Vol. 3, 697-709, December 1994 Cancer Epidemiology, Biomarkers & Prevention 697
Review
Biomarkers in Upper Aerodigestive Tract Tumorigenesis: A Review’
Dong M. Shin,2’3 Walter N. Hittelman, andWaun K. Hong
Departments of Thoracic/Head and Neck Medical Oncology ID. M. S.,
W. K. H.( and Clinical Investigation (W. N. H.J, The University of TexasM. D. Anderson Cancer Center, Houston, Texas 77030
Abstrad
Because therapeutic efforts such as surgery,radiotherapy, and chemotherapy have only marginallyimproved the 5-year survival rate from cancers of theupper aerodigestive trad (including head and neck andlung cancers) over the past 2 decades, chemopreventionhas become an important strategy in reducing the ratesof incidence and mortality of these cancers. However,chemoprevention trials have been hampered by seriousfeasibility problems; they require large numbers ofsubjeds and long-term follow-up for accuratedetermination of cancer incidence and they are verycostly. Because the use of intermediate end points wouldreduce the duration and costs of these studies,biomarkers that could serve as such end points haverecently become a subjed of great interest. With thestrengthening of the assumption that tumorigenesis is amultistep process of transformation from normal tissuesto malignant lesions, there has been a great effort toexamine each of these steps for genetic and/orphenotypic alterations that might be candidates for suchbiomarkers. These candidates include genomic markers,certain specific gene alterations, such as tumorsuppressor genes, oncogenes, growth fadors and theirreceptors, proliferation markers, and differentiationmarkers. In this review, we describe several genomicmarkers, including micronuclei, chromosomalalterations, and specific genetic markers, e.g., the rasgene family, erb Bi, int-2/hsf-1 , and p.5.3 tumorsuppressor gene. We also review the proliferationmarkers, including proliferating cell nuclear antigen, andsquamous cell differentiation markers, includingkeratins, involucrin, and transglutaminase 1 . Thesebiomarker candidates have the potential to be importantadjuncts to the development of new chemopreventiveagents and to the rational design of future interventiontrials. However, we can not overemphasize that thesemarkers need to be validated in clinical trials; only then
Received 3/8/94; revised 7/7/94; accepted 7/7/94.
1 Supported in part by NIH Grant CA-52501 and National Cancer Institute
Core Grant CA-16672.2 Recipient of the American Cancer Society Clinical Oncology CareerDevelopment Award (ACS-CDA-91 -271).1 To whom requests for reprints should be addressed, at The University of
Texas M. D. Anderson Cancer Center, Department of Thoracic/Head andNeck Medical Oncology, Box 80, 1515 Holcombe Boulevard, Houston, TX
77030.
can they replace cancer incidence as the sole end pointfor chemoprevention trials.
Introduction
Epithelial cancers of the upper aerodigestive tract are an
increasingly important public health problem throughout
the world. Despite improvements in surgery, radiotherapy,
and chemotherapy over the past 2 decades, the 5-yearsurvival rates for head and neck and lung cancers have
improved only marginally. New research directions are
cleanly needed. Interest has been renewed in chemopre-vention as a means of reducing the incidence and mortality
ofthese cancers (1-3). Unfortunately, chemoprevention ap-
proaches have been hampered by serious feasibility prob-
lems associated with the conduct of randomized Phase Ill
trials. Investigators have been forced to rely on cancerincidence as the study end point for determining preventive
efficacy. This requires that chemoprevention trials have
large numbers of subjects and long-term follow-up, mea-
sures which make the trials very costly. Therefore, there has
recently been a great surge of interest in defining “biornar-kers” associated with specific stages of the carcinogenic
process, with the goal of using these markers as “ interme-
diate end points” in chemoprevention trials (4, 5). Thisconcept has been well described by Zelen (6), who stated
that intermediate end points, such as biomarkers, wouldmake prevention trials feasible.
Interest in systemic treatments that would prevent can-
cer in the aerodigestive tract, springs from the understand-
ing that the whole epithelial lining of the tract shares both
common carcinogenic exposure and the resulting increasein cancer risk. This entire epithelium was first described as
mucosa “condemned” by carcinogens by Slaughter et a!. (7)in 1953. They coined the term “field cancenization” to
describe the diffuse histologic abnormalities and multifocal
nature of squamous cell carcinomas of the head and neck.
In the case of aerodigestive tract cancers, this hypothesis is
supported clinically by the frequent association of tumorswith prernalignant lesions in the same field, leg., oral eu-
koplakia (8) or bronchial metaplasia and/or dysplasia (9)]
and by the synchronous or metachnonous development ofmultiple primary tumors (iO).
Another concept of carcinogenesis in the aerodigestivetract is that of the multistep tumonigenic process (i 1 ). The
driving force behind this process is thought to be geneticdamage caused by continuous exposure to carcinogens, asevinced by an increased frequency of rnicronuclei in high-risk tissue and in premalignant lesions (12). Eventually,
these genetic alterations give rise to phenotypic changes inthe tissues, such as dysregulation of cell proliferation, dif-ferentiation, and cell loss pathways. These phenotypic al-terations can be driven by alterations of certain specificgenes such as tumor suppressor genes (e.g., p53 and rb),oncogenes (e.g., ras and myc), or the genes for growth
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Normal Cell Premalignant Cell
F
Prollfsratlon
Mrksrs?
Tumor Sup,.sser
1’4.-,1’4_P�Gsnomlc Difl#{149}rsntl#{149}tloii
Madcrs? MrIcrs?
Oncog.n.s/Growth Factors/
or R.c.ptors?
Fig. 1. Illustration ofbiomarker candidates during the premalignant stage of
upper aerodigestive tract carcinogenesis. The driving force behind the car-cinogenesis process is thought to be genetic damage caused by exposure tocarcinogens, such as accumulation of micronuclei or other genomic mark-ers. The alterations of genomic markers can be driven by certain specificgene alterations, such as oncogenes, tumor suppressor genes, and growth
factors and their receptors. These genetic markers or specific gene alterationsmay eventually result in changes in phenotypic markers, including celldifferentiation and/or proliferation.
factors and their receptors (e.g., EGF,4 EGFR, and transform-ing growth factor a). This concept is illustrated in Fig. 1.
These genetic alterations and phenotypic consequencescan be visualized as histologic transitions from normal ep-ithelium to hyperplasia to metaplasia/dysplasia and then tofrank malignancy.
In this review, we describe many of the genotypic and
phenotypic biomanker candidates in detail; any of thesemay be used as intermediate end points in chemopreven-
tion trials in the near future.
Biomarker Definition and Seledion
The process of developing biomarkens from the labora-tory to clinical application in humans requires the appli-cation of new insights to existing principles. The domainof biomarker identification falls primarily to basic scien-tists such as tumor biologists who are, in seeking to apply
markers for early detection of cancer, faced with threefundamental issues (1 3): (a) to provide a clear definitionof the end point for which the putative index is a marker;(b) to identify the type of clinical specimen from whichthe marker can be measured; and (c) to establish an
expected (i.e., normal or background) range of markervariability. For upper aerodigestive tract, on the otherhand, Lippman et a!. (4) proposed the following criteria
for biomarkens in tobacco-related epithelial carcinogen-esis: (a) that their expression in normal tissue be differentfrom that in high-risk tissue; (b) that they can be detected
even in small tissue specimens; (c) that they are ex-pressed in a quantity or pattern that can be correlatedwith the stage of carcinogenesis; and (d) that preclinicalor early clinical data indicate that the condition nepre-
sented by the marker can be modulated by study agents.A paradigm of carcinogenesis presupposes a multistage
model featuring genetic or phenotypic markers for eachstage. Tumor development is characterized by stepwise
4 The abbreviations used are: EGF, epidermal growth factor; EGFR, epider-mal growth factor receptor; TGF, transforming growth factor; ISH, in situ
hybridization; DMBA, 7,1 2-dimethylbenz(a)anthracene; TFR, transferrin re-
ceptor; PCNA, proliferating cell nuclear antigen; NSCLC, non-small cell lungcancer; SCLC, small cell lung cancer; BrdUrd, bromodeoxyuridine.
698 Review: Biomarkers in Upper Aerodigestive Tract Tumorigenesis
Malignant Cell genetic changes from a normal to a malignant cell. Elegantmodels have been worked out by many investigators to
define the multistep nature of the carcinogenesis process(i 4-1 6). In these models, cellular or genetic alterations mayprecede the occurrence of invasive malignancy and, if theydo, can be used as intermediate end points of carcinogen-
esis. Biomarkers of these intermediate end points, then, canbe defined as measurable markers of cellular or molecular
genetic events associated with specific stages of cancerdevelopment. This definition indicates that the risk of can-
cinogenic transformation correlates with the quantitativedegree and pattern of biomanker expression. Because thesearch for biomarkers useful in chemoprevention is just
beginning, no individual biomarker or pattern of biornank-ens has so far been validated through prevention trials as aconclusive predictor of risk.
To be a candidate biomarker in tumonigenesis, themarker should be altered not only in the malignant stages
but also in the premalignant stages of cancer development.In this sense, activation of oncogenes and inactivation of
tumor suppressor genes may emerge as markers becausethey change in the early stage of tumonigenesis in certaintumors although they may occur at the later stages of tumorprogression in different systems (1 7). Activation of certain
oncogenes (i.e., ras, myc, neu, jun. raf, on los) has beenassociated with the development of lung cancer (1 7). How-ever, the activation sequence of these genes, which woulddefine which oncogenes are involved with the earliest stageof carcinogenesis, has not been well established. Other
genes in this category include k-ras, c-myc, and certainspecific chromosome regions. K-ras has been detected fre-
quently in bronchial adenocancinoma tumor tissue (1 8, 19)and has been associated with shortened survival in both
early (20) and advanced stages of disease (21). Overexpres-sion of the c-myc gene has been associated with growthdysnegulation and loss of terminal differentiation in squa-mous cell (22, 23) and small cell tumors (24-26).
Losses of transcription factors or tumor suppressorgenes may become markers of human head and neck and
lung cancers. Probes that detect allelic deletion of specific
chromosomal regions by restriction fragment length poly-morphisms have frequently found loss of heterozygosity
(expression ofonly a single allele) in SCLC on chromosomes3p (1 00%), 1 3q (91 %), and 1 7p (1 00%) (27). The kanyo-types in NSCLC are very complex, but recurrent losses of1 7p, 3p, and 1 1 p (in 67%, 57%, and 48% of cases, respec-tively) suggest that these regions are “hot spots” for geneticalteration (28). Other candidate regions with breakpointsindicating that they are recessive oncogenes include 1 q, 3q,5p, 7p, 16q24, and 21p (29, 30).
Oncogene activation may alter the metabolic balanceamong cell growth, differentiation, and cell loss (1 7). Bycod ing “transcri ption” protei n, oncogenes activate otherkey genes to code for growth deregulation. The shift in the
balance of cell differentiation to growth marks the selectiveclonal expansion that is characteristic of tumor prolifera-
tion. Two critical “early response” transcription factors, fosand fun, seem to be activated whenever mammalian cells
respond to peptide growth factors (31). Bombesin, for ex-ample, a peptide growth factor released by pulmonary neu-
roendocnine cells, has been shown to induce growth andmaturation of human fetal lung tissue in organ cultures (32).
A functional membrane-associated bombesin receptor hasrecently been isolated from human SCLC carcinoma (NCI-
H345) cells (33). This peptide has also been found in the
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 699
Table 1 Classes of biomarker candidates in upper aerodigestive
tract tumors
Genomic markers (general)
Nuclear aberrations (e.g., micronuclei(
DNA content and flow cytometry
Chromosomal alterations
Specific genetic markers
Oncogene alterations Eras family, myc family, erb family
(erbBl, erbB2/I-Ier-2/neu)J
Src family (src, Ick), retinoic acid receptors
Tumor suppressor genes p53, retinoblastoma gene (rb),
3p (unidentified)J
Proliferation markers
Mitotic frequency (MPM-2(
Thymidine labeling index
Nuclear antigens (e.g., PCNA, DNA-polymerase a, Ki-67(
Polyamines, ornithine decarboxylase
Differentiation markers
Cytokeratins
Transglutaminase type I
Involucrin
Cell loss markers (apoptosis)
In situ end labeling of fragmented DNA
Bcl-2 expression
bronchial lavage fluid of asymptomatic chronic heavysmokers (34), suggesting that activation of certain onco-
genes may occur by autocrine loop stimulation. Thus,markers of growth factor expression, insofar as they reflect
activation of certain oncogenes, may hold promise for theearly detection of lung cancer or premalignant lesions.
Since carcinogenesis begins with certain specific genealterations and/on general genomic alterations that causephenotypic changes such as dysnegulation of proliferation
and cell differentiation, we place biomarkers with potentialutility in chemoprevention in five general classes (Table 1):
(a) general genomic markers; (b) specific genetic markers;
(c� proliferation markers; (d) differentiation markers; and (e)cell-loss markers. These biomarkers can reflect relativelyearly and site-specific carcinogenic changes; they also rep-resent changes that are not site specific or that occur overextended periods of time.
General Genomic Markers
Micronuclei. Micronuclei are chromosome or chromatidfragments formed in proliferating cells during clastogenicevents such as DNA damage caused by carcinogens. Thefrequency of micronuclei was widely studied as a genomic
marker in earlier human chemoprevention trials (35-40).
Because micronuclei are easy to find and quantify, they arestill the most studied of all potential biomarkers of interme-
diate end points. The micronuclei in the aerodigestive tractepithelium are formed in the proliferating basal cell layer,which gives rise to suprabasal cells that migrate to the
epithelial surface and can eventually be detected in easilyobtained exfoliated cells. The presence and frequency ofmicnonuclei in tissue are believed to be quantitative reflec-tions of ongoing DNA damage or genetic instability. A
series of pilot trials has shown that high micnonuclei fre-quency correlates with condemned-tissue cancer risk in
individuals (e.g., smokers) who are at high risk for head and
neck, lung, esophageal, or bladder carcinomas (36-39).
Because micronuclei frequency fits all four of the se-
lection criteria already described, it seems probable thatmicronuclei would be a useful intermediate end point bi-
omanken in trials of upper aenodigestive cancer chemopre-vention. However, a critical analysis of the extensive data
on this marker reveals problems with its use as the onlymarker (38). Stich et a!. (37, 39-41) studied chemopreven-
tion in high-risk groups from India and the Philippines whohave remarkably intense and long-term carcinogenic expo-
sure to betel nuts, uniform lifestyles and dietary habits, and
consistently elevated oral micnonuclei frequencies. These
investigators reported that retinol and beta-carotene eachsuppressed micronuclei frequency in more than 90% of
lesions after treatment; however, both the rates of clinical
response to these agents and the rates of suppression of new
lesions differed greatly (39-41). Therefore, this biomarkerby itself may not be useful in screening or monitoring for
active chemopreventive drugs. Another barge placebo-con-trolled chemoprevention trial was conducted in subjects at
high risk for esophageal cancer in Linxian County, People’sRepublic of China (42). The investigators observed signifi-
cant site-specific suppression of micronuclei (esophageal,not buccal) in subjects receiving chernopreventive agents
(netinol, riboflavin, and zinc) (43). Again, the marker was
not validated, since there was no significant reduction in
the number of premalignant esophageal lesions after 1 yearof chemopreventive intervention.
Despite these problems, studies of rnicronuclei as a
biomarker have contributed to the early development ofnatural compounds, i.e., retinoids and carotenoids, as po-
tential chemopreventive agents. Micronuclei frequency,however, indicates only an ongoing process of chromo-somal damage, not accumulated genetic damage. There-
fore, we are studying more specific alterations of geneticand phenotypic markers resulting from DNA damage bycarcinogens.
Chromosomal Alterations in Tumorigenesis. Head andneck cancer provides a unique model system for the study
of tumorigenesis and the development of biomarkens for
several reasons. First, head and neck cancer probably rep-resents a field cancenization process; the whole aerodiges-
tive tract epithelium is repeatedly exposed to carcinogenicinsult (e.g., tobacco and/or alcohol), placing the entire fieldat risk for tumor development (7, 44). The clinical evidencefor this “field hypothesis” is the high frequency of multiple
primary neoplasms in the aerodigestive tract and a higher-than-expected risk of synchronous and metachnonous sec-ond primary tumors (1 0, 1 1 ). Second, head and neck canceris thought to represent a multistep tumonigenesis processwhereby a series of events may occur prior to tumor devel-
opment (45). This is evinced by the presence of premalig-nant lesions adjacent to the tumor (46). Although these
clinical and histologic findings support the notion of fieldcancenization and multistep tumonigenesis in the head and
neck region, biomankers for these processes are lacking.Therefore, we studied the genetic changes in the tissue at
risk by using a multistep tumonigenesis model.Although a variety of cytogenetic changes have been
described for head and neck and lung cancers (46-48), theability to develop a comprehensive list of specific chromo-somal changes has been limited by impediments commonto solid tumor cytogenetic studies, that is, the low frequency
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
700 Review: Biomarkers in Upper Aerodigestive Trad Tumorigenesis
of mitotic figures from direct preparations, suboptimal chro-
mosomal preparations, and multiple complexity of cytoge-netic changes (49). Identification of karyotypic changes inpremalignant lesions by conventional cytogenetic proce-
dunes is technically even more difficult than in tumors andhas seldom been reported (50-52). Moreover, the spatial
cellular distribution of genetic changes in premalignant andmalignant lesions can not be defined by conventional cy-
togenetic techniques, since the tissue is disaggregated in thesingle cell preparations. Recently, ISH techniques have
been developed that allow the direct detection of chromo-somal abnormalities in interphase cells (53-57). Thismethod has been applied to many types of solid tumors byusing tumor cell lines or dissected tumor material (58-61).
More recently, ISH was adapted for use on formalin-fixed,paraffin-embedded tissue sections by using nonisotopic,
chnomosome-specific DNA probes, and enzyme-mediated(e.g., peroxidase) immunohistochemical procedures (62-66). This technique allows direct visualization of chromo-some changes in normal tissue, premalignant lesions, andtumor tissues with preservation of tissue architecture.
To visualize the accumulation of genetic alterationsduring head and neck tumonigenesis, and to determine the
extent of the genetically altered field, we probed 25 squa-mous cell carcinomas of the head and neck and their ad-
jacent premalignant lesions for numerical chromosome ab-
errations by nonisotopic ISH by using chrornosome-specificcentrornenic DNA probes for chromosomes 7 and 1 7. Nor-mal (control) oral epithelium from cancer-free nonsmokers
showed to chromosome polysomy (i.e., cells with three ormore chromosome copies), whereas histologically normalepithelium adjacent to tumors showed squamous cells with
polysornies of chromosomes 7 and 1 7 (67). Moreover, thefrequency of cells with polysomy increased as the tissuespassed from histologically normal epithelium to hyperplasiato dysplasia to cancer. These chromosomal abnormalities inhistologically normal and precancerous regions adjacent tothe tumors support the concept of field cancenization, and
the finding of progressive genetic changes as the tumor
develops supports the concept of multistep tumonigenesis inthe head and neck region (67).
This finding of chromosome number changes in thepremalignant regions near head and neck tumors is notunexpected. Most of the carcinogens that cause upper aeno-digestive tract tumors, including head and neck and lungcarcinomas, are known to cause chromosomal abnormali-ties (68, 69). In fact, many other studies have reportedincreased frequencies of micronuclei at various sites in the
aenodigestive tract, both in individuals exposed to varioustobacco-related carcinogens and in those harboring prema-lignant lesions (37, 39-41). Individuals with head and neck
cancer have also been shown to demonstrate increasedin vitro sensitivity to chromosome-damaging agents, espe-cially those individuals in whom second primary tumors
develop (70, 71). This increased susceptibility to chromo-some breakage might eventually lead to an accumulation ofgenetic alterations seen as chromosome pobysomies in theheavily exposed epithelial tissue.
One of the goals of our study of head and neck carci-nomas and their adjacent premalignant regions was to iden-tify genetic biomarkens that might be useful for assessing riskof tumor development in the high-risk group and that might
serve as intermediate end points in chemoprevention stud-ies. The chromosome polysomies present even in histolog-ically normal epithelium adjacent to tumors, and the in-
creased frequency of chromosome polysomies as the tissues
progressed to carcinomas suggest that the degree of genen-alized chromosome polysomies might be such a genetic
biomarken. The advantage of such a biomarken is that itpermits the sensitive detection of infrequent events that
reflect accumulated genetic damage or genomic instability,
events which are difficult to detect by bulk analysis (e.g.,DNA content analysis). Detecting genomic instability isparticularly important because it reflects an ongoing genetic
process that translates to a higher risk.
The working hypothesis for our future studies could be
that those individuals whose normal or premalignant epi-
thelium exhibits the greatest degree of genetic abnormality
might be expected to be at the highest risk for tumor de-
velopment. Indeed, our preliminary retrospective studies in
patients with oral premalignant lesions suggested that thoseindividuals who exhibited greaten than normal numbers of
chromosomes were at the highest risk for the development
of oral cancer (72). In this study, four of the seven patients
with dysplastic lesions and one of the six patients withhyperplastic lesions showed evidence of chromosomal
polysomies. More strikingly, of the four patients whose
premalignant disease progressed to invasive cancer or car-
cinoma in situ, three of those patients exhibited chromo-
somal polysomies in more than 5% of cells, whereas only
two of the nine patients who did not develop cancer
reached this level of polysomies (72).
Specific Genetic Markers
ras Gene Family. A significant portion of human tumors
from various sites in the body have been shown to containactivated oncogenes of the ras family (Harvey-ras, Kirsten-
ras, and N-ras) (73-76). Oncogenes in the ras family are
forms of the germline proto-oncogenes with specific pointmutations that, when transfected onto NIH/3T3 munine
fibroblasts, induce foci of morphologically altered cells(77-80). Normal ras genes code for proteins of molecular
weights of approximately 21 ,000 that have guanine nude-otide-binding activity and are able to hydrolyze GTP (81).
The proteins encoded by ras possess intrinsic GTPase ac-tivity which eventually leads to their inactivation, but this
inactivation is greatly enhanced by a second protein, calledthe GTPase-activating protein (82). This protein has been
shown to bind to the domain that is involved in the trans-duction of the ras gene signal, the “effector domain” of p21(83, 84).
H-ras isfound to be activated only infrequently, mainly
in thyroid carcinomas (85), and N-ras activations are foundpredominantly in myeloproliferative disorders and in lym-
phomas (85, 86). For unknown reasons, K-ras is particularlyassociated with adenocarcinomas and has been reported tobe activated in pancreatic cancers (87, 88), colonectal can-cers (89, 90), and adenocarcinomas of the lung (91, 92).Rodenhuis et a!. (93) reported that the majority of all rasmutations in lung cancer are found in adenocarcinomas,with frequencies of about 30% in smokers (41 of 141 sam-pIes) and about 5% in nonsmokers (2 of 40 samples). Incontrast, SCLC is not associated with an activated ras on-
cogene (94). Activation of the K-ras gene has been reportedto predict an unfavorable outcome: it identifies a subgroupof patients who have a very poor prognosis despite appan-ently successful surgery for stage I or II tumors (20). Anotherstudy by the National Cancer Institute essentially confirmed
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 701
this observation by anal�’zing the K-ras gene in 52 cell linesderived from patients with lung adenocancinomas (21).
Azuma et a!. used immunohistochemical analysis ofp21 expression in paraffin-embedded squamous cell head
and neck carcinoma tissues, and found that the extent of
expression of p21 was correlated with the degree of tumordifferentiation, clinical stage, and clinical outcome (95). Inthis study, 59 of 1 21 tumor samples reacted to the mono-
clonal antibody (Yl 3-259) against p21 encoded by the v-rasgene of the Harvey munine sarcoma virus (96), whereas oralleukoplakia and normal mucosa samples did not react,
indicating that they did not express this protein. Azuma eta!. also reported that the expression of ras was associatedwith poor prognosis. The c-H-ras gene was also analyzed in67 specimens of lymph node metastases and in 25 speci-mens of primary tumors obtained from 85 untreated pa-tients with head and neck squamous cell carcinoma (97).Ten of 46 (22%) patients who were hetenozygous for thislocus had lost one allele. Polymenase chain reaction de-
tected a mutation at codon 1 2 in only 2 of 54 (3.8%) tumorsand no mutations at codon 61 (97).
Another study reported c-H-ras mutations at codon 12in 2 of 37 squamous cell carcinomas of the head and neck
and one of eight squamous carcinoma cell lines (98), mdi-dating that the incidence of c-H-ras mutations is low in oral
squamous cell carcinoma, at least among white caucasoidpopulations (99-101). It is interesting that oral carcinomas
associated with chewing tobacco (quid) in an Indian pop-ulation had a relatively high frequency of c-H-ras muta-
tions, 20 of 57 (35%) cases, and all mutations were re-stnicted to codons 61.2 (glutamine to arginine) and 12.2
(glycine to valine) (1 02). The striking difference in c-H-rasmutational frequencies in the two populations may have
been due to any of several factors. For example, tobaccosused in India are likely to be from different strains or species
of plant. Also, it is possible that the chewing of tobaccoexposes the oral mucosa to concentrated levels of tobacco
carcinogens for longer periods than does cigarette smoking.The question of genetic predisposition should also be con-
sidered. Finally, there is a possibility that the carcinogen(s)that induces c-H-ras mutations in oral mucosa may bepresent in some component of quid other than tobacco.
A few studies have been reported on ras gene muta-
tions in premalignant lesions. K-ras mutations were found inadenomatous colonic polyps from patient with familial poly-posis coli, indicating that K-ras mutations could be a usefulmarker in cobonic tumonigenesis (103). In patients with
colorectal tumors, K-ras gene mutations were detectable inDNA purified from stool specimens (104). These patientsincluded both those with benign neoplasms (adenoma) and
those with malignant neoplasms in the colonic epithelium.The c-H-ras mutations of premalignant lesions in the headand neck area have not been well explored. In particular,the premalignant lesions induced by the chewing of quid inIndian populations may be genetically altered before theydevelop to frank malignancy. Because ofthe relatively highincidence of c-H-ras mutations in oral squamous cell can-cinomas in Indian populations, it is worthwhile to studythese premalignant lesions to help in the design of a re-search strategy to prevent frank malignancy.
Activation ofone ofthe rasgenes may be useful for riskassessment during tumonigenesis in epithelial tissues. Fun-then studies should be performed to determine whether
expression of the ras gene or its mutations can be modu-lated by chemopreventive agents; if so, ras activation may
also be used as an intermediate end point in chernopreven-
tion trials.
ErbBl, EGFR Gene. The erbBl oncogene was initially dis-covered as one of two oncogenes carried by the avian
erythnoblastosis virus (1 05). The corresponding proto-onco-
gene was found to encode a membrane-associated tyrosine
kinase protein that was eventually identified as the receptor
for EGF(106-i08). On bmndingthemn respective ligands, the
tyrosine kinase activity became stimulated several-fold, as
indicated by enhanced autophosphorylation of the recep-
ton, increased phosphorylation of exogenous substrates
in vitro, and elevated phosphorylation of the tyrosine resi-
dues of several proteins in vivo (1 09). Two cell lines estab-lished from tumors of the head and neck area at different
clinical stages were found to differ in the expression and in thetynosine kinase activity of EGFR (109). The 1483 cells dis-
played a higher plating efficiency and clonogenicity in soft
agar, suggesting that they have a more tumonigenic phenotype
than the 1 83A cells. Analyses of EGFR levels by using Ri
anti-EGFR serum indicated that the i 483 cells expressed
5-fold more receptors than the 1 83A cells. The autophosphor-
ylation activity of both receptors was stimulated by addition of
EGFR to isolated membrane preparations and intact cells,although the EGFR ofthe 1483 cells was much less responsive
to EGF than that of the 1 83A cells (1 10).The two cell lines having different characteristics, even
though they originated from the same poorly differentiatedsquamous cell carcinomas of the head and neck, is an
example of tumor heterogeneity. We can speculate that
even one clonal cell with a high number of EGFR among a
heterogenous population might eventually progress to frankmalignancy. If these particular cbonal cells were detected at
the earlier stages of tumonigenesis, they would be goodtargets for chernopreventive therapy. If those malignant
cells which had a high number of EGFR had increasedgrowth and tumonigenic phenotype, we would expect the
outcome of therapy, and the prognosis, to be poor.To examine the possibility that EGFR expression could
have predictive clinical value in head and neck squamous
cell carcinoma, Santini et a!. measured the EGFR levels intumor tissues (1 1 1 ). In 59 of 60 samples, EGFR levels werehigher in the tumor than in the corresponding normal con-trols. They also found a significant direct correlation be-tween EGFR levels and tumor size and stage. Using immu-nohistochemical and cytometnic techniques, expression of
the Ki-67 antigen, EGFR, the TFR, and DNA pboidy werestudied in 42 fresh samples of head and neck carcinomas.This study suggested that EGFR and TFR are widely distnib-
uted, especially on proliferating cells at the invading tumormargin. In addition, there is a close spatial correlation
between cells that express EGFR and TFR and those thatexpress Ki-67 antigen. Further follow-up is necessary todetermine whether these parameters will be importantprognostic values (1 1 2).
The EGER gene has been found to be amplified in
NSCLC (up to 20% in squamous cell types) (1 1 3-i i 5),whereas the EGFR protein has been shown to be ovenex-
pressed in many NSCLC cells (approximately 90% of squa-mous cell types, 20-75% of adenocancinomas, and infre-
quently in large cell or undifferentiated types) (114, 116,1 1 7). However, EGFR amplification on protein overexpres-sion has not been seen in SCLC cells. The overexpression of
the EGFR protein may reflect the development of an auto-crine growth loop, as EGF or transforming growth factor a
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
702 Review: Biomarkers in Upper Aerodigestive Tract Tumorigenesis
is required by most epithelial cells, including NSCLC cells,
for their growth (1 1 8). These findings have led us to develop
clinical trials that use monoclonal antibodies against EGFR
in the treatment (primary or adjuvant after surgery) inNSCLC (1 1 9, 1 20); these antibodies may also be useful in
the prevention of lung cancer.The expression of EGFR in premalignant lesions has
not been well studied. To determine whether EGFR be-comes ovenexpressed in premalignant lesions and to deter-
mine the consequences of such ovenexpression, we recentlyexamined 36 head and neck squamous cell carcinomas
with adjacent prernalignant lesions and normal control ep-ithelia. Using a monoclonal anti-EGFR antibody for immu-
nohistochemical analysis of paraffin-embedded tissue sec-
tions, the degree of EGFR expression on epithelial cells was
quantitated by computer-assisted image analysis (1 21 ). Thelevel of EGFR expression was significantly higher in the
adjacent normal tissue than in control samples (P= 0.021)
that had never been exposed to tobacco and/or alcohol.
These findings support the hypothesis offield cancenization
(7). We also found another increment of EGFR expression,
at the transition from the dysplastic lesions to squamous cellcarcinomas, in two-thirds of the samples examined(P = 0.001 ). These preliminary results indicated that EGFR
expression could be an important regulatory marker in the
context of the multistep process of head and neck cancerdevelopment (121). To validate EGFR expression as a bi-
omanker of an intermediate end point, we are currently
exploring this marker in a large number of samples
collected in chemoprevention trials.
Int-2/Hst-1 Genes. The hst- 1 gene is one of the most fre-
quently detected transforming genes after the ras gene fam-
ily (122, 123). Because this gene encodes a protein that is
homologous to a fibroblast growth factor and a int-2-en-
coded protein, it is assumed to be a member of the genefamily that is involved in cell growth (124, 125). Both thehst-1 and int-2 genes are mapped to chromosome 1 1 qi 3
(1 26), and their coamplification has been reported in blad-den carcinoma (127), esophageal carcinoma (128), mela-noma (1 29), gastric carcinoma (1 24), and breast carcinoma(1 30). At the same region of chromosome 1 1 qi 3, otherimportant genes (gst-ir, PRAD, and cyclin Dl ) were bothamplified and expressed. Coamplification of the hst-i andint-2 genes in a hepatocellular carcinoma was accompa-nied by amplification of integrated hepatitis B virus DNA(1 31 ). The biological significance of coamplification of thelist-i and int-2 genes is not clear.
The int-2 gene was amplified 3-fold to 5-fold in 5 of 10laryngeal carcinomas and 2-fold to 3-fold in 5 of 1 1 carci-nomas at other sites of the head and neck (1 32). However,
the amplified int-2 gene has not yet shown any overexpres-sion at either the mRNA or the protein level in head and
neck squamous cell carcinomas. Adjacent histologicallynormal tissues from the same patients had only single cop-es of the gene. In a survey of head and neck tumor-derived
cell lines, int-2 was amplified 9-fold in one, but not in three
laryngeal cell lines (1 32). In another report, int-2 was foundto be amplified in two of eight head and neck carcinomas(1 33). Although there was a suggestion that amplification ofint-2 was proportionally correlated with tumor recurrence
and clinical disease progression (1 32), a new study with alarge patient sample is required to determine more precisely
the significance ofthis gene amplification on ovenexpressionin head and neck carcinomas.
To determine the chromosomal location of the ampli-
fied region and when during tumonigenesis the amplifica-
tion process occurred, we examined head and neck squa-
mous cell lines that were established by Sachs et a!. (109)and paraffin-embedded tissue sections from which the cell
lines were established, both by ISH, using both a cosmid
probe for the int-2 gene and a biotin-labeled chromosome
ii painting probe. Three of 10 cell lines exhibited int-2amplification, 2 of which (1386, 1986) were on chromo-
some 1 1 distal to the single copy gene; the third (886 cells)was on another chromosome. A fourth cell line (1 486 cells),
which did not have int-2 amplification, showed only anonamplified single copy int-2 gene on chromosome 1 1.
These findings correlated well with the amplification ofint-2 as detected by Southern blotting. Paraffin blocks of the
source tumors, which contained adjacent premalignant be-sions, were analyzed with the int-2 probe to allow visual-
ization of the timing of amplification during tumonigenesis.
Two of the source tumors (1 386 and 1 986 cells) showed
int-2 amplification in dysplastic and cancer regions,
whereas a third (886 cells) showed amplification at the
hyperplasia-to-dysplasia transition area and at the carci-noma in situ and tumor areas (1 34). These results suggest
that int-2 amplification can occur in premalignant lesionsprior to tumor development and validate our assumption
that tumors develop through clonal evolution in situ in thismodel system (i.e., the changes persisted from premalignant
to malignant cells). Therefore, int-2 amplification has the
potential to be used as a marker for risk assessment and in
chemoprevention trials (1 34).
p53 Tumor Suppressor Gene. The p53 gene, which en-
codes a nuclear protein, has been mapped to the short arm
of chromosome 1 7 (1 7pl 3). The p53 protein was originallyidentified as a nuclear protein that bound to the large T
antigen of the SV4O DNA tumor virus (1 35, 1 36). Althoughthis gene was initially thought to act as a dominant onco-gene, further investigation indicated that a mutant formexisted (1 37). When the p53 gene was tested for its abilityto transform cells, it was discovered that wild-type p53 genecould suppress transformation while the mutant form couldinduce transformation (1 37). However, many different types
of alterations (rearrangement, deletion, insertion, or pointmutation) have been observed to occur at different boca-tions within the p53 gene in a wide variety of cell lines and
human tumors (1 38-1 52).The genetic alteration most frequently found in the p53
gene is a point mutation. Point mutation analyses have beenconfined primarily to exons 5 through 8, where the muta-tions are most frequently found in phylogenetically con-served regions. Hollstein et a!. examined the frequency ofnucleotide changes according to cancer type in a study of280 p53 base substitutions (1 53). They observed some no-
table features: (a) most mutations in colonectal cancers,brain tumors, leukemia, and lymphomas occur at CpG
dinucleotides, which are known to be mutational hot spots,and (b) G to A transitions constitute the majority of colon
tumor mutations (31 of 39 mutations, 79%), whereas no Gto T tnansvensions were observed. More recent studies havedetected C to T tnansvensions in colon tumors (1 54, 1 55). In
contrast, C to T tnansvensions are the most frequent substi-tution in NSCLC (1 7 of 30 mutations, 57%) and liver tumors(14 of 19 mutations, 74%). In addition, the involvement of
ultraviolet light in p53 mutations of squamous cell carci-noma of the skin was suggested by the presence in these
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 703
cells of a CC to U double-base change, which is known tobe induced by ultraviolet light through the formation ofpyrimidine photodimens (1 56). p53 mutational hotspots inhuman hepatocellular carcinoma vary according to the can-
cinogen by which they are induced (i.e., G to T transversionby hepatitis B virus, and G to C transversion by aflatoxin B1)(1 43). Thus, the specific p53 mutation sites may reflect the
different carcinogen background of each tumor.The p53 gene product is believed to function in cell
cycle control (1 57-1 59). At least two stages in the cell cycleare regulated in response to DNA damage, the G1-S and the
G2-M transitions. These transitions serve as checkpoints atwhich cell cycle progression is delayed to allow repair ofDNA damage before the cell enters either S phase, when
damage would result from DNA adduct formation leading
to mutation or chromosomal damage, or M phase, when
chromosome breaks would cause the loss of genomic ma-terial from daughter cells. Checkpoints are believed to besurveillance mechanisms that can detect DNA damage and
active signal transduction pathways and then regulate rep-lication or segregation machinery and, possibly, repair
activities (160).Since there are p53 mutants that do not exhibit the cell
cycle arrest or delay that occurs in wild-type p53 in re-sponse to DNA damage, both the G1-S and G2-M check-
points are known to be under genetic control. There isstrong evidence that p53 inhibits the G1-S transition, be-
cause high levels of pS3 block cell cycle progression at theG1-S checkpoint (1 57). Tumor cells that lack p53 or that
have dominant mutant forms of p53 lack the G1 -S delay thatoccurs on exposure to ionizing radiation (1 61 ). The loss of
a C1 checkpoint in mammalian cells is not associated withincreased sensitivity to the lethal effects of ionizing radia-tion (1 62), but it is associated with an increase in mutational
frequency. The G2-M checkpoint is abolished by mutationof p53 in a number of yeast genes (1 63) and by treatment ofmammalian cells with caffeine (1 64). The epithelium of thehead and neck area is constantly exposed to carcinogens,and p53 may affect how these damaged cells respond to this
insult. We speculate that the upregulation of p53 in re-sponse to damage induced by carcinogens increases itssusceptibility to mutation. There are human lymphoma celllines that have increased sensitivity to irradiation while the
cells are arrested during S phase (165). Therefore, p53mutation is eventually related to cell proliferation.
The normal p53 protein has a very short half-life (6-20
mm), whereas the mutant form has a half-life of up to 6 h.It has been inferred, therefore, that the presence of detect-
able p53 protein implies mutation (166). Iggo et a!. (167)found increased p53 oncoprotein staining in the types of
lung cancers that are associated with smoking. They re-ported elevated p53 protein levels in 1 4 of 1 7 (82%) squa-mous cell carcinomas and in only eight of 21 (38%)nonsquamous cell carcinomas. Similarly, Chiba eta!. (144)
reported finding pS3 mutations in 65% of the lung squa-mous cell carcinomas and in 36% of the nonsquarnous
tumors. The association of smoking with squamous cell
carcinomas of the lung provides further evidence for a linkbetween p53 mutations and smoking. In a similar study ofp53 in SCLC cell lines, D’Amico et a!. (168) found that
1 00% of the SCLC cells had p53 mutations.More recently, inactivating mutations in the p53 gene
have been identified in truly preneoplastic lesions, namely,
Barrett’s esophagus (1 69), a precursor to adenocarcinomasof the esophagus. In bronchial epithelium, the p53 protein
was detected in 0% of normal mucosa, 6.7% of squamousmetaplasia, 29.5% of mild dysplasia, 59.7% of severe dys-
plasia, and 58.5% of carcinomas in situ (1 70). Nees et a!.(171) studied p53 mutations in the respiratory epithelium
either adjacent to or at significant distance from primaryhead and neck tumors. They observed p53 mutations in the
distant epithelia of these patients and concluded that mu-
tation of p53 is an early event in head and neck carcino-
genesis, supporting the field cancenization hypothesis. A
similar observation was also made by Boyle et a!. (1 72).
We studied p53 protein expression in 33 patients withhead and neck squamous cell carcinomas whose tissue
sections contained adjacent normal epithelium, hyperpla-
sia, and/or dysplasia. Fifteen of 33 (45%) head and neck
tumor specimens expressed p53, but none of the normal
controls (tissue speci mens from cancer-free nonsmokers)expressed detectable p53 protein. However, 5 of 24 (2i%)
specimens of normal epitheliurn adjacent to tumors, seven
of 24 (29%) hyperplastic lesions, and nine of 20 (45%)
dysplastic lesions expressed p53 (1 73). We conclude that
p53 expression can be altered in very early phases of head
and neck tumonigenesis. To determine whether increased
expression of p53 was associated with a gene mutation, we
performed a combination of polymerase chain reaction-si ngle-strand conformation polymorph ism and direct
genomic sequencing on one representative case that ex-
pressed a high level of p53 in the tumor but not in its
adjacent normal epithelium. We found that the adjacent
normal epithelium had a wild-type p53 at codon 1 74 in
exon 5. In the tumor sample, however, 1 0 base pairs hadbeen deleted at this position. This finding confirmed that
there was a good correlation between high levels of p53
expression and mutation. Further work is needed, however,to establish the relationship between low levels of p53
expression and gene mutations. We also evaluated p53protein expression in 27 premalignant oral lesions and innormal oral mucosa of eight healthy nonsmoking control
individuals. In 14 of the 27 lesions (52%), pS3 was ex-pressed in more than S% of cells calculated, while therewas no p53 expression in nonsmoking controls. Eight le-sions (all in current or former smokers) had very high base-line levels of p53. The degree of response to 3 months oftreatment with high-dose 1 3-cis retinoic acid correlatedinversely with baseline p53 expression level (1 74). Thispreliminary study suggests that p53 expression may be anexcellent predictor of risk and may serve as an intermediate
biomarker in chemoprevention trials (1 73, 1 74).
Proliferation Markers
It is hypothesized that only those cells with high prolifera-tive activity could be associated with prernalignant andmalignant tissue changes during tumonigenesis. Lee et a!.(1 75) initially used antibodies against the cell proliferationmarkers, including PCNA, and found PCNA to be quite
useful as a marker for proliferating cells, even in routinelyprocessed formal in-fixed, paraffi n-embedded tissue sec-
tions (1 76). PCNA is a Mr 36,000 acidic, nonhistone, nu-clear protein whose expression is associated with the late
G1 5, and early G2 phases of the cell cycle (1 77). It is anauxiliary protein to DNA polymerase 6, which has a 261-amino-acid polypeptide with high aspartic and glutamicacid contents and plays a critical role in the initiation of cellproliferation (1 78, 1 79).
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
704 Review: Biomarkers in Upper Aerodigestive Trad Tumorigenesis
To better understand the relationship between PCNA
expression in tissue and proliferation status, specimens ofhead and neck and cobonectal cancers were excised from
patients after infusion of BrdUrd (1 80). The specimens wereembedded in paraffin and examined. Adjacent tumor sec-tions were analyzed for PCNA expression and for incorpo-rated BrdUrd (using anti-PCNA antibody and anti-BrdUndantibody, respectively). Regions ofturnors that were high inPCNA-positive cells also had high BndUnd uptake, and viceversa. In all cases, the proportion of PCNA-positive cells
was higher than the proportion of BndUrd-positive cells.These results were not surprising, since PCNA is expressed
in most proliferating phases (G1, S, and G2), whereasBndUrd uptake marks only the S phase (1 79).
With the hypothesis that PCNA expression is dysregu-lated in tumors, 1 07 NSCLC tissue sections were examinedfor PCNA expression (1 75). Squamous cell carcinomasshowed the highest proliferative activity, with a mean of40% PCNA-positive cells (range, 2-90%); adenocarcino-mas had a mean of only 5% PCNA-positive cells (range,0-70%), and large cell carcinomas had a mean of 15%(range, 3-80%). The PCNA-positive fraction became pro-
gressively larger in areas of squamous metaplasia and car-cinoma in situ (1 75). To better understand tumorigenesis inhead and neck cancer, we studied 33 formalin-fixed, par-
affin-embedded tissue specimens from five different sites ofhead and neck squamous cell carcinomas that containedadjacent normal epithelium, hyperplasia, and/or dysplasia
(1 81 ). PCNA expression was assessed by semiquantitativescoring in three epithelial layers (basal, panabasal, and su-perficial). The labeling index (the number of positively
stained cells divided by the total number of cells counted)and the weighted mean index of PCNA expression Lthe sumof the number of counted cells multiplied by degree ofintensity (0-3) in each cell and divided by the total number
of counted cells] were calculated to represent the level ofPCNA expression. What was interesting was that normal
epithelium adjacent to the tumor had much more prolifen-ative activity than control epithelium (from cancer-freenonsmokers). Furthermore, PCNA expression increased as
tissues progressed from adjacent normal epithelium tohyperplasia (P < 0.001 ), hyperplasia to dysplasia(P < 0.001), and dysplasia to squamous cell carcinomas(P = 0.065); the total increase in PCNA expression fromadjacent normal epithelium to squamous cell carcinomasranged from 4-fold to 1 0-fold (1 81 ). As the tissue progressedto carcinoma, we observed not only increases in the num-ben of proliferating cells but also in the amount of PCNAexpressed by each labeled cell. These studies indicate thatPCNA could be a useful biomanker for multistep cancino-
genesis in head and neck cancer and that its expressioncould serve as an intermediate end point in chemopreven-tion trials (1 81).
As part of an ongoing chemoprevention trial in which
chronic smokers (�1 5 pack-years) were screened for squa-mous metaplasia before being randomized to receive either1 3-cis netinoic acid on placebo, our group examined PCNA
expression in bronchial biopsy sections obtained from sixstandardized sites in the major bronchial trees (1 80). In thisstudy, 1 65 samples were evaluated for PCNA expressionand histologic status. Among the 81 biopsied specimensshowing histologically normal epithelium, only 12% hadmore than i% PCNA-positive cells, and no specimen hadmore than 5% positive cells. In contrast, 37% (19 of 52) of
the hyperplasia specimens and 50% (1 0 of 20) of the meta-
plasia specimens had more than 1% PCNA-positive cells.Among those biopsied specimens showing dysplasia withmetaplasia, 58% (7 of 1 2) had more than 1 % PCNA-posi-tive cells. These results suggest a significant correlationbetween increases in proliferative activity and histologicalprogression in epithelium at high risk oftumor development(1 80). In a study ofesophageal premalignancy by Yang eta!.(1 82) and a study of subjects at high risk for colon cancer byLipkin et a!. (1 83, 1 84), the patterns of expression of the
proliferation marker and tnitiated thymidine incorporationwere similar to the patterns of PCNA expression we ob-
served in the lung and head and neck cancers. These twopilot studies also suggested that the chemopreventive drug
can suppress this marker. A large scale of chemopreventionstudy incorporating studies of biomarkens for oral prema-lignancy and second primary cancer prevention is ongoing
at M. D. Anderson Cancer Center. We hope we will have
more definite answers to these questions in the near future.
Squamous Cell Differentiation Markers
Upper aerodigestive epithelia differentiate along the squa-mous pathway in cancinogenesis, and understanding of thisprocess is important for developing chemoprevention strat-
egies (185, 186). To establish a preclinical carcinogenesismodel and to determine which markers are important, we
examined several markers, including cytokenatin and TGase1 , in the DMBA-induced hamster buccal pouch model
(1 87, 1 88). We applied DMBA (0.5%) in heavy mineral oilto the hamster buccal pouch three times per week for up to
1 6 weeks. TGase 1 was expressed at a limited level innormal buccal mucosa, at a low level in the basal layer ofhyperplastic lesions, and at a somewhat higher level indysplasia; its expression was markedly increased in squa-mous cell carcinoma (1 87). The cytokeratin assayed in thisstudy were K14 (Mr 55,000), Ki (Mr 67,000), and Ki3 (Mr
47,000). Normal hamster cheek pouch epithelium ex-pressed K1 4 in the basal layer and K1 3 in the suprabasaland differentiated layers; K1 was not detected. In hyperpla-sia, Ki 4 was no longer restricted to the basal layer but was
expressed in differentiated cells. The same pattern was ob-served in dysplasia; Ki4 expression was in squamous cell
carcinomas. However, Ki 3 was preserved in hyperplasticepithelium during all stages of carcinogenesis, includinganaplastic or differentiated areas. Expression of K1 , in con-trast, started as a weak and patchy pattern after 2 weeks ofDMBA treatment but became stronger and more homoge-neous at 8 weeks of treatment. However, Ki was almost
absent in squamous cell carcinomas. We concluded thatthe pattern of kenatin expression could be important tool inthe study of carcinogenesis (188).
Another marker for squamous cell differentiation isinvolucnin, one of the major protein components of conni-
fied envelopes (1 89). This protein undergoes extensivecross-linking by the membrane-associated (particulate) en-
zyme type 1 TGase, which catalyzes the formation of#{128}-(r-glutamyl isopeptide) linkages between protein-bound
glutamine residues and primary amines such as protein-bound lysine (1 90, 1 91 ). These proteins are expressed in the
upper stratum spinosum and conneal layers of the epidermis(191). Squamous differentiation of keratinocytes is usually
accompanied by increases in the levels of involucrin andTGase 1 , but the expression of involucnin precedes theexpression of TGase 1 (1 89, 1 90). Involucnin is expressed inpremalignant lesions and squamous cell carcinomas (192,
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 705
193), and TGase 1 is also expressed in benign and malig-nant neoplasms ofthe skin and in a DMBA-induced hamster
model (187-194). These squamous cell differentiationmarkers were shown to be modulated by netinoic acids in
cell lines (185, 195-197). However, the mechanisms ofdifferentiation of squamous cells and their modulation bynetinoids seem to be very complex; they are beyond thescope of this review and need to be discussed separately.
Biomarkers in Chemoprevention Trials
The goal of clinical chemoprevention trials in the upper
aerodigestive tract is to reduce the incidence of cancerdevelopment in that field of tissue. As described previously,however, the major obstacle to such trials is that the study
end point (i.e., cancer) generally takes years to becomedetectable. It would be ideal to define intermediate end
points that reflect whether chemopreventive agents have aneffect on the tissue at risk (4, 5) and, if so, their mechanismsof action at the tissue level. First, biomankers would beuseful for assessing the risk of tumor development in high-risk tissue and in premalignant lesions. Second, these mark-
ens would allow us to better understand the pathobiology ofclinical response to the chemopreventive treatment. Third,such markers would be good indicators of intermediate end
points in clinical chemoprevention trials, allowing predic-tion of patients’ responsiveness before the final end point,cancer, could be reached. In such trials, we would want toknow whether clinical outcome ofthe treatment is due to (a)reversal of the abnormal clones at the genetic level; (b)phenotypic reversal of the abnormal clones; or (c) partial
suppression of less-affected clones (if two or more distinctclones can be identified in the lesions). These questions
could be answered by examining both genetic markers andphenotypic markers on the tissue samples obtained before
and after chemopreventive therapy. Finally, although strictvalidation of any biomanker as a true intermediate end point
of cancer development may take many years offollow-up inlarge-scale clinical trials, current biomarker candidates arean important adjunct to the development of new chemo-preventive agents and to the national design of future inter-
vention trials.In this review, we described several candidates for
genetic and phenotypic biomankens, all ofwhich need to bevalidated in clinical trials.In the future, validated compre-
hensive panels of biomarkers, indicating early and interme-diate stages of the multistep carcinogenesis process, mayprovide new standard end points and even replace cancerincidence as the sole end point for chemoprevention trials.If this occurs, an early step in preventive studies will be to
determine whether the agents have any activity at all inaffecting the tumonigenesis process. Later on, the active
agents can be evaluated for their effects on tumor formation.This strategy differs from that of large-scale chemopreven-
tive studies; it requires a smaller number of subjects and
may be conducted within a relatively short period of time.
References1 . Meyskens, F. L., Jr. Coming of age: the chemoprevention of cancer.
N. EngI. J.Med., 323: 825-827, 1990.
2. Boone, C. W., Kelloff, G. J., and Malone, W. E. Identification of candidate
cancer chemopreventive agents and their evaluation in animal models and
human clinical trials: a review. Cancer Res., 50: 2-9, 1990.
3. Hong, W. K., Lippman, S. M., Itri, L. M., Karp, D. D., Lee, J. S., Byers,
R. M., Schantz, S. P., Kramer, A. M., Lotan, R., Peters, L. J., Dimery, I. W.,
Brown, B. W., and Goepfert, H. Prevention of second primary tumors withisotretinoin in squamous cell carcinoma of the head and neck. N. EngI. J.
Med., 323:795-801, 1990.
4. Lippman, S. M., Lee, J. S., Lotan, R., Hittelman, W. N., Wargovich, M. I.,and Hong, W. K. Biomarkers as intermediate end points in chemopreventiontrials. J. NatI. Cancer Inst., 82: 555-560, 1990.
5. Schatzkin, A., Freedman, L. S., Schiffman, M. N., and Dawsey, S. M.
Validation of intermediate end points in cancer research. J. NaIl. Cancer
Inst., 82:1746-1752, 1990.
6. Zelen, M. Are primary cancer prevention trials feasible? 1. NaIl. CancerInst., 80: 1442-1444, 1988.
7. Slaughter, D. L., Southwick, H. W., and Smejkal, W. “Field canceriza-
tion” in oral stratified squamous epithelium: clinical implications of multi-centric origin. Cancer (Phila.(, 6:963-968, 1953.
8. Silverman, S. J., Jr., Gorsky, M., and Lozada, F. Oral leukoplakia and
malignant transformation: a follow-up study of 257 patients. Cancer (Phila.),53:563-568, 1984.
9. Auerbach, 0., Stout, A. P., Hammond, E. C., and Garfinkel, L. Changes in
bronchial epithelium in relation to cigarette smoking and in relation to lungcancer. N. EngI. I. Med., 265: 253-267, 1961.
10. Lippman, 5. M., and Hong, W. K. Second malignant tumors in head andneck squamous cell carcinoma: the overshadowing threat for patients withearly-stage disease. Int. J. Radiat. Oncol. Biol. Phys.. 17: 691-694, 1989.
1 1 . Farber, E. The multistep nature of cancer development. Cancer Res., 44:4217-4223, 1984.
1 2. Stich, H. F. Micronucleated exfoliated cells as indicators for genoloxic
damage and as markers in chemoprevention trials. I. Nutr. Growth Cancer,
4:9-18, 1987.
1 3. Stockman, M. S., Gupta, P. K., Pressman, N. l.� and Mulshine, I. L. Con-siderations in bringing a cancer biomarker to clinical application. CancerRes., 52(Suppl.(:27i1s-2718s, 1992.
14. Pitot, H. C. The natural history of neoplastic development: the relation
ofexperimental models to human cancer. Cancer (Phila.), 49: 1206-1211,
1982.
1 5. Harris, C. C., Brash, D. E., Lechner, J. F., et al. Aberrations of growth and
differentiation pathways during neoplastic transformation of human epithe-hal cells. In: T. Kakunaga, T. Sugimura, and L. Tomalis (eds.(, Cell Differen-tiation, Genes and Cancer, pp. 139-148. New York: Oxford UniversityPress, 1988.
16. Slaga, 1. J. Cellular and molecular mechanisms involved in multistage
skin carcinogenesis. Carcinog. Compr. Surv., 11: 1-8, 1989.
1 7. Harris, 0. C., Reddel, R., Modali, R., Lehman, T. A., Iman, D.,
McMenamin, M., Sugimura, H., Weston, A., and Pfeifer, A. Oncogenes andtumor suppressor genes involved in human lung carcinogenesis. Basic Life
Sci., 53:363-379, 1990.
18. Pulciani, S., Santos, E., Lauver, A. V., Long, L. K., Aaronson, S. A., andBarbacid, M. Oncogenes in solid human tumours. Nature (Lond.(, 300:
539-542, 1982.
19. Rodenhuis, S., van de Wetering, M. L., Mooi, W. J., Evers, S. G., vanZandwijk, N., and Bos, J. L. Mutational activation of the K-ras oncogene. A
possible pathogenetic factor in adenocarcinoma ofthe lung. N. EngI. J. Med.,317: 929-935, 1987.
20. Slebos, R. J., Kibbelaar, R. E., Dalesio, 0., et al. K-ras oncogene activa-lion as a prognostic marker in adenocarcinoma of the lung. N. EngI. I. Med.,
323: 561-565, 1990.
21 . Mitsudomi, T., Steinberg, 5. M., Oie, H. K., Mulshine, J. L., Phelps, R.,Viallet, J., Pass, H., Minna, J. D., and Gazdar, A. F. ras Gene mutations in
non-small cell lung cancers are associated with shortened survival irrespec-tive of treatment intent. Cancer Res., 51: 4999-5002, 1991.
22. Field, J. K., Spandidos, D. A., Stell, P. M., Vaughan, E. D., Evans, G. I.,and Moore, J. P. Elevated expression of c-myc oncoprotein correlated with
poor prognosis in head and neck squamous cell carcinoma. Oncogene, 4:1463-1468, 1989.
23. Sarnath, D., Parchal, R., Nair, R., Mehta, A. R., Sanghavi, V., Sumegi, J.,KIm, G., and Deo, M. G. Oncogene amplification in squamous cell cancer
ofthe oral cavity. Jpn. J. Cancer Res. (Gann), 80: 430-437, 1989.
24. Johnson, B. E., Ihde, D. C., Makuch, R. W., Gazdar, A. F., Carney,D. N., Oie, H., Russell, E., Nau, M. M., and Minna, I. D. myc Family
oncogene amplification in tumor cell lines established from small cell lungcancer patients and its relationship to clinical status and course. J. Clin.Invest., 79: 1629-1634, 1987.
25. Birrer, M. 1., Raveh, L., Dosaka, H., and Segal, 5. A transfected L-mycgene can substitute for c-myc in blocking murine erythroleukemia differen-
tiation. Mol. Cell Biol., 9: 2734-2737, 1989.
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
706 Review: Biomarkers in Upper Aerodigestive Trad Tumorigenesis
26. Yokota, J., Wada, M., Yoshida, T., Noguchi, M., Teaki, T., Shimosato, Y.,Sugimura, T., and Terada, M. Heterogeneity of lung cancer cells with respectto the amplification and rearrangement of mycfamily oncogenes. Oncogene,
2:607-611, 1988.
27. Yokota, J., Wada, M., Shimosato, Y., Terada, M., and Sugimura, T. Lossof heterozygosity of chromosome 3 in adenocarcinoma of the lung. Proc.NatI. Acad. Sci. USA, 84:9252-9256, 1987.
28. Miura, I., Siegfried, J. M., Resau, J., Keller, S., Zhou, J. Y., and Testa, J. R.
Chromosome alternations in 21 non-small cell lung carcinomas. GenesChromosomes Cancer, 2: 228-238, 1990.
29. Whang-Peng, J., Knutsen, T., Gazdar, A., Steinberg, S. M., Oie, H.,Linnoila, I., Mulshine, J. L., Nau, M., and Minna, J. D. Nonrandom structural
and numerical chromosome changes in non-small cell lung cancer. GenesChromosomes Cancer, 3: 1 68-i 88, 1991.
30. Beardsley, T. Smart genes. Sci. Am., 265: 87-95, 1 991.
31 . Schutte, J., Minna, J. D., and Birrer, M. J. Deregulated expression ofhuman c-jun transforms primary rat embryo cells in cooperation with an
activated c-Ha-ras gene and transforms rat-la cells as a single gene. Proc.NatI. Acad. Sci. USA, 86: 2257-2261, 1989.
32. Sunday, M. E., Hua, J., Dai, J. B., Nusrat, A., and Torday, J. S. Bombesinincreases fetal lung growth and maturation in utero and in organ culture. Am.J. Respir. Cell Mol. Biol., 3: 199-205, 1990.
33. Kane, M. A., Aguayo, S. M., Portanova, L. B., Ross, S. E., Holley, M.,Kelley, K., and Miller, Y. E. Isolation of the bombesin/gastrin-releasing pep-
tide receptor from human small cell lung carcinoma NCI-H345 cells. I. Biol.Chem., 266:9486-9493, 1991.
34. Aguayo, S. M., Kane, M. A., King, T. E., Jr., Schwartz, M. I., Grauer, L.,
and Miller, T. E. Increased levels of bombesin-like peptides in the lowerrespiratory tract of asymptomatic cigarette smokers. J. Clin. Invest., 84:
1105-1113, 1989.
35. Lippman, S. M., Lee, I. S., and Lotan, R. Chemoprevention of upper
aerodigestive tract cancer: a report of the Third Upper Aerodigestive TractTask Force Workshop. Head Neck Surg., 12: 5-20, 1990.
36. Lippman, S. M., Peters, E., and Wargovich, M. Bronchial micronuclei asa marker of an “early” stage of carcinogenesis in human tracheobronchialepithelium. Int. J. Cancer, 45:811-815, 1990.
37. Rosin, M. P., Dunn, B. P., and Stich, H. F. Use of intermediate endpoints in quantitating the response of precancerous lesions to chemopreven-
live agents. Can. J. Physiol. Pharmacol., 65: 483-487, 1987.
38. Lippman, S. M., and Hong, W. K. Differentiation therapy for head andneck cancer. In: G. Snow, J. R. Clark (eds.), Multimodality Therapy for Head
and Neck Cancer. New York: Verlag Press, in press.
39. Stich, H. F., Rosin, M. P., and Vallejera, M. 0. Reduction with vitaminA and beta-carotene administration of the proportion of micronucleatedbuccal mucosal cells in Asian betel nut and tobacco chewers. Lancet, 1:1204-1206, 1984.
40. Stich, H. F., Rosin, M. P., and Hornby, A. P. Remission of oral leuko-
plakias and micronuclei in tobacco/betel quid chewers treated with beta-
carotene and with beta-carotene plus vitamin A. Int. J. Cancer, 42: 195-199,1988.
41 . Stich, H. F., Hornby, A. P., and Mathew, B. Response of oral leukopla-kias to the administration of vitamin A. Cancer Lett., 40: 93-1 01 , 1988.
42. Munoz, N., Wahrendorf, J., and Bang, L. J. No effect of riboflavin,
retinol, and zinc on prevalence of precancerous lesions of the esophagus.Randomized double-blind intervention study in high-risk population ofChina. Lancet, 2: 111-114, 1985.
43. Munoz, N., Hayashi, M., and Bang, L. J. Effect of riboflavin, retinol, and
zinc on micronuclei of buccal mucosa and esophagus: a randomized dou-
ble-blind intervention study in China. J. NatI. Cancer Inst., 79: 687-691,1987.
44. Gluckman, J. 0., Crossman, J. D., and Donegan, J. 0. Multicentre squa-
mous cell carcinoma of the upper aerodigestive tract. Head Neck Surg., 3:90-96, 1980.
45. Shibuya, H., Hisamitso, S., Shioiri, S., Horiuchi, J., and Suzuki, S.
Multiple primary cancer risk in patients with squamous cell carcinoma of the
oral cavity. Cancer (Phila.), 60: 3083-3086, 1987.
46. Mitelman, F. Catalog of Chromosome Aberrations in Cancer, Ed. 3. NewYork: Alan R. Liss, 1988.
47. un, Y., Higashi, K., Mandahl, N., Heim, S., Wennerberg, J., Biorklund,A., Dictor, M., and Mitelman, F. Frequent rearrangement of chromosomalbands 1 p22 and 1 1 qi 3 in squamous cell carcinoma of the head and neck.Genes Chromosomes Cancer, 2: 198-204, 1990.
48. Osella, P., Carlson, A., Wyandt, H., and Milungky, A. Cytogeneticstudies of eight squamous cell carcinomas of the head and neck. Deletion of
7q, a possible primary chromosomal event. Cancer Genet. Cytogenet., 59:73-78, 1992.
49. Tessier, J. R. The chromosomal analysis of human solid tumors: a triplechallenge. Cancer Genet. Cytogenet., 37: 103-125, 1989.
50. Lee, J. S., Pathak, S., Hopwood, V., Tomasovic, B., Mullins, T. D., Baker,F. L., Spitzer, G., and Neidhart, J. A. Involvement of chromosome 7 inprimary lung tumor and nonmalignant normal lung tissue. Cancer Res., 47:
6349-6352, 1987.
51 . Sozzi, G., Miozzo, M., Tagliabue, E., Caldenone, C., Lombardi, L.,Pilotti, S., Pastorino, U., Pieriotti, M. A., and Porta, G. D. Cytogenetic ab-
normalities and overexpression of receptors for growth factors in normal
bronchial epithelium and tumor samples of lung cancer patients. CancerRes., 51:400-404, 1991.
52. Mertens, F., Jin, Y., Heim, S., Mandahl, N., Jonsson, N., Mertens, 0.,Persson, B., Salesmark, L., Wennerberg, J., and Mitelman, F. Clonal structuralchromosome aberration in nonneoplastic cells of the skin and upper aero-
digestive tract. Genes Chromosomes Cancer, 4: 235-240, 1992.
53. Pinkel, D., Straume, T., and Gray, J. W. Cytogenetic analysis using
quantitative, high sensitivity fluorescence hybridization. Proc. NatI. Acad.
Sci. USA, 83:2934-2938, 1986.
54. Cremer, T., Lichter, P., Borde, J., Ward, D. C., and Manuelidis, L.
Detection of chromosome aberrations in metaphase and interphase tumorcells by in situ hybridization using chromosome-specific library probes.Hum. Genet., 80:235-246, 1988.
55. Hopman, A. H. N., Wiegant, J., Raap, A. K., Landgent, J. E., Van delPloeg, M., and Van Duijn, P. Bicolor detection of two target DNAs by
non-radioactive in situ hybridization. Histochemistry, 85: 1-4, 1986.
56. Van Dekken, H., Pizzolo, J. G., Reuter, V. E., and Melamed, M. R.
Cytogenetic analysis of human solid tumors by in situ hybridization with a setof 1 2 chromosome-specific DNA probes. Cytogenet. Cell Genet., 54:
103-107, 1990.
57. Reid, T., Baldini, A., Rand, T. C., and Ward, D. C. Simultaneous visu-alization of seven different DNA probes by in situ hybridization using
combinational fluorescence and digital imaging microscopy. Proc. NatI.Acad. Sci. USA, 89: 1388-1392, 1992.
58. Cremer, T., Testin, D., Hopman, A. H. N., and Manuelidis, L. Rapidinterphase and metaphase assessment of specific chromosomal changes inneuroectodermal tumor cells by in situ hybridization with chemically mod-ified DNA probes. Exp. Cell Res., 176: 199-220, 1988.
59. Hopman, A. H., Moesker, 0., Smeets, A. W. G. B., Pauwels, P. E.,
Vooijs, G. P., and Ramaekers, F. C. S. Numerical chromosome 1, 7, 9, and
1 1 aberrations in bladder cancer detected by in situ hybridization. CancerRes., 51:644-651, 1991.
60. Hopman, A. H. N., Ramaekers, F. C. S., Raap, A. K., Beck, J. L. M.,Devilee, P., Van Der Ploeg, M., and Vooijs, G. P. In situ hybridization as atool to study numerical chromosome aberrations in solid tumors. Histochem-
istry, 89: 307-316, 1988.
61 . Matsumura, K., Kallioniemi, A., Kallioniemi, 0., Chen, L., Smith, H. S.,Pinkel, D., Gray, J., and Waldman, F. M. Deletion of chromosome 1 7q lociin breast cancer cells detected by fluorescence in situ hybridization. Cancer
Res., 52:3474-3477, 1992.
62. Emmerich, P., Jauch, A., Hofmann, M., Cremer, T., and Walt, H. Inter-
phase cytogenetics in paraffin-embedded sections from human testiculargerm cell tumor xenografts and in corresponding cultured cells. Lab. Invest.,61:235-242, 1989.
63. Arnoldus, E. P. J., Dreef, E. J., Noordermeer, I. A., Verheggen, M. M.,
Thierry, R. F., Peters, A. C. B., Cornelisse, C. J., Van der Ploeg, M., and Raap,A. K. Feasibility of in situ hybridization with chromosome-specific DNAprobes on paraffin wax-embedded tissue. J. Clin. Pathol., 44: 900-904,
1991.
64. Hopman, A. H. N., van Hooren, E., van de Kaa, C. A., Vooijs, P. G. P.,
and Ramaekers, F. C. S. Detection of numerical chromosome aberrationsusing in situ hybridization in paraffin sections of routinely processed bladder
cancers. Mod. Pathol., 4: 503-51 3, 1991.
65. Dhingra, K., Sahin, A., Supak, J., Kim, S. Y., Hortobagyi, G., and
Hittelman, W. N. Chromosome in situ hybridization on formalin-fixed mam-mary tissue using non-isotopic, non-fluorescent probes: technical consider-
ations and biological implications. Breast Cancer Res. Treat., 23: 201-2 10,1992.
66. Kim, S. Y., Lee, J. S., Ro, J. Y., Gray, M. L., Hong, W. K., and Hittelman,W. N. Interphase cytogenetics in paraffin sections of lung tumors by non-isotopic in situ hybridization: mapping genotype/phenotype heterogeneity.Am. J. Pathol., 142: 307-31 7, 1993.
67. Voravud, N., Shin, D. M., Ro, J. Y., Lee, J. S., Hong, W. K., andHittelman, W. N. Increased polysomies of chromosomes 7 and 1 7 duringhead and neck cancer multistage tumorigenesis. Cancer Res., 53:2874-2883, 1993.
68. Lofroth, G. Environmental tobacco smoke: overview of chemical com-position and genotoxic components. Mutat. Res., 222: 73-80, 1989.
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 707
69. Kao-Shan, C. S., Fine, R. L., Whang-Peng, J., Lee, E. C., and Chabner,B. A. Increased fragile sites and sister chromatid exchanges in bone marrowand peripheral blood of young cigarette smokers. Cancer Res., 47:
6278-6282, 1987.
70. Schantz, S. P., Spitz, M. R., and Hsu, T. C. Mutagen sensitivity in headand neck cancer patients: a biologic marker for risk of multiple primarymalignancies. J. NatI. Cancer Inst., 82: 1773-1775, 1990.
71 . Hsu, T. C., Spitz, M. R., and Schantz, S. P. Mutagen sensitivity: a bio-logical marker of cancer susceptibility. Cancer Epidemiol., Biomarkers &
Prey., 1:83-89, 1991.
72. Lee, J. S., Kim, S. Y., Hong, W. K., Lippman, S. M., Ro, J. Y., Gay, M. L.,Batsakis, J. G., Toth, B., Weber, B. S., Martin, J. W., and Hittelman, W. N.Detection of chromosomal aneuploidy in oral leukoplakia, a premalignant
lesion (Abstract). Proc. Am. Soc. Clin. Oncol., 11: 102, 1992.
73. Fugita, J.. Yoshida, 0., Tusas, Y., Rhim, J. S., Hatamaka, M., and
Aaronson, S. A. Ha-ras oncogenes are activated by somatic alterations in
human urinary tract tumors. Nature (Lond.), 309: 464-466, 1984.
74. Eva, A., Tronick, S. R., Gol, R. A., Pierce, J. H., and Aaronson, S. A.Transforming genes of human hematopoietic tumors: frequent detection ofras-related oncogenes whose activation appears to be independent of tumor
phenotype. Proc. NatI. Acad. Sci. USA, 80: 4926-4930, 1983.
75. Santos, F., Martin-Zanca, D., Reddy, E. P., Pierotti, M. A., Della Porta,
G., and Barbacid, M. J. Malignant activation of K-ras oncogene in lungcarcinoma, but not in normal tissue of the same patients. Science(Washington, DC), 223: 661-668, 1984.
76. Bos, J. L. ras Oncogenes in human cancer: a review. Cancer Res., 49:4682-4689, 1989.
77. Shih, C., Shelo, B. F., Goldfarb, M. P., Dannenberg, A., and Weinberg,
R. A. Passage of phenotypes of chemically transformed cells via transfection
of DNA and chromatin. Proc. NatI. Acad. Sci. USA, 76: 571 4-571 8, 1979.
78. Der, C. J., Krontris, T. G., and Cooper, G. M. Transforming genes ofhuman bladder and lung carcinoma cell lines are homologous to the genesof Harvey and Kirsten sarcoma virus. Proc. NatI. Acad. Sci. USA, 79:
3637-3640, 1982.
79. Der, C. J., and Cooper, G. M. Altered gene products are associated withactivation ofcellularABSKgenes in human lungand colon carcinomas. Cell,
32:201-208, 1983.
80. Goldfarb, M. P., Shimizn, K., Perucho, M., and Wigler, M. H. Isolationand preliminary characterization of a human transforming gene from T24bladder carcinoma cells. Nature (Lond.), 296: 405-409, 1982.
81 . Papageorge, A., Lowy, D., and Scolnick, E. Comparative biochemicalproperties of p2i ras molecules coded for by viral and cellular genes.
J. Virol., 44: 509-519, 1982.
82. McCormick, F. ras GTPase activating protein: signal transmitter andsignal terminator. Cell, 56: 5-8, 1989.
83. McCormick, F. The world according to cute GAP. Oncogene, 5:1281-1283, 1990.
84. Hall, A. ras and GAP-Who’s controlling whom? Cell, 61: 921-923,1990.
85. Lemoine, N. R., MayaII, E. S., Wyllie, F. S., Williams, E. D., Goyns, M.,Stringer, B., and Wynford-Thomas, D. High frequency of ras oncogene
activation in all stages of human thyroid tumorigenesis. Oncogene, 4:159-164, 1989.
86. Bos, J. L., Tokzos, D., Marshall, C. L., Verlaan-de Vries, M., Veeneman,G. H., Van der Eb, A. J., Van boom, J. H., Janssen, J. W. G., and Steenvoor-
den, A. C. M. Amino-acid substitutions at codon 1 3 ofthe N-ras oncogene in
human acute myeloid leukemia. Nature (Lond.), 315: 726-730, 1985.
87. Almoguera, C., Shibata, D., Forrester, K., Martin, J., Arrnheim, N., andPerucho, M. Most human carcinomas of the exocrine pancreas contain
mutant c-K-ras genes. Cell, 53: 549-554, 1988.
88. Smit, V. T. H. B. M., Boot, A. J. M., Smits, A. M. M., Fleuren, G. J.,Cornelisse, C. J., and Bos, J. L. K-ras codon 1 2 mutations occur very fre-quently in pancreatic adenocarcinomas. Nucleic Acids Res., 16:7773-7782,
1988.
89. Bos, J. L., Fearon, E. R., Hamilton, S. R., Verlaan-de Vries, M., VanBoom, J. H., Van der Eb, A. J., and Vogelstein, B. Prevalence of ras genemutations in human colorectal cancer. Nature (Lond.), 327: 293-297, 1987.
90. Forrester, K., Almoguera, C., Han, K., Grizzle, W. E., and Perucho, M.Detection of high incidence of K-ras oncogenes during human colon tumor-igenesis. Nature (Lond.), 327: 298-303, 1987.
91 . Slebos, R. J. C., Evers, S. G., Wagenaar, S. S., and Rodenhuis, S. Cellularproto-oncogenes are infrequently amplified in untreated non-small cell lung
cancer (NSCLC). Br. J. Cancer, 59: 76-80, 1988.
92. Rodenhuis, S., Van De Wetering, M. L., Mooi, W. J., Evers, S. G., VanZandwijk, N., and Bos, J. L. Mutational activation of the K-ras oncogene: a
possible pathogenic factor in adenocarcinoma of the lung. N. EngI. J. Med.,
317: 929-935, 1987.
93. Rodenhuis, S., Slebos, R. J. C., Boot, A. J. M., Evers, S. G., Mooi, W. J.,
Wagenaar, S. S., Van Bodegom, P. C., and Bos, J. L. Incidence and possibleclinical significance of K-ras oncogene activation in adenocarcinoma of thehuman lung. Cancer Res., 48: 5738-5741 , 1988.
94. Mitsudomi, T., Viallet, J., Mulshine, J. L., Linoilla, R. I., Minna, J. D., and
Gazdar, A. F. Mutations of genes distinguish a subset of non-small cell lung
cancer cell lines from small-cell lung cancer cell lines. Oncogene, 6:1353-1362, 1991.
95. Azuma, M., Furumoto, N., Kawamata, H., Yoshida, H., Yanagawa, 1.,Yura, Y., Hayashi, Y., Takegawa, Y., and Sato, M. The relation of rasoncogene product p21 expression to clinicopathological status: criteria and
clinical outcome in squamous cell head and neck cancer. Cancer J., 1:
375-380, 1987.
96. Furth, M. E., Davis, L. J., Flenrdelys, B., and Scolnick, E. M. Monoclonalantibodies to the p21 products of the transforming gene of Harvey murine
sarcoma virus and of the cellular ras gene family. 1. Virol., 43: 294-304,1982.
97. Sheng, Z. M., Barrois, M., Klijanienko, J., Micheau, C., Richard, J. M.,and Rio, G. Analysis of the c-Ha-ras-1 gene for deletion, mutation, amplifi-cation and expression in lymph node metastases of human head and neck
cancer. Br. J. Cancer, 62: 398-404, 1990.
98. Rumsby, G., Carter, R. C., and Gusterson, B. A. Low incidence of rasoncogene activation in human squamous cell carcinomas. Br. I. Cancer, 61:
365-368, 1990.
99. Warnakulasuriya, K. A. A. S., Chang, S. E., and Johnson, N. W. Pointmutations in the Ha-ras oncogene are detectable in formalin-fixed tissues of
oral squamous cell carcinomas, but are infrequent in British cases. J. Oral
Pathol. Med., 21:225-229, 1992.
100. Sakai, E., Rikimaru, K., Ueda, M., Matsumoto, Y., Ishii, N., Enomoto,S., Yamamoto, H., and Tsuchida, N. The p53 tumor-suppiessor gene and ras
oncogene mutations in oral squamous cell carcinoma. tnt. J. Cancer, 52:867-872, 1992.
101 . Chang, 5. E., Bhatia, P., Johnson, N. W., Morgan, P. R., McCormick, F.,
Young, B., and Hiorns, L. ras mutations in United Kingdom examples of oralmalignancies are infrequent. Int. J. Cancer, 48: 409-412, 1991.
102. Saranath, D., Chang, S. E., Bhoite, L. T., Panchall, R. G., Kerr, I. B.,
Mehta, A. R., Johnson, N. W., and Deo, M. G. High frequency mutation incodons 1 2 and 61 of Ha-ras oncogene in tobacco-related human oralcarcinomas. Br. J. Cancer, 63: 573-578, 1 991.
103. Farr, C. J., Marshall, C. J., Easty, D. J., Wright, N. A., Powell, S. C., and
Pakeva, C. A study of gene mutations in colonic adenomas from familialpolyposis coli patients. Oncogene, 3: 673-678, 1988.
104. Sidransky, D., Tokino, T., Hamilton, S. R., Kinzler, K. W., Levin, B.,
Frost, P., Vogelstein, B. Identification of ras oncogene mutations in the stool
of patients with curable colorectal tumors. Science (Washington, DC), 256:102-105, 1992.
105. Vennestrom, B., and Bishop, 1. M. Isolation and characterization ofchicken DNA homologous to the two putative oncogenes of avian erythro-
blastosis virus. Cell, 28: 135-143, 1982.
106. Lin, C. R., Chen, W. S., Kruijer, W., Stolarsky, L. S., Weber, W., Evans,R. M., Verma, L. M., Gill, G. N., and Rosenfeld, M. G. Expression cloning of
human EGF receptor complementary DNA: gene amplification and threerelated messenger RNA products in A431 cells. Science (Washington, DC),224: 843-848, 1984.
107. Ullrich, A., Coussens, L., Hayflick, J. S., Dull, T. J., Gray, A., Tam,
A. W., Lee, J., Yarden, Y., Libermann, T. A., Schlessinger, J., Downward, J.,Mayes, E. L. V., Whittle, N., Waterfield, M. D., and Seeburg, P. H. Humanepidermal growth factor receptor cDNA sequence and aberrant expression of
the amplified gene in A431 epidermoid carcinoma cells. Nature (Lond.),309:418-425, 1984.
108. Xu, Y. H., lshii, S., Clark, A. J. L., Sullivan, M., Wilson, R. K., Ma, D. P.,
Roe, B. A., Merlino, G. T., and Pastan, I. Human epidermal growth factorreceptor cDNA is homologous to a variety of RNAs overproduced in A431
carcinoma cells. Nature (Lond.), 309: 806-81 0, 1984.
109. Sachs, P. G., Parnes, 5. M., Gallick, G. E., Mansouri, Z., Lichter, R.,
Satya-Prakash, K. L., Pathak, S., and Parsons, D. F. Establishment and char-acterization of two new squamous cell carcinoma cell lines derived fromtumors of head and neck. Cancer Res., 48: 2858-2866, 1988.
110. Maxwell, S. A., Sacks, P. G., Gutterman, J. U., and Gallick, G. E. Epi-
dermal growth factor receptor protein-tyrosine kinase activity in human celllines established from squamous carcinomas of the head and neck. Cancer
Res., 49:1130-1137, 1989.
1 1 1 . Santini, J., Formento, J. L., Francoual, M., Milano, G., Schneider, M.,Dassonville, 0., and Dernard, F. Characterization, quantification and
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
708 Review: Biomarkers in Upper Aerodigestive Trad Tumerigenesis
potential clinical value of the epidermal growth factor receptor in head andneck squamous cell carcinomas. Head Neck, 1 3: 1 32-i 39, 1 991.
1 1 2. Kearsley, J. H., Furlong, K. L., Cooke, R. A., and Waters, M. J. Animmunohistochemical assessment of cellular proliferation markers in headand neck squamous cell cancers. Br. J. Cancer, 61: 821-827, 1990.
1 1 3. Cline, M. I., and Battifora, H. Abnormalities of proto-oncogenes in
non-small cell lung cancer. Correlations with tumor type and clinical char-acteristics. Cancer (Phila.(, 60: 2669-2674, 1987.
114. Berger, M. S., Gullick, W. J., Greenfield, C., Evans, S., Addis, B. J., andWaterfield, M. D. Epidermal growth factor receptors in lung tumors.
J. Pathol., 152:297-307, 1987.
1 1 5. Shiraishi, M., Noguchi, M., Shimosato, Y., and Sekiya, T. Amplification
of proto-oncogenes in surgical specimens of human lung carcinomas.Cancer Res., 49: 6474-6479, 1989.
1 1 6. Her.dler, F. I., and Ozanne, B. W. Human squamous cell lung cancersexpress increased epidermal growth factor receptors. J. Clin. Invest., 74:
647-651, 1984.
1 1 7. Cerny, T., Barnes, D. M., Hasleton, P., Barber, P. V., Healy, K., Gullick,W., and Thatcher, N. Expression ofepidermal growth factor receptor (EGF-R)
in human lung tumors. Br. J. Cancer, 54: 265-269, 1986.
1 1 8. Johnson, M. D., Gray, M. E., Carpenter, G., Pepinsky, R. B., and Stahl-
man, M. T. Ontogeny of epidermal growth factor receptor and lipocortin-1 in
fetal and neonatal human lungs. Hum. Pathol., 21: 182-191, 1990.
119. Mendelsohn, J., Masui, H., Sunada, H., and MacLeod, C. Monoclonal
antibodies against the receptor for epidermal growth factor as potentialanticancer agents. In: J. Minna, W. M. Kuehl (eds.), Cellular and Molecular
Biology of Tumors and Potential Clinical Applications, p. 307. New York:Alan R. Liss, 1988.
1 20. Perez-Soler, R., Donato, N. J., Shin, D. M., Rosenblum, M. G., Zhang,H-Z., Tornos, C., Brewer, H., Shah, 1., Chan, J. C., Thompson, L. B., Janus,
M., Lee, J. S., Hong, W. K., and Murray, J. L. Tumor epidermal growth factor
receptor (EGFR) studies in patients with non-small cell lung cancer and headand neck cancer treated with EGFR monoclonal antibody RG83852. J. Clin.
Oncol., 12:730-739, 1994.
121. Shin, D. M., Ro, J. Y., Shah, T., Hong, W. K., and Hittelman, W. M.Dysregulation of epidermal growth factor receptor (EGFR) expression in themultistage process of head and neck carcinogenesis. Cancer Res., 54:
3153-3159, 1994.
1 22. Sakamoto, H., Mori, M., Taira, M., Yoshida, T., Matsukawa, S.,
Shimizu, K., Sekiguchi, M., Terada, M., and Sugimura, 1. Transforming genefrom human stomach cancers and a noncancerous portion of stomach
mucosa. Proc. NatI. Acad. Sci. USA, 83: 3997-4001, 1986.
1 23. Yoshida, M. C., Wada, M., Satoh, H., Yoshida, T. Y., Sakamoto, H.,Myagawa, K., Yokota, J., Koda, T., Kakinuma, M., Sugimura, T., and Terada,
M. Human HST2 (HSTF1) gene maps to chromosome band iiql3 andcoamplifies with the int-2gene in human cancer. Proc. NatI. Acad. Sci. USA,
85: 4861-4864, 1988.
1 24. Taira, M., Yoshida, T., Myagawa, K., Sakamoto, H., Terada, M., and
Sugimura, T. cDNA sequence of human transforming gene hst and identifi-cation of the coding squamous required for transforming activity. Proc. NatI.Acad. Sci. USA, 84:2980-2984, 1987.
1 25. Yoshida, T., Miyagawa, K., Odagiri, H., Sakamoto, H., Little, P. F. R.,Terada, M., and Sugimura, T. Genomic sequence of hst, a transforming gene
encoding a protein homologous to fibroblast growth factor and the int-2-encoded protein. Proc. NatI. Acad. Sci. USA, 84: 7305-7309, 1987.
126. Casey, G., Smith, R., McGilliurary, D., Peters, G., and Dickson, C.
Characterization and chromosome assignment of the human homolog of
int-2, a potential proto-oncogene. Mol. Cell Biol., 6: 502-510, 1986.
1 27. Tsutsumi, M., Sakamoto, H., Yoshida, T., Kakizoe, T., Koiso, K.,Sugimura, T., and Terada, M. Coamplification of the hst- 1 and int-2 genes inhuman cancer. Jpn. J. Cancer Res. (Gann), 79:428-432, 1988.
1 28. Tsuda, Y., Tahara, E., Kajiyama, G., Sakamoto, H., Terada, M., andSugimora, 1. High incidence of coamplification of hst-1 and int-2 genes inhuman esophageal carcinomas. Cancer Res., 49: 5505-5508, 1989.
129. Adelaide, J., Matter, M. G., Marics, I., Raybaund, F., Planche, J.,
Lapeyriere, 0. D., and Birnbaum, D. Chromosomal localization of the hstoncogene and its coamplification with the int-2 oncogene in human mela-noma. Oncogene, 2: 41 3-41 6, 1988.
1 30. Ali, I. U., Merlo, G., Callahan, R., and Liderean, R. The amplification
unit of chromosome 11q13 in aggressive primary breast tumors entails thebcl-1, int-2, and hst loci. Oncogene, 4: 89-92, 1989.
1 31 . Hatada, I., Tokino, T., Ochiya, T., and Matsubara, K. Coamplificationof integrated hepatitis B virus DNA and transforming gene hst-1 in a hepa-
tocellular carcinoma. Oncogene, 3: 537-540, 1988.
1 32. Somers, K. D., Cartwright, S. L., and Schechter, G. L. Amplification ofthe int-2 gene in human head and neck squamous carcinomas. Oncogene,
5:915-920, 1990.
1 33. Zhou, D. J., Casey, G., and Cline, M. J. Amplification of human int-2 in
breast cancers and squamous carcinomas. Oncogene, 2: 279-282, 1988.
1 34. Roh, H. 1., Shin, D. M., Lee, J. S., Ro, J. Y., Tainsky, M. A., Hong,
W. K., and Hittelman, W. N. Visualization of int-2 amplification in prema-lignant lesions during head and neck tumorigenesis. Proc. Am. Assoc.Cancer Res., 35: 1 1 7, 1 994 (abstract).
1 35. Lane, D. P., and Crawford, L. V. T antigen is bound to a host protein in
SV-40-transformed cells. Nature (Lond.), 278: 261-268, 1970.
1 36. Linzer, D. H., and Levine, A. J. Characterization of a 54K dalton cel-
lular SV4O tumor antigen present in 5V40-transformed cells and uninfectedembryonal carcinoma cells. Cell, 17: 43-52, 1979.
1 37. Finlay, C. A., Hinds, P. W., and Levine, A. J. The p53 proto-oncogene
can act as a suppressor of transformation. Cell, 57: 1 083-1 093, 1989.
1 38. Baker, S. J., Presinger, A. C., Jessup, J. M., Paraskeva, C., Markowitz, S.,Willson, J. K. V., Hamilton, S., and Vogelstein, B. p53 gene mutations occur
in combination with 1 7p allelic deletions as late events in colorectal tumor-
igenesis. Cancer Res., 50: 7717-7722, 1990.
1 39. Takahashi, T., Suzuki, H., Hida, T., Sekido, Y., Ariyoshi, Y., and Ueda,R. The p53 gene is very frequently mutated in small cell lung cancer with adistinct nucleotide substitution pattern. Oncogene, 6: 1 775-1 778, 1 991.
1 40. Sidransky, D., Eschenbach, A. V., Tsai, Y. C., Jones, P., Summerhayes,I., Marshall, F., Paul, M., Green, P., Hamilton, S. R., Frost, P., and Vogelstein,
B. Identification of p53 mutations in bladder cancer and urine samples.
Science (Washington, DC), 252: 706-709, 1 991.
1 41 . Farrell, P. J., Allan, G. J., Shanahan, F., Vousden, K. H., and Crook, T.
p53 is frequently mutated in Burkett’s lymphoma cell lines. EMBO J., 10:
2879-2887, 1991.
1 42. Bressac, B., Kew, M., Wands, J., and Ozturk, M. Selective G to Tmutations of pS3 gene in hepatocellular carcinoma from southern Africa.Nature (Lond.), 350: 429-431 , 1991.
143. Hsu, I. C., Metcalf, R. A., Sun, T., Welsh, J. A., Wang, N. J., and Harris,
C. C. Mutational hotspot in the p53 gene in human hepatocellular carcino-mas. Nature (Lond.), 350: 427-428, 1991.
1 44. Chiba, I., Takahashi, T., and Nau, M. M. Mutations in the p53 gene are
frequent in primary, resected non-small cell lung cancer. Oncogene, 5:1603-1610, 1990.
145. Hollstein, M. C., Pen, U., Mandard, A. M., Welsh, J. A., Montesano, R.,
Metcalf, R. A., Bak, M., and Harris, C. C. Genetic analysis of human esoph-ageal tumors from two high-incidence geographic areas: frequent p53 basesubstitutions are absence of ras mutations. Cancer Res., 51: 4102-4106,1991.
146. Okamoto, A., Sameshima, Y., Yamada, Y., Teshima, S., Tehima, Y.,
Terada, M., and Yokoda, J. Allelic loss of chromosome 1 7p and p53 muta-lions in human endometrial carcinoma of the uterus. Cancer Res., 51:
5632-5636, 1991.
147. Cole, R. J., Jhanwar, S. C., Novick, S., and Pellicer, A. Genetic alter-ations ofthe p53 gene are a feature of malignant mesothelioma. Cancer Res.,
51:5410-5416, 1991.
148. Bennett, W. P., Hollstein, M. C., He, A., Zhu, S. M., Resau, J. H.,Trump, B. F., Metcalf, R. A., Welsh, J. A., Midgley, C., Jane, D. P., and
Harris, C. C. Archival analysis of p53 genetic and protein alterations in
Chinese esophageal cancer. Oncogene, 6: 1 779-1 784, 1991.
149. Tamura, G., Kihana, T., Nomura, K., Terada, M., Sugimura, T., and
Hirohashi, S. Detection of frequent p53 gene mutations in primary gastriccancer by cell sorting and polymerase chain reaction-single-strand confor-mation polymorphism analysis. Cancer Res., 51: 3056-3058, 1991.
1 50. Osborne, R. J., Merlo, G. R., Mitsudomi, T., Venesio, T., Liscia, D. S.,
Cappa, A. P. M., Chiba, I., Takahashi, T., Nau, M. M., Callahan, R., andMinna, J. D. Mutations in the p53 gene in primary human breast cancers.
Cancer Res., 51:6194-6198, 1991.
1 51 . Kovach, J. S., McGovern, R. M., Cassady, J. D., Swanson, S. K., Wold,L. E., Vogelstein, B., and Sommer, S. S. Direct sequencing from touch prep-arations of human carcinomas: analysis of p53 mutations in breast carcino-mas. J. NatI. Cancer Inst., 83: 1004-1009, 1991.
1 52. Shirasawa, S., Urabe, K., Yanagawa, Y., Toshitani, T., Iwana, T., and
Sasazuki, T. p53 gene mutations in colorectal tumors from patients withfamilial polyposis coli. Cancer Res., 51: 2874-2878, 1991.
153. Hollstein, M., Sidransky, D., Vogelstein, B., and Harris, C. C. p53mutations in human cancers. Science (Washington, DC), 253: 49-53, 1991.
1 54. Ishirka, C., Sato, 1., Gumoh, M., Suzuki, T., Shibata, H., Kanamaru, R.,
Wakui, A., and Yamazaki, T. Mutations of the p53 gene, including an
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
Cancer Epidemiology, Biomarkers & Prevention 709
intronic point mutation in colorectal tumors. Biochem. Biophys. Res. Com-mun., 177:901-906, 1991.
1 55. Shaw, P., Tardy, S., Benito, E., Obrador, A., and Costa, J. Occurrence
of Ki-ras and p53 mutations in primary colorectal tumors. Oncogene, 6:2121-2128, 1991.
1 56. Brash, D. E., Rudolph, J. A., Simon, J. A., Lin, A., McKenna, G. J.,Baden, H. J.P., Halperin, H. J.,and Poten, J.A role for sunlight in skin
cancer. UV-induced p53 mutations in squamous cell carcinomas. Proc. NatI.Acad. Sci. USA, 88: 10124-10128, 1991.
1 57. Kastan, M. B., Zhan, Q., El-Deiry, W. S., Carris, F., Jacks, T., Welsh,W. V., Plunkett, B. S., Vogelstein, B., and Fornance, A. J. A mammalian cell
cycle checkpoint pathway utilizing p53 and GADD4S is defective in ataxia-
telangiectasis. Cell, 71: 587-597, 1992.
1 58. Yin, V., Tainsky, M. A., Bischoff, F. Z., Strong, L. C., and WahI, G. M.Wild-type p53 restores cell cycle control and inhibits gene amplification in
cells with a mutant p53 allele. Cell, 70: 937-948, 1992.
1 59. Livingston, L. R., White, A., Sprouse, J., Livanos, E., Jacks, T., and Tisty,1. D. Altered cell cycle arrest and gene amplification potential accompany
loss of wild-type pS3. cell, 70: 923-935, 1992.
1 60. Hartwell, K. H., and Weinert, T. A. Checkpoints: controls that ensure
the order of cell cycle events. Science (Washington, DC), 246: 629-633,1989.
1 61 . Kastan, M. B., Onykwere, 0., Sidransky, D., Vogelstein, B., and Craig,
R. W. Participation of p53 protein in the cellular response to DNA damage.
CancerRes., 51:6304-6311, 1991.
1 62. Slichenmyer, W. J., Nelson, W. C., Slebos, R. J., and Kastan, M. B. Loss
of a p53-associated G1 checkpoint does not decrease cell survival followingDNA damage. Cancer Res., 53: 4164-4168, 1993.
163. Rowley, R., Subramani, S., and Young, P. G. Checkpoint control in
Schizosaccharomyces pombe. EMBO J., 1 1: 1 335-1 342, 1992.
1 64. Busse, P. M., Bose, S. K., Jones, R. W., and Tolmach, L. J. The action of
caffeine on x-irradiated HeLa cells. Radiat. Res., 76: 292-307, 1978.
165. O’Connor, P. M., Ferris, D. K., White, G. A., Pines, J., Hunter, T.,
Longo, D. L., and Kohn, K. W. Enhancement of x-ray-induced killing during
�2 arrest. Cell Growth Duff., 3:42-52, 1992.
1 66. Lane, D. P., and Benchimol, S. p53: oncogene or anti-oncogene?
Genes Develop., 4: 1-8, 1990.
167. Iggo, R., Gatter, K., Bartek, J., Lane, D., and Harris, A. L. Increased
expression of mutant forms of pS3 oncogene in primary lung cancer. Lancet,335: 675-679, 1990.
168. D’Amico, D., Carbone, D., Mitsudomi, T., Nau, M., Fedorko, J.,
Russell, E., Johnson, B., Buchhagen, D., Bodner, S., Phelps, R., Gazdar, A.,and Minna, I. D. High frequency of somatically acquired p53 mutations insmall cell lung cancer cell lines and tumor. Oncogene, 7: 339-346, 1991.
169. Casson, A. G., Mukhopadhyay, T., Cleary, K. R., Ro, J. Y., Levin, B.,and Roth, J. A. p53 gene mutations in Barrett’s epithelium and esophagealcancer. Cancer Res., 51:4495-4499, 1991.
1 70. Bennet, W. P., colby, T. V., Travis, D. W., Borkowski, A., Jones, R. 1.,Lane, D. P., Metcalf, R. A., Samet, J. M., Takeshima, Y., Gu, J. R., Vahakan-gas, K. H., Soini, Y., Paakko, P., Welsh, J. A., Trump, B. F., and Harris, C. C.p53 protein accumulates frequently in early bronchial neoplasia. Cancer
Res., 53:4817-4822, 1993.
1 71 . Nees, M., Homann, N., Discher, H., AndI, T., Enders, C., Herold-Mende, C., Schuhmann, A., and Bosch, F. X. Expression of mutated p53
occurs in tumor-distant epithelia of head and neck cancer patients: a possiblemolecular basis for the development of multiple tumors. Cancer Res., 53:4189-4196, 1993.
1 72. Boyle, J. 0., Hakim, J., Koch, W., van der Riet, P., Hruban, R. H., Roa,R. A., Correo, R., Eby, Y. I., Ruppert, J. M., and Sidransky, D. The incidenceof p53 mutations increases with progression of head and neck cancer.CancerRes., 53:4477-4480, 1993.
1 73. Shin, D. M., Kim, J., Ro, J. Y., Hittelman, J., Roth, J. A., Hong, W. K.,and Hittelman, W. N. Activation of p53 gene expression in premalignantlesions during head and neck tumorigenesis. Cancer Res., 54: 321-326,
1994.
1 74. Lippman, S. M., Shin, D. M., Lee, J. J., Hittelman, W. N., and Hong,W. K. p53 in oral carcinogenesis and retinoid chemoprevention. Proc. Am.
Soc. Clin. Oncol., 13: 1 71 , 1 994 (abstract).
1 75. Lee, I. S., Ro, J. Y., Sahin, A., Hong, W. K., and Hittelman, W. N.
Quantitation of proliferating cell fraction (pcf) in non-small cell lung cancer(NSCLC) using immunostaining for proliferating cell nuclear antigen (PdNA).
Proc. Am. Assoc. Cancer Res., 31: 22, 1 990 (abstract).
1 76. Lippman, S. M., Lee, J. S., Peters, E., Ro, J., Wargovich, M., Morice, R.,Hittelman, W., and Hong, W. K. Expression of proliferating cell nuclear
antigen (PCNA) correlates with histologic stage of bronchial carcinogenesis.Proc. Am. Assoc. Cancer Res., 31: 168, 1990 (abstract).
1 77. Celis, J. E., and Celis, A. Cell cycle-dependent variations in the distri-
bution of the nuclear protein cyclin proliferating cell nuclear antigen in
cultured cells: subdivision of S phase. Proc. NatI. Acad. Sci. USA, 82:3262-3266, 1985.
1 78. Bravo, R., and Celis, J. E. A search for differential polypeptide synthesis
throughout the cell cycle of HeLa cells. I. Cell Biol., 84: 795-802, 1980.
1 79. Bravo, R., Frank, R., Blundell, P. A., and MacDonald-Bravo, H. Cyclin/PCNA is the auxiliary protein of DNA polymerase alpha. Nature (Lond.),
326:515-517, 1987.
180. Lee, J. S., Lippman, S. M., Hong, W. K., Ro, J. Y., Kim, S. Y., Lotan, R.,
and Hittelman, W. N. Determination of biomarkers for intermediate end
points in chemoprevention trials. Cancer Res. (Suppl.), 52: 2707s-2710s,
1992.
181. Shin, D. M., Voravud, N., Ro, J. Y., Lee, J. S., Hong, W. K., andHittelman, W. N. Sequential increase in proliferating cell nuclear antigen inhead and neck tumorigenesis: a potential biomarker. I. NatI. Cancer Inst., 85:
971-978, 1993.
182. Yang, G. C., Lipkin, M., Yang, K., Wang, G. Q., U, J. Y., Yang, C. S.,
and Winawer, S. Proliferation of esophageal epithelial cells among residentsof Linxian, People’s Republic of China. I. NatI. Cancer Inst., 79: 1 241-1 246,
1987.
183. Lipkin, M. Biomarkers of increased susceptibility to gastrointestinalcancer: new application to studies of cancer prevention in human subjects.
Cancer Res., 48: 235-245, 1988.
1 84. Lipkin, M., Friedman, E., Winawer, S. J., Newmark, H., Blot, W. J., andFraumeni, J. F., Jr. Colonic epithelial cell proliferation in responders and
nonresponders to supplemental dietary calcium. Cancer Res., 49: 248-254,
1989.
1 85. Lotan, R. Effects of vitamin A and its analogs (retinoids) on normal and
neoplastic cells. Biochem. Biophys. Acta., 605: 33-91, 1980.
186. Lippman, S. M., and Meyskins, F. L. Retinoids for the prevention of
cancer. In: T. E. Moon, M. Micozzi (eds.), Nutrition and Cancer Prevention:
The Role of Micronutrients, pp. 243-272. New York: Marcel Dekker, 1989.
187. Shin, D. M., Gimenez, I. B., Lee, J. S., Nishioka, K., Wargovich, M. J.,Thacker, S., Lotan, R., Slaga, T. J., and Hong, W. K. Expression of epidermalgrowth factor receptor, polyamine levels, ornithine decarboxylase activity,
micronuclei, and transglutaminase I in a 7,12-dimethylbenz(a)anthracene-
induced hamster buccal pouch carcinogenesis model. Cancer Res., 50:
2505-2510, 1990.
188. Gimenez-Conti, J. B., Shin, D. M., Bianchi, A. B., Roop, D. R., Hong,W. K., Contri, C. J., and Slaga, T. J. Changes in keratin expression during7,1 2-dimethylbenz(a)-anthracene-induced hamster cheek pouch carcino-
genesis. Cancer Res., 50: 4441-4445, 1990.
1 89. Eckert, R. L. Structure, function, and differentiation of the keratinocyte.Physiol. Rev., 69: 1 31 6-1 346, 1989.
190. Simon, M., and Green, H. Enzymatic cross-linking of involucrin andother proteins by keratinocyte particulates in vitro. Cell, 40: 677-683, 1985.
191 . Thacher, S. M., and Rice, R. H. Keratinocyte-specific transglutaminaseof cultured human epidermal cells: relation to cross-linked envelope forma-lion and terminal differentiation. Cell, 40: 685-695, 1985.
192. Kaplan, M. J., Mills, S. E., Rice, R. H., and Johns, M. E. Involucrin inlaryngeal dysplasia. Arch. Otolaryngol., 1 10: 71 3-71 6, 1984.
193. Murphy, G. M., Flynn, T. C., Rice, R. H., and Pinkus, G. S. Involucrinexpression in normal and neoplastic human skin: a marker for keratinocytedifferentiation. J. Invest. Dermatol., 82: 453-457, 1984.
194. Ta, B. M., Gallagher, G. T., Chakravarty, R., and Rice, R. H. Keratino-
cyte transglutaminase in human skin and oral mucosa: cytoplasmic local-ization and uncoupling of differentiation markers. J. Cell Sci., 95: 63 1-638,1990.
195. Rubin, A. L., and Rice, R. H. Differential regulation by retinoic acidand calcium of transglutaminases in cultured neoplastic and normal humankeratinocytes. cancer Res., 46: 2356-2361, 1986.
196. Lotan, R., Sacks, P. G., Lotan, D., and Hong, W. K. Differential effects
of retinoic acid on the in vitro growth and cell-surface glycoconjugates oftwo human head and neck squamous cell carcinomas. nt. J. Cancer, 40:
224-229, 1987.
197. Xu, X-C., Ro, J. Y., Lee, J. S., Shin, D. M., Hittelman, W. N., Lippman,S. M., Toth, B. B., Martin, J. W., Hong, W. K., and Lotan, R. Differential
expression of nuclear retinoic acid receptors in surgical specimens from headand neck “normal”, hyperplastic, premalignant, and malignant tissues. Proc.
Am. Assoc. Cancer Res., 34: 551 , 1 993 (abstract).
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from
1994;3:697-709. Cancer Epidemiol Biomarkers Prev D M Shin, W N Hittelman and W K Hong review.Biomarkers in upper aerodigestive tract tumorigenesis: a
Updated version
http://cebp.aacrjournals.org/content/3/8/697
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
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://cebp.aacrjournals.org/content/3/8/697To request permission to re-use all or part of this article, use this link
on March 19, 2020. © 1994 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from