pathology of hpv infection at the cytologic and histologic ... of hpv infection at the cytologic...

International Journal of Gynecology and Obstetrics (2006) 94 (Supplement 1), S22---S31

www.elsevier.com/locate/ijgo

CHAPTER 3

Pathology of HPV infection at the cytologic andhistologic levels: Basis for a 2-tiered morphologicclassification system

Thomas C. Wright, Jr.

KEYWORDSSquamous intraepitheliallesion (SIL);Cervical intraepithelialneoplasia (CIN);Cervix;Cancer precursors;Human papillomavirus(HPV)

Abstract Over the last 2 decades the pathogenesis and natural history of cervi-cal cancer has become clearer. As a result, the cytologic and histologic terminol-ogy used to refer to cervical cancer precursors has needed to change. Today werecognize that almost all cervical cancers are due to infection with specific high-risk types of human papillomavirus (HPV). Most women become infected with theseviruses within several years of initiating sexual intercourse and a productive HPV in-fection frequently results in characteristic morphologic changes within the infectedcervical squamous cells. Cells demonstrating the morphologic changes associatedwith a productive HPV infection are referred to as low-grade squamous intraepi-thelial lesions (LSIL) when observed in cytologic specimens and low-grade cervicalintraepithelial neoplasia (CIN 1) when observed in histologic specimens. In somewomen, HPV gene expression becomes unlinked to the state of differentiation ofthe infected epithelial cells and deregulated expression of the early region of theviral genome results in a dramatic increase in expression of two HPV oncoproteins(E6 and E7). This results in loss of normal cell cycle control of the epithelium andgenetic instability. When this occurs the epithelium develops characteristic mor-phologic features, with immature “basaloid-type” squamous cells and mitotic fig-ures in the upper half of the cervical epithelium. Such lesions are felt to represent“true” neoplasia and are referred to as high-grade squamous intraepithelial lesions(HSIL) when observed in cytologic specimens and high-grade cervical intraepithelialneoplasia (CIN 2,3) when observed in histologic specimens.© 2006 International Federation of Gynecology and Obstetrics. Published by ElsevierIreland Ltd. All rights reserved.

1. Introduction

The histopathologic and cytopathologic classifica-tion of a disease should reflect both its clinical be-havior and the current concepts regarding its patho-genesis. Over the last 2 decades our understand-

0020-7292/$ --- see front matter © 2006 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd.All rights reserved.

ing of the pathogenesis of cervical cancer and itsprecursor lesions has increased dramatically. As aresult, the old terminology is being discarded andreplaced by a new terminology that more accu-rately reflects both clinical behavior and pathogen-esis. Changes in terminology are frequently disrup-

Pathology of HPV infection at the cytologic and histologic levels S23

tive. Inevitably, many will question the need for thechange and it takes years for training resources suchas textbooks and management guidelines to be up-dated. The recent change in the histopathologic andcytopathologic classification of cervical cancer pre-cursors seems to have been particularly disruptive,perhaps because so many clinicians are involved indifferent aspects of cervical cancer prevention. Inthis article we review the pathogenesis of cervicalcancer as well as the histopathologic and cytopatho-logic classification and features of cervical cancerprecursors.

2. Older terminologies

From the 1930s though the 1960s, noninvasive intra-epithelial lesions of the cervix were classified us-ing a 4-tiered terminology [1]. The 4 categories ofprecursor were mild dysplasia, moderate dysplasia,severe dysplasia, and carcinoma in situ (Figure 1).It was not widely accepted that dysplastic lesionswere cervical cancer precursors, and carcinoma insitu was believed to be the only “true” precur-sor. Therefore, women with dysplasias of all gradestended to be followed up cytologically, whereastreatment for those with carcinoma in situ was fre-quently a cone biopsy or a hysterectomy. By theearly 1970s, laboratory studies as well as prospec-tive clinical follow-up studies showed that the dys-plasia/carcinoma in-situ terminology reflected nei-ther the clinical behavior nor the pathogenesis ofcervical cancer precursors [2]. The laboratory stud-ies available at that time indicated that the cellular

Figure 1 Dysplasia and carcinoma in situ terminologycompared with the CIN terminology. From the 1950sthrough the 1960s the terms dysplasia and carcinoma insitu were widely used to describe noninvasive intraepi-thelial lesions of the cervix. They were replaced in the1970s with the CIN terminology, which combined severedysplasia and carcinoma in situ into CIN 3.

changes observed in dysplasia and carcinoma in situwere qualitatively similar and remained rather con-stant throughout the histologic spectrum of puta-tive precursor lesions. Both dysplasia and carcinomain situ could be monoclonal proliferations of abnor-mal squamous epithelial cells, and both were fre-quently aneuploid [3]. By electron microscopy andtime-lapse cinematography, all grades of dysplasiaand carcinoma in situ shared features [4]. Moreover,studies of intraobserver and interobserver variabil-ity in the histologic diagnosis of severe dysplasia andcarcinoma in situ demonstrated that pathologistscould not reproducibly differentiate the 2 entities[5]. Perhaps most importantly, prospective follow-up studies demonstrated that dysplastic lesions ofall grades could progress to carcinoma in situ, andthat all, therefore, could progress to cervical can-cer. Based on the information available in the early1970s, Richart [6] introduced the concept that alltypes of precursor lesions to squamous carcinomasof the cervix represented a single disease process.He termed this process cervical intraepithelial neo-plasia (CIN).

At the time of its introduction, CIN was believedto describe a spectrum of histologic and cytologicchanges that shared a common etiology, biology,and natural history. Although the CIN terminologydivided the precursor lesions into 3 groups --- CIN1 for mild dysplasia; CIN 2 for moderate dyspla-sia; and CIN 3 for both severe dysplasia and carci-noma in situ (Figure 1) --- these groups were con-sidered to represent different stages of a single bi-ological continuum. Therefore, inherent in the CINterminology was the belief that CIN lesions of allgrades were “true” cancer precursors and that all,even CIN 1 lesions, had the potential to progress toinvasive cancer if untreated. The CIN terminologywas introduced concordantly with the developmentof simple outpatient ablative techniques for treat-ing CIN lesions, such as cryotherapy and electro-fulguration. Because of this convergence it becamewidely recommended that CIN lesions of all grades,including CIN 1, be treated. The CIN terminology be-came widely adopted for use in both histopathol-ogy and cytopathology. Yet, as our understanding ofthe pathogenesis of cervical cancer and its precur-sor lesions improved, it became clear that the basicpremise underlying the CIN terminology was incor-rect.

3. Pathogenesis of cervical cancer

Today it is recognized that human papillomavirus(HPV) infection plays an essential role in the patho-genesis of cervical cancer [7]. Most women become

S24 T.C. Wright, Jr.

exposed to and infected with HPV shortly after ini-tiating sexual activity. The target cells for an initialHPV infection are the immature basal cells of theepithelium, and HPV is thought to reach these cellsthrough microabrasions or cracks within the epithe-lium. Viral replication is tightly linked to the dif-ferentiation state of the virally infected epithelialcells [8,9]. Replication of the HPV genome is tightlycontrolled by cellular mechanisms within the basalcells, and appears to be linked to cellular repli-cation so that the viral DNA replicates with thehost' s genome. The number of copies of the HPVgenome, as a circular or episomal form, is thus lowin the basal cell' s nucleus, and virally encoded pro-teins are expressed at very low levels. As a result,HPV-infected basal cells show no specific cytologicor histologic changes and cannot be distinguishedfrom uninfected cells. This stage of an HPV infec-tion is referred to as a “latent” or “clinically un-apparent” infection since the woman is HPV DNApositive, but no lesions can be detected, even bymicroscopy.

As HPV-infected epithelial cells differentiate andmove upwards in the epithelium, viral transcriptionof the early region of the HPV genome dramaticallyincreases [8,9]. The early region encodes for a num-ber of proteins, including E1, E2, E6, and E7, whichare important for viral replication. The HPV genomeis relatively small (approximately 8000 base pairs)and does not encode the enzymes required for vi-ral DNA replication. The only proteins provided bythe viral genome that are directly involved in viralreplication are 2 regulatory proteins, E1 and E2, andHPV must therefore rely on the host' s DNA replica-tion machinery for viral DNA synthesis [10]. How-ever, as the epithelial cells differentiate, the cel-lular DNA replication machinery is normally inacti-vated. In order to undergo vegetative DNA ampli-fication in the differentiating epithelial cells, thevirus needs to reactivate the cellular DNA repli-cation machinery. Studies of cultured human ker-atinocytes have shown that the viral E7 protein iscapable of reactivating the cellular DNA replica-tion machinery in differentiated cells [11]. The viralE6 protein also appears to play an important roleby blocking the apoptosis that would normally oc-cur in the differentiated cells [12]. Together, thesechanges provide in the cell the synthetic phase envi-ronment necessary for vegetative viral DNA replica-tion and complete virion formation. Infectious virusis eventually released as the differentiated cells areshed from the epithelium. In most women immu-nity develops against HPV after a period of monthsor years, and productive viral infection ceases [13].These women eventually become HPV DNA nega-tive.

In some HPV-infected women viral gene expres-sion becomes unlinked to the state of differentia-tion of the infected epithelial cells, resulting in achange in the topography of viral gene expressionwithin the epithelium (see Chapter 1) [8,9]. One ofthe results of this other mode of viral infection isthe dramatic increase in the expression of E6 and E7HPV in the lower layers of the epithelium due to theproteins' deregulated expression [8,9]. The deregu-lated expression of HPV proteins E6 and E7 results inthe disruption of normal cell cycle regulation; abro-gation of apoptosis mechanisms; and genetic insta-bility. Genetic instability, which is a characteristicfeature of most malignant neoplasms, occurs earlyin the development of precancers, thereby allow-ing for the stepwise acquisition of multiple muta-tions. Produced by alterations in the mitotic spin-dle apparatus, genetic instability permits aberrantmitotic events that can produce a disequilibrium inthe distribution of chromosomes, leading to changesin the number and structure of chromosomes. It caneventually produce a change in the overall DNA con-tent, referred to as aneuploidy. Genetic instabilityis thought to play a critical role in the developmentof cervical cancer [14].

The association between infection with high-risktypes of HPV and high-grade cervical cancer precur-sors as well as cervical cancer is so strong that ithas prompted the use of HPV DNA testing in theclinical management of women with the “border-line” cytologic abnormalities referred to as atypicalsquamous cells of undetermined significance (ASC-US), and as an adjunct to the cytologic evaluation ofwomen aged 30 years and older who undergo routinecervical screening [15,16]. When HPV DNA testingis used in the management of women with ASC-US,molecular testing for high-risk types of HPV is per-formed, typically with residual fluid after a liquid-based cytologic specimen is processed. If a womantests positive for high-risk HPV DNA she is referredfor colposcopy; and if she tests negative, routinescreening is resumed [16,17]. In women who testpositive for high-risk HPV DNA but are found to becytologically negative, both tests are repeated 6 to12 months after the initial screening. Because thenegative predictive value of being both cytologicallynegative and testing negative for high-risk HPV DNAis greater than 98.8% in all studies, and 99.9% or100% in most, women for whom the results of bothtests are negative do not require rescreening for 3years [15].


4. Two-tiered morphologic classificationsystem of noninvasive HPV-associatedcervical lesions

It used to be believed that the multistep nature ofthe pathogenesis of cervical cancer described abovewas mirrored by a defined sequence of morpho-logic changes, identified as CIN 1, CIN 2, and CIN3, and that these changes reflected the natural his-tory of CIN lesions. Numerous studies have demon-strated over the past several decades that a higherCIN grade correlated with a greater risk that lesionsprogress to a higher grade or to invasive cervicalcancer. A comprehensive review of natural historystudies of CIN lesions was conducted by Ostor (Ta-ble 1) [18]. The investigator concluded that spon-taneous regression during the different follow-upstudies occurred in 57%, 43%, and 32% of cases forCIN 1, CIN 2, and CIN 3 lesions, respectively, andthat persistence occurred in 32%, 35%, and 56% ofthese lesions. Only 1% of CIN 1 lesions and 5% of CIN2 lesions, but more than 12% of CIN 3 lesions, pro-gressed to invasive cervical cancer during follow-up.

Moreover, it is now recognized that lesions thatwould in previous days have been considered high-grade CINs (CIN 2,3) often develop quickly after ini-tial infection with high-risk types of HPV, and thatsuch lesions are not necessarily preceded by lesionswith the morphological appearance of a CIN 1 [19].For example, in a long-term follow-up study of col-lege students, it was found that the median time

Table 1 Summary of the natural history of CINlesions *

Regress Persist Progress to Progress toto CIS to invasion

CIN 1 57% 32% 11% 1%CIN 2 43% 35% 22% 5%CIN 3 32% 56% >12%

Abbreviations: CIN, cervical intraepithelial neoplasia; CIS,carcinoma in situ.* Adapted from reference [18].

Table 2 Comparison of terminologies used for cytologic and histologic findings

Dysplasia/carcinoma CIN terminology Bethesda System Modified CIN terminologyin situ for cytologic studies for histologic studies

Mild dysplasia CIN 1 LSIL Low-grade CIN (CIN 1)Moderate dysplasia CIN 2 HSIL High-grade CIN (CIN 2,3)Severe dysplasia CIN 3Carcinoma in situInvasive cancer Invasive cancer Invasive cancer Invasive cancer

Abbreviations: CIN, cervical intraepithelial neoplasia; HSIL, high-grade squamous intraepithelial lesions; LSIL, low-gradesquamous intraepithelial lesions.

from first detection of HPV DNA to identification ofa histologically confirmed CIN 2,3 lesion was only14.1 months [19]. Similarly, in a long-term follow-up study of a prophylactic vaccine against HPV-16,most CIN 2,3 lesions associated with HPV-16 in theplacebo arm appeared to develop rapidly after theinitial infection with HPV-16 [20]. Therefore, theolder concept that it took years to progress froma CIN 1 to a CIN 2 or CIN 3 lesion appears to be in-correct.

We thus recognize that high-grade CINs (CIN 2,3)often develop quickly after an incident infectionwith high-risk types of HPV, and that most CIN 1 le-sions are not “neoplasia” but simply virus-producinglesions. Moreover, a lesion with the morphologic ap-pearance of CIN 2 does not necessarily have fewerchromosomal alterations than a lesion with the mor-phologic appearance of CIN 3. Therefore, the termi-nology used for noninvasive HPV-associated cervicallesions needs to be changed to better reflect the bi-ology of HPV infections. Instead of classifying non-invasive HPV-associated cervical lesions as a contin-uum of lesions (CIN 1, CIN 2, and CIN 3) reflectinga single underlying biological process at differentstages of development, we need to represent non-invasive HPV-associated cervical lesions as 2 distinctbiological entities (Figure 2) [21]. One represents aproductive viral infection caused by oncogenicallylow- or high-risk types of HPV. These lesions are usu-ally self-limited and spontaneously resolve when thepatient stops shedding HPV. The other represents atrue neoplastic process confined to the epithelium,but with the capacity to progress to invasive cervicalcancer if not treated. These neoplastic lesions areusually histologically high grade; they almost alwaysare associated with high-risk types of HPV; and theyshow deregulation of E6 and E7 expression. Theyalso are monoclonal and have chromosomal alter-ations.

The 2-tiered classification system can account forall noninvasive HPV-associated cervical lesions. Le-sions that are most likely the manifestation of aproductive viral infection are referred to as “low-grade” and those that potentially represent a cervi-


Figure 2 Low-grade/high-grade terminology for noninvasive HPV-associated epithelial lesions. This terminology isbased on the realization that morphologically low-grade lesions generally represent productive HPV infections and donot act as precursors to invasive cervical cancer. In contrast, morphologically high-grade lesions tend to be neoplasticand have the potential to act as precursors to invasive cancer. Low-grade and high-grade lesions are separated on thediagram to emphasize the fact that high-grade lesions can develop de novo.

cal cancer precursor are referred to as “high grade”(Figure 2). However, whether these lesions shouldbe referred to as squamous intraepithelial lesions(SILs) or should remain known as CINs is subjectto controversy among proponents of a 2-tiered ter-minology system. In the United States, where 2-tiered terminology and the Bethesda System havebeen widely adopted for cytologic determination,the terms low-grade squamous intraepithelial le-sion (LSIL) and high-grade squamous intraepitheliallesion (HSIL) are used in cytologic diagnosis [22].However, most pathologists continue to use CIN 1to refer to lesions likely to manifest a productive vi-ral infection and CIN 2,3 for lesions that manifest aneoplastic process [23].

It is important for clinicians to recognize thathistopathologic and cytologic appearance are, un-fortunately, relatively poor predictors of the bio-logic potential of a given lesion. Based on bothbiomarkers and biologic behavior over long-termfollow-up, it is now clear that some histologicallyand cytologically low-grade lesions have the bio-logic characteristics of a high-grade lesion [24]. Thisis particularly true for low-grade lesions associatedwith high-risk types of HPV. For example, low-gradelesions associated with HPV-16 and HPV-18 are in-variably monoclonal proliferations of cells, a fea-ture of neoplasia and not of a productive viral in-fection [25]. Conversely, some lesions that could beclassified as “high-grade” based on their histologicand cytologic appearance have the biologic charac-teristics of low-grade lesions. This is especially truefor lesions with the histologic appearance of CIN 2.

Almost half of these lesions are self-limited, andhave been observed to regress spontaneously withinshort periods without any treatment [24]. However,most of these lesions are associated with high-risktypes of HPV and many are aneupoloid and demon-strate loss of heterozygosity (LOH) at specific chro-mosomal loci. These are characteristic features ofCIN 3 lesions and invasive cervical cancer [26]. Ahistopathologic diagnosis of CIN 2 is also poorly re-producible. Because a considerable proportion of le-sions with the histologic features of CIN 2 have thebiologic features of neoplasia, they are classified ashigh-grade in the 2-tiered terminology [21,22,26].

5. Histologic and cytologic features ofnoninvasive HPV-associated cervical lesions

5.1. Low-grade productive HPV infections

During the productive viral infection stage, largenumbers of complete viral genomes are produced,which are subsequently packaged with the capsidproteins into infectious virions. Characteristic cyto-logic changes occur in the squamous cells in whichproductive HPV infection is taking place. These aremultinucleation, nuclear enlargement, nuclear hy-perchromasia, irregular nuclear outlines, and perin-uclear halos. Squamous epithelial cells demonstrat-ing these cytologic features are referred to as koilo-cytes, a word derived from the Greek koilos, whichmeans hole (Figure 3). These changes are recogniz-able in both cytologic and histologic specimens. In


Classic cytologic features low-grade CIN:• Cellular enlargement• Multinucleation• Nuclear hyperchromasia• Nuclear irregularity• Presence of perinuclear halos

Classic cytologic and histologic features high-grade CIN:• Immature basaloid-type cells with high nucleus to cytoplasm ratio• Immature basaloid-type cells in the upper half of the epithelium• Mitoses occurring in the upper half of the epithelium• Irregular, hyperchromatic nuclei• Abnormal mitotic figures

addition to the cytologic changes occurring in in-dividual squamous cells, architectural changes alsooccur in a squamous epithelium in which produc-tive HPV infection is taking place. The squamousepithelium then frequently becomes acanthotic (orthickened), and sometimes forms papillary projec-tions with central fibrovascular cores (Figure 4A andB). Lesions with this histologic appearance are re-ferred to as low-grade CINs (CIN 1), whereas speci-mens showing the cytologic features of HPV effectsare referred to as LSILs.

In-situ hybridization techniques can demonstratethe presence of HPV DNA in koilocytes within tis-

Figure 3 Low-grade squamous intraepithelial lesion(LSIL). Epithelial cells removed using a spatula or broomfrom the more differentiated layers of the squamousepithelium of the cervix demonstrate the characteristicfeatures of a productive human papillomavirus (HPV) in-fection. They show multinucleation, perinuclear halos,nuclear enlargement, and hyperchromasia. Cells withthese features are frequently referred to as koilocytes.

Figure 4 Low-grade cervical intraepithelial neoplasia(CIN 1). (A) During productive HPV infection the cervi-cal squamous epithelium becomes somewhat thickened,or acanthotic. (B) At higher magnification the cytologiceffects of a productive HPV infection are obvious, withmultiple koilocytes.


Figure 5 In-situ hybridization of HPV in a low-gradeCIN (CIN 1). In-situ hybridization has been performedwith a probe against the DNA of high-risk types of HPV.Cells that have blue nuclei have large numbers of HPVgenomes indicating that a productive viral infection istaking place. Most of the productively infected cells arein the more differentiated layers of the epithelium. Al-though this lesion has the morphologic features of a low-grade lesion, it is caused by infection with a high-risktype of HPV.

sue sections (Figure 5), and electron microscopicstudies have confirmed the presence of completelyformed viruses in koilocytes [27]. However, the pres-ence of perinuclear clearing, or halos, in cells with-out significant nuclear atypia is seen in a vari-ety of non-neoplastic disorders as well as in HPVinfection. These disorders include infections withTrichomonas organisms, Gardnerella vaginalis, andCandida organisms. Moreover, perinuclear clearingcan occasionally be a feature of the atrophic epithe-lial changes found in the cervices of postmenopausalwomen or of women with immature squamous meta-plasia. Before the recognition that all CINs are as-sociated with HPV infection, cervical condylomasand low-grade CINs (CIN 1) were generally classi-fied as separate histopathologic entities. With thewidespread application of HPV typing to cervical le-sions, it has become evident that it is impossibleto distinguish, on histologic grounds alone, betweenflat cervical condylomas and low-grade CIN (CIN 1)lesions associated with “low oncogenic risk” types of

HPV and those associated with “high oncogenic risk”types of HPV [26]. Since clinical and histologic cri-teria do not allow differentiation among low-gradelesions on the basis of their oncogenic potential,distinctions between flat cervical condylomas andlow-grade CIN (CIN 1) lesions are no longer clinicallymeaningful and have been dropped.

Over a period of several months, the vast ma-jority of HPV-infected persons develop an effectiveimmune response against the type of HPV that theyharbor [13]. The shedding of HPV DNA spontaneouslystops once the immune response has developed.This occurs within 2 years of infection in approx-imately 90% of women. When the shedding stops,the cytopathic effects of productive HPV infectiondisappear. In natural history studies, women havetended to remain HPV DNA negative after they havestopped shedding a given type of HPV. However, itis still unknown what proportion of the women whospontaneously stop shedding HPV DNA actually clearthe infection from their bodies, and in what propor-tion HPV is maintained in an episomal form and lowcopy numbers in the basal cells of the mucosa.

5.2. High-grade neoplastic HPV-associated lesions

Lesions that have the histologic features of a neo-plasia and may be precursor lesions are referredto as high-grade CINs (CIN 2,3), whereas cytologi-cal specimens with these features are referred toas HSILs. Although HPV-induced cytopathic effectsare often present, they are usually less prominentin high-grade CIN (CIN 2,3) than in low-grade CIN(CIN 1). High-grade CIN (CIN 2,3) is characterizedby immature basaloid cells and mitotic figures thatextend into the upper half of the epithelium (Figure6). The immature basaloid cells usually have nuclearcrowding, pleomorphism, and loss of normal cell po-larity. They have a minimal amount of cytoplasmand therefore a high nucleus to cytoplasm ratio.Another feature found in most high-grade lesionsis the presence of abnormal mitotic figures (Figure7). Most invasive cervical cancers have chromosomalabnormalities and LOH at specific chromosomal loci.In addition, many are aneuploid. This suggests thataneuploid intraepithelial lesions with LOH at theseloci are at risk for progressing to invasive carcino-mas if left untreated, and studies have documentedthat the percentage of aneuploid CIN lesions withLOH at the specific sites increases with CIN grade[28---30]. In one study that used an imaging systemto measure the DNA content of Feulgen-stained cer-vical cells, it was found that 33% of mildly dysplasticlesions were aneuploid [28]. The percentage of ane-uploid lesions increased to 75% for moderate dys-plasias and 90% for severe dysplasias and carcino-


Figure 6 High-grade cervical intraepithelial neoplasia(CIN 2,3). Immature basaloid-type cells extend into themiddle layers of the epithelium. These cells have en-larged hyperchromatic nuclei giving them a high nucleusto cytoplasm ratio. Mitoses are also occurring in the up-per half of the epithelium.

mas in situ. Similar results have been obtained us-ing other techniques to assess DNA content. As 85%of CINs with aneuploid DNA patterns contain abnor-mal mitotic figures, the best histologic predictor ofan aneuploid lesion is the presence of abnormal mi-totic figures [31].

Another characteristic feature of high-gradelesions is overexpression of p16INK, a cyclin-dependent kinase inhibitor involved in the controlof the cell cycle [32]. The intracellular expressionof p16INK is increased upon the binding of high-risk,HPV-derived E7 oncoproteins to the retinoblastomagene product. A number of studies have demon-strated that staining with p16INK is very uncommonin normal cervical squamous epithelium, but thata small proportion of CIN 1 lesions, a higher pro-portion of CIN 2 lesions, and the vast majority ofCIN 3 lesions and cancers stain positively for p16INK(Figure 8) [33---36]. Moreover, prospective follow-upstudies suggest that p16INK-positive lesions behaveas precursor lesions. In one prospective study withwomen having low-grade CINs (CIN 1), the rate ofregression of p16INK-negative lesions was found tobe higher (71%) than that of p16 INK-positive lesions(38%) [37]. Although some of the p16INK-negative

Figure 7 Abnormal mitotic figure. A characteristic fea-ture of high-grade CIN (CIN 2,3) is aneuploidy. Abnormalmitotic figures are good markers for aneuploidy; theirpresence in a lesion indicates that the lesion is neoplas-tic.

Figure 8: Immunohistochemical staining for p16INK.High-grade CIN (CIN 2,3) lesions almost invariablydemonstrate overexpression of p16INK. p16INK overex-pression is shown by the dark-brown staining of the ep-ithelium.

low-grade CINs (CIN 1) progressed to high-gradeCINs (CIN 2,3) in that study, the p16INK-positive le-sions were much more likely to progress. In anotherprospective study, 44% of the women with lesionsfound to be p16INK positive on biopsy and classi-fied as “not CIN 2,3” by consensus histopathologicdiagnosis were subsequently diagnosed with CIN 2,3[33]. Therefore, there is considerable data suggest-ing that p16INK immunostaining can help identifywhich high-grade CIN (CIN 2,3) lesions will behaveas precursor lesions.


6. Conclusions

Over a decade ago, S.J. Gould thus described thesignificance of taxonomy to the sciences:

“The much maligned science of taxonomy,the ordering and classification of organisms,takes a culturally imposed backseat to themore interventionist and generalizing workstyle of experimentation and quantification.But taxonomy should be viewed as one of themost fundamental, and probably most noble,of scientific pursuits --- for what can be morebasic than the parsing of nature' s rich andconfusing complexity? Our categories, moreover,record our modes of thought, and taxonomytherefore teaches us as much about our mentalfunctioning as about nature' s variety.” [38]

This review of the evolution of the terminolo-gies used for noninvasive HPV-associated lesions ofthe cervix provides greater insight into how pathol-ogists, clinicians, and scientists view the pathogen-esis of cervical cancer than into the actual morpho-logic expression of the lesions. While the latter haveremained stable, each change in terminology, fromthe dysplasia/carcinoma in situ terminology of the1950s, through the spectrum of CIN introduced byRichart in the late 1960s, to the Bethesda System' slow-grade/high-grade lesion terminology, has beenbased on new scientific information that has alteredour fundamental views of the pathogenesis of cervi-cal cancer.

References

[1] Reagan JW, Ng ABP, Wentz WB. Concepts of genesis anddevelopment in early cervical neoplasia. Obstet GynecolSurvey 1969;24:860---74.

[2] Richart RM. Natural history of cervical intraepithelial neo-plasia. Clin Obstet Gynecol 1968;10:748---84.

[3] Fu YS, Braun L, Shah KV, Lawrence WD, Robboy SJ. Histo-logic, nuclear DNA, and human papillomavirus studies ofcervical condylomas. Cancer 1983;52:1705---11.

[4] Richart RM, Lerch V. Time lapse cinematographic observa-tions of normal human cervical epithelium, dysplasia andcarcinoma in situ. J Natl Cancer Institute 1966;37:317.

[5] Crocker J, Fox H, Langley FA. Consistency in the histologi-cal diagnosis of epithelial abnormalities of the cervix uteri.J Clin Pathol 1968;21:67---70.

[6] Richart RM. Cervical intraepithelial neoplasia: a review.In: Sommers SC, ed. Pathology Annual. East Norwalk, CT:Appleton-Century-Crofts; 1973:301---28.

[7] Bosch FX, de Sanjose S. Human papillomavirus and cervicalcancer --- burden and assessment of causality. J Natl CancerInst Monogr 2003(31):3---13. Chapter 1.

[8] Stoler MH, Rhodes CR, Whitbeck A, Wolinsky SM, ChowLT, Broker TR. Human papillomavirus type 16 and 18 gene

expression in cervical neoplasias. Hum Pathol 1992;23(2):117---28.

[9] Durst M, Glitz D, Schneider A, zur Hausen H. Human papil-lomavirus type 16 (HPV 16) gene expression and DNA repli-cation in cervical neoplasia: analysis by in situ hybridiza-tion. Virology 1992;189(1):132---40.

[10] Chow LT, Broker TR. Papillomavirus DNA replication. Inter-virology 1994;37(3---4):150---8.

[11] Munger K, Howley PM. Human papillomavirus im-mortalization and transformation functions. Virus Res2002;89(2):213---28.

[12] Horner SM, DeFilippis RA, Manuelidis L, DiMaio D. Re-pression of the human papillomavirus E6 gene initi-ates p53-dependent, telomerase-independent senescenceand apoptosis in HeLa cervical carcinoma cells. J Virol2004;78(8):4063---73.

[13] Onda T, Carter JJ, Koutsky LA, et al. Characterization ofIgA response among women with incident HPV 16 infec-tion. Virology 2003;312(1):213---21.

[14] Snijders PJ, Steenbergen RD, Heideman DA, Meijer CJ.HPV-mediated cervical carcinogenesis: concepts and clini-cal implications. J Pathol 2006;208(2):152---64.

[15] Wright TC, Jr., Schiffman M, Solomon D, et al. Interim guid-ance for the use of human papillomavirus DNA testing as anadjunct to cervical cytology for screening. Obstet Gynecol2004;103(2):304---9.

[16] Wright TC, Jr., Cox JT, Massad LS, Twiggs LB, Wilkin-son EJ. 2001 consensus guidelines for the managementof women with cervical cytological abnormalities. JAMA2002;287(16):2120---9.

[17] Wright TC, Jr., Lorincz A, Ferris DG, et al. Reflex humanpapillomavirus deoxyribonucleic acid testing in womenwith abnormal Papanicolaou smears. Am J Obstet Gynecol1998;178(5):962---6.

[18] Ostor AG. Natural history of cervical intraepithelial neopla-sia: a critical review. Int J Gynecol Pathol 1993;12:186---92.

[19] Winer RL, Kiviat NB, Hughes JP, et al. Development andduration of human papillomavirus lesions, after initial in-fection. J Infect Dis 2005;191(5):731---8.

[20] Mao C, Koutsky LA, Ault KA, et al. Efficacy of humanpapillomavirus-16 vaccine to prevent cervical intraepithe-lial neoplasia: a randomized controlled trial. Obstet Gy-necol 2006;107(1):18---27.

[21] Wright TC, Kurman RJ. A critical review of the morpho-logic classification systems of preinvasive lesions of thecervix: the scientific basis for shifting the paradigm. In:Lacey C, ed. Papillomavirus Reviews: Current research onpapillomaviruses. Leeds, UK: Leeds Medical Information;1996:215---25.

[22] Solomon D, Davey D, Kurman R, et al. The 2001 BethesdaSystem: terminology for reporting results of cervical cytol-ogy. JAMA 2002;287(16):2114---9.

[23] Wright TC Jr, Cox JT, Massad LS, Carlson J, Twiggs LB,Wilkinson EJ. 2001 consensus guidelines for the manage-ment of women with cervical intraepithelial neoplasia. AmJ Obstet Gynecol 2003;189(1):295---304.

[24] Melnikow J, Nuovo J, Willan AR, Chan BK, Howell LP. Nat-ural history of cervical squamous intraepithelial lesions: ameta-analysis. Obstet Gynecol 1998;92(4 Pt 2):727---35.

[25] Park TJ, Richart RM, Sun X-W, Wright TC. Associationbetween HPV type and clonal status of cervical squa-mous intraepithelial lesions (SIL). J Natl Cancer Inst1996;88(66):355---8.

[26] Wright TC, Ferenczy AF, Kurman RJ. Precancerous lesionsof the cervix. In: Kurman RJ, ed. Blaustein' s Pathology ofthe Female Genital Tract. 5th ed. New York, NY: SpringerVerlag; 2002:253---354.


[27] Laverty CR, Russell P, Hills E, Booth N. The significanceof noncondylomatous wart virus infection of the cervicaltransformation zone: a review with discussion of two illus-trative cases. Acta Cytol 1978;22:195---201.

[28] Fu YS, Huang I, Beaudenon S, et al. Correlative study ofhuman papillomavirus DNA, histopathology and morphom-etry in cervical condyloma and intraepithelial neoplasia.Int J Gynecol Pathol 1988;7:297---307.

[29] Chung TK, Cheung TH, Lo WK, et al. Loss of het-erozygosity at the short arm of chromosome 3 in mi-crodissected cervical intraepithelial neoplasia. Cancer Lett2000;154(2):189---94.

[30] Narayan G, Pulido HA, Koul S, et al. Genetic analysisidentifies putative tumor suppressor sites at 2q35-q36.1and 2q36.3-q37.1 involved in cervical cancer progression.Oncogene 2003;22(22):3489---99.

[31] Winkler B, Crum CP, Fujii T, et al. Koilocytotic lesions ofthe cervix: the relationship of mitotic abnormalities to thepresence of papillomavirus antigens and nuclear DNA con-tent. Cancer 1984;53:1081---7.

[32] Cho NH, Kim YT, Kim JW. Alteration of cell cycle in cervi-cal tumor associated with human papillomavirus: cyclin-dependent kinase inhibitors. Yonsei Med J 2002;43(6):722---8.

[33] Wang SS, Trunk M, Schiffman M, et al. Validation ofp16INK4a as a marker of oncogenic human papillomavirusinfection in cervical biopsies from a population-based co-hort in Costa Rica. Cancer Epidemiol Biomarkers Prev2004;13(8):1355---60.

[34] Kalof AN, Evans MF, Simmons-Arnold L, Beatty BG, CooperK. p16INK4A immunoexpression and HPV in situ hy-bridization signal patterns: potential markers of high-grade cervical intraepithelial neoplasia. Am J Surg Pathol2005;29(5):674---9.

[35] Agoff SN, Lin P, Morihara J, Mao C, Kiviat NB, Koutsky LA.p16(INK4a) expression correlates with degree of cervicalneoplasia: a comparison with Ki-67 expression and detec-tion of high-risk HPV types. Mod Pathol 2003;16(7):665---73.

[36] Klaes R, Friedrich T, Spitkovsky D, et al. Overexpressionof p16(INK4A) as a specific marker for dysplastic and neo-plastic epithelial cells of the cervix uteri. Int J Cancer2001;92(2):276---84.

[37] Negri G, Vittadello F, Romano F, et al. p16INK4a expressionand progression risk of low-grade intraepithelial neoplasiaof the cervix uteri. Virchows Arch 2004;445(6):616---20.

[38] Gould SJ. Science without taxonomy is blind. Sciences1994(March/April):38---9.

pathology of hpv infection at the cytologic and histologic ... of hpv infection at the cytologic...

Documents