Patients with Localized Gastric Cancer

Despite a worldwide prevalence of nearly one million new cases annually (1), drug development in gastric cancer has lagged and the prognosis for patients with gastric cancer remains poor. Conventional therapy for metastatic gastric cancer remains palliative, with a median survival for metastatic disease of less than one year (2–4). New therapeutic targets for gastric cancer are needed.

Preclinical data suggest that the hepatocyte growth factor (HGF)/MET pathway may represent a therapeutic target for gastric adenocarcinoma (5). The MET proto-oncogene, located on the 7q31 locus, encodes the receptor tyrosine kinase MET, also known as the MET or HGF receptor (6, 7). The binding of HGF to its receptor, MET, results in C-terminus receptor tyrosine phosphorylation and receptor activation. MET receptor targets include activation of the phosphoinositide 3-kinase-Akt/protein kinase B (PI3K-Akt), mitogen-activated protein kinase (MAPK), and phospholipase Cγ pathways, all of which suppress apoptosis, promote tumor cell survival, gene transcription, angiogenesis, cellular proliferation, migration, mitosis, and differentiation (8). In tumors, MET oncogene dependence occurs when the MET tyrosine kinase becomes constitutively active, resulting in gain of function due to MET mutation [found in hereditary papillary renal carcinomas (9) and lung cancers (10)] or MET amplification [reported in gastric, esophageal (11), and lung cancers (12)]. MET mutations are exceedingly rare in gastric cancer (13–16).

Earlier reports describe MET amplification in approximately 20% of gastric tumors (17–21) and MET protein overexpression [assessed by immunohistochemistry (IHC)] in approximately 50% of advanced gastric cancers (22–24). MET amplification and overexpression may herald aggressive tumor biology and worse clinical outcome (8, 17, 22, 24). MET protein overexpression correlates with increased depth of tumor invasion and increased metastatic potential (23, 24). On the basis of evidence that MET dysregulation contributes to the growth and progression of gastric adenocarcinoma and that MET inhibition may be an attractive new target for the treatment of gastric adenocarcinoma, this study was conducted to determine the incidence of MET expression and amplification in gastric cancer in a uniform population of U.S. patients with locally advanced gastric adenocarcinoma.

The Memorial Sloan-Kettering Institutional Review Board approved the study. All patients provided written informed consent for participation on National Cancer Institute–sponsored protocol of a preoperative chemotherapy study (NCI 5700) between June 2003 and November 2005. Thirty-eight patients provided tissue for MET assessment. Patient clinical characteristics included sex, age, histology, and pathologic stage.

FISH was conducted on formalin-fixed, paraffin-embedded tissues. DNA probes for MET (bacterial artificial chromosome clone RPC11-163C9; Invitrogen Life Technologies) and the centromere of chromosome 7 were directly labeled via nick translation with SpectrumRed (MET) and SpectrumGreen (centromere of chromosome 7) fluorophores, respectively. Slides were prepared by using standard cytogenetic techniques. The slides were denatured in 70% formamide/2× SSC for 5 minutes at 72°C and dehydrated in 70%, 85%, and 100% ethanol. The slides were then hybridized in 50% formamide, 2× SSC, Cot-1 DNA, and 50 ng of each probe at 37°C in a humid chamber overnight. After washing in 2× SSC/0.3% NP-40 at 72°C for 2 minutes, the slides were air-dried, counterstained with 0.2 μmol/L 4′,6-diamidino-2-phenylindole (DAPI), and cover slipped. The signals were visualized with a Nikon Eclipse fluorescence microscope containing SpectrumRed (MET locus), SpectrumGreen (centromere), and DAPI filters (Nikon Instruments). A total of 200 interphase cells were analyzed from each sample. The Metasystem software (Digital Scientific) was used for capturing the images.

Quantitative PCR for analysis of MET genomic amplification primers and probes for MET and 18s rRNA were obtained from Applied Biosystems. Primer and probe sequences for MET were (5′-3′): F-GGAGCCAAAGTCCTTTCATCTGTAA, RGCAATGGATGATCTGGGAAATAAGAAGAAT, and FAM-CCGGTTCATCAACTTC. Reactions were done in triplicate under standard thermocycling conditions by using 10 ng of genomic DNA, primers at 900 nmol/L, and probes at 250 nmol/L. Levels of expression of MET mRNA are reported as relative copies that are normalized against 18S rRNA expression (25).

All specimens were fixed in 10% phosphate-buffered formalin and embedded in paraffin. For each case, all available hematoxylin and eosin–stained sections were reviewed, and a representative tissue block was selected for additional studies. Lauren classification was used to classify tumors according to histologic type. Standard ABC peroxidase techniques were used for IHC that was carried out on 4-μm paraffin sections of formalin-fixed, paraffin-embedded resected gastric cancer specimens. The following antibodies were used: anti-MET (C-12) from Santa Cruz Biotechnology and phospho-MET [(p-MET) Y1234/Y1235] from Cell Signaling Technology. A pathologist coded MET and p-MET expression as the percentage of positive tumor cells (scale 0%–100%) with staining intensity from 0 to 3+. Positive IHC expression is defined as 25% or more staining with intensity 2 or 3+. The reference pathologist (L.T.) reviewed all IHC MET and p-MET stains.

MET mRNA (PCR) and protein (IHC) expression was correlated with histology, tumor location, and treatment response by using the Fischer exact and Wilcoxon rank-sum tests.

Table 1 shows clinical characteristics of the patients. The patient cohort consists predominantly of middle to distal stomach tumors (71%), with a similar number of patients with Lauren's diffuse (38%) and intestinal tumors (40%). Fifty-seven percent of the patients had locally advanced stage III or IV (occult peritoneal disease) tumors.

FISH analysis of MET was successful in all 38 (100%) tumor specimens. Table 2 presents the fractions of cells with MET/CEP7 signal in each tumor specimen. The presence of more than 2 gene-specific signals (red) accompanied by the same number of chromosome 7 centromere-specific signals (green) was regarded as indication of polysomy of chromosome 7 (CEP7; Fig. 1). Eleven (33%) tumors were polysomic for chromosome 7 and displayed equal numbers of copies of MET and CEP7 (range, 3–8 copies). Nine of these 11 tumors had high MET IHC. MET amplification (defined as MET/CEP7 ratio > 2) was not identified in this sample set.

Table 2 summarizes individual tumor MET and p-MET IHC, MET mRNA PCR, and FISH results.

Quantitaive PCR was done in 15 tumor specimens and matched normal gastric mucosa. Relative MET mRNA expression was significantly higher for tumor than for normal (9.9 vs. 3.0, P = 0.008), and this was due to high relative MET mRNA expression in Lauren's intestinal histology versus normal (P = 0.02; Table 3).

Resected tumor specimens of 38 patients were examined by IHC for MET, and 30 specimens were tested by IHC for p-MET. Table 4 summarizes the tumor characteristics of patients in MET IHC–positive and -negative groups. Positive MET staining by IHC was associated with Lauren intestinal histology (P = 0.006). Five of 7 (71%) cases with high MET mRNA expression (above median) were noted to have high MET IHC (Fig. 2). Two of 30 (6.6%) gastric cancer specimens revealed positive staining for p-MET (Table 2, Fig. 3). Of these, one specimen was MET IHC positive without an increase in MET mRNA expression (Table 3).

The result summary is presented in Table 5. Although MET IHC positivity was relatively common, increased MET phosphorylation was rare and MET amplification was not seen.

MET amplification is believed to occur frequently in gastric cancer (5, 26). Previous studies noted MET amplification in up to 50% of gastric cancer cell lines (5, 17) and up to 20% of patients' gastric tumor samples (17–21). However, in this evaluation, we identified MET gene amplification in none of 38 locally advanced gastric adenocarcinomas that comprised the study set. We did observe 30% of gastric tumors with multiple MET gene copy numbers as a result of polysomy 7. It is known that breast tumors with an increased HER2 gene copy number as a result of polysomy 17 behave as HER2-negative tumors (27). This phenomenon, therefore, suggests that gastric cancers with MET polysomy are unlikely to be MET driven.

The presence of MET amplification in prior studies (mostly from Japan) could possibly be linked to differences in tumor biology between Asian and Western patients. Distal stomach tumors are more common in Japan and have a favorable prognosis compared with proximal stomach and gastroesophageal junction (GEJ) tumors, which are more common in U.S. patients (5-year overall survival rates of approximately 60% vs. 20%; refs. 2, 4, 28–34). A recent study compared survival following resection of 2,357 Korean and U.S. patients. Even when evaluated by multivariate analysis, correcting for validated prognostic factors (35), Korean gastric cancer patients had improved survival over U.S. patients, suggesting that differences in tumor biology cannot be excluded (36). The distinct tumor biology of gastric cancer subtypes (37–40) and specific host genetic variations among ethnic groups (30, 41–43) might contribute to the difference in survival; although treatment approaches and mass screening programs in Japan also add to survival variability (44).

Another possible explanation for the discrepancy of our study with previous reports is that prior studies were conducted by using the Southern blot technique, which overestimated the incidence of MET amplification because it could not discriminate polysomy 7 from MET amplification. FISH methodology is technically more standardized and less affected by tissue variables, and it has replaced Southern blot in modern clinical diagnostic molecular pathology. In 1998, Hara and colleagues used FISH to examine 154 primary gastric tumors from Japanese patients and found that 6 (4%) tumors were MET amplified (45). It is likely that MET amplification does occur in gastric cancer but at a rate substantially lower than that commonly reported in the literature.

Smolen and colleagues have shown that gastric cancer cell lines with high-level amplification of MET are extraordinarily susceptible to the selective MET tyrosine kinase inhibitor (TKI) PHA-665752. Treatment with MET TKI resulted in massive apoptosis in 5 of 5 MET-amplified (FISH) gastric cancer cell lines and none of the 12 MET-negative cell lines (5). With such a dramatic benefit in preclinical models of MET-driven gastric cancer, the success of MET-targeted therapy (31) in gastric cancer will depend on correctly selecting the patient population whose tumors depend on MET for growth and development.

In our analysis, MET protein and mRNA expression were commonly encountered (in 63% and 50% of resected tumor specimens, respectively) and are concordant with reported literature (23, 24, 46). The significance of the increase in MET expression on a transcriptional level is unclear, especially when considering the absence of MET amplification or MET tyrosine kinase activity. In breast cancer, multiple copies of chromosome 17 (location of the HER2 gene) has been associated with increased HER2 oncoprotein staining by IHC, without HER2 gene amplification (47).

In the absence of MET amplification, MET activation in gastric cancer might be related to deregulation of microRNA that is related to the MET gene as an alternative pathway that is associated with aggressiveness of the gastric tumors. MicroRNAs are a class of small, noncoding RNAs that regulate gene expression, and they are increasingly implicated in the pathogenesis of cancer (48). Migliore and colleagues have shown that miR-34b, miR-34c, and mir-199a can decrease MET expression on the protein and RNA level and impair MET-mediated invasive growth in a gastric cell line that has MET amplification (49). Unique microRNA signatures are associated with different histologic subtypes, pattern of progression, and prognosis in gastric cancer (50). It is possible that certain microRNAs, the nature of which remains to be investigated, can increase MET protein expression (in the absence of MET gene amplification).

Our study shows the lack of MET amplification and pathway activation in Western patients with gastric cancer. In a separate study of unselected advanced gastric cancer patients, single-agent MET TKI (GSK1363089) failed to show antitumor activity (51). We conclude that MET-driven gastric cancers are rare in the Western population. Future studies of MET inhibitors will, therefore, require better patient selection and trial design.

M.A. Shah obtained a commercial research grant from GlaxoSmith Klein. Y.Y. Janjigian is a consultant/advisory board member of Roche/Genentech and obtained a commercial research grant from Boehringer-Ingelheim.

This study was sponsored in part by the DeGregorio Family Foundation for Gastric and Esophageal Cancer Research and ASCO Career Development Award (to M.A. Shah).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.