Molecular and cellular alterations in tobacco smoke-associated larynx cancer

Ž .Mutation Research 445 1999 259–274www.elsevier.comrlocatergentox

Community address: www.elsevier.comrlocatermutres

Molecular and cellular alterations in tobacco smoke-associatedlarynx cancer

K. Szyfter a,), Z. Szmeja b, W. Szyfter b, K. Hemminki c, J. Banaszewski b,R. Jaskuła-Sztul a, J. Louhelainen c

a Institute of Human Genetics, Polish Academy of Sciences, ul. Strzeszynska 32, Poznan 60-479, Poland´ ´b Clinic of Otolaryngology, K. Marcinkowski Medical UniÕersity, Poznan, Poland´

c Center for Nutrition and Toxicology, Karolinska Institute, Huddinge, Sweden

Received 15 November 1998; accepted 15 December 1998

Abstract

Tumours of head and neck belong to the most frequent types of cancer world-wide. In Poland, mortality from larynxcancer among males has been continuously increasing during the last decades up to 8.4 deaths per 100,000 men in 1993,which exceeds epidemiological records from other countries. The aetiology of laryngeal cancer is strongly associated withexposure to carcinogens present in tobacco smoke. The review describes a sequence of molecular and cellular events from

Ž .carcinogenic exposure, DNA adduct formation, detection of mutations in the p53 gene, loss of heterozygosity LOH inchromosomal loci encoding the p53 and p16 genes, and loss of control of the cell cycle. The section concerning DNAadducts includes a discussion of the role of such confounders as exogenous exposure, the age and sex of the subject, anddisease progression. The significance of genetic factors as individual risk determinants is discussed in relation tobleomycin-induced chromosome instability and in connection with the occurrence of defects in genes encoding detoxifyingenzymes. The question concerning the substantial difference between men and women in larynx cancer morbidity andmortality remains open, even when the significantly higher adduct formation in male DNA compared with female materialwas taken into account. Preliminary experiments suggest a role of the frequently observed loss of the Y-chromosome.q 1999 Elsevier Science B.V. All rights reserved.

Keywords: Laryngeal cancer; Tobacco smoking; DNA adduct; p53 mutation; Chromosome alteration; Genetic risk

AbbreÕiations: HNCs, head and neck cancers; scc, squamouscell carcinoma; PAH, polycyclic aromatic hydrocarbons; BPDE,

Ž .benzo a pyrene-diolepoxide; SHBG, sex hormone-binding globu-lin; PCNA, proliferating cell nuclear antigen; GST, glutathione-S-transferase; NAT, N-acetyltransferase; EH, epoxide hydroxylase;LOH, loss of heterozygosity; SSCP, single-strand conformationpolymorphism; TLC, thin-layer chromatography; HPLC, high-per-formance liquid chromatography; FISH, fluorescence in situ hy-bridization; brc, breaks per cell

) Corresponding author. Tel.: q48-61-8233011; fax: q48-61-8233235; E-mail: [email protected]

1. Introduction

ŽTumours of head and neck excluding those of the.oesophagus belong to the most common types of

cancer, taking the sixth position in the list of thew xmost frequent cancers world-wide 1 . It is estimated

that there are roughly 500,000 new cases annually of

1383-5718r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.Ž .PII: S1383-5718 99 00131-X

( )K. Szyfter et al.rMutation Research 445 1999 259–274260

Ž .head and neck cancer HNC in the world. Besidesmedical aspects, there are many matters concerningmolecular epidemiology and biology of HNC to bestudied with adequate molecular methods. A modelof multistep carcinogenesis preceding clinicallyrecognised HNC, among which is cancer of thelarynx, can be presented by the following scheme:

The above scheme explains the interest that re-searchers take in investigating particular steps of thisprocess and in studying the relationships betweenthem. In Poznan, Poland, a project was established in´1992, which focused on DNA damage in tobaccosmoke-associated larynx cancer. Shortly afterwards,this research effort turned into a multilaboratorycooperation aiming at the investigation of multistagecarcinogenesis in the larynx and its interindividualvariability and genetic background. The present arti-cle is an overview of our results and conclusionsconfronted with the findings of other investigators.

2. Larynx cancer epidemiology and aetiology

HNCs are recognised as a common categoryw xprimarily because of their anatomic contiguity 2 .

Tumours can be located in many different sitesincluding mouth, oral cavity, nasopharynx, larynx,

Žoesophagus there is, however, no common agree-.ment in including oesophagus in this category and

other locations, with larynx appearing to be the mostfrequent target for tumourigenesis in the head and

neck region. Brain and eye tumours are classifiedoutside the HNC category. In fact, even havingexcluded brain and eye, HNC represents aetiologicaland morphological diversity. On the other hand, one

w xcan find several common features of HNC 2 . First,the vast majority of HNCs develop in the sameanatomical entity — the squamous epithelial liningof the upper aerodigestive tract. So it may be prefer-able to speak about head and neck squamous cell

Ž .carcinoma HNSCC . Second, most HNCs appear tobe resistant to cytostatic drugs, and chemotherapy isgenerally followed by severe, undesired side-effects.Hence, treatment of HNC usually comprises surgeryor radiotherapy or a combination of both. Next, theprognosis of diagnosed HNC is poor, as is evidentfrom a 5-year survival coefficient below 55%. Itshould be added that despite some progress in earlydiagnosis and subsequent treatment, the survival stillremains low. Third, it was hypothesised that multipleprimary head and neck tumours share a common

w xclonal origin 3,4 ; however, there is no generalw xagreement on this point 5 . Finally, the aetiology of

HNC is strongly associated with tobacco smokingw xand alcohol consumption 6–9 , an aspect that needs

to be discussed in more detail.Tobacco and alcohol consumption are mentioned

together, as smokers often tend to be also drinkersand vice versa. Moreover, the exposures to thesefactors seem to act synergistically. It should benoted, however, that this is not the case for lungcancer, where alcohol drinking does not contribute toan increased risk. The risk of cancer associated withcombined exposure to tobacco and alcohol is morepronounced in the upper than in the lower part of thelarynx, which can be explained by limited penetra-tion of the respiratory tract with volatile componentsof beverages. A well-marked link between alcoholconsumption and larynx cancer risk is only seen withstrong alcoholic drinks such as vodka, brandy,whisky, etc. The cancer risk attributable to tobacco

w xsmoking 10 is related to the quality of the tobacco,cigarettes and cigars used. Tobacco products withhigh tar and nicotine levels have been proven tocause a higher increase of larynx cancer risk for thesmokers than do the products with lower tar andnicotine levels. The protection of the respiratory tractthrough the use of filter cigarettes is beneficial fortobacco smokers. On the other hand, fibres can be

( )K. Szyfter et al.rMutation Research 445 1999 259–274 261

released from the filters and penetrate into the respi-ratory tissue, thus becoming a further factor initiating

w xtumorigenesis 11 .Tobacco smoking combined with alcohol con-

sumption is undoubtedly the main causative agent inlaryngeal cancer. Other human exposures to exoge-nous, chemical genotoxicants include environmentalpollution and hazardous working conditions associ-ated with such industries as metallurgy, petrochem-istry, dye production, or occupations such as varnish-

w xing, asphalt laying and others 12 . Besides, there areindications of a viral origin for some types of HNCw x13,14 .

Some geographical differences were reported con-cerning the occurrence of particular types of HNCw x1 . Nasopharyngeal cancer is very rare in the popu-lation world-wide except in China and to some ex-tent, in southeast Asia. Oral cavity tumours arecommon among betel chewers, which is a habit inthe Indian subcontinent. The occurrence of laryngealcancer is not randomly distributed in Europe: it isvery rare in Scandinavia and moderately frequent inMediterranean countries; in Poland, Hungary andformer Czechoslovakia, death rates from larynx can-

w xcer in males belong to the highest in the world 15 .In Poland, mortality from larynx cancer among maleshas been continuously increasing from 2.7 deaths per100,000 in 1963, through 7.5r100,000 in 1989 and

w x8.4r100,000 in 1993 16,17 , while the mortalityamong females is almost constant at a level below0.5r100,000. A working hypothesis to explain theextraordinarily high morbidity and mortality fromlarynx cancer in Poland has put forward a combina-

Ž .tion of all the previously mentioned factors: i ahigh consumption of poor quality, non-filter

Žcigarettes the number of cigarettes per capita sold inPoland is lower only than that in Mexico and China

. Ž .— the world leaders in tobacco smoking ; ii heavyenvironmental pollution resulting from forced indus-

Ž .trialisation under communist reign; and iii difficultworking conditions in many branches of industryw x6,15 . It must be concluded that larynx cancer asboth a health and a social problem in Poland is achallenge itself to researchers.

Another peculiarity of larynx cancer is the re-markably disproportionate distribution of its morbid-ity and mortality between men and women, as men-tioned above for the Polish situation: in the world

population, the malerfemale proportion of mortalityw xdue to larynx cancer is about 8–10:1 1,16,18 .

3. Damage of genetic material induced by geno-toxicants

3.1. DNA adducts in larynx cancer subjects

The presence of carcinogen:DNA adducts in hu-man cells is recognised as a proof of a previous

w xexposure to a carcinogen 19,20 . A demonstration ofaromatic DNA adducts in human laryngeal tissue

w xwas first described in 1993 21,22 when DNA fromthis tissue was analysed by means of the 32 P-postla-

w xbelling technique 23,24 , which was commonly usedfor detection and quantification of different types ofDNA adducts. ‘Bulky’ aromatic DNA adducts resultfrom exposure to polycyclic aromatic hydrocarbonsŽ .PAHs , nitrohydrocarbons and aromatic amines. Allof the above compounds require metabolic activationby cytochrome P450-dependent enzymes before they

w xreact with DNA 10,23 . The significance of theactivation step of tobacco smoke genotoxicants inthe case of larynx cancer was confirmed by Degawa

w xet al. 25 , who found a correlation between thelevels of CYP1A1, 2C and 3A4 and the amounts ofaromatic DNA adducts in the human larynx. Anassociation of the DNA adducts observed by 32 P-postlabelling with PAH was deduced from co-migra-tion of the main spot seen in the thin-layer chro-

Ž .matography TLC chromatogram with the standard2 w xadduct, BPDE-N -dGMP 21,22 . A more direct dif-

ferentiation between DNA adducts derived from PAHand aromatic amines, based on variants of the 32 P-postlabelling technique exploring nuclease-P1 diges-tion or butanol extraction for adduct enrichment, wasnot commonly used at that time. The recent follow-up

Žstudy J. Banaszewski and L. Moller, manuscript in¨.preparation provides proof for the observed adducts

to be mostly PAH:DNA adducts, as shown by meansof a 32 P-postlabellingrhigh-performance liquid chro-

Ž .matography HPLC protocol with a number ofPAH:DNA adduct standards.

More recently, N 7-alkylguanine was found inDNA of larynx tissue, and the authors providedevidence that N-nitrosoamines present in tobaccosmoke are capable, after metabolic activation, to


alkylate DNA, which includes the formation of N 7-w xalkylguanine in the mucosa of the larynx 26 .

With regards to carcinogenesis and DNA adducts,most research papers are concerned with the levelsof DNA adducts in biopsies of cancer subjects com-pared to those of non-cancer donors as studied in

w x w x w xbreast 27 , bronchus 28 , and colon 29 . In larynxcancer patients, DNA adducts were detected in lar-ynx tumour cells, in sections of larynx histopatholog-ically recognised as normal and in peripheral blood

w xlymphocytes 22,26,30 . As expected, the level ofDNA adducts in blood cells was lower than that in

Žlarynx tissue directly exposed to tobacco smoke Ta-.ble 1 . The proportion of DNA adduct levels in

tumour vs. non-tumour cells of the same organ is aninteresting matter. Rapid proliferation of tumour cellsfacilitates the DNA repair process through a tran-scription-coupled repair mechanism on the one hand,but there are many reports of defective DNA repair

w xin tumour tissue 31 . Comparing tumour biopsiesand non-tumour larynx cells, we have found a signif-

w xicantly higher level of aromatic DNA adducts 30and a non-significantly higher level of N 7-al-

w xkylguanine 26 in the former. When studying breastw xbiopsies, Li et al. 27 found a fully comparable

pattern of aromatic DNA adducts in tumour cells andin adjacent normal tissue, but the majority of tu-mours showed lower levels of DNA adducts thannormal adjacent tissues. The only possible interpreta-

tion is that the whole organ is subject to DNA adductformation, and the presence of DNA adducts is notthe only reason for entering multistage carcino-genesis. This conclusion has been strengthened sincethe aromatic DNA adducts were analysed in the

Ž .interarytenoid area non-tumour . This region of lar-ynx was taken under investigation when in 3800subjects treated for laryngeal cancer at Clinics ofOtolaryngology of K. Marcinkowski Medical Uni-versity in the years 1946–1996, only one case wasfound with primary tumour location. The level ofaromatic DNA adducts in the interarytenoid areaslightly exceeded that in tumour as well as non-

Žtumour tissue in the group of 33 subjects Banaszew-.ski et al., manuscript submitted .

The lack of pronounced differences in DNAadduct level in laryngeal tumour and non-tumourtissues may reflect the heterogeneous structure of thetumour. It reminds of a variable pattern of chromoso-mal aberrations in various sections of laryngeal tu-mours. The differences of chromosomal aberrationswere demonstrated by microsatellite marker hy-bridization and interpreted in terms of field cancer-

w xization 32 . It could also be used to explain avariability of DNA adduct level in tumour material.

The intertissue comparison of DNA adduct levelsleads to one of the fundamental questions in molecu-lar epidemiology related to the applicability of surro-gate cells instead of the target tissue. An obvious

Table 1Average DNA adduct levels in tissues of larynx cancer subjects shown as number of adducted nucleotides per 108 normal nucleotidesThe numbers of samples analysed are given in parentheses.

7Ž . Ž .Grouprtissue Aromatic DNA adducts x"SD N -alkylguanines x"SD

Larynx cancer subjectsŽ . Ž .Larynx tumour — all 5.7"5.1 41 26.2"38.0 44Ž . Ž .Smokers 6.5"5.5 30 28.2"36.3 39Ž . Ž .Non-smokers 3.7"0.7 4 11.3"5.9 5Ž . Ž .Larynx non-tumour — all 4.7"4.9 36 22.7"19.9 33Ž . Ž .Smokers 4.8"5.1 30 24.1"18.9 31Ž . Ž .Non-smokers 5.8"3.0 4 15.3"2.0 3Ž . Ž .Leukocytes — all 2.1"1.8 12 13.1"5.6 9Ž . Ž .Smokers 2.3"2.1 11 13.6"48 8

Non-smokers nd nd

Non-cancer controlsŽ . Ž .Leukocytes — all 1.3"2.1 12 9.2"5.9 15

Ž .Smokers nd 9.7"5.9 10Ž .Non-smokers nd 5.1"2.8 5


advantage of the use of surrogate cells is an easynon-invasive way of aspiration. Its validity, though,should first be established by determining a relation-ship between the levels of a given biomarker in bothkinds of tissue. There is an almost universal agree-ment on the use of peripheral white blood cells assurrogate cell type, but it should be noted that somepapers report a lack of correlation between the levelsof aromatic DNA adducts in white blood cells and

w x w xlung 33 or bronchial tissue 28 . In our study, wehave found a strong correlation in larynx cancersubjects between DNA adduct levels in larynx tu-mour cells and in blood leukocytes, and a slightlyweaker correlation between adduct levels in non-

w xtumour larynx cells and blood leukocytes 30 .

3.2. Confounders of the DNA adduct leÕel in larynxbiopsies

Tobacco smoking is recognised as a maincausative factor in larynx cancer aetiology. Hence,attention was paid to the relationship between theextent of tobacco smoking and the DNA adduct levelw x26 . When subjects were divided into non-smokersŽincluding ex-smokers who already decided to quit

.smoking at least 5 years before surgery and smokerscategorised further according to number of cigarettessmoked per day, the average levels of aromaticadducts and N 7-alkylguanine increased, as shown inFig. 1. This is in accordance with the data of otherauthors, who have found a positive correlation be-tween tobacco smoking and DNA adduct levels in

w x w xwhite blood cells 34 , bronchial tissue 28 , and lungw x35 .

Some other potential confounders modulating thelevel of DNA adducts in human laryngeal cells werestudied. A weakly positive correlation was estab-

Žlished between the subjects’ age ranging from 38 to. 7 w x78 years and the level of N -alkylguanine 26 . In

another study, we reported that an increase of theN 7-alkylguanine level was typically found in a group

Žof male subjects of over 70 years of age unpub-. 7lished . Lower efficiency of N -alkylguanine re-

moval from senescent tissue as compared to youngerw xtissue was observed also by Gaubatz and Tan 36 in

mouse kidney cells. Altogether, these observationsfit the general phenomenon of a reduced capacity of

w xthe aged organism to remove DNA lesions 37 .Any link between DNA adduct level and cancer

progression may have prognostic value. Average lev-

Ž 8 7Fig. 1. The effect of tobacco smoking on DNA adduct formation in laryngeal tissues. DNA adduct levels per 10 nucleotides or 107 .nucleotides for aromatic DNA adducts and N -alkyl-dGMP, respectively are shown as a function of the number of cigarettes smoked per

day.


els of N 7-alkylguanine were calculated separatelyfor groups of larynx cancer subjects categorised ac-

w xcording to the TNM staging system 38 . Althoughthe attempt was unsatisfactory because of the smallnumber of patients in the early stage of tumourigene-sis, there still were changes observed in the N 7-al-kylguanine levels in relation to the tumour growth,reaching a maximum at the T3 stage. Metastasis tothe adjacent lymph nodes is usually followed by aremarkable worsening of the patients’ general health,which is not reflected by changes in the N 7-al-kylguanine levels. We take it as another proof forDNA adducts playing a significant role in the early

w xstage of carcinogenesis 38,39 ; nonetheless, theirsignificance in late stages of cancer cannot be ex-cluded.

We have also compared DNA adduct levels inlarynx tissues of subjects with primary and recurrentlarynx tumours. The average levels of aromatic DNAadducts and N 7-alkylguanine both in tumour and innon-tumour larynx tissue were higher in recurrent

Ž .biopsies Fig. 2 . One can speculate then on persis-tent impairment of DNA repair occurring during thecourse of tumorigenesis.

3.3. Sex-related damage of genetic material

The question concerning the cause of the substan-tial difference between men and women in larynxcancer morbidity and mortality remains open. A fewhypotheses were raised and experimentally tested.The first one is related to exogenous exposure, andrefers to the more frequent tobacco smoking andalcohol consumption in the male than in the femalepopulation. However, in leading industrialised coun-tries, it has been observed that adoption by men andwomen of similar lifestyle habits is not followed bydrastic changes in larynx cancer epidemiology. Analternative explanation of this sex difference sug-gests involvement of hormonal regulation of larynxfunction, but sufficient experimental data have notbeen provided yet.

We have compared the levels of aromatic DNAadducts and N 7-alkylguanine in tumour and non-tumour larynx tissue in the groups of moderate to-

Ž .bacco smokers, both male and female Fig. 3 . Forboth types of tissue, the DNA adduct levels werehigher in males. The differences were more pro-

Ž 8 7Fig. 2. Average DNA adduct levels per 10 nucleotides or 10nucleotides for aromatic DNA adducts and N 7-alkyl-dGMP, re-

.spectively in laryngeal tissues in the groups of subjects withprimary or recurrent tumours.

nounced in the case of N 7-alkylguanine, reaching amalerfemale ratio of 3.7 in larynx tumour tissuew x40 . These results are in contrast with those of

w xRyberg et al. 41 , who reported higher levels ofaromatic DNA adducts in lung biopsies from femalelung cancer patients compared to male patients, afteradjusting for tobacco smoke exposure. On the other

w xhand, Phillips et al. 42 , studying bronchial biopsiesfrom patients undergoing pulmonary surgery, havefound that normal bronchial epithelial cells frommales contained 1.9-fold more aromatic DNA adductsthan cells from females. In the latter case, the resultswere not adjusted for tobacco smoke exposure. Alsoin nasal mucosa tissue from non-cancer smokersw x43 , the level of aromatic DNA adducts was 3-foldhigher in males than in females; in non-smokers, the

Ž .difference was smaller 1.8-fold . The analysis ofDNA adducts in nasal mucosa was performed on asmall group of subjects.

We think that the discrepancy between men andwomen with regards to DNA adduct levels in lung,bronchus, nasal cavity and larynx reflects a differ-ence between lower and upper respiratory tracts.This statement should be considered with caution,however, as exposure to tobacco smoke is not takeninto account in the same way by the authors of thecited papers. In any case, tobacco smoke exposure in


Ž 8 7Fig. 3. Sex differences of average DNA adduct levels per 10 nucleotides or 10 nucleotides for aromatic DNA adducts and7 .N -alkyl-dGMP, respectively in laryngeal tissues.

the upper respiratory tract is direct and almost topi-cal, in contrast to exposure conditions in the lowerrespiratory tract. Differences in activities of carcino-gen-metabolising enzymes should be taken into ac-count as well. Nevertheless, the morbidity differ-ences found at the epidemiological level seem to berecognisable also by methods used in molecular epi-demiology.

Parallel to DNA adduct studies, we have alsoinvestigated a possible association between larynxcancer morbidity and the concentration of sex hor-

w xmones in blood serum 44 . The serum concentra-tions of testosterone and sex hormone-binding globu-

Ž .lin SHBG were compared with the average DNAadduct levels in the patient groups divided accordingto age. The only significant result was the associa-tion of a decreased level of SHBG with increasedmorbidity in males over 50 years old. Altogether, wefailed to find, however, any significant correlationbetween the parameters studied. Hence, we haveconcluded that testosterone and SHBG concentrationin serum are not suitable markers of male predisposi-

w xtion to laryngeal cancer 40 .In order to explain sex differences in larynx can-

cer morbidity, we have conducted an analysis ofnumerical aberrations of sex chromosomes. The fre-quent loss of the Y-chromosome reported in HNCwas interpreted as a process accompanying aging

w x45,46 . On the other hand, loss of the Y-chro-mosome was claimed to be directly associated with

w xHNC 47 . We decided to study the latter possibilityconcerning a linkage between Y-chromosome lossand larynx cancer by comparing the presence of thischromosome in larynx tumour and non-tumour cellsand in peripheral blood lymphocytes in a smallgroup of larynx cancer subjects. For cytogeneticanalysis of lymphocytes, a full classical karyotypingfollowing 72-h culture of mitogen-stimulated cellswas applied. Larynx cells were only incubated for 6h and then analysed by use of the fluorescence in

Ž .situ hybridization FISH technique, with a probespecific for the Y-chromosome centromere. As shownin Table 2, frequent numerical aberrations of the

ŽY-chromosome were seen only in tumour tissue un-.published . Although it is still not clear whether the

observed aberration is a marker of cancer predisposi-tion and has prognostic value, it seems worthwhile toexplore Y-chromosome disintegration further. Thiswork is in progress.

( )3.4. Mutations and loss of heterozygosity LOH inp53 and p16 tumour suppressor genes

The DNA adduct analysis described above con-cerns the whole genome. It is well-known that many


Table 2Numerical aberration of the Y-chromosome in larynx cancer shown by full karyotyping in PBL and by the FISH technique in laryngeal cells

Ž . Ž . Ž .Subject Blood lymphocytes karyotype Larynx non-tumour cells FISH signal Larynx tumour cells FISH signal

1 46, XY single none2 46, XY single single3 46, XY single none4 46, XY single double5 46, XY single single6 46, XY single none

lesions along the genome remain without effect,whereas some are crucial for carcinogenesis. Thelatter group includes lesions located in proto-onco-

w xgenes and tumour suppressor genes 48 . Mutationsin the p53 tumour suppressor gene are frequent in

w x w xHNC 49,50 but not in nasopharyngeal cancer 51 .Using single-strand conformation polymorphism

Ž .SSCP , we have analysed mutations in samplesw xderived from 40 larynx cancer subjects 52 . An extra

band was found in 15 samples, among which sixtransitions and seven transversions localised in exons

Ž .5–8 of the p53 gene were detected Table 3 . Someof these mutations were found in codons known tobe hotspots for mutagenesis in HNC. On the basis of

w xlung cancer findings 53 , we assume that at least sixof the detected mutations could be attributed togenotoxic activity of tobacco smoke. We have found

Žfour additional T™A transversions not yet de-.scribed in larynx cancer that have not been con-

nected with tobacco smoking. Although the groupunder study comprised many alcohol abusers, theoccurrence of T™A transversions in associationwith alcohol exposure seems unlikely, as acetalde-

Ž .hyde a metabolite of ethanol tends to interact with

Table 3Mutations found in the p53 gene in larynx tumour tissue

Type Number Suggested origin

Transition A ™G 1Transversion A ™T 1

Ž .Transition C™T 2 tobacco smoke PAHŽ .Transition G™A 2 tobacco smoke N-nitrosoaminesŽ .Transversion G™T 2 tobacco smoke PAH

Transversion T™A 4 ?Transition T™C 1

exocyclic groups of A, G, and C but not T. Hence,alcohol consumption could contribute to the induc-tion of the mutations observed, and in this wayincrease larynx cancer risk, but it does not explainthe origin of the frequent T™A transversions.

Overexpression of the p53 gene and accumula-tion of the p53 protein are regular responses of

w xlaryngeal cells to DNA damage 54,55 . The muta-tions in p53 gene do not necessarily lead to aninhibition of gene expression. There is always achance that the remaining copy of the wild-type genewould compensate through synthesis of unchangedprotein. However, many studies have documented

Ž .the complete deletion of one copy and mutation s inthe other copy in such genes as p53 and p16. ThisLOH in regions encoding tumour suppressor genes isa severe defect blocking the function of relevantproteins. We have examined 48 samples from larynxcancer subjects for the occurrence of LOH and mi-crosatellite instability in chromosome 9p near thep16 locus, in 17p at the p53 locus and in 17p at thelocus D17S520, which is not involved in carcino-

w xgenesis 56 . By use of highly polymorphic mi-crosatellite markers in a polymerase chain reactionŽ .PCR -based technique, it was found that the highest

Žlevel of LOH was present at the p53 locus range:. Ž .33–45% followed by the p16 locus 28–38% and

Ž .the D17SS520 locus 22% . The data suggest animportant protective role of p53 in laryngeal cancer,exceeding that of p16. Similar results have been

w xobtained by Gonzales et al. 57 . On the other hand,some papers claim a substantial role for the p16protein in suppressing growth of head and neck

w xcarcinoma cells 58,59 . Recent findings associateLOH at the p16 protein with an early phase ofcarcinogenesis. Although most of the tumours weanalysed could be classified as T3NM and T4NM,


the finding concerning more frequent LOH at thep53 location than at p16 seems to be correct.

At this point, it is worth adding that studies ontumour suppressor genes and their alterations are of

w xclinical significance 60 . The discovery of LOH atp53 has become a starting point to design individualradiotherapy regimens. A gene-therapy protocol,based on the introduction of the wild-type p53 geneby recombinant adenovirus technology in order tosuppress growth of head and neck tumour cells, has

w xbeen tested in a mouse model 61,62 .

3.5. The loss of cell cycle control

Tumour suppressor genes control cell prolifera-tion via the mechanism of cell cycle regulation.Dysfunction of these genes allows a cell to enteruncontrolled proliferation that is a part of neoplastictransformation.

The accumulation of p53 protein during progres-sive stages of carcinogenesis, i.e., from anaplasia anddysplasia to clinically developed cancer, has already

w xbeen reported 54,55 . With the use of histochemicalstaining, we have determined levels of p53 protein inlarynx tumour biopsies obtained from 120 donors in

w xdifferent stages of disease progression 63 . The levelof p53 protein increased with tumour staging. In thesame set of samples, we have determined two proteinmarkers of cell proliferation, proliferating cell nu-

Ž .clear antigen PCNA and Ki67, the levels of whichalso showed a gradual increase. This means that thelarynx cells, which are in the G phase under physio-0

logical conditions, have entered proliferation, givingrise to their uncontrolled multiplication and tumourgrowth. Our data are consistent with those of

w xLavieille et al. 64 who have studied a very similarmodel. The determination of protein markers for cell

w xproliferation can also be applied to prognostics 65or, more specifically, to the prediction of occult

w xmetastases in laryngeal cancer 66 . We further con-clude that a considerable fraction of the p53 proteinis the product of a mutated gene unable to exert itsproper function in cell cycle control. This is partly

Žconfirmed by the results described above Section.3.4 , as all 40 samples analysed for p53 mutations

w x52 were obtained from subjects belonging to aŽ .larger group 120 persons analysed for proliferation

w xmarkers 63 .

4. The significance of genetic factors in larynxcancer

Larynx cancer is not recognised as a hereditarytype of cancer mainly because of its clear aetiologyassociated with exogenous exposure to tobaccosmoke as the main causative agent. On the otherhand, not all tobacco smokers and not even all heavysmokers develop any type of cancer. Interindividualvariability in susceptibility to tobacco smoke car-cinogens can hide genetic predisposition. Carcino-genesis is a multistage process involving manynon-genetic modulators that affect individual suscep-

w xtibility. From the work of Copper et al. 67 , there isconvincing evidence that unspecified genetic factorsplay a role in HNSCC. In this large retrospectivestudy, the occurrence of cancer in over 600 first-de-gree relatives of HNC patients is compared with thatin an unrelated control group of the same size. It wasfound that cancers of the respiratory tract and theupper part of digestive tract are 3.1-fold more fre-

Ž .quent in first-degree relatives 8-fold in siblingsthan in controls. These findings have been confirmed

w xin reports on familial HNC in Brazilian 68 andw xCanadian 69 populations. Studies in this field are

focused on the individual response to genotoxicants,which depends on the genetically determined activi-

w xties of xenobiotic metabolising enzymes 70,71 .Hence, the heterogeneous response to genotoxic ex-posure as deduced from interindividual differencesof a given marker level within the group of subjectsexposed to similar doses of genotoxicant can berecognised as an indirect indication of genetic poly-morphism. In the course of our studies, interindivid-ual differences were already noticed, which wouldindicate potential genetic variability.

4.1. Variable pattern of DNA adduct leÕels

The analysis of DNA adduct levels revealed largew xinterindividual differences 22,26,30,38 , as shown in

Table 4. Large differences have been noted for aro-matic DNA adducts and N 7-alkylguanine in all tis-sues studied. The calculations concern only a groupof moderate cigarette smokers, but exclusion of heavysmokers and non- and ex-smokers did not lead to a


Table 4Interindividual variation in the level of DNA adducts in biopsiesof laryngeal subjects

Tissue Average Range of n-foldlevel results variation

Aromatic DNA adductsLarynx tumour 5.7"5.1 0.3–22.1 73.7Larynx non-tumour 4.7"4.9 0.1–23.8 238.0Leukocytes 2.1"1.8 0.7–8.4 12.0

7N -alkylguaninesLarynx tumour 26.2"38.0 0.3"224.5 748.3Larynx non-tumour 22.7"19.9 0.4"64.9 162.3Leukocytes 13.1"5.6 1.5"26.5 17.7

Numbers indicate average levels of adducted nucleotides per 108

normal nucleotides.

drastic narrowing of the range of interindividualresults. The differences in DNA adduct levels have

w xbeen critically discussed by Hemminki 72 in thew xlight of the assumption of Harris 48 , who explained

the actual level of DNA adducts as a result ofvariable efficacy of counteracting processes ofmetabolic activation, detoxification and DNA repair.

Having analysed two types of DNA adducts inbiopsies of the same donors, we have attempted tocorrelate their levels. Fig. 4 shows the interindividualvariation of DNA adduct levels found in tumourbiopsies. Aromatic DNA adducts arise from expo-sure to PAH, whereas N 7-alkylguanines are formedby interaction of N-nitrosoamines with DNA. Corre-

w xlation coefficients were found to be very low 26 ,which means that formation and removal of bothtypes of DNA adducts proceed independently undercontrol of entirely different enzymes. The activationof PAH is mediated by CYP1A1, while N-nitro-soamines are activated by CYP2D6, CYP2E1 andother cytochrome P450-dependent enzymes, all of

w xthem encoded by genes in different loci 48 . Differ-ences also emerge in the DNA repair process. Aro-matic DNA adducts are removed by nucleotide exci-sion repair, while ‘small’ DNA alkylations can beprocessed in various ways. Therefore, when aromaticadducts and DNA alkylations are formed under thesame conditions of complex exposure, these lesionsare formed and removed independently according tothe genetically determined activity of enzymes in-

w xvolved 48,70 .

4.2. SensitiÕity to bleomycin-induced chromosomebreaks

Genomic instability in cancer comprises bothcleavage at hidden chromosome breakpoints emerg-

w xing under the action of a mutagen 73 and apparentchromosome rearrangements which arise during thecourse of carcinogenesis and subsequently give rise

w xto aberrant cell function 74 .We have undertaken a study on chromosome

sensitivity to bleomycin exposure in vitro in linewith the suggestion to apply the test for the identifi-cation of subjects characterized by an increased risk

w xto develop HNC 71,72 . Established in the labora-w xtory of Spitz et al. 76 and since then known as the

‘bleomycin test’, the method has been used through-w xout our study 77 . The concept of the method is to

estimate the number of chromatid breaks per cellŽ .brc induced by bleomycin in peripheral bloodlymphocytes proliferating in vitro. In agreement with

w xthe results of other authors 75,78 , we found thatbleomycin-induced damage in 35 larynx cancer sub-

Ž .jects average brcs0.72 significantly exceeded thatŽ .in the matched controls brcs0.42 . The number of

spontaneous chromatid breaks was also higher inŽ .larynx cancer subjects brcs0.20 than in healthy

Fig. 4. Individual levels of aromatic DNA adducts and N 7-alkyl-dGMP in laryngeal tumour biopsies of larynx cancer subjects.


Ž .controls brcs0.17 , although the difference didnot reach the level of significance. Extending thestudy group to 72 subjects and 60 controls hasgenerally confirmed these findings, and has alsodemonstrated the genetic heterogeneity that is en-countered during analysis of a range of individualresults. As can be seen from the distribution patterns

Ž .of individual bleomycin-induced breaks Fig. 5 , la-ryngeal cancer subjects and controls suggestivelydiffer with respect to chromosome instability. An-other relevant point is that brc indices )0.8, or

w x)1, classified according to Hsu 75 to be indicativeŽ .of genuine instability brc or overinstability, re-

spectively, were found in the group of larynx cancerŽsubjects but never in controls Dabrowski et al.,

.manuscript submitted .

4.3. Genotypes of detoxifying enzymes

A considerable part of metabolically activatedxenobiotic material is detoxified and removed fromthe cell and finally removed through urine or bile.We have assumed that differences in the activity ofdetoxifying enzymes could modulate a biologicallyactive dose of carcinogen more effectively than dif-ferences in activating enzymes. The significance ofsome detoxifying enzymes as modifiers of genetic

w xrisk of cancer has been already described 71,79 .

We have decided to study the genotype of thepolymorphic detoxifying enzyme, glutathione-S-

Ž .transferase GST , instead of estimating the enzy-w xmatic activity itself 80 . GST occurs in one microso-

mal and four cytosolic isoforms encoded by differentw xgenes at different loci 81 . To determine the distri-

bution of defective genes potentially responsible forpoor detoxification, we have employed a PCR-based

w xmethod 82,83 . A group of 179 larynx cancer sub-jects was compared with 180 controls with respect tothe distribution of GST M1 and GST T1 genotypes.

Ž .The results show that: i the percentage of ‘null’Žgenotypes in the control group 57.7% for GST M1,

.21.7% for GST T1 is comparable to other dataw x Ž .concerning Caucasians 79 ; and ii none of the

studied GST ‘null’ genotypes is significantly over-represented in larynx cancer subjects. At this point,the conclusion is that none of the genes studiedseparately plays a role in individual susceptibility todevelop larynx cancer. When tobacco smoking wastaken into account, however, the percentage of GSTM1 ‘null’ and GST T1 ‘null’ genotypes showed atendency to increase in the following order: heavy

Žsmokers 46.3.1 and 19.4% for GST M1 and GST. ŽT1, respectively , moderate smokers 51.1 and

. Ž .16.0% and non-rex-smokers 55.6 and 22.2% . Thisresult at least seems to indicate that defective GSTM1 activity cannot be excluded as a determinant of

Fig. 5. Distribution of spontaneous and bleomycin-induced chromosome breaks in peripheral blood lymphocytes proliferating in vitro.


Table 5Occurrence of ‘risk’ genotypes of detoxifying enzymes in larynx cancer subjects

Ž . Ž . Ž .Genotype Larynx cancer subjects ns214 Bladder cancer subjects ns45 Controls ns240

Ž . Ž . Ž .n % OR 95% CI n % OR 95% CI n % OR 95% CI

Ž . Ž . Ž .GST M1 y 111 51.5 0.9 0.8–1.6 25 55.6 1.0 0.5–1.9 132 55.0 1.0Ž . Ž .GST M3 BrB 2 0.9 0.3 0.1–1.5 4 8.9 3.9 1.0–11.3 7 2.9 1.0

Ž . Ž . Ž .GST T1 y 39 18.2 0.7 0.8–2.1 7 15.5 1.6 0.7–3.8 55 22.9 1.0Ž . Ž .GST P1 ‘slow’ 31 14.5 1.3 0.7–1.4 11 24.5 2.2 1.1–5.5 29 12.1 1.0Ž . Ž .NAT2 ‘slow’ 112 52.3 1.0 0.7–1.9 28 62.2 1.4 0.–2.8 128 53.3 1.0Ž . Ž .EPHX1 ‘low’ a 30 14.0 1.1 0.6–2.0 10 22.2 2.1 0.8–5.6 34 14.2 1.0Ž . Ž .EPHX1 ‘low’ b 102 47.8 1.3 0.8–1.9 21 46.7 1.4 0.6–3.3 102 43.8 1.0

risk of smoking-related larynx cancer. On the otherhand, in the case of larynx cancer, high exposure totobacco smoke carcinogens may mask the signifi-cance of genetic factors in larynx cancer risk.

The data and conclusions described above wereconfirmed in a study conducted among a group

Žextended to 217 cases and 240 controls R. Jaskula-.Sztul, 1998, PhD thesis . Because of the recognised

association of GST ‘null’ genotypes with a bladderw xcancer risk 84 , blood samples collected from blad-

der cancer patients were included as a positive con-trol. Apart from inclusion of larger groups into thestudy, the analyses were extended to GST M3, GST

Ž .P1, N-acetyltransferase 2 NAT2 and epoxide hy-Ž .droxylase EH . The cases and control groups were

large enough to demonstrate a significant overrepre-sentation of the following ‘risk’ genotypes: GST P1and EPHX1 in larynx cancer subjects and GST M3,GST P1, NAT2 and EPHX1 in bladder cancer pa-

Ž .tients Table 5 . The involvement of different geno-types as determinants of the risk of the cancer typesunder study may reflect differences between the

Ž .direct almost topical exposure to tobacco smoke inthe larynx and the indirect exposure in the urinarybladder. The study concerning detoxifying enzymes

Ž .confirmed: i a tendency of the gene defects toŽoccur in combination which was the case for GST

.M1 ‘null’ and NAT2 ‘slow’ in larynx cancer ; andŽ .ii a masking of the genetic factor by heavy tobaccosmoking.

A further analysis of GST genotypes explored thesignificance of the demonstrated combined occur-

w xrence of ‘risk’ genotypes 85 . The aim of the workwas to correlate a biomarker of exposure to tobaccosmoke with the multiplicity of gene deficiences, bycalculating the levels of aromatic DNA adducts inthe various groups divided according to the numberof ‘risk’ genotypes. It was shown that the meanDNA adduct level increased with the number of

Ž .defective genes Table 6 . Possibly, a deficiency intwo or more pathways of detoxification could berecognised as a real risk factor in laryngeal cancer.

The final conclusion from this part of the investi-gations is that the role of a genetic factor as a

Table 6Ž 8 .Mean levels of aromatic DNA adducts per 10 normal nucleotides in laryngeal mucosa calculated in groups of subjects according to

multiplicity of defects of detoxifying enzymes genes

Multiplicity of gene defect Number of subjects Tumour Non-tumour

None 9 25.8"12.0 36.6"34.2Single 18 30.8"28.6 33.4"40.1Double 10 48.8"44.4 38.8"49.3Triple 4 58.4"59.1 68.9"59.7


Table 7Summary of molecular and cellular alterations studied in laryngeal cancer

Stage Phenomena studied

Tobacco smoke carcinogens PAH, N-nitrosoamines, dose effect2 7Ž .DNA adducts aromatic DNA adducts BPDE-N -dGMP , N -alkyl-dGMP; dose effect,

targetrsurrogate tissue correlation, confoundersŽ .Gene mutations p53 G™T, G™A, C™T, T™A

Chromosome aberrations LOH at p53 and p16; loss of YAltered gene function p53 abortive expressionUncontrolled cell proliferation PCNA, Ki67Pathology laryngeal scc

determinant of larynx cancer risk is difficult todemonstrate on the basis of genotypes of detoxifyingenzymes.

5. Conclusions

An extensive review on the molecular biology ofw xHNC was published in 1995 by Koch 86 . The

author emphasized the frustration of physicians whoare getting an increasingly better understanding ofthe sequence of events in HNC on the one hand,while at the same time, the basic knowledge con-tributes relatively little to clinical practice. We areafraid that this is still the case. Nevertheless, wewant to present a further approximation to under-standing the various steps from carcinogenic expo-sure to a clinically recognised laryngeal tumour.Altogether, the schematic sequence of events pre-sented in Fig. 1 on the basis of the studies performedcan now be illustrated with some clear-cut findingscompiled in Table 7. Moreover, the magnitude ofchanges induced by exogenous factors is modulatedby host factors. The influence of a factor, such as asubject’s sex, is still difficult to explain. Concerningthe significance of genetic factors, no conclusivedata have been provided yet to demonstrate the roleof a specific gene or group of genes.

We strongly believe that a better understanding ofthe molecular biology and epidemiology of laryngealcancer will contribute to improvement of the pa-tients’ treatment, which is still far from satisfactory.

Acknowledgements

The authors would like to express their gratitudeto the colleagues from Poznan, Huddinge, Helsinki,´Bergen and Linkoping who contributed through ex-¨perimental work and discussion of the knowledge ofbiology of larynx cancer.

References

w x1 D.M. Parkin, E. Laara, C.S. Muir, Estimates of the world-wide frequency of sixteen major cancers in 1980s, Int. J.

Ž .Cancer 41 1988 184–197.w x2 Ch.W. Cummings, J.M. Fredrickson, L.A. Hacker, Ch.J.

Ž .Krause, D.E. Schuller Eds. , Otolaryngology — Head andNeck Surgery, Vol. 3, 2nd edn., Mosby Year Book, St.Louis, 1993.

w x3 M.J. Worsham, S.R. Wolman, T.E. Carey, R.J. Zarbo, M.S.Benninger, D.L. van Dyke, Common clonal origin of syn-chronous primary HNSCC: analysis by tumour karyotypes

Ž .and FISH, Hum. Pathol. 26 1995 251–261.w x4 G.C. Bedi, W.H. Westra, E. Gabrielson, W. Koch, D. Sidran-

sky, Multiple head and neck tumours: evidence for a com-Ž .mon clonal origin, Cancer Res. 56 1996 2484–2487.

w x5 T.E. Carey, Field cancerization: are multiple primary cancersŽ .monoclonal or polyclonal?, Ann. Med. 28 1996 183–188.

w x6 W. Zatonski, H. Becher, J. Lissowska, J. Wahrendorf, To-´bacco, alcohol, and diet in the etiology of laryngeal cancer: apopulation-based case–control study, Cancer Causes and

Ž .Control 2 1991 3–10.w x7 S.Y. Choi, H. Kahyo, Effect of cigarette smoking and alco-

hol consumption in the aetiology of cancer of the oral cavity,Ž .pharynx and larynx, Int. J. Epidemiol. 20 1991 878–885.

w x8 J.L. Freudenheim, S. Graham, T.E. Byer, J.R. Marshall, B.G.Haughey, M.K. Swanson, G. Wilkinson, Diet, smoking andalcohol in cancer of the larynx: a case-control study, Nutr.

Ž .Cancer 17 1992 33–45.w x9 S.P.Schantz, N.K.Lee, Tobacco carcinogenesis and head and

neck cancer, in: J.A. Werner, B.M. Lippert, H.H. Ruppert


Ž .Eds. , Head and Neck Cancer — Advances in Basic Re-search, Elsevier, Amsterdam, 1996, pp. 9–20.

w x10 IARC, Tobacco Smoking, IARC Monographs on the Evalua-tion of the Carcinogenic Risk of Chemicals to Humans, Vol.38, IARC, Lyon, 1986.

w x11 J.L. Pauly, H.A. Allaart, M.I. Rodriquez, R.J. Streck, Fibresreleased from cigarette filters: an additional health risk to

Ž .smoker?, Cancer Res. 55 1995 253–258.w x12 J.E. Muscat, E.L. Wynder, Tobacco, alcohol, asbestos and

occupational risk factors for laryngeal cancer, Cancer 69Ž .1992 2244–2251.

w x13 K. Anwar, K. Nakakuki, H. Naiki, M. Inuzuka, ras genemutations and HPV infection are common in human laryn-

Ž .geal carcinoma, Int. J. Cancer 53 1993 22–28.w x14 Z. Chen, K.A. Storthz, E.J. Shillitoe, Mutations in the long

control region of human HPV in oral cancer cells, and theirŽ .functional consequences, Cancer Res. 57 1997 1614–1619.

w x15 J. Lisowska, W. Zatonski, Tumours of the larynx in Poland´in comparison to other European countries. Etiological hy-

Ž .potheses, Otolaryngol. Pol. 46 1992 211–216.w x16 W. Zatonski, J. Tyczynski, J. Didkowska, Malignant tumours´ ´

of the larynx in Poland in the years 1963–1989, Otolaryngol.Ž .Pol. 46 1992 203–210.

w x17 Z. Wronkowski, M. Zwierko, Malignant tumours in Poland,in: Clinical Oncology, UICC, Polish edition by PWN, War-saw, 1994, addendum.

w x Ž .18 L. Tomatis Ed. , Cancer: Causes, Occurrence and Control,IARC Scientific Publications, IARC, Lyon, 1990, pp. 52–64.

w x19 K. Hemminki, DNA adducts, mutations and cancer, Carcino-Ž .genesis 14 1993 2007–2012.

w x20 D.H. Phillips, DNA adducts in human tissues: biomarkers ofexposure to carcinogens in tobacco smoke, Environ. Health

Ž . Ž .Perspect. 104 4 1996 453–458, Suppl.w x21 S.J. Stern, M. Degawa, M.V. Martin, F.P. Guengerich, R.K.

Kaderlik, K.F. Ilett, R. Breau, M. McGhee, D. Montague, B.Lyn-Cook, F.F. Kadlubar, Metabolic activation, DNAadducts, and h-ras mutations in human neoplastic and non-

Ž .neoplastic laryngeal tissue, J. Cell. Biochem. 17F 1993129–137, Suppl.

w x22 K. Szyfter, Z. Szmeja, W. Szyfter, Analysis of DNA adductsŽ .in larynx cancer patients, Otolaryngol. Pol. 47 1993 496–

501.w x 3223 D.H. Phillips, Detection of DNA modifications by P-post-

Ž .labelling assay, Mutat. Res. 378 1997 1–12.w x24 K. Szyfter, J. Kruger, P. Ericson, C. Vaca, A. Forsti, K.¨ ¨

Hemminki, 32 P-postlabelling analysis of DNA adducts inhumans: adduct distribution and method improvement, Mu-

Ž .tat. Res. 313 1994 269–276.w x25 M. Degawa, S.J. Stern, M.V. Martin, F.P. Guengerich, P.F.

Fu, K.F. Ilett, R.K. Kaderlik, F.F. Kadlubar, Metabolic acti-vation and carcinogen–DNA adduct detection in human lar-

Ž .ynx, Cancer Res. 54 1994 4915–4919.w x26 K. Szyfter, K. Hemminki, W. Szyfter, Z. Szmeja, J. Ba-

naszewski, M. Pabiszczak, Tobacco smoke-associated N 7-al-kylguanine in DNA of larynx tissue and leucocytes, Carcino-

Ž .genesis 17 1996 501–506.w x27 D. Li, M. Wang, K. Dhindhra, W.N. Hittelman, Aromatic

DNA adducts in adjacent tissues of breast cancer patients:Ž .clues to breast cancer aetiology, Cancer Res. 56 1996

287–293.w x28 B. Schoket, S. Kostic, I. Vincze, Determination of smoking-

related DNA adducts in lung-cancer and non-cancer patients,Ž .in: D.H. Phillips, M. Castegnaro, H. Bartsch Eds. , Postla-

belling Methods for Detection of DNA Adducts, IARC,Lyon, 1993, pp. 315–319.

w x29 A. Pfohl-Leszkowicz, Y. Grosse, V. Carriere, P.H. Cugnenc,A. Berger, F. Carnot, Ph. Beaune, I. de Waziers, High levelsof DNA adducts in human colon are associated with colorec-

Ž .tal cancer, Cancer Res. 55 1995 5611–5616.w x30 K. Szyfter, K. Hemminki, W. Szyfter, Z. Szmeja, J. Ba-

naszewski, K. Yang, Aromatic DNA adducts in larynx cancerŽ .biopsies and leukocytes, Carcinogenesis 15 1994 2195–

2199.w x31 E.C. Friedberg, Relationships between DNA repair and tran-

Ž .scription, Ann. Rev. Biochem. 65 1996 15–42.w x32 J. Califano, P. van der Riet, W. Westra, H. Nawroz, G.

Clayman, S. Piantadosi, R. Corio, D. Lee, B. Greenberg, W.Koch, D. Sidransky, Genetic progression model for head andneck cancer: implications for field cancerization, Cancer Res.

Ž .56 1996 2488–2492.w x33 F.J. van Schooten, M.J.X. Hillebrand, F.E. van Leeuven, N.

van Zandwijk, H.M. Jansen, L. den Engelse, E. Kriek,PAH–DNA adducts in white blood cells from lung cancerpatients: no correlation with adduct levels in lung, Carcino-

Ž .genesis 13 1992 987–993.w x34 R.A. Baan, M.-J.S.T. Steenwinkel, P.T.M. van den Berg, R.

Roggeband, J.H.M. van Delft, Molecular dosimetry of DNAdamage induced by PAH; relevance for exposure monitoring

Ž .and risk assessment, Hum. Exp. Toxicol. 13 1994 880–887.w x35 R.C. Garner, J. Cuzick, D. Jenkins, D.H. Phillips, A. Hewer,

M.M. King, M.N. Routledge, Linear relationship betweenDNA adducts in human lung and cigarette smoking, in: H.

Ž .Vainio, M. Sorsa, A.J. McMichael Eds. , Complex Mixturesand Cancer Risk, IARC, Lyon, 1990, pp. 421–426.

w x36 J.W. Gaubatz, B.H. Tan, Age-related studies on the removalof 7-methylguanine from DNA of mouse kidney tissue fol-lowing N-methyl-N-nitrosourea treatment, Mutat. Res. 295Ž .1993 81–91.

w x37 V.A. Bohr, R.M. Anson, DNA damage, mutation and fineŽ .structure DNA repair in aging, Mutat. Res. 338 1995

25–34.w x38 W. Szyfter, Z. Szmeja, K. Szyfter, J. Banaszewski, A. Kruk-

Zagajewska, K. Hemminki, Estimation of aromatic DNAadducts level in larynx tumour in relation to cancer staging,

Ž .Otolaryngol. Pol. 49 1994 410–417.w x39 K. Szyfter, K. Hemminki, R. Jaskula-Sztul, J. Banaszewski,

W. Szyfter, Z. Szmeja, Significance of DNA adduct studiesin the early phase of carcinogenesis, Period. Biolog. 96Ž .1994 345–347.

w x40 K. Szyfter, A. Kruk-Zagajewska, Z. Szmeja, W. Kozak, M.Kujawski, J. Banaszewski, J. Wojtowicz, Sex-associated dif-´ferential susceptibility of larynx to tobacco smoke carcino-

Ž .gens, Otolaryngol. Pol. 52 1998 153–158.w x41 D. Ryberg, A. Hewer, D.H. Phillips, A. Haugen, Different


susceptibility to smoking-induced DNA damage among maleŽ .and female lung cancer patients, Cancer Res. 54 1994

5801–5803.w x42 D.H. Phillips, B. Schoket, A. Hewer, E. Bailey, S. Kostic, I.

Vincze, Influence of cigarette smoking on the levels of DNAadducts in human bronchial epithelium and white blood cells,

Ž .Int. J. Cancer 46 1990 569–575.w x43 M. Peluso, E. Amasio, S. Bonassi, A. Munnia, F. Strupa, S.

Parodi, Detection of DNA adducts in human nasal mucosa32 Ž .tissue by P-postlabelling, Carcinogenesis 18 1997 339–

344.w x44 Z. Szmeja, A. Zagajewska-Kruk, W. Kozak, W. Szyfter, J.

Wojtowicz, A correlation between clinical stage of laryngeal´cancer and concentrations of sex hormones in blood, Oto-

Ž .laryngol. Pol. 50 1996 21–31.w x45 Y. Jin, F. Mertens, N. Mandahl, S. Heim, C. Olegard, J.

Wennerberg, A. Biorklund, F. Mitelman, Chromosome ab-¨normalities in 83 head and neck squamous cell carcinomas:influence of culture conditions in karyotypic pattern, Cancer

Ž .Res. 53 1993 2140–2146.w x46 Ch. Jin, Y. Jin, J. Wennerberg, J. Akervall, B. Grenthe, N.

Mandahl, S. Heim, F. Mitelman, F. Mertens, Clonal chromo-some aberrations accumulate with age in upper aerodigestive

Ž .tract mucosa, Mutat. Res. 374 1997 63–72.w x47 M. Worsham, M.J. Benninger, R.J. Zarbo, T.E. Carey, D.L.

van Dyke, Deletion 9p22pter and loss of Y as primarychromosome abnormalities in a squamous cell carcinoma of

Ž .the vocal cord, Genes, Chromosomes Cancer 6 1993 58–60.w x48 C.C. Harris, Chemical and physical carcinogenesis: advances

Ž .and perspectives for 1990s, Cancer Res. 51 1990 5023–5044, Suppl.

w x49 K.D. Sommers, M.A. Merrick, M.E. Lopez, L.S. Incognito,G.L. Schechter, G. Casey, Frequent p53 mutations in head

Ž .and neck cancer, Cancer Res. 52 1992 5997–6000.w x50 J.C. Law, T.L. Whiteside, S.M. Gollin, J. Weissfeld, L.

El-Ashmavy, S. Srivastava, R.J. Landreneau, J.T. Johnson,R.E. Ferrel, Variation of p53 mutational spectra betweencarcinoma of the upper and lower respiratory tract, Clin.

Ž .Cancer Res. 1 1995 763–768.w x51 Ch.H. Spruck III, Y.C. Tsai, D.P. Huang, A.S. Yang, W.M.

Rideaut III, M. Gonzales-Zulueta, P. Choi, K.-W. Lo, M.C.Yu, P.A. Jones, Absence of p53 mutations in primary na-

Ž .sopharyngeal carcinomas, Cancer Res. 52 1992 4787–4790.w x52 L.F. Zhang, K. Hemminki, K. Szyfter, W. Szyfter, P. Sode-¨

rkvist, p53 mutations in larynx cancer, Carcinogenesis 15Ž .1996 2949–2951.

w x53 P. Biggs, W. Warren, S. Venitt, M.R. Stratton, Does agenotoxic carcinogen contribute to human breast cancer? Thevalue of mutational spectra in unravelling the aetiology of

Ž .cancer, Mutagenesis 8 1993 275–283.w x54 K. Anwar, K. Nakakuki, H. Imai, M. Inuzuka, Over-expres-

sion of p53 protein in human laryngeal carcinoma, Int. J.Ž .Cancer 53 1993 952–956.

w x55 W.P. Bennett, T.V. Colby, W.D. Travis, A. Borkowski, R.T.Jones, D.P. Lane, R.A. Metcalf, J.M. Samet, Y. Takeshima,J.R. Gu, K.H. Vahakangas, Y. Soini, P. Paakko, p53 protein¨ ¨ ¨¨ ¨

accumulates frequently in early bronchial neoplasia, CancerŽ .Res. 53 1993 4817–4822.

w x56 J. Louhelainen, K. Szyfter, W. Szyfter, K. Hemminki, Lossof heterozygosity and microsatellite instability in larynx can-

Ž .cer, Int. J. Oncol. 10 1997 247–252.w x57 M.V. Gonzales, M.F. Pello, C. Lopes-Larrea, C. Suarez, M.J.

Menendez, E. Coto, LOH and mutation analysis of the p16Ž . Ž .9p21 and p53 17p13 genes in squamous cell carcinoma

Ž .of the head and neck, Clin. Cancer Res. 1 1995 1043–1049.w x58 W.H. Liggett jr., D.A. Sewell, J. Rocco, S.A. Ahrendt, W.

Koch, D. Sidransky, p16 and p16b are potent growth sup-pressor of head and neck squamous cell carcinoma cells in

Ž .vitro, Cancer Res. 56 1996 4119–4123.w x59 A.L. Reed, J. Califano, P. Cairns, W.H. Westra, R.M. Jones,

W. Koch, S. Ahrendt, Y. Eby, D. Sewell, H. Nawroz, J.Bartek, D. Sidransky, High frequency of p16 inactivation inhead and neck squamous cell carcinoma, Cancer Res. 56Ž .1996 3630–3633.

w x60 C.C. Harris, p53 tumour suppressor gene: from the basicresearch laboratory to the clinic — an abridged historical

Ž .perspective, Carcinogenesis 17 1996 1187–1198.w x61 T.-J. Liu, W.-W. Zhang, D.L. Taylor, J.A. Roth, H. Goepfert,

G.L. Clayman, Growth suppression of human head and neckcancer cells by the introduction of a wild-type p53 gene via

Ž .a recombinant adenovirus, Cancer Res. 54 1994 3662–3667.w x62 S.M. Overholt, T.-J. Liu, D.L. Taylor, M. Wang, A.K.

El-Naggar, S. Shillitoe, K. Adler-Storthz, L. St. John, W.-W.Zhang, J.A. Roth, G.L. Clayman, Head and neck squamouscell growth suppression using adenovirus-p53-FLAG; a po-tential marker for gene therapy trials, Clin. Cancer Res. 3Ž .1997 185–191.

w x63 W. Golusinski, J. Olofsson, Z. Szmeja, K. Szyfter, W.Szyfter, W. Biczysko, K. Hemminki, Alteration of p53 genestructure and function in laryngeal squamous cell cancer,

Ž . Ž .Eur. Arch. Otorhinolaryngol. 254 1 1997 S133–S137,Suppl.

w x64 J.P. Lavieille, E. Brambilla, C. Riva-Lavieiile, E. Reyt, R.Charachon, C. Brambilla, Immunohistochemical detection ofp53 protein in preneoplastic lesions and squamous cell carci-

Ž .noma of the head and neck, Acta Otolaryngol. StockholmŽ .115 1995 334–339.

w x65 R.M. Tomasino, E. Daniele, V. Bazan, V. Morello, V.Tralongo, R. Nuara, C. Nagar, M. Salvato, F. Ingria, S.Restivo, G. Dardanomi, A. Vecchione, A. Russo, Prognosticsignificance of cell kinetics in laryngeal squamous cell carci-

Ž .noma: clinicopathological associations, Cancer Res. 55 19956103–6108.

w x66 A. Franchi, O. Gallo, V. Boddi, M. Santucci, Prediction ofoccult neck metastases in laryngeal carcinoma: role of PCNA,MIB-1, and E-cadherin immunohistochemical determination,

Ž .Clin. Cancer Res. 2 1996 1801–1808.w x67 M.P. Copper, A. Jovanovic, J.J.P. Nauta, B.J.M. Braakhuis,

N. de Vries, I. van der Waal, G.B. Snow, Role of geneticfactors in the aetiology of squamous cell carcinoma of thehead and neck, Arch. Otolaryngol. Head Neck Surg. 121Ž .1995 157–160.


w x68 W.D. Foulkes, J.S. Brunet, L.P. Kowalski, S.A. Narod, E.L.Franco, Family history of cancer is a risk factor for scc of thehead and neck in Brazil; a case-control study, Int. J. Cancer

Ž .63 1995 769–773.w x69 W.D. Foulkes, J.S. Brunet, W. Sieh, M.J. Black, G. Shenouda,

S.A. Narod, Familial risk of scc of head and neck: retro-Ž .spective case-control study, Br. Med. J. 313 1996 716.

w x70 F. Janot, L. Massad, V. Ribrag, I. de Waziers, Ph.H. Beaune,B. Luboinski, O. Parise Jr., A. Gouyette, G.G. Chabot,Principal xenobiotic metabolizing enzyme systems in humanhead and neck squamous cell carcinoma, Carcinogenesis 14Ž .1993 1279–1283.

w x71 A. Hirvonen, Combination of susceptible genotypes and indi-vidual responses to toxicants, Environ. Health Perspect. 105Ž .1997 755–758.

w x72 K. Hemminki, Interindividual variation in the level of DNAand protein adducts in humans, in: Ecogenetics, Chapman &Hall, London, 1991, pp. 217–234.

w x73 T.D. Tlsty, A. Briot, A. Gualberto, I. Hall, S. Hess, M.Hixon, D. Kuppuswamy, S. Romanov, M. Sage, A. White,

Ž .Genomic instability and cancer, Mutat. Res. 337 1995 1–7.w x74 J.K. Wiencke, M.R. Spitz, In vitro chromosomal assays of

mutagen sensitivity and human cancer risk, Cancer Bull. 46Ž .1994 238–246.

w x75 T.C. Hsu, Genetic predisposition to cancer with specialreference to mutagen sensitivity, In Vitro Cell Dev. Biol. 23Ž .1987 591–603.

w x76 M.P. Spitz, J.J. Fueger, S. Halabi, S.P. Schantz, D. Sample,T.C. Hsu, Mutagen sensitivity in upper aerodigestive tractcancer: a case-control analysis, Cancer Epidemiol., Biomark-

Ž .ers Prev. 2 1993 329–333.w x77 P. Da̧browski, K. Pia̧tkowski, S. Kita, K. Szyfter, Z. Szmeja,

W. Szyfter, Chromosome sensitivity to bleomycin-induceddamage in larynx cancer as a marker of genetic risk, Oto-

Ž .laryngol. Pol. 52 1998 245–250.w x78 J. Cloos, B.J.M. Braakhuis, I. Steen, M.P. Cooper, N. de

Vries, J.J.P. Nauta, G.B. Snow, Increased mutagen sensitiv-ity in head and neck scc patients, particularly those with

Ž .multiple primary tumours, Int. J. Cancer 56 1994 816–819.w x79 A. D’Erricco, A. Taioli, X. Chen, P. Vineis, Genetic

metabolic polymorphisms and the risk of cancer: a review ofŽ .literature, Biomarkers 1 1996 149–173.

w x80 J. Seidegard, G. Ekstrom, The role of human glutathionetransferases and epoxide hydrolases in the metabolism of

Ž . Ž .xenobiotics, Environ. Health Perspect. 105 4 1997 791–798, Suppl.

w x81 V. Bongers, G.B. Snow, B.J.M. Braakhuis, The role ofglutathione-S-transferases in head and neck scc, Oral Oncol.

Ž .31B 1995 349–354.w x82 R. Jaskula-Sztul, A. Hirvonen, K. Husgafvel-Pursiainen, Z.

Szmeja, W. Szyfter, K. Szyfter, Glutathione-S-transferasegenotypes as modifiers of individual susceptibility to smok-ing-related larynx cancer, in: J.A. Werner, B.M. Lippert,

Ž .H.H. Ruppert Eds. , Head and Neck Cancer — Advances inBasic Research, Elsevier, Amsterdam, 1996, pp. 29–33.

w x83 R. Jaskula-Sztul, M. Reinekainen, K. Husgafvel-Pursiainen,Z. Szmeja, W. Szyfter, K. Szyfter, A. Hirvonen, Glutathione-S-transferase M1 and T1 genotypes and susceptibility to

Ž .smoking related larynx cancer, Biomarkers 3 1998 149–155.w x84 J. Brockmoller, R. Kerb, N. Drakoulis, B. Staffeld, I. Roots,¨

GST M1 and its variants A and B as host factors of bladdercancer susceptibility: a case-control study, Cancer Res. 54Ž .1994 4103–4111.

w x85 K. Szyfter, R. Jaskula-Sztul, J. Banaszewski, L. Moller, W.¨Szyfter, Z. Szmeja, Analysis of genotypes of detoxifyingenzymes demonstrates a significance of genetic factor in

Ž .laryngeal cancer, Wsp. Onkol. 7 1998 159–161, in Polish.w x86 W.M. Koch, Molecular biology of head and neck tumours,

in: Ch.W. Cummings, J.M. Fredrickson, L.A. Hacker, Ch.J.Ž .Krause, D.E. Schuller Eds. , Otolaryngology — Head and

Neck Surgery, 2nd edn., Mosby, St. Louis, 1995, pp. 35–50.

Molecular and cellular alterations in tobacco smoke-associated larynx cancer

Documents