[CANCER RESEARCH 62, 589–596, January 15, 2002]

Met Protein Expression Level Correlates with Survival in Patients with Late-stageNasopharyngeal Carcinoma

Chao-Nan Qian, Xiang Guo, Brian Cao, Eric J. Kort, Chong-Chou Lee, Jindong Chen, Ling-Mei Wang,Wei-Yuan Mai, Hua-Qing Min, Ming-Huang Hong, George F. Vande Woude, James H. Resau, and Bin Tean Teh1

Laboratories of Cancer Genetics [C-N. Q., X. G., J. C., B. T. T.], Monoclonal Antibody Production [B. C.], Analytical, Cellular, and Molecular Microscopy [E. J. K., J. H. R.],and Molecular Oncology [C-C. L., L-M. W., G. F. V. W.], Van Andel Research Institute, Grand Rapids, Michigan 49503; Department of Nasopharyngeal Carcinoma, CancerCenter, Sun Yat-sen University of Medical Sciences, Guangzhou 510060, People’s Republic of China [C-N. Q., X. G., W-Y. M., H-Q. M., M-H. H.]; and Department ofRadiology/Radiation Oncology, Baylor College of Medicine, Houston, Texas 77030 [W-Y. M.]

ABSTRACT

Met tyrosine kinase, the receptor for HGF/SF, is important in variouscellular functions, including proliferation, mitogenesis, formation ofbranching tubules, angiogenesis, and tumor cell invasion and metastasis.However, the role of Met/HGF signaling pathway in nasopharyngealcarcinoma (NPC) has not been evaluated. In this study, we determined theexpression profile and clinical correlation of Met/HGF in 66 cases ofadvanced NPC and the activation mechanisms of Met receptor in fiveNPC cell lines. Immunofluorescent staining and quantitative image anal-ysis showed that the Met protein was expressed throughout the tumorsand normal nasopharyngeal epithelia. Compared with NPC, the Metexpression level was higher in columnar nasopharyngeal epithelium butlower in squamous nasopharyngeal epithelium. The normal interstitialstromal tissue expressed the lowest level of Met protein. HGF was detectedmainly in the normal interstitial tissue surrounding the tumor. Met pro-tein expression level was inversely correlated with patients’ survival time;the correlation coefficient was �0.319 (P � 0.009). The mean survival timewas 118 months in low Met expression group versus 52 months in highexpression group (P � 0.0004). The cumulative 5-year survival rate was77.68% in low expression group versus 38.24% in high expression group.The clinical stage was also significantly more advanced in high Metexpression group. In the multivariate analysis, both clinical stage and Metprotein expression level were independent prognostic indicators for pa-tient survival. All of the five NPC cell lines tested did not express hgfmRNA but expressed met mRNA, and tyrosine phosphorylation of Metprotein was mainly induced by exogenous HGF stimulation in these cells.No mutation was found in the tyrosine kinase and the juxtamembranedomains of Met receptor in the five NPC cell lines tested. These resultsindicate that: (a) high Met protein expression level correlates with poorersurvival in late-stage NPC and serves as an independent prognostic indi-cator; and (b) the Met receptor in NPC is activated by its paracrine ligandHGF from the interstitial tissues rather than by an autocrine loop or itsactivating mutation.

INTRODUCTION

NPC2 has its highest incidence rate in Southern China and is one ofthe most common cancers in Southeast Asia (1–3). The etiologicalfactor is believed to be the interaction between genetic susceptibilityand environmental factors, especially EBV (4, 5). The WHO devel-oped a three-tier histological classification of NPC based on thedegree of differentiation (6). These are squamous cell carcinoma,differentiated nonkeratinizing carcinoma, and undifferentiated carci-noma. In the endemic area, �95% of NPCs belong to the latter twotypes. But in nonendemic Western countries, squamous cell carcino-

mas account for 30–50% of NPC (7). Besides the epidemiologicalsimilarity, the serological and the clinical characteristics are alsosimilar between differentiated nonkeratinizing carcinoma and undif-ferentiated carcinoma. In these two types of NPC, there is intenseleukocyte infiltration, which predominantly comprises T lymphocytesand accessory cells (2, 8, 9). In addition to its rapid growth behavior,NPC has a great tendency to invade adjacent regions and metastasizeto regional lymph node and distant organs. As a result, NPC is oftenlocally advanced or already has spread to lymph nodes at time ofdiagnosis. Together these factors account for the high rate of treatmentfailure despite the tumor cells’ radiosensitivity (1, 10, 11).

The Met tyrosine kinase receptor, whose ligand is HGF/SF (12),was identified originally as a transforming gene activated by rear-rangement in cells treated with a chemical carcinogen (13, 14). TheMET protooncogene contains 21 exons which give rise to four do-mains (15, 16); the extracellular domain (exons 2–13), the transmem-brane domain (exon 13), the juxtamembrane domain (exon 14), andthe tyrosine kinase domain (exons 15–21). Missense mutations in thetyrosine kinase domain and in the juxtamembrane domain have beenfound to be tumorigenic and able to induce metastasis by constitu-tively activating the Met tyrosine kinase activity (17–24).

The Met protein is a tyrosine kinase receptor synthesized as a Mr

�170,000 precursor that is glycosylated and cleaved to create themature protein (25, 26). As a cell membrane receptor, the Met proteinis composed of two subunits: an extracellular � chain of Mr �50,000and a transmembrane � chain of Mr �145,000 (25). There is a singlemultifunctional docking site located in the tail of the � chain. Itsligand, HGF/SF, is a heterodimer glycoprotein composed of a Mr

�69,000 � chain and a Mr �34,000 � chain. After the stimulation ofHGF, the Met docking site is phosphorylated on two adjacent tyrosineresidues leading to the recruitment and activation of multiple trans-ducers and signaling pathways involving proliferation, mitogenesis,formation of branching tubules, angiogenesis, and tumor cell invasionand metastasis (27, 28). Although Met and HGF/SF are expressedpredominantly in cells of epithelial and mesenchymal origin, respec-tively (29, 30), Met overexpression has been reported in a variety ofcancers and cell lines from the tissues of breast (31, 32), ovary (33),thyroid (34), pancreas (35), stomach (36), brain (37), prostate (38),and endometrium (39). However, its expression profile in NPC hasnot been studied.

In the present study, we investigated the expression profiles ofMet/HGF in NPC, the clinical correlation of Met with advanced NPC,and the activation mechanism of Met receptor in NPC cell lines.

MATERIALS AND METHODS

Patients and Tissue Samples. A total of 66 patients with histologically,clinically, and radiographically diagnosed late-stage NPC from the CancerCenter, Sun Yat-sen University of Medical Sciences, between December 1982and December 1996 was characterized. There were 48 males and 18 femaleswith an average age of 46.4 years (range, 20–70 years). The WHO histologicalclassification system was used for pathological stratification. Standard curative

Received 6/22/01; accepted 11/14/01.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 with18 U.S.C. Section 1734 solely to indicate this fact.

1 To whom requests for reprints should be addressed, at Laboratory of Cancer Genet-ics, Van Andel Research Institute, 333 Bostwick NE, Grand Rapids, MI 49503. Phone:(616) 234-5296; Fax: (616) 234-5297; E-mail: bin.teh@vai.org.

2 The abbreviations used are: NPC, nasopharyngeal carcinoma; SO, styloid process-occipital foramen; HGF/SF, hepatocyte growth factor/scatter factor; VEGF, vascularendothelial growth factor; TTBS, Tween 20 in Tris buffered saline; CI, confidenceinterval; VARI, Van Andel Research Institute.

589

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

radiotherapy was applied to the patients with local or locoregional diseases.Platinum-based concomitant and/or adjuvant chemotherapies combined withalleviate radiotherapy were used for the patients with distant disseminationNPC. The median follow-up time was 79 months (range, 1–163 months).Survival time was calculated from the date of treatment until the time of deathor the most recent follow-up if the patient was alive. At the end of the studyperiod (December 1996), 37 patients (56%) had died as a result of recurrenceof their disease. The patients and tumor-related characteristics are shown inTable 1. The formalin-fixed, paraffin-embedded archived primary tumor bi-opsy tissues of all patients before treatment were retrieved and used in theanalyses. Consecutive 4-�m paraffin sections were prepared for all tissues.One set of the sections was used for routine H&E staining and were reviewedand restratified using WHO histological classification. The remaining sectionswere used for further immunofluorescent staining.

Clinical Staging System. The disease stages of all patients were classifiedor reclassified according to the 1992 NPC staging system of China (40). Thestaging system is characterized according to the following model: T, primarytumor: T1, limited to the nasopharynx; T2, involvement of the nasal cavity,oropharynx, soft palatine, anterior cervical vertebrae soft tissue, and parapha-ryngeal space extension before the SO line (the SO line is between the styloidprocess and the midpoint on the posterior edge of the great occipital foramen);T3, extension over the SO line, involvement of the anterior or posterior cranialnerves alone, the base of the skull, the pterygoprocess zone, and the pterygo-palatine fossa; T4, involvement of both anterior and posterior cranial nerves,parabasal sinus, cavernous sinus, orbit, infratemporal fossa, and direct invasionof the first or second cervical vertebrae; N, regional lymph node involvement:N0, no enlarged lymph nodes; N1, greatest dimension of upper neck lymphnode �4 cm, movable; N2, lower neck lymph node or greatest lymph nodedimension between 4 and 7 cm; N3, supraclavicular lymph node, lymph nodegreatest dimension �7 cm, fixed, or skin infiltration (the border between theupper neck and the lower neck is the inferior margin of the cricoid cartilage);M, distant metastasis: M0, absence of distant metastasis; M1, presence ofdistant metastasis; staging: Stage I, T1N0M0; Stage II, T2N0-N1M0, T0-T2N1M0; Stage III, T3N0-N2M0, T0-T3N2M0; Stage IVa, T4N0-N3M0,T0-T4N3M0; Stage IVb, M1.

Cell Culture, HGF, and Antibodies. In addition to the primary biopsies,two well-differentiated NPC cell lines, CNE-1 and HK-1, were used, as wellas three poorly differentiated NPC cell lines: CNE-2, HONE-1, and SUNE-1(41–45). S-114 is a cell line transformed from NIH 3T3 cells that stably

coexpresses human Met and HGF, resulting in autoactivation of Met receptor(46, 47). We used this cell line for positive control in Northern blot analyses.The five NPC cell lines were maintained in DMEM, supplemented with 10%fetal bovine serum, whereas S-114 was cultured in DMEM, supplemented with8% calf serum. Penicillin (100 IU/ml) and 100 �g/ml streptomycin were addedto the culture media.

Human recombinant HGF/SF was purified from the supernatant of trans-formed NIH 3T3 cells as described previously (47). HGF/SF concentrationswere presented as scatter units per ml; five scatter units are equivalent to �1ng of protein.

The anti-hHGF monoclonal antibody was prepared in the Lab of Mono-clonal Antibody Production, VARI (46). For immunofluorescent studies, theanti-hHGF monoclonal antibody was used at 20 �g/ml in blocking buffer (10%BSA and 5% donkey serum in PBS); the polyclonal rabbit anti-hMet antibody(C-28; Santa Cruz Biotechnology, Santa Cruz, CA.) was used at 4 �g/mlblocking buffer. Negative control antibodies were nonspecific mouse IgG2b

(R&D Systems, Minneapolis, MN; catalogue no. MAB004) and normal rabbitIgG (R&D Systems; catalogue no. AB-105-C) used at 10 �g/ml. Secondaryantibodies were 8 �g/ml rhodamine-conjugated donkey antimouse immuno-globulins and 12 �g/ml FITC-conjugated donkey antirabbit immunoglobulins(Jackson ImmunoResearch Laboratories, West Grove, PA) diluted in blockingbuffer. For immunoprecipitation, mouse monoclonal antihuman Met antibody(Upstate Biotechnology, Lake Placid, NY; Info-ID 05237) was used as 0.5�l/sample. For Western blot analysis, mouse monoclonal antihuman Metantibody (Upstate Biotechnology; Info-ID 05238) was used in the dilution of1:2,500, mouse monoclonal antiphosphotyrosine antibody (Info-ID 05321)was used in the dilution of 1:1,000, and goat antimouse horse-radish peroxi-dase-conjugated secondary antibody (Pierce, Rockford, IL) was diluted to1:10,000.

Immunofluorescent Double Staining. The sections for immunofluores-cent double staining were deparaffinized, rehydrated, and then boiled in 10 mM

sodium citrate buffer (pH 6.0) for 10 min to increase antigen exposure.Nonspecific antibody-binding sites in the tissue sections were blocked byincubation in blocking buffer for 30 min at room temperature. The sectionswere incubated overnight at 4°C with both mouse anti-hHGF and rabbitanti-hMet antibodies. For negative control staining, nonspecific mouse IgGand rabbit IgG were used in place of the primary antibodies. The sections werethen washed three times with PBS containing 0.5% Tween 20 to moveunbound primary antibodies followed by incubation in fluorochrome-conju-gated secondary antibodies for 30 min at room temperature. The sections werefinally rinsed with PBS containing 0.5% Tween 20 three times to removeunbound secondary antibodies and mounted in Gel/Mount permanent aqueousmounting medium (Biomeda Corp, Foster City, CA).

Quantitative Fluorescent Analyses. The techniques and methods of quan-titative fluorescent analyses have been described previously (21–23, 48).Briefly, a Zeiss LSM410 confocal laser scanning microscope configured witha 25-mW argon internal HeNe laser was used to image the fluorescent-stainedsections with lines 488 and 543 for FITC and rhodamine excitation, respec-tively. The 635 red laser was used to produce Nomarski images. The fluores-cent intensity was captured via the microscope’s photomultiplier tube as a512 � 512 pixel image file with an image depth of 8 bits (i.e., each pixel wasstored as an intensity ranging from 0 to 255). The images were quantified fromthe digital microscopy images using software developed by Laboratory ofAnalytical, Cellular, and Molecular Microscopy, VARI.3 Analysis of eachsample was made with the same laser intensity, brightness, power, and scanrate. The regions of interest in each image were defined using Paint Shop Prosoftware version 5.01 (Jasc Software, Minneapolis, MN). The average fluo-rescence intensity of each region of interest was calculated using Image ProPlus software. The area with intensity �20 was excluded because that corre-sponded to only glass or paraffin without any tissue. For each tissue, two tothree images were obtained, and the average value was used as the detectionvalue.

Northern Blot Analysis. The cDNA probe for human hgf was derivedfrom the coding region for HGF � chain (accession no. X16323) by PCRamplification using the primer pairs: forward primer, 5�-TGTGGGTGAC-CAAACTCCTGCC-3� (nucleotides 136–167) and reverse primer, 5�-TCG-

3 Internet address: http://www.sharedpath.org/software.

Table 1 Patients and tumor-related characteristics in low and high Metexpression groups

Unless otherwise noted, the values in this table show the number of patients with thepercentage of patients within a subgroup in parenthesis. Ps are results from the compar-isons between the two groups.

Characteristic Low Met expression High Met expression P

Total Patients 32 34Sex NSa

Male 25 (78.1%) 23 (67.6%)Female 7 (21.9%) 11 (23.4%)

Age (mean yr; range) 44 (24–69) 48.7 (20–70) NSHistologic type (WHO type) NS

Differentiated nonkeratinizing 3 (9.4%) 4 (11.8%)Undifferentiated carcinoma 29 (90.6%) 30 (88.2%)

Primary tumor (T) stage NST1 1 (3.1%) 1 (2.9%)T2 8 (25%) 9 (26.5%)T3 22 (68.8%) 16 (47.1%)T4 1 (3.1%) 8 (23.5%)

Regional lymph node (N) NSN0 0 (0%) 1 (2.9%)N1 6 (18.8%) 5 (14.7%)N2 21 (65.6%) 14 (41.2%)N3 5 (15.6%) 14 (41.2%)

Distant metastasis (M) NSM0 31 (96.9%) 30 (88.2%)M1 1 (3.1%) 4 (11.8%)

Clinical staging 0.005III 25 (78.1%) 13 (38.2%)IVa 6 (18.8%) 17 (50.0%)IVb 1 (3.1%) 4 (11.8%)

Mean survival (month; 95% CI) 118 (95, 141) 52 (35, 69) 0.0004a NS, no significance.

590

MET EXPRESSION IN NASOPHARYNGEAL CARCINOMA

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

CAGTTGTTTCGTTTTGGC-3� (nucleotides 1596–1616). The cDNA probefor human met was derived from the coding region containing exons 15 and 16(accession no. X54559.1) by PCR amplification using the primer pairs:forward primer, 5�-ATCTCAGAACGGTTCATGCC-3� (nucleotides 3239–3258) and reverse primer, 5�-CATCATTGTCCAACAAAGTC-3� (nucleotides3479–3498). PCR products were then separated on a 0.8% agarose gel andpurified using Qiaex II gel extraction kit (Qiagen, Inc., Valencia, CA; cata-logue no. 20021). Human �-actin cDNA was purchased from Ambion, Inc.(Austin, TX, catalogue no. 7424–10 �g). Total cellular RNA was prepared bya rapid extraction method using the TRIzol reagent according to the manufac-turer’s instructions (Life Technologies, Inc.) and subjected to electrophoresisthrough 1% formaldehyde-agarose gels followed by transfer to Hybond-XLnylon membranes (Amersham Pharmacia Biotech UK, Ltd., Little Chalfont,England). Blots were prehybridized and hybridized at 65°C using Rapid-Hybbuffer (Amersham Pharmacia Biotech UK, Ltd.) according to the manufactur-er’s instruction. The cDNA probes were labeled using the random primelabeling system Rediprime II (Amersham Pharmacia Biotech UK, Ltd.).Hybridization occurred in the present of [�-32P]dCTP-labeled cDNA probesfor hgf, met, or �-actin. The membrane was washed at room temperature with2 � SSC (150 mM sodium chloride and 15 mM sodium citrate), 0.05% SDS for40 min, and at 50°C in 0.1 � SSC, 0.1% SDS for 30 min. And then the membranewas exposed to Kodak X-Omat film using intensifying screens at �80°C for4–72 h.

Immunoprecipitation and Western Blot Analysis. The cells with 80%confluence were starved with serum-free DMEM supplemented by 0.1% BSAfor 12 h. Human recombinant HGF was then at the concentration of 100unit/ml 10 min before harvest. The cells were washed twice with ice-cold PBSand then lysed on ice for 20 min in a lysis buffer containing 50 mM Tris-HCl(pH 7.4), 1% NP40, 0.25% Na-deoxycholate, 150 mM sodium chloride, 1 mM

EDTA, 1 mM phenylmethylsulfonyl fluoride, 10 �g/ml aprotinin, 10 �g/mlleupeptin, and 1 mM sodium vanadate. The insoluble debris was removed bycentrifugation at 14,000 � g for 15 min in 4°C. A total of 300 �g/300 �l lysatefrom each cell line was incubated with 0.5 �l of antihuman Met antibodyovernight in 4°C, and the immunocomplexes were precipitated with proteinG-Sepharose (Amersham Pharmacia Biotech AB, Uppsala, Sweden). Theimmunoprecipitated proteins were washed three times with lysis buffer andseparated using precast 10% Tris-glycine acrylamide gels (Introgen and LivingScience; EC60752). The proteins were electrophoretically transferred from thegels to nitrocellulose membrane. The membranes were blocked with 3% nonfatdry milk in TTBS [10 mM Tris-HCl (pH 8), 150 mM sodium chloride, and0.05% Tween 20] for 1 h at room temperature and then incubated with primaryantibodies (anti-hMet or antiphosphotyrosine) diluted in 10 ml of TTBS-BSA (TTBS with 2% BSA) for 1 h at room temperature. After being washedwith TTBS, the membranes were incubated with secondary antibody diluted inTTBS-BSA and again washed with TTBS. An enhanced chemiluminescencekit (Pierce) was used to detect the protein. The membranes were finallyexposed to films from 30 s to 3 min.

MET Mutation Screening in NPC Cell Lines. All of the five NPC celllines were collected for mutation screening in exons 14 and 16–21 of METgene. The genomic DNA was prepared by using the TRIzol reagent accordingto the manufacturer’s instructions. Sequencing primers for these exons arelisted in Table 2. PCR was performed in a 50-�l reaction volume containing50 ng of DNA; 20 mM Tris-HCl (pH 8.4); 50 mM KCl; 1.5 mM Mg Cl2; 0.2 �M

each primer; 0.2 mM dATP, dGTP, dCTP, and dTTP each; and 2 units of TaqDNA polymerase (Life Technologies, Inc.). Amplification was carried out ina programmable thermal cycler (GeneAmp PCR system 9700; Perkin-Elmer)at the following settings: after a denaturation at 94°C for 5 min, samples wereamplified for 35 cycles at 94°C, 30 s; 55°C-58°C, 30 s; and 72°C, 45 s, witha final extension at 72°C for 10 min. After amplification, all of the PCR

products were subjected to purification using Microcon YM-100 column(Amicon and Millipore) following the manufacturer’s manual. Purified PCRproducts were subjected to direct DNA sequencing using ABI PRISM BigDyeTerminator cycle sequencing ready reaction kit (PE Applied Biosystems).

Statistical Analysis. Mann-Whitney U test was used to compare the meansof fluorescent intensity among different tissues. Spearman’s rank-sum correlationcoefficients were calculated to study the correlation between patient survival timeand fluorescent intensity of Met protein. Estimates of cumulative survival werecalculated using Kaplan-Meier method. Observed differences in survival wereanalyzed by Log-rank test. Multivariate analysis was performed by Cox’s propor-tional hazards regression model. Student’s t test was used to compare the differ-ence of age between the two groups. To test the association between Met expres-sion level and other clinicopathological characteristics, Fisher’s exact test wasperformed. All tests were two tailed. A P � 0.05 was defined as statisticallysignificant.

RESULTS

Expressions of Met and HGF Proteins in NPC Biopsy Tissues.With either the H&E-stained images alone or the Nomarski and H&Ecombinations, the tumor tissues, nontumor columnar and squamousepithelial tissues, and normal nonepithelial interstitial tissues could beeasily identified by standard histopathology criteria. Among these 66primary tumor tissues, 12 of them contained normal, adjacent, anduninvolved nasopharyngeal columnar epithelia, 6 cases containeduninvolved nasopharyngeal squamous epithelia, and 55 cases con-tained normal nonepithelial interstitial tissues surrounding the tumors.Specific fluorescent staining was obtained in all 66 cases. Met proteinwas expressed throughout the cytoplasm of most of the tumor cellsand epithelial cells (Figs. 1 and 2). Fluorescent intensity was detectedand evaluated in four kinds of tissues: primary tumor, columnarnasopharyngeal epithelium, squamous nasopharyngeal epithelium,and normal nonepithelial interstitial tissue. The results showed that thehighest expression level of Met protein was in nasopharyngeal co-lumnar epithelium, followed by NPC. However, the Met intensitylevel in NPC was significantly higher than squamous epithelium andnormal interstitial tissue. The marginal significance of the differencein Met expression level between squamous epithelium and interstitialtissue may be because less cases of squamous epithelial tissue werestudied. The Met protein expression levels among these four kinds oftissues were shown in Fig. 3. HGF/SF was detected mainly in theinterstitial tissues surrounding the tumor (Fig. 1). Only scattered spotsin the tumor tissues were stained with HGF/SF antibodies. The ex-pression pattern of Met/HGF in normal, adjacent nasopharyngealcolumnar epithelium was different compared with that of normalsquamous epithelium. As shown in Fig. 1, Met expression level waslower in the tumor cells compared with the expression level in thecolumnar epithelial cells, whereas it was higher in the neoplastic cellscompared with squamous epithelium in Fig. 2.

The Correlation of Met Protein Expression with Clinical Pa-rameters. The expression level of Met protein in NPC was signifi-cantly correlated with patients’ survival time (P � 0.009); the corre-lation coefficient was �0.319, indicating that higher levels of Metexpression was correlated with shorter survival time. A 4-fold tablemethod (49) was used to calculate the sensitivity and specificity of

Table 2 Primer sequences for MET mutation screening

Exons Forward primers (5�-3�) Reverse primers (5�-3�)

14 CCATGATAGCCGTCTTTA CAATGTCACAACCCACTG16 ATTAAATGTTACGCAGTGCTAAC GGTTGCAAACCACAAAAGTAT17 GTATTCACTGTTCCATAATGAAGT GATGGCTGGCTTACAGCTAGTT18 AACAGTAGATGCTTAGTTTATGCT AACAGATTCCTCCTTGTCACTT19 TTCTATTTCAGCCACGGGTAAT ATGAAAGTAAAAGAGGAGAAACTC20 TATGTATGGTCACATCTCTCAC CCTTTGAAGGCAGGCATTTCTGT21 ACAGAAATGCCTGCCTTCAAAGG CGGAGCGACACATTTTACGTTCAC

591

MET EXPRESSION IN NASOPHARYNGEAL CARCINOMA

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

each cutoff point of Met staining intensity in predicting patients’survival time, which was defined as short-term survival if the survivaltime was �36 months or long-term survival if the survival time was�36 months. The cutoff value of 94 was chosen to distinguish highand low expression, as the largest product of sensitivity multiplied byspecificity was obtained at this level. Table 3 showed the patientclassification using the cutoff value of 94 and survival time. Thesensitivity and specificity of Met expression in predicting survivaltime were 72 and 65%, respectively, in this classification. Separationof the patients into low Met expression group (n � 32) and high Metexpression group (n � 34) under this cutoff point was shown in Table1. Kaplan-Meier survival analysis showed that in low Met expressiongroups, the mean survival time was 118 months, 95% CI, 95–141.However, the mean survival time in high expression group was only52 months, 95% CI, 35–69 (Fig. 4). Log-rank test showed that thesurvival time was significantly different between these two groups(P � 0.0004); the low Met expression group had better survival,whereas the high Met expression group had worse survival. The

cumulative 5-year survival rate was 77.68% in low expression group,but it was only 38.24% in high expression group. The clinical stagewas significantly later in high Met expression group (P � 0.005). Nosignificant difference had been found in age, gender, histologicalstratification, T stage, N stage, or M stage between the two Metexpression groups. Besides Met expression level, N stage, M stage,and clinical stage were also significantly correlated with survival inKaplan-Meier analysis and Log-rank test (for N stage, P � 0.015; forM stage and clinical stage, P � 0.001). Therefore, the initial multi-variate model included Met expression level, N stage, M sage, clinicalstage, and their interaction terms (to control for the possible effect ofcovariance between those terms with significant correlation coeffi-cients). No interaction terms were significant, except for that of Mstage and clinical stage, which was incalculable because of theircomplete congruence. Therefore, M stage was removed from the finalmodel along with its interaction terms. The remaining interactionterms and the N stage were removed from the model using thestepwise backwards elimination method and a retainment threshold of

Fig. 1. A, H&E staining image of an NPC biopsytissue. Lymphocyte infiltration (L) is shown sur-rounding the tumor (T). B, C, and D, immunofluores-cent images from the adjacent serial section to thesection in image A. Met is stained with the C-28polyclonal rabbit anti-hMet antibody and secondarilyby FITC-conjugated donkey antirabbit antibody.HGF is stained by the monoclonal mouse anti-hHGFantibody and secondarily by rhodamine-conjugateddonkey antimouse antibody. In B, the tissue is excitedby laser line 488 for FITC (green). Met is expressedthroughout the tumor. This image was used for fur-ther intensity calculation. In C, the tissue is excited bylaser 543 for rhodamine-conjugated secondary anti-body (red). HGF is detected predominantly in theinterstitial tissue surrounding the tumor. D, mergedgreen/red (Met/HGF) image. E, H&E staining imageof another NPC case with tumor (T) and normalnasopharyngeal columnar epithelia (C). F, immun-ofluorescent image indicating Met (green) and HGF(red) from the adjacent section to the section in imageE. The normal columnar epithelia (C) express higherlevels of Met compared with that of tumor (T).

592

MET EXPRESSION IN NASOPHARYNGEAL CARCINOMA

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

P � 0.05. The resulting final model was shown in Table 4, indicatingthat the clinical staging and Met protein expression level were inde-pendent prognostic indicators for survival.

Expression of Met and HGF mRNA in NPC Cell Lines. Tofurther clarify the expression profiles of met and hgf in NPC, five NPCcell lines were used for the mRNA evaluation. All of the five NPC celllines expressed met mRNA but did not express hgf in Northern blotanalysis (Fig. 5). These results were consistent with the finding in theimmunofluorescent staining study of primary tumors, implying that anautocrine HGF-Met loop was not involved in the activation of the Metreceptor in NPC.

Response of Met Receptor to HGF Stimulation. The paracrineactivation mechanism of Met receptor in NPC was verified by exog-enous stimulation of HGF in NPC cells. In all of the five NPC celllines, Met tyrosine phosphorylation could be induced by exogenousHGF stimulation (Fig. 6). Although some extent of autophosphoryl-ation of Met protein occurred in CNE-2, it also responded to HGF asmore phosphorylated Met protein was detected after the HGF stimu-lation.

Fig. 4. Kaplan-Meier survival analysis showed that the cumulative 5-year survival ratewas 77.68% in the low Met protein expression group (n � 32), but it was only 38.24%in the high expression group (n � 34). The mean survival time in the low Met proteinexpression group was 118 months (95% CI, 95–141), but it was only 52 months (95% CI,35–69) in the high expression group.

Table 3 The 4-fold table data classified according to patients with short-term andlong-term survival and met expression levels

With this table, the sensitivity and specificity to predict poor survival in NPC patientson the cutoff point of 94 can be calculated as: sensitivity � 21/(21 � 8) � 72%;specificity � 24/(13 � 24) � 65%.

Met expression(fluorescent intensity)

Short-term survival(patient no.)

Long-term survival(patient no.)

�94 21 13�94 8 24

Fig. 2. A, H&E staining image of a tissue sectioncontaining nasopharyngeal squamous epithelium(S) and NPC in situ (T). B, immunofluorescentimage from the adjacent section to the section inimage A. Met is in green, and HGF is in red. Theexpression level of Met protein in the neoplasticcells is higher than in the nonmalignant squamousepithelial cells. The HGF expression is negligible inthis section.

Fig. 3. The means of fluorescent intensity of Met staining in normal nasopharyngealcolumnar epithelia, NPC, normal nasopharyngeal squamous epithelia, and normal non-epithelial interstitial tissues were 126.40 20.93, 102.28 32.03, 71.08 18.16, and56.19 14.75, respectively. Mann-Whitney U test showed: columnar epithelia versusNPC, P � 0.008; columnar epithelia versus squamous epithelia, P � 0.001; columnarepithelium versus nonepithelial interstitial tissue, P � 0.001; NPC versus squamousepithelia, P � 0.018; NPC versus nonepithelial interstitial tissue, P � 0.001; andsquamous epithelia versus nonepithelial tissues, P � 0.059. Columns, means; bars, SDs.

593

MET EXPRESSION IN NASOPHARYNGEAL CARCINOMA

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

Mutation Detection in MET Oncogene. The possibility of acti-vating mutation in MET oncogene was investigated by sequencing thesix exons of MET gene coding for juxtamembrane and tyrosine kinasedomains of the protein. No mutation of MET oncogene was found inall of the five NPC cell lines. However, polymorphisms were found inexons 20 and 21 in HONE-1. They were Asp1304Asp (GAC3 GAT)in exon 20 and Pro1382Pro (CCG 3 CCA) in exon 21.

DISCUSSION

Although NPC has remarkable radiosensitivity and chemosensitiv-ity, the 5-year survival rate in late-stage NPC patients is still �30%,especially in patients with bulky regional lymph node involvement (1,40). Distant metastasis of NPC is more frequent than most of otherhead and neck cancers (1), and it remains the most common reason fortreatment failure despite improvement in the treatment modality inrecent years by combining radiotherapy with chemotherapy (50).Efforts to find any effective biomarker to predict the prognosis of thepatients with late-stage NPC are therefore warranted to improvethe cure rate. In the present study, we explored for the first time thecorrelation between Met protein expression and survival time in 66cases of late-stage NPC investigated. We found that Met proteinexpression was inversely correlated with survival time. The patientswith higher Met protein expression had shorter survival time. The Metprotein expression level was also significantly correlated with theclinical staging, with more cases of stages IVa and IVb in high Metexpression group (Table 1). Met expression was not correlated withprimary tumor development in our study. Its correlations with the Nand M staging also did not reach statistical significance, which may beexplained by the reason that the sample size might not be large enoughto reach statistical significance, especially in the M staging, becauseonly five cases with distant metastasis were studied. Moreover, bothMet protein expression level and clinical staging were independentprognostic factors in multivariate analysis. Prospective clinical studyis necessary to confirm that Met protein is one of the reliable clinicalpredictors of outcome for individual patients with late-stage NPC. Itis likely that combination of Met with other biomarkers, such as

circulating VEGF (51) and EBV DNA (52), may improve both thesensitivity and specificity in predicting individual patient’s prognosis.

Mutations in MET oncogene have been found to be tumorigenic andable to induce metastasis by constitutively activating the Met tyrosinekinase activity. Missense mutations in exons 16–21 coding for thetyrosine kinase domain were discovered in hereditary and sporadichuman papillary renal carcinoma, ovarian cancer, and hepatocellularcarcinoma (17–23). Another missense mutation in exon 14 coding forthe juxtamembrane domain with a regulatory site was reported re-cently in primary gastric adenocarcinoma (24). In the present study,we screened exons 14 and 16–21 for MET mutation in five NPC celllines. We did not find any mutation in these exons; only polymor-phisms were found in exons 20 and 21 in HONE-1. Therefore,mutation in the exons of MET oncogene coding for the juxtamem-brane and the tyrosine kinase domains is not involved in the activationof Met receptor in NPC. Interestingly, we found some extent ofautophosphorylation of Met receptor in one NPC cell line: CNE-2.The mechanism of this constitutive activation is unclear. The possi-bility of the existence of novel activation mechanism(s) in this cellline is worthwhile for further investigation.

The role of HGF/SF in many tumors is as a stroma-derived factorthat promotes invasion and metastasis of cancer cells (53–55). But insome cancers, an autocrine HGF-Met loop is involved (56, 57). In thepresent study, HGF/SF was found to be expressed neither in NPCtissues nor in NPC cell lines. However, HGF/SF was abundantlydetected in the interstitial tissues surrounding NPC, in which lympho-cyte has been reported to be the predominant cell type (2). The strongphosphorylation response of Met protein in NPC cells to the exoge-nous HGF stimulation, which itself is an indicator of activation ofmultiple signal transduction pathways, implied that HGF acted as aparacrine factor in NPC development. The influence of the interstitialtissue, especially the lymphocytic infiltration in NPC tissues, is con-troversial. Although correlation of rapid tumor development and dis-tant metastasis with reduced lymphocyte infiltration has been reported(8), others found that lymphocytic infiltration was not statisticallyrelated to a better survival. (9) But a higher density of dendritic cellsand monocytes/macrophages in NPC tissues has been shown to cor-relate with good prognosis, fewer lymph node metastases, and fewerdistant metastases (8, 9). Further study should clarify the specific celltype that is secreting HGF/SF and its impact on the promotion anddevelopment of NPC.

The tissue origin of NPC is not very clear. Although NPC is mostlyfound in the regions covered by columnar nasopharyngeal epithelium,it possesses certain characteristics of both columnar and squamousepithelia (58, 59). Moreover, it has been reported that NPC can bemultifocal in origin (60). Our results showed that the expression levelof Met in NPC was lower than in columnar nasopharyngeal epitheliabut higher than in squamous epithelia. The variance in the expressionlevel of Met protein among normal epithelial tissues and NPC tissue

Fig. 5. Northern blot analysis was performed to evaluate the mRNA expressions of hgf(A) and met (B) in five NPC cell lines compared with positive control cell line S-114.�-actin expression was probed to monitor the equal mRNA sample loading.

Fig. 6. Phosphorylation of Met receptor in NPC cell lines under the stimulation ofHGF. The cells were starved under serum-free media overnight before HGF was added tothe media. Then the proteins were extracted for immunoprecipitation (IP) using anti-hMetantibody. The precipitated proteins were then assayed by Western blot (WB) analysesusing antiphosphotyrosine (anti-P-tyr) antibody and anti-hMet antibody.

Table 4 Analysis of 66 cases of NPC by Cox’s proportional hazards regression model

Factors � SE P

Clinical staging 1.3224 0.3443 0.0001Met expression level 0.7936 0.3926 0.0432

594

MET EXPRESSION IN NASOPHARYNGEAL CARCINOMA

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

may be in part attributable to different transcription activity. Theexploration to clarify the mechanism that shifts Met/HGF signalingpathway from differentiation in normal epithelium to the proliferationand the invasion in malignant cells would undoubtedly be helpful tofully understand the role of Met/HGF in NPC development.

Angiogenesis is a critical process in the development and metasta-sis of many solid tumors, including NPC (51, 61). We found previ-ously that serum VEGF level was elevated in the patients withmetastatic NPC (50), and vasculature density in primary NPC wascorrelated with distant metastasis (61). In vivo experiments showedthat antiangiogenic agents could not only suppress the tumor growthof NPC cell lines (62, 63) but also enhance the therapeutic effect ofthe conventional cytotoxic drug 5-fluorouracil, which is one of thefirst line drugs in chemotherapy of NPC (64). HGF/SF is a well-known up-regulator of angiogenesis via activating the Met/HGF sig-naling pathway (65). HGF was reported to act in concert with VEGFto promote human endothelial cell survival and tubulogenesis (66).Furthermore, in a recent study, HGF was found to be able to promoteexpression of angiogenesis factors interleukin-8 and VEGF in headand neck cancer cell lines through both mitogen-activated protein/extracellular signal-regulated kinase kinase- and phosphatidylinositol3�-kinase-dependent pathways (67). It will be worthwhile to studywhether blocking the Met/HGF pathway would benefit the tumorcontrol in NPC patients via inhibiting angiogenesis in addition toinhibiting tumor invasion and proliferation.

In summary, this study demonstrates that the expression level ofMet protein in NPC stained immunofluorescently and evaluated quan-titatively by confocal microscopy is inversely correlated with thesurvival time and directly correlated with clinical stage of patientswith late-stage NPC. Both clinical staging and Met expression levelare independent prognostic factors in the multivariate analysis. HGFacted as a paracrine factor in the activation of Met receptor in NPCcells. Met protein may be a potential biomarker to predict poorprognosis in NPC patients, and blocking Met/HGF signaling pathwaymay be beneficial in the treatment of NPC.

ACKNOWLEDGMENTS

We thank Bree Buckner and Eric A. Hudson, Laboratory of Analytical,Cellular, and Molecular Microscopy, VARI, for help in tissue staining andimage analysis. We also thank Dr. Yu-Wen Zhang, Laboratory of MolecularOncology, VARI, for help in developing cDNA probe for human hgf.

Note Added in Proof

After submission of the present manuscript, Horikawa et al. (68) published their work

in American Journal of Pathology reporting a close association of Met expression with

cervical lymph node metastasis in NPC.

REFERENCES

1. Fandi, A., Altun, M., Azli, N., Armand, J. P., and Cvitkovic, E. Nasopharyngealcancer: epidemiology, staging, and treatment. Semin. Oncol., 21: 382–397, 1994.

2. Vokes, E. E., Liebowitz, D. N., and Weichselbaum, R. R. Nasopharyngeal carcinoma.Lancet, 350: 1087–1091, 1997.

3. Gallimore, A. P. Nasopharyngeal carcinoma. Clin. Oncol., 7: 388–393, 1995.4. Chio, P. H., Sven, M. W., Huang, D. P., Lo, K. W., and Lee, J. C. Nasopharyngeal

carcinoma: genetic changes, Epstein-Barr virus infection, or both. A clinical andmolecular study of 36 patients. Cancer (Phila.), 72: 2873–2878, 1993.

5. Ali, H., and Al-Sarraf, M. Nasopharyngeal cancer. Hematol. Oncol. Clin. N. Am., 13:837–847, 1999.

6. Shanmugaratnam, K., and Sobin, L. H. Histological typing of tumors of the upperrespiratory tract and ear. In: World Health Organization International HistologicalClassification of Tumors, Ed. 2, pp. 32–33. Berlin: Springer, 1991.

7. Shanmugaratnam, K., Chan, S. H., de The, G., Goh, J. E. H., Khor, T. H., Simons,M. J., and Tye, C. Y. Histopathology of nasopharyngeal carcinoma: correlations withepidemiology, survival rates, and other biological characteristics. Cancer (Phila.), 44:1029–1044, 1979.

8. Zhong, Y. S., Zhang, C. Q., Zhang, F., Ruan, J. B., Chen, M. Y., Feng, K. T., and Yu,Z. F. Infiltrating lymphocytes and assessory cells in nasopharyngeal carcinoma. Jpn.J. Cancer Res., 84: 900–905, 1993.

9. Gallo, O., Bianchi, S., Giannini, A., Gallina, E., Libonati, G. A., and Fini-Storchi, O.Correlations between histopathological and biological findings in nasopharyngealcarcinoma and its prognostic significance. Laryngoscope, 101: 487–493, 1991.

10. Lee, A. W. M., Poon, Y. F., Foo, W., Law, S. C. K., Cheung, F. K., Chan, D. K. K.,Tung, S. Y., and Ho, J. H. C. Retrospective analysis of 5037 patients with nasopha-ryngeal carcinoma treated during 1976–1985: overall survival and patterns of failure.Int. J. Radiat. Oncol. Biol. Phys., 23: 261–270, 1992.

11. Heng, D. M., Wee, J., Fong, K. W., Lian, L. G., Sethi, V. K., Chua, E. T., Yang, T. L.,Khoo Tan, H. S., Lee, K. S., Lee, K. M., Tan, T., and Chua, E. J. Prognostic factorsin 677 patients in Singapore with nondisseminated nasopharyngeal carcinoma. Cancer(Phila.), 86: 1912–1920, 1999.

12. Bottaro, D. P., Rubin, J. S., Faletto, D. L., Chan, A. M., Kmiecik, T. E., VandeWoude, G. F., and Aaronson, S. A. Identification of the hepatocyte growth factorreceptor as the c-met proto-oncogene product. Science (Wash. DC), 251: 802–804,1991.

13. Cooper, C. S., Park, M., Blair, D. G., Tainsky, M. A., Huebner, K., Croce, C. M., andVande Woude, G. F. Molecular cloning of a new transforming gene from a chemi-cally transformed human cell line. Nature (Lond.), 311: 29–33, 1984.

14. Dean, M., Park, M., Le Beau, M. M., Robins, T. S., Diaz, M. O., Rowley, J. D., Blair,D. G., and Vande Woude, G. F. The human met oncogene is related to the tyrosinekinase oncogenes. Nature (Lond.), 318: 385–388, 1985.

15. Duh, F. M., Scherer, S. W., Tsui, L. C., Lerman, M. I., Zbar, B., and Schmidt, L. Genestructure of the human MET proto-oncogene. Oncogene, 15: 1583–1586, 1997.

16. Villa-Moruzzi, E., Puntoni, F., Bardelli, A., Vigna, E., De Rosa, S., and Comoglio,P. M. Protein tyrosine phosphatase PTP-S binds to the juxtamembrane region of thehepatocyte growth factor receptor Met. Biochem. J., 336: 235–239, 1998.

17. Jeffers, M., Schmidt, L., Nakaigawa, N., Webb, C. P., Weirich, G., Kishida, T., Zbar,B., and Vande Woude, G. F. Activating mutations for the met tyrosine kinase receptorin human cancer. Proc. Natl. Acad. Sci. USA, 94: 11445–11450, 1997.

18. Jeffers, M., Fiscella, M., Webb, C. P., Anver, M., Koochekpour, S., and VandeWoude, G. F. The mutationally activated Met receptor mediates motility and metas-tasis. Proc. Natl. Acad. Sci. USA, 95: 14417–14422, 1998.

19. Schmidt, L., Duh, F. M., Chen, F., Kishida, T., Glenn, G., Choyke, P., Scherer, S. W.,Zhuang, Z., Lubensky, I., Dean, M., Allikmets, R., Chidambaram, A., Bergerheim,U. R., Feltis, J. T., Casadevall, C., Zamarron, A., et al. Germline and somaticmutations in the tyrosine kinase domain of the MET proto-oncogene in papillary renalcarcinomas. Nat. Genet., 16: 68–73, 1997.

20. Schmidt, L., Junker, K., Weirich, G., Glenn, G., Choyke, P., Lubensky, I., Zhuang, Z.,Jeffers, M., Vande Woude, G., Neumann, H., Walther, M., Linehan, W. M., and Zbar,B. Two North American families with hereditary papillary renal carcinoma andidentical novel mutations in the MET proto-oncogene. Cancer Res., 58: 1719–1722,1998.

21. Schmidt, L., Junker, K., Nakaigawa, N., Kinjerski, T., Weirich, G., Miller, M.,Lubensky, I., Neumann, H. P., Brauch, H., Decker, J., Vocke, C., Brown, J. A.,Jenkins, R., Richard, S., Bergerheim, U., et al. Novel mutations of the MET proto-oncogene in papillary renal carcinomas. Oncogene, 18: 2343–2350, 1999.

22. Olivero, M., Valente, G., Bardelli, A., Longati, P., Ferrero, N., Cracco, C., Terrone,C., Rocca-Rossetti, S., Comoglio, P. M., and Di Renzo, M. F. Novel mutation in theATP-binding site of the MET oncogene tyrosine kinase in a HPRCC family. Int. J.Cancer, 82: 640–643, 1999.

23. Park, W. S., Dong, S. M., Kim, S. Y., Na, E. Y., Shin, M. S., Pi, J. H., Kim, B. J.,Bae, J. H., Hong, Y. K., Lee, K. S., Lee, S. H., Yoo, N. J., Jang, J. J., Pack, S., et al.Somatic mutations in the kinase domain of the Met/hepatocyte growth factor receptorgene in childhood hepatocellular carcinomas. Cancer Res., 59: 307–310, 1999.

24. Lee, J. H., Han, S. U., Cho, H., Jennings, B., Gerrard, B., Dean, M., Schmidt, L., Zbar,B., and Vande Woude, G. F. A novel germ line juxtamembrane met mutation inhuman gastric cancer. Oncogene, 19: 4947–4953, 2000.

25. Giordano, S., Ponzetto, C., Di Renzo, M. F., Cooper, C. S., and Comoglio, P. M.Tyrosine kinase receptor indistinguishable from the c-met protein. Nature (Lond.),339: 155–156, 1989.

26. Giordano, S., Di Renzo, M. F., Narsimhan, R. P., Cooper, C. S., Rosa, C., andComoglio, P. M. Biosynthesis of the protein encoded by the c-met proto-oncogene.Oncogene, 4: 1383–1388, 1989.

27. Ponzetto, C., Bardelli, A., Zhen, Z., Maina, F., dalla Zonca, P., Giordano, S.,Graziani, A., Panayotou, G., and Comoglio, P. M. A multifunctional docking sitemediates signaling and transformation by the hepatocyte growth factor/scatter factorreceptor family. Cell, 77: 261–271, 1994.

28. Stuart, K. A., Riordan, S. M., Lidder, S., Crostella, L., Williams, R., and Skouteris,G. G. Hepatocyte growth factor/scatter factor-induced intracellular signalling. Int. J.Exp. Pathol., 81: 17–30, 2000.

29. Stoker, M., Gherardi, E., Perryman, M., and Gray, J. Scatter factor is a fibroblast-derived modulator of epithelial cell mobility. Nature (Lond.), 327: 239–242, 1987.

30. Sonnenberg, E., Meyer, D., Weidner, K. M., and Birchmeier, C. Scatter factor/hepatocyte growth factor and its receptor, the c-met tyrosine kinase, can mediate asignal exchange between mesenchyme and epithelia during mouse development.J. Cell Biol., 123: 223–235, 1993.

31. Altstock, R. T., Stein, G. Y., Resau, J. H., and Tsarfaty, I. Algorithms for quantitationof protein expression variation in normal versus tumor tissue as a prognostic factor incancer: Met oncogene expression and breast cancer as a model. Cytometry, 41:155–165, 2000.

32. Tsarfaty, I., Alvord, W. G., Resau, J. H., Altstock, R. T., Lidereau, R., Bieche, I.,Bertrand, F., Horev, J., Klabansky, R. L., Keydar, I., and Vande Woude, G. F.

595

MET EXPRESSION IN NASOPHARYNGEAL CARCINOMA

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

Alteration of Met protooncogene product expression and prognosis in breast carci-nomas. Anal. Quant. Cytol. Histol., 21: 397–408, 1999.

33. Huntsman, D., Resau, J. H., Klineberg, E., and Auersperg, N. Comparison of c-metexpression in ovarian epithelial tumors and normal epithelia of the female reproduc-tive tract by quantitative laser scan microscopy. Am. J. Pathol., 155: 343–348, 1999.

34. Di Renzo, M. F., Olivero, M., Ferro, S., Prat, M., Bongarzone, I., Pilotti, S., Belfiore,A., Costantino, A., Vigneri, R., Pierotti, M. A., and Comoglio, P. M. Overexpressionof the c-MET/HGF receptor gene in human thyroid carcinomas. Oncogene, 7:2549–2553, 1992.

35. Ebert, M., Yokoyama, M., Friess, H., Buchler, M. W., and Korc, M. Coexpression ofthe c-met proto-oncogene and hepatocyte growth factor in human pancreatic cancer.Cancer Res., 54: 5775–5778, 1994.

36. Di Renzo, M. F., Narsimhan, R. P., Olivero, M., Bretti, S., Giordano, S., Medico, E.,Gaglia, P., Zara, P., and Comoglio, P. M. Expression of the MET/HGF receptor innormal and neoplastic human tissues. Oncogene, 6: 1997–2003, 1991.

37. Jung, W., Castren, E., Odenthal, M., Vande Woude, G. F., Ishii, T., Dienes, H. P.,Lindholm, D., and Schirmacher, P. Expression and functional interaction of hepato-cyte growth factor-scatter factor and its receptor c-met in mammalian brain. J. CellBiol., 126: 485–494, 1994.

38. Humphrey, P. A., Zhu, X., Zarnegar, R., Swanson, P. E., Ratliff, T. L., Vollmer, R. T.,and Day, M. L. Hepatocyte growth factor and its receptor (c-met) in prostaticcarcinoma. Am. J. Pathol., 147: 386–396, 1995.

39. Wagatsuma, S., Konno, R., Sato, S., and Yajima, A. Tumor angiogenesis, hepatocytegrowth factor, and c-met expression in endometrial carcinoma. Cancer (Phila.), 82:520–530, 1998.

40. Min, H., Hong, M., Ma, J., Zhang, E., Zheng, Q., Zhang, J., Zhang, J., Zhang, F., Su,Y., and Qiu, F. A new staging system for nasopharyngeal carcinoma in China. Int. J.Radiat. Oncol. Biol. Phys., 30: 1037–1042, 1994.

41. Teng, Z. P., Ooka, T., Huang, D. P., and Zeng, Y. Detection of Epstein-Barr virusDNA in well and poorly differentiated nasopharyngeal carcinoma cell lines. VirusGenes, 13: 53–60, 1996.

42. Yao, K. T., Zhang, H. Y., Zhu, H. C., Wang, F. X., Li, G. Y., Wen, D. S., Li, Y. P.,Tsai, C. H., and Glaser, R. Establishment and characterization of two epithelial tumorcell lines (HNE-1 and HONE-1) latently infected with Epstein-Barr virus and derivedfrom nasopharyngeal carcinomas. Int. J. Cancer, 45: 83–89, 1990.

43. Zhang, S. Z., Gao, X. K., and Zeng, Y. Cytogenetic studies on an epithelial cell linederived from poorly differentiated nasopharyngeal carcinoma. Int. J. Cancer, 31:587–590, 1983.

44. Huang, D. P., Ho, J. H., Poon, Y. F., Chew, E. C., Saw, D., Lui, M., Li, C. L., Mak,L. S., Lai, S. H., and Lau, W. H. Establishment of a cell line (NPC/HK1) from adifferentiated squamous carcinoma of the nasopharynx. Int. J. Cancer, 26: 127–132,1980.

45. Wu, Y. T., Wang, H. M., Yang, X. P., Li, M. Z., Chen, J., Fang, Y., Jian, S. W.,Zhang, L., Wu, Q. L., and Zhang, W. T. Establishment and characterization of axenografted nasopharyngeal carcinoma and corresponding epithelial cell line. Chin. J.Cancer, 14: 83–86, 1995.

46. Cao, B., Su, Y., Oskarsson, M., Zhao, P., Fisher, R. J., Wang, L. M., and VandeWoude, G. F. Neutralizing monoclonal antibodies to hepatocyte growth factor/scatterfactor (HGF/SF) display antitumor activity in animal models. Proc. Natl. Acad. Sci.USA, 98: 7443–7448, 2001.

47. Rong, S., Oskarsson, M., Faletto, D., Tsarfaty, I., Resau, J. H., Nakamura, T., Rosen,E., Hopkins, R. F., III, and Vande Woude, G. F. Tumorigenesis induced by coex-pression of human hepatocyte growth factor and the human met protooncogene leadsto high levels of expression of the ligand and receptor. Cell Growth Differ., 4:563–569, 1993.

48. Klineberg, E., Tsarfaty, I., Alvord, W. G., Sathyanarayana, B. K., and Resau, J. H.Correction and quantification of normal differentiation in human epithelium: appli-cation for Optimas 4.0 image analysis program. Cell Vis., 3: 402–406, 1996.

49. Shapiro, D. E. The interpretation of diagnostic tests. Stat. Methods Med. Res., 8:113–134, 1999.

50. Al-Sarraf, M., LeBlanc, M., Giri, P. G., Fu, K. K., Cooper, J., Vuong, T., Forastiere,A. A., Adams, G., Sakr, W. A., Schuller, D. E., and Ensley, J. F. Chemoradiotherapy

versus radiotherapy in patients with advanced nasopharyngeal cancer: Phase IIIrandomized Intergroup study 0099. J. Clin. Oncol., 16: 1310–1317, 1998.

51. Qian, C. N., Zhang, C. Q., Guo, X., Hong, M. H., Cao, S. M., Mai, W. Y., Min, H. Q.,and Zeng, Y. X. Elevation of serum vascular endothelial growth factor in malepatients with metastatic nasopharyngeal carcinoma. Cancer (Phila.), 88: 255–261,2000.

52. Lo, Y. M., Chan, A. T., Chan, L. Y., Leung, S. F., Lam, C. W., Huang, D. P., andJohnson, P. J. Molecular prognostication of nasopharyngeal carcinoma by quantita-tive analysis of circulating Epstein-Barr virus DNA. Cancer Res., 60: 6878–6881,2000.

53. Seslar, S. P., Nakamura, T., and Byers, S. W. Regulation of fibroblast hepatocytegrowth factor/scatter factor expression by human breast carcinoma cell lines andpeptide growth factors. Cancer Res., 53: 1233–1238, 1993.

54. Nakamura, T., Matsumoto, K., Kiritoshi, A., Tano, Y., and Nakamura, T. Inductionof hepatocyte growth factor in fibroblasts by tumor-derived factors affects invasivegrowth of tumor cells: in vitro analysis of tumor-stromal interactions. Cancer Res.,57: 3305–3313, 1997.

55. Jiang, W., Hiscox, S., Matsumoto, K., and Nakamura, T. Hepatocyte growth factor/scatter factor: its molecular, cellular, and clinical implications in cancer. Crit. Rev.Oncol. Hematol., 29: 209–248, 1999.

56. Rong, S., Jeffers, M., Resau, J. H., Tsarfaty, I., Oskarsson, M., and Vande Woude,G. F. Met expression and sarcoma tumorigenicity. Cancer Res., 53: 5355–5360, 1993.

57. Chattopadhyay, N., Butters, R. R., and Brown, E. M. Agonists of the retinoic acid-and retinoid X-receptors inhibit hepatocyte growth factor secretion and expression inU87 human astrocytoma cells. Brain Res. Mol. Brain Res., 87: 100–108, 2001.

58. Zong, Y. S., Lin, H., Choy, D. T., Sham, J. S., Wei, W., Chan, K. H., and Ng, M. H.Nasopharyngeal carcinoma and lymphoinfiltration. Oncology (Basel), 48: 290–296,1991.

59. Michaels, L., and Hyams, V. J. Undifferentiated carcinoma of the nasopharynx: alight and electron microscopical study. Clin. Otolaryngol., 2: 105–114, 1977.

60. Sham, J. S., Wei, W. I., Zong, Y. S., Choy, D., Guo, Y. Q., Luo, Y., Lin, Z. X., andNg, M. H. Detection of subclinical nasopharyngeal carcinoma by fibreoptic endos-copy and multiple biopsy. Lancet, 17: 371–374, 1990.

61. Qian, C. N., Min, H. Q., Liang, X. M., Zheng, S. S., and Lin, H. L. Primary study ofneovasculature correlating with metastatic nasopharyngeal carcinoma using computerimage analysis. J. Cancer Res. Clin. Oncol., 123: 645–651, 1997.

62. Qian, C. N., Min, H. Q., Lin, H. L., Hong, M. H., and Ye, Y. L. Primary study inexperimental antiangiogenic therapy of nasopharyngeal carcinoma with AGM-1470(TNP-470). J. Laryngol. Otol., 112: 849–853, 1998.

63. Qian, C. N., Min, H. Q., Lin, H. L., Feng, G. K., Ye, Y. L., Wang, L. G., and Kuang,Z. J. Anti-tumor effect of angiogenesis inhibitor TNP-470 on the human nasopha-ryngeal carcinoma cell line NPC/HK1. Oncology (Basel), 57: 36–41, 1999.

64. Qian, C. N., Min, H. Q., Lin, H. L., and Hong, M. H. Combination of angiogenesisinhibitor TNP-470 with cytotoxic drugs in experimental therapy of nasopharyngealcarcinoma. Ann. Otol. Rhinol. Laryngol., 109: 641–645, 2000.

65. Rosen, E. M., and Goldberg, I. D. Regulation of angiogenesis by scatter factor. EXS(Basel), 79: 193–208, 1997.

66. Xin, X., Yang, S., Ingle, G., Zlot, C., Rangell, L., Kowalski, J., Schwall, R., Ferrara,N., and Gerritsen, M. E. Hepatocyte growth factor enhances vascular endothelialgrowth factor-induced angiogenesis in vitro and in vivo. Am. J. Pathol., 158: 1111–1120, 2001.

67. Dong, G., Chen, Z., Li, Z. Y., Yeh, N. T., Bancroft, C. C., and Van Waes, C.Hepatocyte growth factor/scatter factor-induced activation of MEK and PI3K signalpathways contributes to expression of proangiogenic cytokines interleukin-8 andvascular endothelial growth factor in head and neck squamous cell carcinoma. CancerRes., 61: 5911–5918, 2001.

68. Horikawa, T., Sheen, T. S., Takeshita, H., Sato, H., Furukawa, M., and Yoshizaki, T.Induction of c-Met proto-oncogene by Epstein-Barr virus latent membrane protein-1and the correlation with cervical lymph node metastasis of nasopharyngeal carci-noma. Am. J. Pathol., 159: 27–33, 2001.

596

MET EXPRESSION IN NASOPHARYNGEAL CARCINOMA

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from

2002;62:589-596. Cancer Res   Chao-Nan Qian, Xiang Guo, Brian Cao, et al.   Patients with Late-stage Nasopharyngeal CarcinomaMet Protein Expression Level Correlates with Survival in

  Updated version

  http://cancerres.aacrjournals.org/content/62/2/589

Access the most recent version of this article at:

   

   

  Cited articles

  http://cancerres.aacrjournals.org/content/62/2/589.full.html#ref-list-1

This article cites 66 articles, 18 of which you can access for free at:

  Citing articles

  http://cancerres.aacrjournals.org/content/62/2/589.full.html#related-urls

This article has been cited by 24 HighWire-hosted articles. Access the articles at:

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  .permissions@aacr.orgDepartment at

To request permission to re-use all or part of this article, contact the AACR Publications

Research. on February 22, 2016. © 2002 American Association for Cancercancerres.aacrjournals.org Downloaded from