Hindawi Publishing CorporationInternational Journal of DentistryVolume 2012, Article ID 540561, 7 pagesdoi:10.1155/2012/540561Review ArticleOral Leukoplakia as It Relates to HPV Infection: A ReviewL. Feller and J. LemmerDepartment of Periodontology and Oral Medicine, University of Limpopo, Medunsa Campus, Medunsa, South AfricaCorrespondence should be addressed to L. Feller, lfeller@ul.ac.zaReceived 13 July 2011; Revised 2 September 2011; Accepted 19 December 2011Academic Editor: Neil S. NortonCopyright 2012 L. Feller and J. Lemmer. ThisisanopenaccessarticledistributedundertheCreativeCommonsAttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.Leukoplakia is the most commonpotentiallymalignant lesionof the oral cavityandcanbe categorisedaccordingtoitsclinical appearanceashomogeneousornonhomogenous. Tobaccoandarecanutuse, eitheraloneorincombinationarethemost commonriskfactors for oral leukoplakia, but someoral leukoplakias areidiopathic. Someleukoplakias arisewithineldsof precancerizedoral epitheliuminwhichthekeratinocytesmaybeat dierent stagesof cytogenetictransformation.Leukoplakias may unpredictably regress, may remain stable, or may progress to carcinoma. There is a greater risk of carcinomatoustransformation of idiopathic leukoplakia, of non-homogenous leukoplakia, of leukoplakia aecting the oor of the mouth; theventrolateral surfaceofthetongueandthemaxillaryretromolarandadjoiningsoftpalate(collectivelycalledhigh-risksites),of leukoplakia with high-grade epithelial dysplasia, and of leukoplakia in which the keratinocytes carry cytogenetic alterationsassociated with carcinomatous transformation. Although there appears to be some link between human papillomavirus (HPV)and oral leukoplakia, there is little evidence to support a causal relationship either between HPV infection and oral leukoplakia orbetween HPV-infected leukoplakic keratinocytes and their carcinomatous transformation.1. IntroductionLeukoplakia is the most common potentially malignantlesion of the oral cavity [1, 2]. Leukoplakia is a term describ-ing a white lesion of the oral mucosa that cannot be char-acterizedclinicallyormicroscopicallyasanyotherdenedoraldiseaseentity[3,4].AtaWorldHealthOrganisation(WHO)workshopheldin2005, itwasrecommendedthatoral leukoplakia be dened as a white plaque of questionablerisk having excluded (other) known diseases or disorders thatcarrynoincreasedriskforcancer[57].Oralleukoplakianeedstobedistinguishedfromotherpredominantlywhitekeratotic lesions including frictional keratosis and stomatitisnicotina, which do not have malignant potential [1, 2, 58].About 7090% of oral leukoplakias are related to smok-ing and areca nut use, either alone or in combination, andthere is a direct relationshipbetweenthe frequency andthedurationof cigarette, pipe, or cigar smokingandtheprevalence of oral leukoplakia [8, 9]. The factors implicatedin the pathogenesis of idiopathic leukoplakia are unknown.However it is possible that infection of the oral epithe-liumwith human papillomavirus (HPV) and excessiveconsumption of alcoholic beverages may be associated withoral leukoplakia, but there is little evidence of a causalrelationshipbetweeneitherHPVinfectionoralcohol, andoral leukoplakia [6].It has been suggested that a denitive diagnosis oforal leukoplakiamust beestablishedbyhistopathologicalexclusion of other keratotic oral lesions that are recognisedas specic entities, and by exclusion of any aetiological agentsother than tobacco/areca nut use [7].It istheopinionof theauthorsthat thesecriteriaareunrealistically limiting, ignoring as they do the possibleroles of HPV, alcohol, chronic inammation, and low-grade chronic frictional trauma in the pathogenesis of oralleukoplakia. Referring to the WHO denition of 2005 above,it is dicult tounderstandhowsucha denitioncouldgainanyuseful currencysinceit is soexclusivethat itleaves no rational guidance for the everyday diagnosis of oralleukoplakia.2. Oral Leukoplakia: Clinical AspectsAccordingtoits clinical appearance, oral leukoplakiacanbecategorisedintotwomainclinical types: homogeneous2 International Journal of Dentistryand non-homogeneous. Either type may occur as an isolatedlesion or as multiple lesions. The leukoplakic lesion can varyin size from a few millimetres to several centimeters [1, 2, 512].Homogeneous leukoplakia is a uniformly white atplaquewithasmoothor relativelysmoothsurface; non-homogeneous leukoplakia may be nodular or verrucoushaving a wrinkled or corrugated surface or may be a minglingof white and red areas termed erythroleukoplakia [7, 10, 11].The clinical appearance of oral leukoplakia may changeover time. Some homogeneous lesions may become larger, ornon-homogeneous,butmostoralleukoplakiaswillremainstable or will regress, while some few will undergo carcino-matous transformation [1315].Oral erythroplakia, of all precancerous oral lesions,carriesthegreatest threat of malignant transformation. Ithas avelvety-redappearance, andabout 50%of all casesof erythroplakiaarealreadysquamouscell carcinomataatthe time of diagnosis. The erythroplakic component of oralerythroleukoplakia is identical to erythroplakia [16].Proliferative verrucous leukoplakia, considered to beeitheraclinical subtypeof non-homogeneousoral leuko-plakia or to be a distinct clinical entity, is not strongly associ-ated with smoking, is characterized by multiple leukoplakiclesions that aect wide areas of the oral epithelium, and mayprogress either to verrucous carcinoma or to squamous cellcarcinoma[1720]. Inmost cases, proliferativeverrucousleukoplakiaisrecognisedonlylateinitscoursesinceinitsinitial stages it is identical to an isolated leukoplakia [1724].3. Epidemiology of Oral LeukoplakiaThedatafromepidemiologicalstudiesonoralleukoplakiaisinconsistent, mostprobablyowingtodierencesincaseselection criteria (house-to-house surveys, hospital surveys,age, gender, race, ethnicity, and tobacco use) and in method-ology (diagnostic criteria, time of follow-up and whether ornot the leukoplakia had previously been treated) [1, 10, 11].Estimates of theglobal prevalenceof oral leukoplakiarange from 0.5% to 3.46%, and of the rates of carcinomatoustransformation of oral leukoplakia from 0.7% to 2.9% [25].OralleukoplakiaismoreprevalentinIndiawherepersonssmoke andpractice the habit of tobaccoandareca nutchewing more than elsewhere [26].Oral leukoplakia is usually diagnosed in middle age, andits prevalence increases with age. About 10% of oral leuko-plakias are idiopathic and the greater part of the remaining90%isassociatedwiththeuseof tobacco/arecanut [26].Males are more often aected than females probably owingtothegreaterprevalenceoftobaccousebymales[8]. Thebuccalmucosaisaectedin25%ofcases,themandibulargingiva in 20%, the tongue in 10%, the oor of the mouth in10%, and other oral sites account for the remainder [13].The literature on the relationship between race and oralleukoplakiaissparse. InaSouthAfricanstudyofarchivedhistopathological material, 86%of oral leukoplakias werefrom whites, 9% from blacks, and 5% from Asians, despitethefactthatthevastmajorityofSouthAfricansareblack[27]. Thisisnoteasytoexplain. Bearinginmindthatthestudy was on histopathological material, an explanation maybethatblackstendtopostponeseekingmedicaltreatmentuntil the leukoplakia has already undergone carcinomatoustransformation, thus skewingthestatistics awayfromtheleukoplakia[27], orbecauseblackpeopleinSouthAfricamay smoke less than white people.4. Epithelial Dysplasia and Oral LeukoplakiaThe reported prevalence of epithelial dysplasia in oral leuko-plakia ranges from5%to 25%[8]. Dysplasia is more frequentin non-homogeneous than in homogeneous leukoplakia[11], and it is probable that dysplasia is the histopathologicalexpression of genomic and molecular alterations in a eld ofkeratinocytes [28, 29].Thepresenceof epithelial dysplasiaisamarkerof themalignant potential of oral leukoplakia, andthe risk ofanindividual leukoplakiclesiontoprogresstocarcinomaincreases withthe increase of the grade of the epithelialdysplasia [11, 12, 14, 30].However,somedysplasticoralleukoplakiascanremainstable or even regress, while some oral leukoplakias withoutepithelial dysplasiawill indeedprogress tocarcinoma[5,11, 12]. Inonestudy36%of dysplasticoral leukoplakiasprogressedtosquamous cell carcinoma, but asubstantialproportionof 16%of oral leukoplakiaswithout epithelialdysplasia at the time of initial biopsyalsoprogressedtocarcinoma[31]. Theriskof progressiontocarcinomaofleukoplakias with moderate and severe dysplasia is estimatedtobetwiceas great as for oral leukoplakias withsimpleepithelial hyperplasia or with mild dysplasia [14].Treatment of dysplastic oral leukoplakia by excision, bylaserorbycryosurgery, orbytopical orsystemicchemo-therapy does not eliminate either the risk of relapse orrecurrence, or the risk of carcinomatous transformation[1,5,6].Theestimatedrecurrencerateoforalleukoplakiamay be as high as 30% [32], and squamous cell carcinomadevelops at 12%of sites of treated leukoplakia [15]. In a studyinvestigating the pattern of carcinomatous transformation oforal leukoplakia and oral erythroplakia, 36% of carcinomatadeveloped at the same site, 49% at contiguous sites, and 15%at oral sites remote from the preexisting lesions [33].It isevident fromthisdatathat somecasesof treatedoral leukoplakia are unpredictably destined to recur ortoundergocarcinomatous transformation, andthat thereare not yet any diagnostic methods available (clinical,histological or molecular), to condently identify these cases[1].Epithelial dysplasia in oral leukoplakia is a useful markerof the risk of carcinomatous transformation and is an impor-tant guide to clinical management [1, 11, 28]. However, sincedysplasiacanremainstableforlongperiods, itcannotbeused with condence as a predictor of carcinomatous trans-formation [14, 29]. Moreover, as the histological exercise ofgradingofepithelialdysplasiaishighlysubjectivewith lowinterpersonal and intrapersonal reproducibility [16, 34, 35],andasanincisional biopsycannotberepresentativeofanInternational Journal of Dentistry 3entire lesion [34, 36], a histopathological report of any degreeof epithelial dysplasia or of the absence of epithelial dysplasiamust be viewed with caution.5. The Natural Course and the MalignantPotential of Oral LeukoplakiaThe progression of oral leukoplakia to carcinoma is unpre-dictable but is relatively infrequent with an estimated overallriskoflessthan2%peryear[5, 6, 13, 26]; ifprogressionoccurs, it maytakeafewmonthsormanyyears[8]. Thecarcinomatous transformationof oral leukoplakia is notpredictablyassociatedwithtobaccosmoking[33], andthefrequency of carcinomatous transformationof idiopathicleukoplakia is higher than that of tobacco-associated leuko-plakia [11, 26].Inpopulations where smoking, the use of smokelesstobacco, reverse smoking, and the use of areca nut arevery prevalent, most squamous cell carcinomata arise frompreexistingleukoplakias;whileinpopulationswith alowerprevalence of these habits, most squamous cell carcinomataarise de novo innormal-looking epithelium[26]. It hasbeensuggestedthat squamous cell carcinoma arising denovousuallyrunsamoreaggressivecourseandhasalessfavourable prognosis than squamous cell carcinoma arisingfrompreexistingleukoplakia[12, 26], but arecent studyhas demonstrated little dierence [37]. Sometimes squamouscell carcinoma arises de novo inclose proximity tooralleukoplakias [8].Non-homogeneous leukoplakia has a greater risk of car-cinomatous transformation(2025%) thanhomogeneousleukoplakia (0.65%) [11, 13]. Most leukoplakias eitherremain stable or will regress [13, 15]. However, if proliferativeverrucous leukoplakia is considered as a distinct entity, mostsuch cases progress to carcinoma [18, 24].The rates of progression of large oral leukoplakias(>5 mm)andofleukoplakiasatsitesinthemouthknownto be at most risk of developing carcinoma (oor ofmouth, ventrolateral surface of tongue, and maxillaryretromolar/soft palateregion)aregreaterthanforsmallerleukoplakias or for leukoplakias at other sites in the mouth[1, 11, 13, 33, 38]. The increasedriskof carcinomatoustransformationoforal leukoplakiaathigh-risksitesisnotentirely a functionof the degree of dysplasia. It is alsodependent upon as yet undened characteristics of thelocationoftheleukoplakiasincetherateofcarcinomatoustransformation of dysplastic leukoplakias at high-risk sites isgreater than the rate of transformation of equally dysplasticleukoplakias at other sites [38].There is evidence that clearly suggests that some leuko-plakiasarisefromcytogeneticallyalteredtransformedker-atinocytes withinelds of precancerizedoral epithelium.Keratinocytes of oral leukoplakia show cytogenetic changesincludingalterations inthep53tumour suppressor gene,aberrations intheir DNAcontent, andloss of heterozy-gosity (LoH) at chromosomal regions of candidate tumoursuppressor genes [20, 33, 3945]. LoHat either 3por at9poccurs frequentlyinkeratinocytes of oral leukoplakiaand is associated with carcinomatous transformation ofthese lesions [33, 41, 42, 44, 45]. Additional cytogeneticalterations tothekeratinocytes intheprecancerizedeldreferredtoabove may result inthe evolutionof one orseveral keratinocytes containing a complete set of cytogeneticalterations of a cancerous phenotype, and in the subsequentdevelopment of squamous cell carcinoma [41, 46, 47].However, some precancerous oral leukoplakias in whichcytogenetic alterations in the keratinocytes cannot bedemonstrated, nevertheless undergocarcinomatous trans-formation [41, 42]. The pathogenic mechanisms that bringabout the progressive transformation of these keratinocytestocarcinomatouscellsareyet tobeelucidated. Most oralleukoplakias arebenigninnatureandwill remainstableorwill regress[28, 32]. Theseleukoplakiasprobablyhaveadierent aetiopathogenesis toprecancerous leukoplakiasand probably do not have the cytogenetic characteristics ofprecancerous leukoplakias. What is certain, however, is thatleukoplakias withmalignant potential andthose withoutmalignant potential cannot be distinguished clinically [1].6. HumanPapillomavirus andOral LeukoplakiaHuman papilloma viruses (HPVs) are strictly epitheliotropicandinfect eithercutaneousormucosal squamousepithe-lium, dependingupontheirgenotype[48, 49]. Thosethatinfect mucosal epithelium have been categorized into high-risk types (e.g., HPV-16, 18, 31, 33, and 35) based on theirepidemiological associationwithcarcinomaof the cervixuteri,orintolow-risktypes(HPV-6,11, 13, and32)[50].Thesecategorieshavebeenuniversallyadoptedforuseinstudies of the oncogenic signicance of HPV infection at allanatomical regions of the upper aerodigestive tract.Low-riskHPVgenotypes havebeenimplicatedinthepathogenesis of the benign oral proliferative epitheliallesions, squamous cell papilloma, commonwart (verru-cousvulgaris), condylomaacuminatum, andfocal epithe-lial hyperplasia(Heckdisease); whilehigh-risktypeshavebeenassociatedwithprecancerousandcancerousoralandoropharyngeal epithelial lesions [49, 5156].There is an extreme variation in the reported prevalenceof HPV infection in oral precancerous and cancerous lesionsranging from0%to 100%[57, 58]. This is owing todierences insampling andHPVdetectionmethods, todierences in ethnicity, geographic locations, and sample sizeofthesubjectsexamined, andtotheinappropriategroup-ingtogetherof dierent lesionsfromdierent anatomicallocations of the mucosa of the upper aerodigestive tract[49, 51, 5864].ManystudiesinvestigatingtheassociationofHPVandsquamous cell carcinoma of the mucosa of the upperaerodigestive tract usedPCRtechniques for detectionofHPVDNAwithout alsoquantifyingtheDNAviral load.PCR can detect extremely small fragments of DNA that mayrepresent either contamination of the sample or biologicallyinsignicant HPV infection [51, 57, 58]. These ndings havebeen reported as if they were pathogenically signicant.4 International Journal of DentistryWhetherthesearelegitimatendingsoraretheresultsofinconsistenciesanderrorsinmethodology,severalHPVgenotypeshavebeendetectedinprecancerousorallesions.High-risk HPV genotypes, in particular HPV-16, have beenreported to be the most prevalent in oral leukoplakias,including proliferative verrucous leukoplakia [55, 65]. Otherreports implicated low-risk rather than high-risk HPVgenotypes in oral leukoplakia [54, 63], and yet others assertthat oral leukoplakiais coinfectedwithavarietyof HPVgenotypes[49, 55, 66]. Inameta-analysisofdatafrom94studiesofatotalof4580specimens, MillerandJohnstone[63] determined that the likelihood of HPV being detectedinprecancerousoral lesionsis2to3timesgreaterandinoral squamouscell carcinomais4to5timesgreaterthaninnormaloralmucosa.TheprevalenceofHPVinnormaloral mucosa, innondysplastic leukoplakias, indysplasticleukoplakiasandinotherprecancerousintraepithelial oralneoplasms, and in oral squamous cell carcinoma is likely tobe 10%, 20.2%, 26.2%, and 46.5%, respectively [63].This suggests that there may be some link between HPVinfectionandoral precancerousandcancerouslesions. AsE6andE7oncoproteinsofhigh-riskHPVgenotypeshavethecapacitytomediatecarcinomatous transformationofinfectedkeratinocytesbyinactivatingcellularp53andRbtumour suppressor pathways [50, 52], HPV may play eitheran oncogenic or a co-oncogenic role in some HPV-infectedprecancerous and cancerous epithelial neoplasms.In fact, HPV-16 has been found to be causally associatedprimarily with squamous cell carcinoma of the palatal tonsils[6770]inasubsetofsubjectswhoareyounger, consumeless tobacco, are more engaged in high-risk sexual behaviour(great number of lifetime sexual partners andpracticingoral-genital sex), have higher HPV-16 serum antibody titers,andhaveabetterdisease-freesurvival andoverall survivalrates thansubjects withHPV-cytonegative oropharyngealsquamous cell carcinoma [51, 64, 6770]. The cells of HPV-cytopositive oropharyngeal carcinoma in these subjects havea distinct molecular prole [50]. The cells of squamouscell carcinoma causally associated with HPV express E6/E7oncoproteins. They frequently demonstrate viral integrationwithin the cellular genome with the presence of intactE6gene. Theyexhibit highviral load, reducedexpressionof Rbproteins, functional overexpressionof p16INK4A,unmutatedp53gene, andlossof heterozygosity(LoH)atchromosomal loci 3p, 9pand17pis infrequent [51, 6874]. In contrast, HPV-cytonegative oropharyngeal squamouscell carcinomaischaracterizedbyp53genemutations, byfrequent LoHat 3p, 9p, and17p, bydecreasedlevels ofp16INK4A and by normal or increased levels of Rb proteins[50, 51, 71].Recent meta-analyses and comprehensive studies [60, 62,67, 75]showlittleornocausal associationbetweenHPVand oral squamous cell carcinoma in contrast to the strongassociation between HPV and oropharyngeal squamous cellcarcinoma. HPV-16cytopositiveoral squamouscell carci-noma is characterized by a low viral load, by infrequent viralintegration, andthecancerouscellsseldomcontainactivetranscriptional E6/E7 mRNA [72, 76]. However, it is possiblethat, inHPV-cytopositiveoral squamouscell carcinomatathat do not express E6/E7 mRNA, E6/E7 oncoproteins maywell have participated or have had a complementary role inthe initial transformation, but then phased out [77].Withoropharyngeal carcinomaasthemodel, acausalassociation between HPV and cancerization in oral epithe-liumislikelyif thecellsof thelesioncontainHPVDNAexpressing E6 and/or E7 mRNA [71], if there is viralintegrationwithinthecellulargenome[74],andifthereisahighviral load(>1copiespercell). Alimitedbiologicalsignicanceof thevirus intheprocess of transformationcan be deduced if there is a low-copy number (