vascular endothelial growth factor is a marker of tumor ...or invasive scc of the head and neck who...

Vascular Endothelial Growth Factor Is a Marker of TumorInvasion and Metastasis in Squamous Cell Carcinomas ofthe Head and Neck1

Edward R. Sauter, Mark Nesbit,James C. Watson, Andres Klein-Szanto,Samuel Litwin, and Meenhard Herlyn2

The Wistar Institute, Philadelphia, Pennsylvania 19104 [E. R. S.,M. N., M. H.], and Fox Chase Cancer Center, Philadelphia,Pennsylvania 19111 [J. C. W., A. K-S., S. L.]

ABSTRACTAngiogenesis has been linked to increased metastasis

formation and decreased overall survival in patients withvarious tumors, including and neck squamous cell carcino-mas (HNSCC). Vascular endothelial growth factor (VEGF)is a key regulator of angiogenesis. In the present study, weevaluated VEGF expression and microvessel density (MVD),a quantitative means of angiogenesis, in both experimentaland clinical models of HNSCC. Analysis of VEGF RNAexpression in cell lines of keratinocyte origin [HNSCC, facialskin squamous cell carcinoma (SCC), and transformed butnontumorigenic keratinocytes] and normal skin keratino-cytes revealed two VEGF transcripts corresponding to pro-teins of 165 and 121 amino acids in length, with the tran-script for the 165-amino acid species predominating. Six ofeight SCC cell lines showed increased levels of one or bothtranscripts, and seven SCC cell lines and the transformedkeratinocyte cell line showed increased protein expression.We then evaluated VEGF protein expression in human headand neck specimens containing normal epithelium (n 5 10),dysplasia or carcinomain situ (CIS; n 5 15), early invasiveSCCs (n 5 9), advanced primary SCCs (n 5 10), lymphnode metastases (n 5 3), and s.c. tumors or cysts (n 5 7)formed in severe combined immunodeficient mice. IntenseVEGF staining was found in the majority of advanced pri-mary SCCs, lymph node metastases, and human SCCs insevere combined immunodeficient mice, whereas no dyspla-sia, CIS, or early SCCs showed intense immunostain. Ahighly significant increase (P5 0.0001) in VEGF expressionwas seen in the advanced SCCversusdysplasias and CISlesions, as was the difference between SCCversusnormal

epithelium from nonsmokers (P 5 0.01). VEGF expressionin advanced primary cancers was greater (P 5 0.002) and, inearly cancers, marginally greater (P 5 0.05) than adjacentnormal mucosa. MVD increased with the progression ofpreinvasive disease (P5 0.04). VEGF expression and MVD(both, P 5 0.003) were directly associated with tumor ag-gressiveness in experimental tumors. These findings suggesta role for VEGF in both clinical and experimental HNSCC.

INTRODUCTIONAngiogenesis is required for solid tumor growth (1) and

facilitates tumor progression and metastasis (2–4). Angiogene-sis is thought to be regulated by a number of growth factors,including bFGF,3 TGF-b (5), and VEGF. Whereas tumor-derived bFGF and TGF-b have both autocrine and paracrinefunctions, VEGF seems to be produced by human tumors solelyfor the stimulation of tumor vascularization. VEGF is a disul-fide-linked dimeric glycoprotein that increases blood vesselpermeability, endothelial cell growth, proliferation, migration,and differentiation (6). VEGF shares significant amino-acidsequence homology with platelet-derived growth factor (7).Four different transcripts that have been identified encode pro-teins of 206, 189, 165, and 121 amino acids through alternativesplicing (8). Only the 165- and 121-amino acid proteins aresecreted to a significant extent.

Although VEGF and MVD have documented importancein both the development and progression of tumors in multiplemodels (6), their roles in HNSCC carcinogenesis are not clear.In other tumor systems, it seems that the timing (earlyversuslate in tumor development and progression) of increased expres-sion of VEGF is tumor-specific. In the breast [in which in-creased VEGF expression is seen in ductal CIS (9)], in the ovary[in which increased expression has been identified in borderlinetumors (10)], and in the vulva [in which stage III vulval intra-epithelial neoplasias show increased expression (11)], VEGFseems to be up-regulated before tumor invasion. In other tumortypes, including colon carcinomas [in which increased VEGFexpression is associated with the metastatic phenotype (12)] andprostate cancer [in which enhanced expression is associatedwith invasive carcinoma (13)], VEGF up-regulation seems tooccur later in progression.

VEGF can be constitutively expressed by human tumors, orit can be induced by other growth factors and cytokines includ-Received 5/4/98; revised 9/2/98; accepted 10/8/98.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisementin accordance with 18 U.S.C. Section 1734 solely toindicate this fact.1 Supported in part by NIH Grants DE-00380, CA-25874, and CA-47159.2 To whom requests for reprints should be addressed, at The WistarInstitute, 3601 Spruce Street, Philadelphia, PA 19104. Phone: (215)898-3950; Fax: (215) 898-0980; E-mail: [email protected].

3 The abbreviations used are: bFGF, basic fibroblast growth factor;SCC, squamous cell carcinoma; HNSCC, head and neck SCC; VEGF,vascular endothelial growth factor; MVD, microvessel density; SCID,severe combined immunodeficiency disease; LNM, lymph node metas-tasis/metastases; CIS, carcinoma(s)in situ; TGF, transforming growthfactor; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

775Vol. 5, 775–782, April 1999 Clinical Cancer Research

Research. on November 8, 2020. © 1999 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

http://clincancerres.aacrjournals.org/

ing TGF-a, epidermal growth factor, and phorbol esters (14).Changes in environmental conditions such as episodes of hy-poxia seem to be significant for VEGF induction (15). Thus,expression of VEGF may depend on a variety of endogenousand exogenous factors that can vary greatly during tumor pro-gression. Studies in non-SCC lesions have demonstrated a directcorrelation between VEGF expression and MVD (16), whichsuggests that VEGF is the predominant growth factor for tumorvascularization. High MVD is associated with increasing tumoraggressiveness as demonstrated by Weidneret al. (2), whofound a nearly linear relationship between microvessel countsand metastatic potential. However, this close relationship be-tween MVD and metastatic potential has not been detecteduniversally in human tumors (17), nor do all human tumorsdemonstrate increased expression of VEGF. These differencesin observations may reflect: (a) high tumor vascularization indysplasia or biologically early malignancy, leading to insignif-icant increases in VEGF with progression; (b) inconsistency ofVEGF expression because hypoxia may occur only intermit-tently, with the tumors removed at times of low VEGF expres-sion; or (c) changes in the microenvironment that lead to de-creased expression of VEGF-inducing growth factors andcytokines.

In the present study, we evaluated VEGF expression andMVD in head and neck epithelial lesions containing cysts,dysplasia, CIS, early and advanced SCC (LNM) and in cell linesrepresenting human precancerous, primary and metastatic can-cer, and metastatic stages. VEGF expression generally increasedwith progression, and the most aggressive lesions showed thehighest expression levels. These findings suggest a role forVEGF in HNSCC that drives the progression of HNSCC towardan invasive and aggressive phenotype.

MATERIALS AND METHODSTissue and Cell Lines

Tissue from 10 adult nonsmokers with histologically nor-mal epithelium, as well as from 34 adult subjects with dysplasiaor invasive SCC of the head and neck who underwent surgicalexcision at the Fox Chase Cancer Center (Philadelphia), wasevaluated. Specimens included 9 moderate dysplasias, 6 CIS, 9stage I or II (early) SCCs, and 10 stage III or IV (advanced)invasive carcinomas. LNM from 3 of the 10 advanced invasivecarcinomas were also evaluated. None of the subjects withbiopsies demonstrating only normal mucosa or dysplasia re-ceived preoperative chemotherapy/radiation to the oral cavity.Two of nine subjects with Stage I/II tumors received preoper-ative chemo- and/or radiation therapy, and 1 of 10 subjects withStage III/IV tumors received preoperative chemo- and/or radi-ation therapy. In addition, nine cell lines derived from squamousepithelium (six HNSCC lines—FaDu, Det562, A253, SCC 4,SCC 9, SCC 25; two lines from SCCs of the facial skin—SCC12, SCC 13; and one spontaneously transformed line from skinepithelium— HaCaT) were also studied. Cell lines FaDu,Det562, and A253 were obtained from the American TypeCulture Collection (Rockville, MD); SCC 4, SCC 9, SCC 12,SCC 13, and SCC 25 were kindly provided by Dr. J. Rheinwald(Boston, MA); and HaCaT was a gift of N. Fusenig (Heidelberg,Germany). Normal human epidermal keratinocytes were iso-

lated from neonatal foreskins and grown in serum-free kerati-nocyte medium supplemented with 5 ng/ml epidermal growthfactor and 50mg/ml bovine pituitary extract (Keratinocyte-SFM, Life Technologies, Inc., Grand Island, NY). Keratinocyteswere obtained when 70–80% confluent using trypsin/EDTAand RNA, and protein extracts were made from passages 2–4.All of the tumor cell lines were grown in serum-free MCDB201/L15 (4:1) medium (Sigma, St. Louis, MO) supplementedwith 5 mg/ml insulin.

Experimental TumorsMonolayer cells at 70–80% confluence were trypsinized

and counted, and 23 106 cells in 50ml of culture medium wereinjected s.c. in the dorsum of SCID mice. Each group of fivemice was monitored twice weekly for the development of tu-mors. Mice were killed when tumors reached a minimum diam-eter of 5 mm, or if no tumor was observed at 100 days aftertumor cell injection. Tumors were excised, fixed overnight in10% neutral buffered formalin, and embedded in paraffin.

RNA Isolation and Northern Blot AnalysisRNA was isolated from epidermal keratinocytes and cell

lines using TRIzol reagent (Life Technologies, Inc.) accordingto the manufacturer’s instructions. Total RNA concentrationwas determined based on absorbance. Total RNA (15mg) fromeach specimen was loaded onto 1.2% agarose-formaldehydegels containing 10 ng/ml ethidium bromide, subjected to dena-turing electrophoresis, transferred to Hybond-N1nylon mem-branes (Amersham Life Science, Arlington Heights, IL), andprehybridized in 63SSC, 23Denhardt’s solution, and 0.5%SDS. Membranes were probed with an [a-32P]dATP-labeledfragment containing the entire open reading frame of the humanVEGF gene. After washing, blots were exposed to Kodak X-OMAT AR film with an intensifying screen at280°C.

Protein Isolation and Western Blot AnalysisCell cultures grown to 70–80% confluence were lysed

with an SDS-based single lysis buffer. The lysate was centri-fuged at 13,000 rpm, and the supernatant was collected. Totalprotein concentration of the soluble cell extract was determinedusing the Pierce BCA Protein Assay Reagent Kit (Rockford,IL). Total protein (100mg) of each sample was labeled in thepresence of SDS before the addition of 2-mercaptoethanol andapplied to a discontinuous 10% polyacrylamide gel for electro-phoretic separation. Proteins were then electrophoretically trans-ferred to a polyvinylidene difluoride membrane, which wasplaced in blocking solution (5% nonfat dry milk), probed withVEGF-specific antibody (1:200 mouse monoclonal Ab-3, Cal-biochem, Cambridge, MA) followed by a phosphatase-conju-gated goat antimouse IgG (1:1000; Jackson ImmunoResearch,West Grove, PA), and washed. The substrates 5-bromo-4-chloro-3-indolyl phosphate and nitroblue tetrazolium wereadded for signal detection.

Quantitation of Secreted VEGFThe concentration of VEGF protein in supernatants that

were collected 3 days after culture medium change of 1.03 105

immortalized cells was determined. Normal keratinocyte me-

776 VEGF Increases with SCC Progression



dium was not collected because it contains: (a) growth factorsnot present in the SCC medium that stimulates VEGF (18); and(b) a relatively high percentage of senescent keratinocyteswhose cell membranes are degrading, which results in the meas-urement of cellular VEGF in addition to secreted VEGF. AVEGF ELISA kit (R&D Systems, Minneapolis, MN) was usedto measure VEGF protein levels in the SCC supernatants ac-cording to the manufacturer’s instructions.

ImmunohistochemistryParaffin-embedded, formalin-fixed normal epithelium,

dysplasias, CIS, invasive SCCs, LNM from primary HNSCC,and tumors from SCC cell lines in SCID mice were cut in5-mm-thick sections and placed on poly-l-lysine-coated glassslides. After boiling with distilled water for 10 min, slides wereincubated with VEGF antibody Ab-3 (1:50). Reactions weredetected using the Super Sensitive detection system with thesubstrate Fast Red (Biogenex, San Ramon, CA); levamisole wasadded to the substrate to block endogenous alkaline phosphataseactivity. Skeletal muscle in the section served as a positivecontrol, and a corresponding tissue section without primaryantibody Ab-3 served as a negative control.

To determine MVD, slides were incubated sequentiallywith a mouse monoclonal antibody to Factor VIII-related anti-gen (Dako Corp., Carpinteria, CA), a horse antimouse IgG(1:200 dilution; Vector Labs, Burlingame, CA), and avidin-biotin-peroxidase complex (1:25 dilution; Vector Labs). Thechromagen 3939-diaminobenzidine was used to visualize anti-body binding, and microvessels were counted as describedbelow.

Each evaluation of VEGF and Factor VIII was performedsimultaneously and agreed upon by at least two of the investi-gators (E. R. S., J. C. W., and A. K-S.) using a double-headedlight microscope. VEGF staining was scored as 05 none; 15weak; 25 moderate; or 35 intense.

Image Cytometric Quantitation of Factor VIIIImmunostain

In normal epithelium and preinvasive lesions, the dermisbelow the epithelial lesion was evaluated. For invasive lesions,peritumoral vessels were counted. After the area of interest wasidentified, individual microvessels were counted in 3–5 (asavailable for any given lesion) noncontiguous, randomly se-lected fields (3200) using a computerized image analyzer(Roche Pathology Work Station; 0.104 mm2 per field). Anybrown-staining endothelial cell or endothelial cell cluster thatwas clearly separate from adjacent microvessels, tumor cells,and other connective-tissue elements was considered a single,countable microvessel. Vessel lumens, although usually present,were not necessary in defining a structure as a microvessel, andred cells were not used to define a vessel lumen. All of thecounts were performed simultaneously and in agreement by twoinvestigators (J. C. W. and A. K-S.) without knowledge of thepatient’s diagnosis. MVD was expressed as number of mi-crovessels per mm2. Because MVD is often heterogeneousamong tumors, average MVD was determined after the lowestand highest counts were omitted; this gave values more repre-sentative of the histology as a whole and improved reproduc-ibility.

Statistical AnalysisVEGF Expression in Clinical Specimens. Fisher’s ex-

act test (two-sided) was used to test the hypothesis that VEGFimmunoreactivity increased with increasing lesion severity.Specimens with Stage I/II cancer did not have different stainingintensity than specimens with normal histology, whereas StageIII/IV specimens stained more intensely for VEGF (P 5 0.01).There was not a significant difference identified between lesionswith moderate dysplasiaversusCIS, between moderate dyspla-sia versus Stage I/II invasive cancer, or between moderatedysplasia and CISversusStage I/II invasive cancer. The differ-ence between moderate dysplasia and CISversusinvasive can-cer (Stage I-IV) was marginally significant (P5 0.05), whereasspecimens with Stage III/IV cancer had significantly higherstaining intensity (P 5 0.0001) than specimens containing mod-erate dysplasia and CIS. VEGF levels in histologically normalmucosa adjacent to lesions did not differ from lesion levels forsamples taken for either moderate dysplasia or CIS. However,VEGF level in Stage I/II lesions marginally exceeded that inadjacent mucosa (P5 0.05) and significantly exceeded adjacentmucosa in Stage III/IV lesions (P 5 0.002).

VEGF Expression in Experimental Tumors. Cellswith VEGF intensity of 2 are less aggressive than those withVEGF intensity of 3. In mice with 2 million cell innocula,neither mouse with a VEGF intensity of 2 developed a tumorwithin the 100-day observation period. Each of the 4 mice witha 2-million-cell innocula with VEGF intensities of 3 developeda tumor within the observation period. We ranked the number ofdays until tumor could be observed. The VEGF 2 mice thusreceived ranks 5 and 6, whereas the VEGF 3 mice receivedranks 1, 2, 3, and 4. Thus, the rank sum of VEGF 2 mice in thisgroup is 51 6 5 11. On the hypothesis that the VEGF 2 mouseranks are chosen randomly from the numbers 1, 2, 3, 4, 5, and6, the chance of this observation or one more extreme (there arenone more extreme) is 1/15. This result is not statisticallysignificant in its own right but was combined with our resultsfrom 5 mice inoculated with 10 million cells.

As Table 2 shows, one VEGF 2 mouse in the 10-million-cell category did not develop a tumor within the 100-day period.The other mouse did develop a tumor at 89 days. All of the 3VEGF 3 mice in this group developed a tumor within 14 days ofinoculation. Again, ranks assigned to the VEGF 2 mice wouldbe 4 and 5. VEGF 3 mice are assigned ranks 1, 2, and 3. Thus,the rank sum of the VEGF 2 mice in this group is 41 5 5 9,again the most extreme that is possible for this group. Thechance of this rank sum, assuming random assignment, is 1/10.

An overall assessment of statistical significance is obtainedby summing the two rank sums from the two VEGF 2 groups.This gives an overall rank sum of 111 9 5 20, the highest itcould possibly be. There is only one way to arrive at so large arank-sum total, and its probability under the null hypothesis ofrandom assignment is 1/150, with a significant association (P 50.003; Fisher’s exact test) between VEGF level and tumoraggressiveness.

MVD. The Wilcoxon rank-sum test was used to deter-mine the significance of increased MVD comparing histologi-cally abnormalversus adjacent normal epithelium and thencomparing lesion subtypes (normal epithelium, dysplasias, early

777Clinical Cancer Research



SCCs, advanced SCCs, LNM, and tumors formed in SCIDmice). Lesions with CIS had a higher MVD (P 5 0.04) thanlesions with less severe dysplasia or than normal epithelium.Fisher’s exact test was used to evaluate the association betweenthe rate of formation of experimental tumors and MVD. Theassociation was significant (P 5 0.003).

RESULTSVEGF RNA and Protein Expression

Cell Lines. Northern blot analysis (Fig. 1) demonstrateda 2-fold or greater increase in VEGF RNA (compared withGAPDH) in six of the eight cells using normal keratinocytes asa control. Western blotting (Fig. 2) demonstrated a higher levelof VEGF protein in all of the SCC cell lines than in normalhuman keratinocytes. All of the cell lines had been cultured inserum-free medium with insulin as the only mitogen. The lowestprotein level was found in HaCaT cells, which are spontane-ously immortalized but nontumorigenic keratinocytes (19). AnELISA was performed on day-3 supernatants of each immortal-ized cell line, with VEGF concentrations ranging from 121 to1817 pg/ml. SCC 9 and HaCaT cells had the lowest VEGFsecretion levels, and Det 562 had the highest.

Normal Epithelium and Lesions from Clinical Speci-mens. Lesions in the progressive steps of SCC,i.e.,dysplasia,CIS, early SCC (stages I and II), advanced primary SCC (stagesof III and IV), and metastatic lesions in lymph nodes, as well asnormal adjacent epithelium, were tested for VEGF expression

by immunostaining (Table 1). In normal epithelium from non-smokers as well as in normal epithelium adjacent to dysplasiasand SCC, expression was weak or absent in all of the specimens.Median staining intensity in lesions was moderate in dysplasias,CIS, and early primary SCC, but intense in advanced primaryand metastatic SCC (Table 1; Fig. 3). The differences in VEGFstaining intensity for advanced SCCversusdysplasias and CISlesions were highly significant (P5 0.0001), as was the differ-ence between SCCversusnormal epithelium from nonsmokers(P 5 0.01). When VEGF expression was compared in lesionsadjacent to histologically normal mucosa, expression was sig-nificantly greater in the Stage III/IV tumors (P 5 0.002), mar-ginally greater in the Stage I/II tumors (P 5 0.05), and notdifferent in the preinvasive lesions. No significant differenceswere found between LNM and advanced SCC or tumors grownin SCID mice.

Tumors and Cysts in SCID Mice. Seven tumors orcysts generated in SCID mice showed moderate-to-intense cy-toplasmic VEGF staining, with five tumors intensely staining(Table 2). The two specimens with moderate intensity either didnot form a tumor (HaCaT cells) or required a larger inoculum toform a tumor (103 106 cells instead of 23 106 cells for SCC12). By contrast, the cell lines with the highest VEGF expres-sion (SCC 13, Det 562, A253, and FaDu) showed the most rapidtumor growth in SCID mice. Tumor formation was acceleratedwhen 103 106 instead of 23 106 cells were inoculated. Theassociation between VEGF expression and rapid tumor growthwas significant (P 5 0.003). Intensity of VEGF staining alsocorrelated for vessel formation (P 5 0.003). The highest meanMVD was seen for FaDu and A253, the two cell lines that had(a) the highest VEGF expression in Western analysis (Fig. 2);(b) intense expression in the developing s.c. tumor; and (c) theshortest latency period until tumors were visible after injecting2 3 106 cells (Table 2).

MVD and Invasive PotentialComparison of MVD in dysplasiaversusCIS revealed a

significantly increased MVD (P 5 0.04) in CIS (median MVD,173 vessels/mm2) compared with moderate-severe dysplasia(median MVD, 112 vessels/mm2) and normal epithelium (me-dian MVD, 88 vessels/mm2; Fig. 4). On the other hand, therewas no association between peritumoral MVD and disease pro-

Fig. 1 Northern blot of VEGF in normal (NL) keratinocytes (used as abaseline) and SCCs (SCC 4, SCC 9, SCC 12, SCC 13, SCC 25, A253,DET 562, andFaDu) compared with GAPDH control.

Fig. 2 Western blot analysis of VEGF in normal keratinocytes (NL),immortal but nontumorigenic keratinocytes (HaCaT), and SCCs (SCC 9,SCC 12,SCC 13,SCC 25,A253,Det 562, andFaDu).

Table 1 VEGF immunostaining intensity in head and neck lesionsa

Histology of lesionNo. of

specimens

Median (mean6 SD) stainingintensity

Normalepithelium Lesion

Normal epithelium 10 1.0 (1.46 0.5)Moderate dysplasia 9 1.0 (1.336 0.50) 2.0 (1.566 0.53)CIS 6 1.0 (0.836 0.75) 1.5 (1.506 0.55)Preinvasive 15 1.0 (1.136 0.64) 2.0 (1.536 0.52)Stages I and II 9 1.0 (1.336 0.58) 2.0 (1.866 0.54)Stages III and IV 10 1.0 (0.606 0.53) 3.0 (2.406 0.53)LNM 3 N/Ab 3.0 (2.676 0.58)

a Stain intensity scored as: 05 no stain; 15 weak; 25 moderate;3 5 intense.

b N/A, not applicable.




gression, with no significant difference in MVD of stage I andII lesions (median MVD 140, vessels/mm2) versusstage III andIV lesions (median MVD, 123 vessels/mm2; Fig. 4). Preinvasivelesions were not compared with invasive lesions because thelatter were evaluated for peritumoral vessels, regardless of theirrelationship to the epithelium. An example of MVD staining infour different specimens is indicated in Fig. 5.

DISCUSSIONTo evaluate a complete model of HNSCC carcinogenesis,

we analyzed VEGF expression and MVD in clinical specimensranging from normal epithelium of nonsmokers through prein-

vasive lesions, primary invasive and metastatic cancer. Ad-vanced lesions (stage III and IV SCC and SCC formed in SCIDmice) showed a significant increase in VEGF expression ascompared with early lesions (dysplasias to early invasive SCC)and normal mucosa, and compared with adjacent histologicallynormal mucosa. Northern and Western analyses of VEGF ex-pression in experimental HNSCC carcinogenesis revealedhigher VEGF expression in all of the HNSCC cell lines than innormal keratinocytes, and HNSCC cells implanted in SCID

Fig. 3 Immunohistochemicalstaining of VEGF.A, moderatedysplasia (staining intensityscore 5 0; no staining abovebackground);B, CIS (stainingintensity 5 2; moderate inten-sity); C, stage I SCC (stainingintensity5 2); andD, SCC xe-nograft (staining intensity5 3;intense staining).

Table 2 Association of VEGF with MVD and tumor aggressivenessin experimental tumors and cystsa from human squamous cell lines

Cell lineVEGF

intensityb

Days to visible tumor or cystMeanMVDd2 3 106 cellsc 10 3 106 cellsc

HaCaT 2 81 (cyst) 55 (cyst) 0.3SCC 12 2 no tumor 89 3.3SCC 13 3 89 N/Ae 6.7Det 562 3 89 14 6.7FaDu 3 25 14 28.0A253 3 74 14 49.3

a All of the cell lines formed tumors except HaCaT, which formeda cyst.

b Cellular staining compared with negative control: 05 negative;1 5 weak; 25 moderate; 35 intense.

c Number of cells injected at baseline.d Mean number of vessels counted in three separate 0.83-mm2

areas of tumor.e N/A, not available.

Fig. 4 MVD (vessels/mm2) in normal epithelium, preinvasive (Dys-plasiaandCIS), and invasive (Stages I-IV) human HNSCC.p, statisticalsignificance in MVD between CISversusnormal epithelium and dys-plasia.




mice formed tumors readily only in the lines with intense VEGFexpression. Increased VEGF transcript levels were detected infour of nine cell lines evaluated (SCC 13, A253, Det 562,FaDu). Tumors that formed from these four cell lines wheninjected into SCID mice all showed intense VEGF staining, andtumors formed after a low tumor cell inoculum in each case.Taken together, these data suggest that VEGF expression isup-regulated in both clinical and experimental HNSCC, and thatthe increased expression occurs primarily in association with theability of clinical tumors to metastasize and of experimentaltumors to readily form tumors in immunodeficient mice.

Angiogenesis can be measured in a variety of ways, in-cluding the measurement of MVD and the expression of growthfactors known to be important in stimulating angiogenesis,including bFGF, platelet-derived growth factor, and VEGF.Among the various growth factors thought to be important in theangiogenic process, VEGF is generally regarded as the mostlikely candidate for the induction of tumor growth because it isa soluble secreted endothelial cell mitogen and its receptor(KDR) is selectively expressed by activated endothelium (20). Inexperimental tumors, administration of anti-VEGF monoclonalantibody decreased the MVD and suppressed tumor growth(21).

Recent reports document the expression of VEGF both inHNSCC cell lines (22, 23) and in clinical specimens (24–27).What is not clear from these reports is the role of VEGF in eitherexperimental or clinical HNSCC carcinogenesis. In one of the

reports (27), subjects with normal mucosa, dysplasia, and inva-sive cancer were evaluated, but neither the degree of dysplasia(moderate, severe, or CIS) nor the stage of the invasive SCCwas clear. Tumor metastases were apparently not evaluated. Theauthors concluded that VEGF expression is increased in high-grade dysplasia and SCC compared with normal epithelium,although no statistical support was described. A separate reportusing a polyclonal antibody to VEGF (25) documents an asso-ciation of VEGF expression with LNM compared with non-metastatic SCC.

We considered the use of image analysis to quantitateVEGF immunostain but elected not to perform the procedure fora variety of reasons. First, image analysis has a number ofproblems that are difficult to circumvent. The problems include:(a) only a portion of a cell may be on a slide because of thethickness of the tissue section. This will affect the intensityscoring; (b) even if sections from tissue sections are cut at thesame setting on the same microtome, the thickness of the sec-tions can vary, potentially altering intensity scoring. This resultsin loss of precision in image analysis measurements that couldneutralize its perceived advantages over the semi-quantitativeimmunostain method; (c) many tumor cells are nonspherical andvary in size and shape, which limits the ability to quantitateimmunostain intensity (28); (d) staining heterogeneity—eithertrue or induced—during tissue fixation and processing requiresthe subjective selection of fields, leads to inter-observer varia-bility, and limits the reproducibility of image analysis results

Fig. 5 MVD staining patternin A, normal epithelim;B, CIS;C, stage I (early stage) SCC;and D, stage III (advanced)SCC.Arrows, examples of ves-sels.




(29); and (e) although image analysis is an accepted method todetermine the DNA content within a cell, quantitation of proteincontent using image analysis is less well accepted.

Although a minority of subjects received preoperative che-moradiation therapy, this treatment did not significantly influ-ence VEGF expression. VEGF staining intensity was scored as1 and 1, respectively, for the two Stage I/II specimens that hadreceived prior treatment. The median score for all of the StageI/II tumors was 1. VEGF intensity was 2 for the Stage III/IVspecimen that had received prior treatment, compared with amedian of 3 for the group.

MVD in invasive primary carcinomas has shown promiseas a reliable prognostic marker in a number of tumor systems,including both node-positive and node-negative breast cancer(30). A variety of markers have been used to measure MVD.One of the more common, Factor VIII-related antigen, is ex-pressed mainly on large blood vessels and also on the endothe-lial cells of lymph vessels (31). This staining of lymph vesselsmay be an advantage when evaluating HNSCC, given the pro-pensity of these tumors to metastasize through lymphatics. An-other marker, platelet/endothelial cell adhesion molecule, hasthe reported advantage of staining a higher number of smallblood vessels as compared with Factor VIII (32). However,neither of these two markers is specific for tumor-stimulatedblood vessels. An endothelial cell proliferation antigen, endog-lin, may prove more specific for tumor-stimulated blood vesselgrowth (33).

In our normal epithelium and preinvasive lesions, stromalvessels deep in the basement membrane below the preinvasivelesion were measured. On the other hand, peritumoral vesselswere counted in our tissue sections containing HNSCC. Thus,comparison of preinvasive with invasive lesions did not seempossible, and only normalversusdysplasiaversusCIS and onlyearly versusadvanced SCC were compared. MVD was higher(P 5 0.04) in CIS compared with dysplasia and normal mucosa(Fig. 4). Although there was no association between peritumoralMVD and disease stage, the effects of earlier treatment mayhave influenced MVD, especially in the advanced lesions. Ra-diation therapy, for example, is known to reduce the network ofsmall blood vessels in tissue, leading to ischemia, atrophy, andoccasionally necrosis (34). Unfortunately, complete clinical in-formation is not available for most of the advanced lesions, butwe do know that the preinvasive and early invasive lesions werenot treated with radiation or chemotherapy before resection.This is consistent with the findings of Inoueet al. (25) thatMVD is not a prognostic marker in oral SCC. On the other hand,we found a direct association between MVD and experimentalSCC tumor formation (Table 2). Two prior studies evaluatingexperimental tumors of different origin (20, 35) reported similarfindings.

We observed an increase in MVD in preinvasive lesionswithout a concomitant increase in VEGF, which was ob-served only much later in squamous cell carcinogenesis. It iswell known that the induction of angiogenesis frequentlyprecedes the formation of malignant tumors (36), whichsuggests that angiogenesis may be rate-limiting not only fortumor progression but also for the onset of malignancy.Skobeet al. (37), who analyzed the role of VEGF in bothinitiating and sustaining angiogenesis in malignant skin ker-

atinocytes, demonstrated that, during the first 2 weeks aftergrafting benign and malignant variants of HaCaT cells on acollagen matrix, both types of cells express similar levels ofVEGF mRNA. On the other hand, increased VEGF mRNAwas detected 2– 6 weeks after grafting in the more advancedlesions. The authors concluded that VEGF is crucial in sus-taining, but not initiating, angiogenesis by malignant squa-mous cells, and that other angiogenic factor(s) mediatesearly-stage tumor angiogenesis. Both our clinical and ourexperimental data support this hypothesis.

In summary, VEGF expression seems to be associated withprogression to a more aggressive phenotype both in experimen-tal and in clinical systems involving HNSCC, with clinicalevidence of lymphatic spread and with more rapid developmentof tumors in immunodeficient mice. Up-regulation of MVDseems to occur early in HNSCC carcinogenesis, although thisobservation awaits confirmation in a larger study.

ACKNOWLEDGMENTSWe thank the members of the Wistar Editorial Department for

assistance with preparing the article.

REFERENCES1. Folkman, J. What is the evidence that tumors are angiogenesisdependent? J. Natl. Cancer Inst.,82: 4–6, 1990.2. Weidner, N., Semple, J. P, Welch, W. R., and Folkman, J. Tumorangiogenesis and metastasis—correlation in invasive breast carcinoma.N. Engl. J. Med.,324: 1–8, 1991.3. Russel, D. S., and Rubinstein, L. J. Glioblastoma multiforme.In:D. S. Russel and L. J. Rubinstein (eds.), Pathology of Tumors of theNervous System, Ed. 5, pp. 219–232. London: Edward Arnold, 1989.4. Weinstat-Saslow, D., and Steeg, P. S. Angiogenesis and colonizationin the tumor metastatic process: basic and applied advances. FASEB J.,8: 401–407, 1994.5. Rifkin, D. B., Kojimba, S., Abe, M., and Harpel, J. G. TGF-b:structure, function and formation. Thromb. Haemostasis,70: 177–179,1993.6. Senger, D. R., van de Water, L., Brown, L. F., Nagy, J. A., Yeo,K-T., Yeo, T-K., Berse, B., Jackman, R. W., Dvorak, A. M., andDvorak, H. F. Vascular permeability factor (VPF, VEGF) in tumorbiology. Cancer Metastasis Rev.,12: 303–324, 1993.7. Keck, P. J., Hauser, S. D., Krivi, G., Sanzo, K., Warren, T., Feder, J.,and Connolly, D. T. Vascular permeability factor, an endothelial cellmitogen related to PDGF. Science (Washington, DC),246: 1309–1312,1989.8. Tischer, E., Mitchell, R., Hartman, T., Silva, M., Gospodarowicz, D.,Fiddes, J. C., and Abraham, J. A. The human gene for vascular endo-thelial growth factor. Multiple protein forms are encoded through alter-native exon splicing. J. Biol. Chem.,266: 11947–11954, 1991.9. Brown, L. F., Berse, B., Jackman, R. W., Tognazzi, K., Guide, A. J.,Dvorak, H. F., Senger, D. R., Connolly, J. L., and Schnitt, S. J.Expression of vascular permeability factor (vascular endothelial growthfactor) and its receptors in breast cancer. Hum. Pathol.,26: 86–91,1995.10. Abu-Jawdeh, G. M., Faix, J. D., Niloff, J., Tognazzi, K., Manseau,E., Dvorak, H. F., and Brown, L. F. Strong expression of vascularpermeability factor (vascular endothelial growth factor) and its receptorsin ovarian borderline and malignant neoplasms. Lab. Invest.,74: 1105–1115, 1996.11. Bancher-Todesca, D., Obermair, A., Bilgi, S., Kohlberger, P.,Kainz, C., Breienecker, G., Leodolter, S., and Gitsch, G. Angiogenesisin vulvar intraepithelial neoplasia. Gynecol. Oncol.,64: 496–500, 1997.12. Takahashi, Y., Kitadai, Y., Bucana, C. D., Cleary, K. R., and Ellis,L. M. Expression of vascular endothelial growth factor and its receptor,




KDR, correlates with vascularity, metastasis, and proliferation of humancolon cancer. Cancer Res.,55: 3964–3968, 1995.13. Ferrer, F. A., Miller, L. J., Andrawis, R. I., Kurtzman, S. H.,Albertsen, P. C., Laudone, V. P., and Kreutzer, D. L. Vascular endo-thelial growth factor (VEGF) expression in human prostate cancer: insitu and in vitro expression of VEGF by human prostate cancer cells.J. Urol.,157: 2329–2333, 1997.14. Detmar, M., Yeo, K.-T., Nagy, J. A., Van De Water, L., Brown,L. F., Berse, B., Elicker, B. M., Ledbetter, S., and Dvorak, H. F.Keratinocyte-derived vascular permeability factor (vascular endothelialgrowth factor) is a potent mitogen for dermal microvascular endothelialcells. J. Invest Dermatol.,105: 44–50, 1995.15. Shima, D. T., Deutsch, U., and D’Amore, P. A. Hypoxic inductionof vascular endothelial growth factor (VEGF) in human epithelial cellsis mediated by increases in mRNA stability. FEBS Lett.,370: 203–208,1995.16. Obermair, A., Kohlberger, P., Bancher-Todesca, D., Tempfer, C.,Sliutz, G., Leodolter, S., Reinthaller, A., Kainz, C., Breitenecker, G.,and Gitsch, G. Influence of microvessel density and vascular perme-ability factor/vascular endothelial growth factor expression on prognosisin vulvar cancer. Gynecol. Oncol.,63: 204–209, 1996.17. Westphal, J. R., van’t Hullenaar, R. G., van der Laak, J. A.,Cornelissen, I. M., Schalkwijk, L. J., van Muijen, G. N., Wesseling, P.,de Wilde, P. C., Ruiter, D. J., and de Waal, R. M. Vascular density inmelanoma xenografts correlates with vascular permeability factor ex-pression but not with metastatic potential. Br. J. Cancer,76: 561–570,1997.18. Frank, S., Hubner, G., Breier, G., Longaker, M. T., Greenhalgh,D. G., and Werner, S. Regulation of vascular endothelial growth factorexpression in cultured keratinocytes: implications for normal and im-paired wound healing. J. Biol. Chem.,270: 12607–12613, 1995.19. Breitkrevtz, D., Boukamp, P., Ryle, C. M., Stark, H-J., Roop, D. R.,and Fusenig, N. E. Epidermal morphogenesis and keratin expression inc-Ha-ras-transfected tumorigenic clones of the human HaCaT cell line.Cancer Res.,51: 4402–4409, 1991.20. Vries, C., Escobedo, J. A., Ueno, H., Houck, K., Ferrara, N., andWilliams, L. T. The fms-like tyrosine kinase, a receptor for vascularendothelial growth factor. Science (Washington, DC),255: 989–991,1992.21. Kim, K. J., Li, B., Winer, J., Armanini, M., Gillet, N., Phillips,H. S., and Ferrara, N. Inhibition of vascular endothelial growth factor-induced angiogenesis suppresses tumor growthin vivo. Nature (Lond.),362: 841–844, 1993.22. Benefield, J., Petruzzelli, G. J., Fowler, S., Taitz, A., Kalkanis, J.,and Young, M. R. Regulation of the steps of angiogenesis by humanhead and neck squamous cell carcinomas. Invasion Metastasis,16:291–301, 1996.23. Petruzzelli, G. J., Benefield, J., Taitz, A. D., Fowler, S., Kalkanis,J., Scobercea, S., West, D., and Young, M. R. Heparin-binding growthfactor(s) derived from head and neck squamous cell carcinomas induceendothelial cell proliferations. Head Neck,19: 576–582, 1997.

24. Salven, P., Heikkila, P., Anttonen, A., Kajanti, M., and Joensuu, H.Vascular endothelial growth factor in squamous cell head and neckcarcinoma: expression and prognostic significance. Mod. Pathol.,10:1128–1133, 1997.

25. Moriyama, M., Kumagai, S., Kawashiri, S., Kojima, K., Kakihara,K., and Yamamoto, E. Immunohistochemical study of tumor angiogen-esis in oral squamous cell carcinoma. Oral Oncol.,33: 369–374, 1997.

26. Eisma, R. J., Spiro, J. D., and Kreutzer, D. L. Vascular endothelialgrowth factor expression in head and neck squamous cell carcinoma.Am. J. Surg.,174: 513–517, 1997.

27. Denhart, B. C., Guidi, A. J., Tognazzi, K., Dvorak, H. F., andBrown, L. F. Vascular permeability factor/vascular endothelial growthfactor and its receptor in oral and laryngeal squamous cell carcinomaand dysplasia. Lab. Invest.,77: 659–664, 1997.

28. Bacus, J. W., and Bacus, J. V. A method of correcting DNA ploidymeasurements in tissue sections. Mod. Pathol.,7: 652–664, 1994.

29. Esteban, J. M., Kandalaft, P. L., Mehta, P., Odom-Maryon, T. L.,Bacus, S., and Battifora, H. Improvement of the quantification ofestrogen and progesterone receptors in paraffin-embedded tumors byimage analysis. Am. J. Clin. Pathol.,99: 32–38, 1993.

30. Weidner, N., Folkman, J., Pozza, F., Bevilacqua, P., Allred, E. N.,Moore, D. H., Meli, S., and Gasparini, G. Tumor angiogenesis: a newsignificant and independent prognostic indicator in early-stage breastcarcinoma. J. Natl. Cancer Inst.,84: 1875–1887, 1992.

31. Parums, D. V., Cordell, J. L., Micklem, K., Heryet, A. R., Galter,K. C., and Mason, D. Y. JC70: a new monoclonal antibody that detectsvascular endothelium associated antigen on routinely processed tissuesections. J. Clin. Pathol.,43: 752–757, 1990.

32. Horak, E. R., Leek, R., Klenk, N., LeJeune, S., Smith, K., Stuart, N.,Greenall, M., Stepniewska, K., and Harris, A. L. Angiogenesis, assessedby platelet/endothelial cell adhesion molecule antibodies, as indicator ofnode metastases and survival in breast cancer. Lancet,340: 1120–1124,1992.

33. Wang, J. M., Kumar, S., Pye, D., van Agthoren, A. J., Krupinski, J.,and Hunter, R. D. A monoclonal antibody detects heterogeneity invascular endothelium of tumours and normal tissues. Int. J. Cancer,54:363–370, 1993.

34. Hellman, S. Principles of radiation therapy.In: V. T. DeVita, S.Hellman, and S. A. Rosenberg (eds.), Cancer: Principles and Practice ofOncology, Ed. 4, pp. 248–275. Philadelphia: J. B. Lippincott, 1993.

35. Kondo, S., Asano, M., and Suzuki, H. Significance of vascularendothelial growth factor/vascular permeability factor for solid tumorgrowth, and its inhibition by the antibody. Biochem. Biophys. Res.Commun.,194: 1234–1241, 1993.

36. Hanahan, D., and Folkman, J. Patterns and emerging mechanisms ofthe angiogenic switch during tumorigenesis. Cell,86: 353–364, 1996.

37. Skobe, M., Rockwell, P., Goldstein, N., Vosseler, S., and Fusenig,N. E. Halting angiogenesis suppresses carcinoma cell invasion. Nat.Med., 3: 1222–1227, 1997.




1999;5:775-782. Clin Cancer Res Edward R. Sauter, Mark Nesbit, James C. Watson, et al. Head and NeckInvasion and Metastasis in Squamous Cell Carcinomas of the Vascular Endothelial Growth Factor Is a Marker of Tumor

Updated version

http://clincancerres.aacrjournals.org/content/5/4/775

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/5/4/775.full#ref-list-1

This article cites 35 articles, 7 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/5/4/775.full#related-urls

This article has been cited by 12 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

[email protected] at

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

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/5/4/775To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/content/5/4/775.full#ref-list-1

http://clincancerres.aacrjournals.org/content/5/4/775.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



vascular endothelial growth factor is a marker of tumor ...or invasive scc of the head and neck who...

Documents