Michael J. Campion
Cervical cancer remains worldwide the second most common cancer among women and is, uniquely among human cancers, entirely attributable to infection. It accounts for 15% of all female cancers and a tenth of all female cancer deaths (1). It is the most common cancer among women in many developing countries, constituting 20% to 30% of female cancers (2,3). Societal impact is accentuated by the young average age at death (4). This is due to inadequate screening leading to late detection and the absence of standard treatment options. In developed Western countries, cervical cancer accounts for only 4% to 6% of female cancers (5,6). This difference largely reflects the impact of mass screening using cervical cytologic methods (7,8,9).
The primary goal of cervical screening is to prevent cervical cancer. This is achieved by the detection, eradication, and follow-up of preinvasive cervical lesions (10,11,12). The ability to detect preinvasive cervical disease, coupled with comparatively easy access to the cervix for screening and assessment, have contributed greatly to the understanding of cervical carcinogenesis and to the definition of the precursor lesions to cervical cancer. Modern understanding that almost all cervical cancers are caused by persistent infection with approximately 15 types of human papillomavirus (HPV) has led to promising new approaches to cervical cancer prevention (13,14,15).
Classification of Preinvasive Cervical Disease (Precancer)
The proposal that invasive squamous carcinoma of the cervix arises through progression of a preinvasive lesion as opposed to a de novo event was initially postulated by Schauenstein in 1908 (16). The term “carcinoma in situ” was later introduced to describe cancerous changes confined to the epithelium (17).
The Dysplasia Terminology
Although referred to earlier by Papanicolaou and Traut (18), Reagan and Hamonic in 1956 (19) described cytologic differences between “carcinoma in situ” and a group of “less anaplastic” lesions, for which they introduced the term dysplasia (20). In 1975, the World Health Organization defined dysplasia as a “lesion in which part of the epithelium is replaced by cells showing varying degrees of atypia.” Dysplastic changes were graded as mild, moderate, and severe, but precise guidelines for these subdivisions were not defined, and grading always remained highly subjective (21,22,23,24).
A dual terminology for epithelial abnormalities of the cervix developed, leading to irrational treatment policies. If a diagnosis of “dysplasia” was made, this was considered a nonspecific change, and the patient was subjected to a cone biopsy. If the diagnosis of “carcinoma in situ” was made, this was considered a “preinvasive cancer,” and the patient underwent an obligatory hysterectomy (23,24,25).
Cervical Intraepithelial Neoplasia
Invasive squamous cell carcinoma of the cervix was demonstrated to be the end result of progressive intraepithelial dysplastic atypia occurring within the metaplastic epithelium of the cervical transformation zone (26). The classification of lesions from mild dysplasia to carcinoma in situ did not truly reflect either the morphologic or biologic continuum of preinvasive cervical disease. The diagnosis was highly subjective and was not reproducible. After pioneering research into the natural history of cervical cancer precursors, Richart(27) proposed the term “cervical intraepithelial neoplasia” (CIN) to describe the biologic spectrum of cervical preinvasive squamous disease. Three grades of CIN were described, specifically, CIN 1 (mild dysplasia), CIN 2 (moderate dysplasia), and CIN 3 (severe dysplasia/carcinoma in situ). This system was consistent with biologic evidence that strongly implied a single process of cervical squamous carcinogenesis (27,28,29,30,31,32,33).
Forty years' experience with the CIN terminology, coupled with recent advances in the understanding of the role of human papillomavirus (HPV) in the causation of cervical neoplasia (32,33,34,35,36,37,38), has led to critical reappraisal of this model. It has also led to further reclassification of the terminology for reporting cytologic abnormalities consistent with preinvasive disease (39,40,41,42,43,44,45,46). The CIN grading is very subjective. No reproducible cytologic or histologic distinction at the lower end of the CIN continuum exists between CIN 1 and HPV infection alone. Both interobserver and intraobserver consistency in diagnosis are poor (45,46,47,48). Separating CIN 2 from CIN 3 is again often not reproducible (49,50,51,52). In reality, the two critical questions in the assessment of the cervical epithelium are: (i) do the changes represent a cancer precursor; and (ii) is the lesion invasive cancer?
In histologic terms, CIN 3 is clearly established as a bona fide cancer precursor. CIN 3 is a reliable and reproducible morphological diagnosis, with undifferentiated cells with fixed genetic abnormalities replacing almost the full thickness of the cervical epithelium (48,49). CIN 3 is reliably distinguished from recently acquired HPV infection and is a genuine surrogate marker of subsequent cancer risk. There still exists uncertainty in relation to the progressive potential of less severe dysplastic lesions (53,54).
CIN 1 is increasingly viewed as an insensitive histologic marker of HPV infection. The diagnosis includes errors of processing and interpretation of colposcopically directed biopsies (48). Standardized for positivity for a given high-risk HPV type, a diagnosis of CIN 1 does not predict a meaningfully higher risk of CIN 3 than does a negative biopsy (53).
Histologically confirmed CIN 1 lesions confer a lower risk of developing cervical cancer than does a Pap smear report of low-grade squamous intraepithelial lesion (53). Recent data indicate considerable heterogeneity in the microscopic diagnosis, biology, and clinical behavior of CIN 2 lesions (54). CIN 2 can be produced by noncarcinogenic HPV types and is equivocal in cancer potential (55). Some represent acute HPV infection with a more severe microscopic appearance and are destined to regress. Others are incipient precancer and will persist and progress with an attendant high risk of future invasion if left untreated. The clinical dilemma remains the inability to reliably predict those lesions less severe than CIN 3 that are at greatest risk of progression to cancer and those that are likely to regress. New molecular markers hold promise in this regard (56,57,58), but prospective validation to determine risk of invasion is unethical.
Cervical Precancer - Modern Concepts
The stepwise progression of increasingly severe cervical intraepithelial neoplasia to invasive cancer, implicit in the CIN continuum, remains an important histopathological concept to assist clinical management. Two decades of epidemiologic and preventive research focused on the HPV cervical cancer model have revolutionized our understanding of cervical precancer (15).
HPV infection is a broad transition state between normal and precancer. A defined precancerous lesion remains the target of screening and preventive treatment programs and represents a genuine surrogate for cancer risk (59). With increasing detection of smaller and less serious lesions in cervical screening programs and the inclusion of such lesions in the precancer continuum, the likelihood of diagnosing a precancerous lesion as a surrogate predictor of invasive cancer risk has declined. This has altered the effect and assessment of cervical cancer prevention programs.
CIN 3, particularly full thickness carcinoma in situ, shares the same HPV-type spectrum and cofactor profile as invasive cancer. CIN 3 lesions demonstrate the same aneuploid DNA content and genetic instability as seen in invasive cancer. CIN 3 is the most certain surrogate marker of cancer risk. Some CIN 3 lesions are small; some will regress, particularly after biopsy and the eventual risk of invasion for such lesions is less certain. At this time there is no reliable predictor of CIN 3 lesions likely to progress to cancer and as such all are managed as definite precancer.
CIN 2 demonstrates greater heterogeneity in biology and definition (54). It can be caused by low-risk HPV types rarely found in cancer and has a greater regression potential. A diagnosis of CIN 2 is not a reliable surrogate for cancer risk. Although of equivocal malignant potential, in the absence of reliable predictors of risk of progression, CIN 2 lesions tend to be managed as precancer to provide a further safety margin against development of cancer.
A histological diagnosis of low-grade cervical intraepithelial lesions (HPV infection/CIN 1) is increasingly viewed as not representing precancer. Persistence of oncogenic HPV types is strongly linked to precancer (60,61,62). Incident HPV infection might not be associated with any microscopic abnormality while most low-grade abnormalities will regress (63,64), particularly among young women (65). Only a fraction of precancers arise from HPV infection in the absence of mild or equivocal microscopic abnormalities.
High-grade lesions are commonly found within a broader field of low-grade disease, suggesting that CIN 3 may develop in high-risk HPV-infected epithelium independent of and within a CIN 1 lesion, rather than as a classical stepwise progression. The reported progressive potential of low-grade lesions is small but definite, varying from 12% to 33%(25,26,66,67,68).
Understanding the Cervical Transformation Zone
The cervix and vagina are derived from the müllerian ducts and are initially lined by a single layer of müllerian-derived columnar epithelium. At 18 to 20 weeks of gestation, this columnar epithelium lining the vaginal tube is colonized by the upward growth of stratified squamous epithelium derived from cloacal endoderm.
Original Squamocolumnar Junction
The junction in fetal life between the stratified squamous epithelium of the vagina and ectocervix, and the columnar epithelium of the endocervical canal is called the original squamocolumnar junction (69). Original squamous epithelium extends from Hart's line or the mucocutaneous, vulvovaginal junction to the original squamocolumnar junction. The position of the original squamocolumnar junction is variable, lying on the ectocervix in 66%, within the endocervical canal in 30%, and on the vaginal fornices in 4% of female infants (70). The position of the original squamocolumnar junction determines the extent of cervical squamous metaplasia (70,71). Squamous metaplasia is a pivotal process in cervical carcinogenesis. Embryogenesis, in determining the distribution of native squamous and columnar epithelia, is an important early influence in determining future risk of neoplastic transformation (Fig. 8.1).
New Squamocolumnar Junction
The volume of the cervix alters throughout a woman's life in response to hormonal stimulation (69,70). Increased estrogen secretion, particularly with puberty and with the first pregnancy, causes an increase in cervical volume and an eversion of endocervical columnar epithelium to an ectocervical location (70). This eversion of columnar epithelium onto the ectocervix is called an ectropion. An ectropion is often mistakenly referred to as an erosion.
The estrogen surge of puberty results in the establishment of lactobacilli as part of the normal flora of the vagina. These microorganisms produce lactic acid, reducing the vaginal pH to 4 or less (70). Everted endocervical columnar epithelium is exposed in the postpubertal years to the acidity of the vaginal environment. Damage to the everted columnar epithelium caused by vaginal acidity results in proliferation of a stromal reserve cell underlying the columnar epithelium. This results in replacement of the columnar epithelium with an immature, undifferentiated, stratified, squamous, metaplastic epithelium (70,71). Immature squamous metaplasia then undergoes a maturation process, producing a mature, stratified squamous metaplastic epithelium distinguishable only with difficulty from the original squamous epithelium.
Figure 8.1 Location of squamocolumnar junction at various times in a woman's life. CE, columnar epithelium; SE, squamous epithelium; SCJ, squamocolumnar junction.
The original linear junction between squamous and columnar epithelium is replaced by a zone of squamous metaplasia at varying degrees of maturation. At the upper or cephalad margin of this zone is a sharp demarcation between epithelium, which appears morphologically squamous, and villous epithelium, which appears colposcopically columnar. This colposcopic junction is called the new squamocolumnar junction.
The Transformation Zone
The transformation zone is defined as that area lying between the original squamocolumnar junction and the colposcopic new squamocolumnar junction (28,29). The initial clinical assessment for most women is in the postpubertal years. Mature squamous metaplastic epithelium has often replaced the distal or caudad limit of the columnar epithelium. As the transformation zone matures, the original squamocolumnar junction becomes impossible to delineate. Only the presence of Nabothian follicles and gland openings hint at the original columnar origin of mature squamous metaplasia.
Cervical neoplasia almost invariably originates within the transformation zone. Understanding squamous metaplasia is the key to understanding the concepts of the cervical transformation zone and its role in cervical carcinogenesis (Fig. 8.2). For reasons that are poorly understood, persistent HPV infection causes cancers mainly at the transformation zone between different kinds of epithelia (e.g., cervix, anus, and oropharynx) (2). Carcinogenic HPV infection is equally common in the cervical and vaginal epithelia (72). Cervical cancer is the second most common cancer among women while vaginal cancer is rare. This reflects the pivotal importance of the metaplastic epithelium of the transformation zone in cervical carcinogenesis (73).
Figure 8.2 The anatomy of the transformation zone.
Squamous metaplasia is a permanent process but is not continuous. It occurs in “spurts,” with greatest activity during puberty and the first pregnancy. During the maturation phase, the columnar villi fuse, losing the distinctive appearance of columnar epithelium and producing a myriad of cytologic, colposcopic, and histologic appearances. The process fluctuates in response to hormonal influences but ultimately produces a mature, glycogenated squamous epithelium. The presence of a subepithelial inflammatory infiltrate in biopsy specimens of immature squamous metaplasia may lead to a histologic misdiagnosis of chronic cervicitis. The presence of such inflammatory white cells is a normal part of the metaplastic process and is not a response to an infectious organism. A histologic diagnosis of “chronic cervicitis” is often misleading and should not be accepted as a satisfactory explanation for an abnormal Papanicolaou (Pap) smear (Fig. 8.3).
If the new squamocolumnar junction is seen in its entirety in the absence of premalignant disease, the incidence of squamous disease above or cephalad to the new squamocolumnar junction is virtually nil, and the colposcopic examination of the cervix is described as satisfactory. If the new squamocolumnar junction is not seen in its entirety, the colposcopic examination is described as unsatisfactory. The transformation zone further defines the distal limit of high-grade glandular intraepithelial neoplasia, the precursor lesion to invasive adenocarcinoma of the cervix.
Upper Limit of Squamous Metaplasia
The new squamocolumnar junction is an unstable boundary. Serial colposcopic assessments of the cervix frequently show the new squamocolumnar junction to have moved cephalad. Careful colposcopic assessment of columnar villi immediately above the new squamocolumnar junction reveals opaque, opalescent tips and early villous fusion (Fig. 8.4). Histologic study of colposcopically directed biopsy specimens reveals reserve cell hyperplasia and early immature squamous metaplasia occurring in epithelium, which appears colposcopically columnar. This early immature squamous metaplasia can extend as far as 10 mm above the new squamocolumnar junction.
The immature metaplastic epithelium cephalad to the new squamocolumnar junction is not included in the modern definition of the transformation zone but represents the epithelium at greatest risk for future neoplastic transformation. During dynamic phases of metaplasia, occurring particularly with puberty and the first pregnancy, the immature metaplastic cells are actively phagocytic (70). The most critical phase is the initiation of squamous metaplasia at puberty and in early adolescence.
Figure 8.3 Histology of immature squamous metaplasia (chronic cervicitis).
Figure 8.4 Colposcopic appearance of immature squamous metaplasia.
Age of coitarche is an important epidemiologic variable in determining risk of cervical neoplasia (74,75,76,77,78). The lifetime risk for development of cervical cancer is increased 26-fold if age at first intercourse is within 1 year of menarche, as opposed to 23 years of age or older (76). Potential carcinogens in the vaginal environment at times of active metaplasia can deviate early metaplastic transformation along a neoplastic pathway. Mature metaplastic epithelium exposed to the same mutagen is at less risk of neoplastic transformation.
Human Papillomaviruses and Cervical Neoplasia
Extensive molecular biologic and epidemiologic research confirms certain HPV types to be carcinogenic in humans (13,14,15,79,80,81,82,83,84,85). Infections with HPV cause approximately 5% of the global burden of human cancers and at least 500,000 deaths annually (1). Infection with specific HPV types is necessary for the development of the vast majority of cervical cancers (>99.7%) and the immediate precursor lesion (CIN 3) (85). The magnitude of the association between HPV and cervical cancer is higher than the association between smoking and lung cancer. The four major steps in the development of cervical cancer are (i) infection of the metaplastic epithelium of the transformation zone with one or more carcinogenic HPV types; (ii) viral persistence rather than clearance reflecting the host immune response; (iii) clonal progression of persistently infected epithelium to cervical precancer (CIN 3); and (iv) invasion. The precise molecular events which lead HPV-infected cervical cells to invade are unknown, but this causal model is supported by a large volume of epidemiological and laboratory data.
Taxonomy and Biology
Papillomaviruses are small, nonenveloped, double-stranded DNA viruses encased in a 72-sided icosahedral protein capsid. The HPV genome consists of circular, double-stranded DNA of approximately 7,900 nucleotide base pairs. Papillomaviruses are a divergent group of evolutionarily related viruses with similar biologic characteristics but enormous differences in species specificity, site of predilection, and oncogenic potential (86,87). More than 100 types of HPV have been fully sequenced.
HPV types are divided into phylogenetic trees based on DNA sequence and protein homologies which assist in understanding HPV classification and behaviour (88,89). Four major groups are recognized: two infecting genital skin, generally acquired through sexual intercourse, and two infecting nongenital skin acquired from shed virus (87). At each site, one group causes warts which rarely become malignant and the other includes oncogenic types associated with induction of cancer. Within the four broad groups of HPVs, genetic sequencing defines clades of viruses which produce similar pathology (87). Viruses of clades A7 and A9 are most commonly associated with anogenital cancers (14). In addition to those that infect only humans, there is a large number of other species-specific papillomaviruses affecting other mammalian species, including cattle, horses, sheep, dogs, rabbits, monkeys, pigs, and deer.
The HPV genome is usually maintained as a stable viral episome, independent of the host cell genome, in the nucleus of infected cells. It codes for only eight genes (90). In some high-grade CIN lesions, and more frequently in cervical cancer, HPV genomes are covalently bonded or integrated into the host chromosomes (91,92,93,94). This integration event occurs at random within the host cell genome but is highly specific in relation to the viral genome, involving the E1 and E2 genes, with important consequences for regulation of viral gene expression (Fig. 8.5) (95,96). This integration of the HPV genome into the host genome is associated with invasive cancer and might serve as an important biomarker to distinguish HPV infection from precancer (97). Viral integration may not be necessary for invasion as not all women with invasive cancer have measurable integration (94,98). Continued transcriptional activity of HPV oncogenes is required to maintain the cancer (99).
The late genes, L1 and L2, the sequences of which are highly conserved among all papillomaviruses, encode the common capsid proteins. These viral proteins reflect late viral gene expression and are exclusively present in well-differentiated keratinocytes (100). Both proteins play an important role in mediating efficient virus infectivity.
Figure 8.5 The HPV genome and its expression within the epithelium. The HPV genome consists of approximately 8,000 base pairs of single-stranded, circular DNA. HPV genes are designated as E or L according to their expression in early or late differentiation stage of the epithelium: E1, E2, E5, E6, and E7 are expressed early in the differentiation, E4 is expressed throughout, and L1 and L2 are expressed during the final stages of differentiation. The viral genome is maintained at the basal layer of the epithelium, where HPV infection is established. Early proteins are expressed at low levels for genome maintenance (raising the possibility of a latent state) and cell proliferation. As the basal epithelial cells differentiate, the viral life cycle enters successive stages of genome amplification, virus assembly and virus release, with a concomitant shift in expression patterns from early genes to late genes, including L1 and L2, which assemble into viral capsid. (Reproduced with permission from Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder S. Human papillomavirus and cervical cancer. The Lancet 2007; 370;890-907).
The proteins encoded by the E6 and E7 genes of high-risk HPV types, particularly HPV 16 (clade A9) and 18 (clade A7), are directly involved in cellular transformation in the presence of an active oncogene (84,101). E6 and E7 are the primary HPV oncoproteins with numerous cellular targets (90,101,102,103). Both E6 and E7 proteins can immortalize primary keratinocytes from cervical epithelium and can influence transcription from viral and cellular promoters (104). The activity of these viral oncoproteins results in genomic instability, leading to the malignant phenotype. E6 proteins of high-risk HPV types bind the tumor suppressor protein p53 (105,106). This induces ubiquitination and degradation of p53, removing the p53-dependent control of the host cell cycle (107,108,109,110,111). The role of E6 as an antiapoptotic protein is of key significance in the development of cervical cancer as it compromises the effectiveness of the cellular DNA damage response and allows the accumulation of secondary mutations to go unchecked.
E6 also increases telomerase activity in keratinocytes through increased transcription of the telomerase catalytic subunit gene (hTERT) through induction of c-myc (112,113). Telomerase activity is usually absent in somatic cells, leading to shortening of telomeres with successive cell divisions and to eventual cell senescence. E6 mediation of telomerase activity may predispose to long-term infection and the development of cancer. Recently E6 and E7 viral oncogenes have been shown to antagonize BRCA-mediated inhibition of the hTERT promoter (114).
The E7 gene product is a nuclear phosphoprotein that associates with the product of the retinoblastoma gene (pRb), which is a tumor suppressor gene important in the negative control of cell growth (115,116,117). E7 is the primary transforming protein. Degradation of p53 by E6 and the functional inactivation of pRb by E7 represent the main mechanisms whereby expression of HPV E6 and E7 oncoproteins subverts the function of the negative regulators of the cell cycle (118,119,120). Deregulated expression of the viral oncogenes is a predisposing factor to the development of HPV-associated cancers.
The products of the E2 gene are involved in transcriptional regulation of the HPV genome. The process of HPV integration into the cellular genome, which occurs in some high-grade CIN lesions and most invasive cervical cancers, disrupts the E2 gene (84). This results in increased levels of E6 and E7 expression, correlating with increased immortalization activity (84,122,123,124,125).
Both E6 and E7 proteins are expressed at low levels in the process of HPV infection. At some undefined point in neoplastic transformation and the development of precancer, E6 and E7 expression is deregulated by genetic or epigenetic influence. This leads to overexpression in CIN 3 lesions. Aberrant expression of high-risk viral oncogenes can predispose to the development of cervical cancer but their expression alone is not sufficient (90). HPV-mediated oncogenesis requires accumulation of additional genetic mutations over time. The average age of women with invasive cervical cancer is 50 years, whereas the mean age for women with CIN 3 is 28 years. This suggests a long precancerous state in most cases of invasive cancer that allows the accumulation of secondary genetic mutations. These mutations can occur randomly but may also reflect the influence of cofactors such as smoking carcinogens and hormonal influences. This is discussed later in this chapter.
Human Papillomavirus Type-Specific Disease Pattern
Differing genomic nucleotide sequences of specific HPV types are responsible for the specific anatomic tropism of each HPV type. There are more than 30 genital HPV types. These transfect the mucous membranes of the genital tract most efficiently, but may also be present in the keratinized epithelium of the vulva, perineum, penis, and anorectal areas. Genital HPV types are also occasionally associated with oropharyngeal, conjunctival, and subungual lesions. The genital HPV types are divided into groups based on the frequency of association with malignant tumors and presumed oncogenic potential. The International Agency for Research on Cancer (IARC) reassessed the carcinogenicity of HPV in 2005 and revised these groupings so that the list of high-risk HPV's includes thirteen types, specifically HPV types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 66 (86).
Low-risk HPV types, particularly HPVs 6 and 11 (clade A10), are associated with condylomata acuminata of the genital tract in both sexes. HPV 6 and 11 are also detected alone in low-grade cervical lesions (exophytic condylomata acuminata, subclinical HPV infection, CIN 1 and some CIN 2 lesions). Using more reliable HPV detection techniques, not a single cervical cancer has been shown to be associated with low-risk HPV types, and HPVs 6 and 11 in particular. These viruses do not appear to induce malignant transformation; they are unable to integrate into the human genome. The E6 and E7 proteins of “lowrisk” HPV types only weakly bind p53 and pRb and thus do not immortalize keratinocytes in vitro.
Human papillomavirus 16 is the HPV type universally detected with greatest frequency in high-grade intraepithelial neoplasia and invasive cancers. HPV 16 is associated with 50% of cervical squamous cancers and more than 30% of adenocarcinomas (14,126,127). It is present in more than 80% of high-grade cervical, vaginal, vulvar, perianal, and penile preinvasive lesions. It is detected in more than 25% of low-grade cervical lesions, 40% of subclinical vulvar HPV infections, and 10% of genital condylomata acuminata, particularly the recalcitrant lesions (14,127,128,129).
HPV 18 is the second most common (25%) HPV type in invasive cervical cancer, but is uncommon (5%) in preinvasive cervical lesions (130,131). The association of HPV 18 with aggressive adenocarcinomas, particularly in younger women, and the underrepresentation of this viral type in preinvasive lesions have raised concerns that HPV 18 may be associated with “rapid-transit” cancers that escape reliable cytologic detection (132,133). Although this remains a controversial issue (134), epidemiologic and molecular data support the hypothesis (135,136). HPV 18 DNA is detected 2.6 times more frequently in invasive cervical cancers occurring within 1 year of a negative smear (133). The average age of patients with HPV 18-containing cancers is 8 to 12 years younger, and recurrence rates are higher (45% vs. 16%) than for patients with HPV 16-containing cancers (137). The lack of HPV 18-induced squamous precancerous lesions contributes to the failed detection of glandular and endocervical lesions in well-screened populations with the attendant increased proportion of adenocarcinoma.
Human Papillomavirus and Cervical Cancer: A Causal or Casual Association
Although the true prevalence of cervical HPV infection is unknown, it is the most common sexually transmitted infection, with most sexually active women younger than 35 years of age exposed (138). The 2-year cumulative incidence for first time genital HPV infection for young women is 32% (139). Incidences are similar for virgins and nonvirgins from the time of acquisition of a new partner. For monogamous women, the risk of acquiring HPV after sexual debut is 46% at 3 years (140). Smoking, oral contraceptive use, and report of a new male sexual partner are predictive of incident infection. Male condom usage is not protective (141). Infection in virgins is rare, but any nonpenetrative sexual contact is associated with an increased risk (139). Basal cells of the cervical epithelium are inoculated with the virus at sites of microtrauma. HPV prevalence and incidence peak in women under 20 years of age and decline in women over 30, secondary to HPV clearance, with most women in the world exposed to one or more genital HPV types during their sexual life (138,142).
Most HPV infections, whether associated with a cytological abnormality or not, are transient, usually being cleared or suppressed by cell-mediated immunity, within several months to 2 years (143). The median time to clearance of HPV infection in an immunocompetent woman is 6 to 18 months (average of 12 months) with 90% of women clearing a specific HPV-type after 2 years of observation (144). Humoral and cellular immune responses to genital HPV types are difficult to detect, possibly because the virus is nonlytic to infected cells and does not spread systemically. Viral clearance is not often associated with reappearance of the same HPV type. Occasionally the same HPV type will reappear (145). It is unclear whether this represents reinfection or resurgence from a latent state in the basal cells of the epithelium with very low viral copy numbers and without late viral expression.
The high rate of HPV infection among immunosuppressed HIV-infected women supports the concept of latency (146,147,148). However, HIV-infected women with normal CD4 counts demonstrate increased high-risk HPV persistence compared with HIV-uninfected women, suggesting HPV clearance is generally impaired in this group (149). Secondary peaks of HPV infection in older and postmenopausal women suggest the possibility of reactivation of a latent reservoir due to senescence of cell-mediated immunity, although this could also be explained by sexual behavior (of women or partners).
The longer a specific HPV type persists in the epithelium, the lower the probability of clearance within a defined period, and the greater the risk of precancer (144). Clinical strategies which require repeated specific HPV type detection over 12 months or longer before referral for diagnostic workup take advantage of this natural history to separate transient infections from persistent lesions which carry a greater risk to the patient (61,150,151).
Only persistent high-risk HPV infection of the cervical epithelium appears to trigger neoplastic progression (150,151,152,153,154) (Table 8.5) (Fig. 8.6). There is no accepted definition of clinically significant persistence. Follow-up strategies suggest persistence beyond 1 year and certainly beyond 2 years defines a greater risk to patients (151). The small proportion of highrisk HPV infections (about 10%) which persist for several years is strongly linked to a high risk of developing CIN 3 (89). The risk factors for HPV persistence and progression to CIN 3 are not fully understood. HPV type is the strongest factor affecting risk of viral persistence (89). HPV 16 is highly carcinogenic with an absolute risk of CIN 3 approaching 40% at 3 to 5 years (38). The risks of progression to CIN 3 with other HPV types are several-fold lower. More than 20% to 30% of women with cervical HPV associated disease, regardless of degree of CIN, have more than one HPV type (155). Women infected with multiple HPV types have a further increased risk but it is not clear if the risk is equal to or greater than the cumulative risk associated with each of the individual HPV types. It is also unknown whether infection with multiple HPV types interferes with persistence of a given HPV type or with risk of progression to CIN 3 (145,155,156).
Viral burden or the amount of HPV DNA in the cervical epithelium appears to have an independent effect on CIN incidence as a surrogate for HPV persistence, but does not independently predict future risk of CIN 3 or cancer (157). Low viral load is associated with normal microscopic findings and a low risk of neoplastic progression. High viral loads do not generally imply an increased prospective risk of cancer, except for HPV 16 (158,159). Recently acquired low-grade cervical lesions contain some of the highest viral loads, analogous to condylomata acuminata, and frequently regress (160). Cervical cancers have low amounts of intact virus, probably associated with genomic integration. Infection with multiple HPV types is a further complicating factor as this will affect viral load but cervical cancer typically is a monoclonal event related to a specific single HPV type. The role of viral load in the natural history of HPV infection is unclear and viral load measurement is not clinically useful (15).
Figure 8.6 Major steps in the development of cervical cancer. Incident HPV infection is best measured by molecular tests. Most HPV infections show no concurrent cytological abnormality. Approximately 30% of infections produce concurrent cytopathology, usually non-classical (equivocal) changes. Most HPV infections clear within 2 years. Ten percent persist for 2 years and are highly linked to development of precancer. (Reproduced with permission from Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder S. Human papillomavirus and cervical cancer. The Lancet 2007;370;890-907).
The modal time from HPV infection to CIN 3 is 7 to 15 years, with infection occurring in the late teens or early twenties and CIN 3 diagnosis peaking at 25 to 30 years (127,161). The average age of diagnosis of CIN 3 depends on average societal age of first intercourse, a proxy for initial HPV exposure, and age of onset and intensity of screening. The time from HPV infection to the development of CIN 3 can be short, often within 5 years (162). CIN 3 has been diagnosed within 2 years of coitarche and CIN 2-3 has been documented to rapidly develop within several months of an incident HPV infection (78,163). The biological significance and risk of invasion associated with these early CIN 3 lesions is uncertain. It is plausible that many might regress but prospective follow-up of these lesions is unethical.
CIN 3 lesions are a homologous population of aneuploid lesions, mostly associated with oncogenic HPVs, and are genuine cancer precursors. The transit time to invasive cancer is variable, taking as little as 12 to 18 months or as long as several decades. Most cervical abnormalities do not transform to invasive cancer. The median age of women with invasive cancer is two to three decades later than for CIN 3 and is older still in poorly screened communities. Invasive cervical cancers detected through screening are often earlier stage and are detected among women at least 10 years older than those with CIN 3. This suggests generally a long average transit time for CIN 3 to invasive cancer. Rapid onset invasive cancers, often occurring in young women are rare, but are sometimes fatal and have a disproportionate impact on screening and preventive strategies (15).
Cervical neoplasia can be viewed as the result of a complex interplay between a “seed,” that is, high-risk HPV types, and a “soil,” that is, the immature, metaplastic epithelium of the cervical transformation zone. Exposure to specific high-risk HPV types, in the presence of cofactor activity, may deviate the metaplastic process along a neoplastic pathway.Disease expression begins at the new squamocolumnar junction. The initial abnormality produced is usually a low-grade cervical lesion. Such lesions represent a heterologous mixture of genuine cancer precursors and benign HPV infections (164). The most critical step in cervical carcinogenesis is not acquisition of an HPV infection but progression to CIN 3. HPV infection alone is necessary but not sufficient to induce carcinoma in an immunocompetent host. HPV infection with oncogenic viral types is much more common than cervical neoplasia, indicating the necessity of cofactors in the process of cervical carcinogenesis (165).
Cofactor Interaction with Human Papillomavirus
Plausible cofactors in cervical and lower genital tract carcinogenesis include the use of tobacco products, infection by other microbial agents, specific vitamin deficiencies, hormonal influences, and immunosuppression.
Cigarette smoking has been demonstrated to be a risk factor for cervical and vulvar carcinoma (167,168,169,170,171,172). An increased risk of developing a high-grade squamous intraepithelial lesion (HSIL) has been demonstrated among high-risk HPV positive women who smoke or who are passive smokers. The detection of high levels of genotoxic breakdown products of cigarette smoke—including nicotine, cotinine, hydrocarbons, and tars—in cervical secretions of smokers and the demonstration of mutagenic activity of these products in cervical cells, similar to that observed in lung cells, point to an important role for these compounds in cervical carcinogenesis. The association between smoking and HPV persistence is less consistent.
Cigarette smoking influences epithelial immunity by decreasing the numbers of antigen-presenting Langerhans cells in the genital epithelium (173,174). Cervical HPV infection and CIN are associated with diminished numbers of intraepithelial Langerhans cells. Such local immunologic depletion could favor viral persistence, contributing to malignant transformation. Cigarette smoke concentrates have been demonstrated in vitro to transform HPV-16-immortalized endocervical cells (169), although no increased risk of adenocarcinoma has been identified in association with use of tobacco products. The increased risk associated with passive smoking is as strong as that observed in association with personal cigarette smoking (167,168). The high levels of nitrosamines inhaled in passive smoking may be relevant.
Smoking cessation promotes resolution of HPV-associated abnormalities in Pap smears with probable enhancement of cellular immunity.
Infection by Other Microbial Agents
Genital HPV infection and cervical neoplasia are more common among individuals who have had multiple sexual partners or whose partner has had multiple sexual partners (175,176). An increased incidence of other sexually transmitted diseases has been reported in association with genital HPV infection and cervical neoplasia (176). Disruption of epithelial integrity and reparative metaplasia associated with acute cervicitis that is due to Chlamydia trachomatis, Neisseria gonorrhoeae, herpes simplex virus (HSV), orTrichomonas vaginalis may increase susceptibility to genital HPV infection. No clear picture has emerged from epidemiologic studies addressing these associations (177,178,179,180,181). The role of chronic inflammation due to coinfection with Chlamydia trachomatis is also unclear but chlamydial infection appears to be associated with increased persistence of high-risk HPV infection (182). Chlamydial infection in HPV-positive women is also associated with development of high-grade CIN and invasive cancer suggesting a possible cofactor role (183,184).
Sex Hormonal Influences
Condylomata acuminata increase rapidly in size and number in pregnancy. This could suggest that maternal estrogen status is permissive for HPV replication, although it may reflect the immunosuppressive effect of pregnancy. Increased detection of HPV DNA in cervical cytologic samples in pregnancy, including detection of oncogenic HPV types in up to 27% of pregnant women, suggests hormonally induced active viral replication (185,186).
CIN and cervical cancer are more frequently found in women with increased parity (187,188,189) and in women on oral contraceptives independent of sexual activity (189,190,191,192,193,194,195). Epidemiologic studies have shown an increased risk of CIN in long-term oral contraceptive pill (OCP) users, rising to twofold for 5 or more years of OCP use. OCP-induced folate deficiency with reduced metabolism of mutagens is a proposed mechanism for increased risk (196). Prospective follow-up of high-risk HPV-positive women does not demonstrate an increased risk of CIN 3 among OCP users (189). There has been no demonstrable clinical value to ceasing oral contraceptives in the management of HPV-associated disease.
Protective benefits of barrier contraception remain unclear (197). An earlier association between condom use and decreased persistence of high-risk and low-risk HPV types and reduced risk of progression to CIN 3 among high-risk HPV-positive women has been confirmed in recent studies (198,199). Increased HPV clearance and low-grade CIN regression with condom use has also been reported (200).
Exogenous and Endogenous Immunosuppression
Iatrogenic induction of immunosuppression in renal transplant recipients increases the rate of CIN to 16 times that of the general community (201). The risk of CIN and cervical cancer is increased in human immunodeficiency virus (HIV)-infected women, and failure rates of treatment for preinvasive lesions are increased (202,203,204,205,206). HPV prevalence and persistence are increased in HIV-positive women. HIV infection has a lesser influence on the probability of developing invasive cancer but the risk is increased (206). Systemic immune suppression from diseases such as Hodgkin's disease, leukemia, and collagen vascular diseases are associated with an increased incidence and recalcitrancy of HPV-associated disease.
Dietary deficiencies of vitamin A or beta-carotene may increase the risk of CIN and cervical cancer (208). Higher dietary consumption of vitamins A, C, and E as well as β-carotene and increased circulating levels of certain micronutrients may be protective against cervical neoplasia (209,210,211). Higher levels of dietary vitamin A and carotenoids, reflected in higher dietary vegetable intake and increased circulating levels of cis-lycopene, are associated with a greater than 50% reduction in persistence of high-risk HPV DNA (211). There is a possible protective association between higher folate levels and risk of high-grade CIN. Low socioeconomic status appears to remain a risk factor among HPV infected women even when standardized for available medical care.
Human Papillomavirus Vaccines
The HPV vaccine is a major scientific and public health advance in the prevention of cervical cancer (212,213). Most invasive cervical cancers and CIN lesions are attributable to high-risk HPV infection (15). It is estimated that the annual financial burden of HPV-associated disease in the United States is over $5 billion and 90% of this is spent in the follow-up of abnormal Pap smears and treatment of precancerous lesions (215,216). A prophylactic or therapeutic vaccine against HPV has the potential to have a substantial impact on HPV infection and rates of CIN and cervical cancer (217). HPVs cannot be grown in the laboratory so standard approaches to vaccine development, such as inactivation of live virus or development of attenuated virus, are not possible. Prophylactic vaccines have been produced using recombinant DNA technology. HPV prophylactic vaccines, designed to prevent HPV infection, are based on viruslike particle (VLP) technology developed through the pioneering research of Zhou and Fraser in Brisbane, Australia (218,219).
Fraser describes this process as the major capsid protein of the virus being expressed in eukaryotic cells and self assembling to produce empty viral capsids (212). VLPs are three-dimensional structures similar to papillomavirus particles and produced by expression of L1 and L2 HPV viral capsid proteins. These DNA-free VLPs are empty capsids and contain no oncogenic or infectious materials. Viruslike particles resemble the virus immunologically and induce HPV type-specific antibody on administration (220,221).
The immunogenicity of HPV involves presentation of the major capsid protein L1 to the immune system. L1 VLP vaccines induce strong cell-mediated as well as humoral immune responses (222,223,224). Such vaccines have demonstrated prophylactic efficacy in animal models (225) and then in human clinical trials (220,221, reviewed 213). Two vaccine products have been developed, a quadrivalent vaccine incorporating HPVs 16, 18, 6, and 11 (Gardasil; approved by the U.S. Food and Drug Administration [FDA] in June 2006) and a bivalent vaccine incorporating HPVs 16 and 18 (Cervarix; currently under review by FDA).
Clinical trials demonstrate that HPV vaccines are effective and safe (220,221,226,227,228,229). For ethical and scientific reasons, surrogate end points were prevention of HPV acquisition, long-term viral persistence, development of genital neoplasia and genital warts as opposed to development of cervical cancer. Studies were largely undertaken among sexually active women 16 to 25 years of age although recent studies have been extended to include women up to 45 years (e.g., FUTURE III). Age groups where new HPV infection is uncommon include male and female children ages 9 to 15 years where antibody levels produced are on average substantially higher than in 16- to 24-year-old women (230). Women ages 25 to 45 years also have a lower level of new HPV infection and the antibody response is lower (231,232), but the quadrivalent vaccine does prevent 90% of new high-risk HPV infections in this 24- to 45-year-old female age group (233).
The quadrivalent vaccine demonstrates, following exact clinical protocols, 100% efficacy in preventing condylomata and vulvovaginal precancerous lesions and 98% efficacy in preventing high-grade cervical lesions among young sexually active women (212,213). In trial populations which include less strenuously defined criteria, such as including women with known infection or disease associated with vaccine types prior to vaccination, the quadrivalent vaccine is 44% effective in preventing CIN 2-3 and 73% effective in preventing condylomata and other HPV associated vulvovaginal lesions (213). In strict protocol studies, the bivalent vaccine demonstrates 100% efficacy in preventing persistent high risk HPV cervical infection and development of vaccine type CIN lesions (220,234,235).
For women recruited under less strict protocol conditions, who received at least one vaccine dose and who tested negative for high-risk HPV types or had no cytologic abnormality at a screening visit within 90 days of enrollment, the quadrivalent vaccine efficacy was 94% in preventing persistent high-risk HPV infections at 12 months and 96% in preventing abnormal Pap smears (236). No therapeutic effects have been demonstrated in women with vaccinetype existing HPV infection or HPV-related disease, with lesions regressing or progressing at similar rates in vaccine and placebo recipients (237).
Follow-up studies of women vaccinated with both vaccines have demonstrated sustained efficacy for at least 5 years. Three administrations of vaccine induce peak antibody levels 20 times higher than those seen with natural HPV infection (212). Antibody levels fall significantly in the first 2 years after immunisation but remain above those stimulated by natural infection for at least 5 years (238). These levels are associated with continued protection against infection. A modeling study suggests antibody levels will remain above those associated with natural infection for 12 years or more (239). Immunological memory is retained and a single booster dose given 60 months post-completion of the HPV vaccination protocol has been shown to produce a strong anamnestic increase in antibody titres, not seen in nonimmune subjects, with continued sustained efficacy typical of many vaccines (238).
Immunization with HPV 16 and 18 VLPs provides some protection against other highrisk HPV types and development of associated disease. The bivalent HPV 16/18 vaccine has been demonstrated to provide significant protection against HPV 45 infection and partial protection against HPV 33 infection (220). The quadrivalent (HPV 6/11/16/18) vaccine showed 27% protection against new high-grade cervical lesions associated with 10 nonvaccine HPV types (212). The real level of cross protection with current vaccines is not sufficient at this time to justify change to current clinical practice and vaccinated women should continue conventional screening for cervical neoplasia (240,241).
The immunogenicity and clinical efficacy of HPV vaccines in immunosuppressed women and in women with chronic disease is uncertain but is under study. The quadrivalent vaccine appears safe in pregnant women with no increased rates of spontaneous pregnancy loss or fetal abnormalities compared to placebo risks. The vaccine is classified Category B in terms of pregnancy risk and vaccination during pregnancy is not recommended (240,242). Women who become pregnant during the vaccination schedule should be reassured that adverse consequence for the pregnancy related to vaccination is most unlikely.
No serious adverse affects attributable to vaccination were seen in placebo controlled trials (227,228,234,243,244,245). Local reactogenicity at the immunization site, systemic malaise and fever were slightly more common than with placebo but did not lead to discontinuation of the vaccination schedule. Redness and swelling at the vaccination site was increased with the bivalent vaccine adjuvanted with monophosphoryl Lipid A compared to placebo (234). Since vaccine licensure, over 12 million doses of the quadrivalent vaccine have now been given to young women. The vaccine adverse events reporting service indicates no rare, serious adverse events occurring with greater frequency among vaccine recipients than might be expected in the age-matched unvaccinated population (212,213). Fainting after vaccination is the most common adverse event.
The quadrivalent HPV vaccine is currently licensed in over 80 countries and the bivalent vaccine in 2 countries. Vaccination is approved for young women up to the age of 26 years.The greatest public health benefit is achieved by vaccination of girls and young women prior to sexual initiation as the vaccines are prophylactic (246). The U.S. Advisory Committee on Immunization Practices (ACIP) recommends routine vaccination with the quadrivalent HPV vaccine of all 11- to 12-year-old girls, “catch-up” vaccination of all 13- to 26-year-old girls and women (240). Modeling suggests this will have the most immediate impact on prevention of CIN detected through existing screening programs (212,247,248). Vaccination of 9 to 10-year-old girls should be at the providers' discretion.
Vaccination of young sexually active women may still provide some protection. Routine performance of Pap smears or HPV DNA testing prior to vaccination is not recommended, although such screening may be appropriate for sexually active women. Cervical cancer screening should continue for the immunized population to screen for disease caused by nonvaccine HPV types, to monitor the continued efficacy of the vaccination program (which may not be 100%), and to screen HPV infected women as the vaccine is not therapeutic.
The optimal vaccine dosing schedule is three doses at 0, 2, and 6 months. Accelerated delivery schedules over 4 months are being used in some countries. The U.S. Advisory Committee on Immunization Practices (ACIP) recommends (240):
All studies examining cost effectiveness suggest that vaccinating 12-year-old girls against HPV is cost effective (247,248,249,250,251,252), followed by either annual screening starting at age 18 or even 3 yearly screening starting at age 25. The costs of lost productivity due to cervical cancer related illness and death, which will further improve cost effectiveness of prevention, are not factored into these studies. Cost savings are achieved in reducing abnormal Pap smear results, colposcopy referrals, cervical biopsies and treatment procedures, as well as reducing the costs of diagnosis and treatment of genital warts. The costs, both financial and personal, associated with the complications of treatment procedures among young women, including the obstetric morbidity related to impaired cervical function in pregnancy such as premature delivery, low birth weight babies, and premature rupture of membranes will also be reduced. Divergent findings have been reported in assessment of the cost effectiveness of vaccinating females and males (247,251), though no efficacy data exists yet for vaccination of young men.
The HPV vaccine can be administered with other age-appropriate vaccines such as those for meningococcus, tetanus, diphtheria, and pertussis. Lactating and immunocompromised women can receive the vaccine but it is not recommended in pregnancy due to insufficient data supporting safety (240). Most cervical disease in older women results from persistence of preexisting HPV infection. The vaccines do not treat existing HPV infection or associated disease. New high-risk HPV infections occur throughout life but with reduced frequency. Cross-protection against infection with other nonvaccine HPV types is partial (253) or nonexistent (221). The benefit of vaccination in preventing cervical cancer falls with increasing age (241,246). The relevant benefit of cervical cancer screening increases. At a still undetermined age, the benefits and cost effectiveness of screening may outweigh those of vaccination.
The reduction in cervical cancer risk with vaccination will be determined by the number of oncogenic HPV types included in the vaccine, the proportion of the population vaccinated, the durability of protection and the continued compliance by women and providers with screening guidelines. The possible 70% reduction in risk of cervical cancer associated with the current vaccines cannot be assessed for several decades, as young girls currently vaccinated begin to reach the median age of cervical cancer diagnosis of 48 years. Premature relaxation of existing cancer control measures could see cervical cancer rates rise offsetting some of the benefits of vaccination.
Widespread HPV vaccination will decrease the prevalence of CIN 2-3 in the population, causing the positive predictive value of Pap smear screening tests for detection of precancer to decrease and negative predictive value to increase. With a lowered incidence of disease in the community, there will be a need for more sensitive screening tests. In this context, HPV 16 and HPV 18 type-specific screening tests are likely to alter screening recommendations for vaccinated and unvaccinated women in the future. At this time, vaccinated women should continue to be screened according to recommendations, with close monitoring of changes in test performance characteristics.
The development of an effective HPV vaccine could provide a powerful strategy for the control of cervical cancer in resource-poor regions and underserved populations in developed countries. Cytologic screening programs have been relatively ineffective in poor countries throughout the world and have underperformed among minority groups in developed regions. The reasons for this failure are complex. They reflect racial, ethnic, and socioeconomic disparities in participation in the screening program, in addition to the relative insensitivity of a single Pap smear, and the multivisit model for diagnosis and treatment.
The success of HPV vaccination programs in poor regions will depend on cost of vaccine, HPV epidemiology, sociocultural environment, and the commitment of national and international health organizations to the logistical challenges of implementing widespread HPV vaccination programs weighed against other health priorities. An international effort is currently underway to make HPV vaccination globally available in an affordable manner.
The politics surrounding a vaccine directed against a sexually transmitted disease and aimed at prevention of a related disease in women alone, possibly years later, are extremely challenging (254,255). Concerns have been raised that protecting girls and young women from later development of HPV-associated disease by vaccination may increase unsafe sexual behavior and premature sexual activity (“behavioral disinhibition”) (256,257,258,259). This has been seen as a potential barrier to vaccine acceptance and implementation, as well as to parental and provider acceptance. The basis for these concerns is the perception that fear of HPV infection is a motivation for abstinence or for safe sex practices among adolescents. In reality, the understanding of HPV as a sexually transmitted disease is extremely limited among male and female adolescents (260) and even among adults (261). Overall, HPV vaccine acceptance among young women, parents and providers is very high (259) influenced by the vaccines' high efficacy and safety, the risks and severity of the targeted diseases and the support by professional organizations. Controversial issues remain surrounding mandatory vaccination, affordability, access, impact on screening behavior and influence on sexual behavior (262,263,264,265,266).
According to the American Cancer Society's Guideline for HPV Vaccine Use (240), the limitations of current vaccines are: (i) the vaccines do not protect against all carcinogenic HPV types; (ii) the vaccines do not treat existing HPV infections; (iii) the duration of protection and the length of protection required to prevent cancer are unknown; (iv) the cost of vaccination and the possible need for booster doses may limit vaccine use among poor and uninsured women; and (v) the three-dose regimen may not be feasible in poor and medically underserved populations.
The Guideline suggests further research is required for: (i) long-term data on duration of vaccineinduced immunity; (ii) vaccine safety data including reproductive toxicity and side effects of HPV vaccination with other adolescent vaccines; (iii) registry data to assess vaccine coverage; (iv) surveillance data to assess population-based vaccine efficacy; (v) data for population-based and lesion-based changes in type-specific HPV prevalence across the full spectrum of HPVassociated disease, and (vi) qualitative and quantitative data on vaccine acceptability and the impact on sexual behavior and screening practices.
Vaccination against a subset of high-risk HPV types may not have the same long-term protective effect as destruction or excision of the transformation zone among women with established CIN lesions. Prevention will avoid the trauma, costs, and morbidity associated with the detection, diagnosis, and management of CIN lesions. There is a critical need to understand the likely impact of vaccination on screening, including the performance characteristics of cytology and high-risk HPV-DNA testing as well as on women's screening behavior and providers recommendations regarding screening. The combination of vaccination of adolescents and young women with primary HPV DNA screening in older women could substantially decrease the incidence of cervical cancer whilst requiring a limited lifetime number of patient visits. Such an approach needs validation before widespread adoption (Fig. 8.7).
Figure 8.7 HPV Vaccination and the Natural History of HPV Infection in Cervical Cancer Prevention. Peak prevalence of transient infections with carcinogenic types of HPV (blue line) occurs among women during their teens and 20s after initiation of sexual activity. Peak prevalence of cervical precancer occurs approximately 10 years later (green line). Peak prevalence of invasive cancer occurs at 40 to 50 years of age (red line). The conventional model of cervical-cancer prevention is based on repeated rounds of cytologic examination, including Papanicolaou smears, and colposcopy (small blue arrows). Alternative strategies include HPV vaccination of adolescents (large beige arrow), one or two rounds of HPV screening at the peak ages of treatable precancerous conditions and early cancer (large reddish-brown arrows), or both. (Reproduced with permission from Schiffman M, Castle PE. The promise of global cervical-cancer prevention. N Engl J Med 2005;353;2101-2104).
Screening for Cervical Neoplasia
Incidence and mortality rates for cervical cancer in the United States have steadily decreased since the 1950s (267). Although the incidence of cervical cancer in Western countries was beginning to decline before the introduction of screening efforts, the significant decreases in cervical cancer incidence and mortality can be largely attributed to the success of widespread screening (267,268,269,270). The Pap smear is widely recognized as the most costeffective cancer screening test yet devised, and serves as a model for screening for other malignancies.
A cohort effect for cervical cancer incidence and mortality has been clearly demonstrated (271). Women who entered their early reproductive years at times of great social upheaval, such as during World Wars I and II, remained at high risk for cervical neoplasia all their lives. Women in their early reproductive years in the two decades after the end of World War II, a period of reversion to very traditional sexual and social mores in many Western countries, appear to have been at low risk for development of cervical cancer. There has been an increasing incidence of cervical cancer in young women in many Western countries since the 1970s in spite of dramatic increases in diagnosis and treatment of preinvasive cervical disease (5,271,272,273).
Test Performance Characteristics of Conventional Cervical Cytologic Screening
The accuracy of the Pap smear has never been tested in a prospective, double-blinded study. Only relatively recently has the accuracy of the Pap smear been questioned (273), although it has long been apparent that it has a definite false-negative rate for invasive cancer and its precursors (274,275,276,277,278).
Sensitivity of Cervical Cytologic Screening
Sensitivity levels reported by experts under research conditions are not reproducible in routine clinical practice. Reasonable test performance using a competent laboratory results in falsenegative rates of 15% to 30% for high-grade lesions (HGLs) (CIN 2 to 3) (279,280,281,282). Falsenegative rates for invasive cervical cancer can be even higher, approaching 50% in some series, because of obscuring effects of blood, inflammatory exudate, and necrotic debris. In Western countries, many women in whom invasive cancer develops have never been screened but up to 50% have been screened yet still develop cancer (283,284). This occurs more frequently among younger women with invasive cancer and reflects the inherent suboptimal sensitivity of conventional cytologic screening (Fig. 8.8).
A false-negative cytologic result occurs when the smear report does not predict the presence of any grade of cervical neoplasia. This consists of “true” false-negative results (70%) and laboratory errors (30%) (285,286). True false-negative smears are free of abnormal cells, even on review of the slide, in the presence of histologically proven cervical disease.The main factors contributing to the false-negative rate are (i) specimen collection; (ii) laboratory error; and (iii) deficiencies in laboratory quality assurance mechanisms.
Specimen Collection The accuracy of cytologic diagnosis is highly sensitive to sample-to-sample variation in number of cells per smear. The quality of sample taking is the major factor contributing to this variability. The cervix may desquamate unpredictably. A large, fourquadrant HGL may fail to provide representative cells despite conscientious sampling.
The patient should be informed to refrain from douching or using tampons or intravaginal medications for at least 48 hours before the scheduled examination. She should also avoid intercourse for 48 hours before the visit and should reschedule if menstrual bleeding occurs. Best results are obtained by paired use of the Ayre's spatula and cytobrush or sampling devices that adequately sample both endocervix and ectocervix.
Figure 8.8 Sensitivity and specificity of screening test as reciprocal ratios.
Laboratory Error One-third of false-negative smear reporting is attributable to laboratory error (280, 285). In response to medical and media pressure to address this problem, there has been a significant increase in the number of smears reported as showing minor abnormalities. The effect of this has been to decrease the specificity of cytologic methodology without significantly increasing the sensitivity of the test for high-grade lesions and cancer (282,286,287,288,289).
Quality Assurance Laboratories in the United States are required to rescreen 10% of randomly selected negative cases. This strategy has been of uncertain value because it provides limited assurance of quality given the relatively low prevalence of high-grade lesions and cancer. The Health Care Financing Agency has restricted the number of cervical smears that a cytopathologist can evaluate to 80 slides per day.
Specificity of Cervical Cytologic Screening
Historically, the primary aim of cervical screening was the detection of clinically occult cervical cancer. High specificity was required at the cost of reduced sensitivity. The recognition that cytologic screening prevents cervical cancer by detection of preinvasive disease has shifted this balance, favoring increased sensitivity. Cytologic criteria for HGLs and invasive cancer were formulated when specificity was demanded, and competent laboratories operate with a very low false-positive rate, usually between 2% and 5%, for the diagnosis of high-grade disease (290). The specificity of cytologic screening has been eroded by cytologic overcall of low-grade disease (270). Colposcopic assessment of women with low-grade cytologic abnormalities reveals no disease in as many as 30% of cases (287).
Liquid-Based, Thin-Layer Cervical Cytology
A meta-analysis of 28 studies in which conventional cytology was evaluated for accuracy as a screening test reported a mean sensitivity and specificity of 58% and 69%, respectively (281). A large U.S. study concluded that estimates of the sensitivity of the conventional smear were biased in most studies (287). Based on the least biased studies, they concluded the sensitivity of conventional cytology was 51%, much lower than generally believed.
Sampling and preparation errors are responsible for more than 70% of false-negative Pap smears (289). Up to 80% of cervical cells are discarded with collection devices used in taking conventional smears (288,290). When abnormal cells are present on the slide, they may be difficult to identify and interpret in conventional smears because of the obscuring effects of air-drying artifact, excess blood, mucus, and inflammatory debris, or areas of thick cellularity where there may be insufficient permeation of the fixative.
When previous negative Pap smears of women diagnosed with cancer of the cervix are reviewed, many are shown to have been falsely reported as negative (283,284,291). To address these issues, a new slide preparation method applied to gynecologic specimens was developed. The cervical sample is taken in the routine manner using conventional sampling devices. Instead of smearing the sample onto a glass slide, the collection device is rinsed in a vial containing 20 mL of a buffered alcohol liquid preservative.
Liquid-Based Cytology Techniques
The most widely researched of the liquid-based cytology (LBC) techniques is the ThinPrep method (Cytyc Corporation, Boxborough, MA; approved by the Food and Drug Administration [FDA] in May 1996) (292). The slide preparation technique is automated. Slide evaluation is usually performed by cytotechnicians/cytologists, but automated image analysis technology also may be used (Fig. 8.9). Another LBC technique, SurePath (TripPath Imaging, Burlington, NC; approved by FDA in May 2003), is available for gynecologic use. A Dutch analysis suggested further evaluation of Surepath LBC was necessary (293), whereas the United Kingdom evaluation of both technologies concluded that there was insufficient evidence to recommend one LBC product over another (294,295,296).
Studies assessing the ThinPrep method include both split-sample studies (295,296,297,298), in which a conventional smear is first prepared and then the remainder of the specimen is rinsed into a vial for thin-layer preparation, and direct-to-vial studies (299,300,301,302), in which the thin-layer sample is used as a replacement for routine cytology. These initial studies suggested a substantial increase in detection of biopsy-confirmed, high-grade cervical abnormalities, ranging from 16% to 100%. The same studies showed a significant decrease in “unsatisfactory” smear reports. The ThinPrep Pap test was FDA approved as a replacement for the conventional Pap smear on the basis that the test was significantly more effective than the conventional smear for the detection of low-grade and more severe cervical abnormalities in a variety of populations.
Figure 8.9 Comparison of (A) standard Papanicolaou smear with (B) mono-layer preparation.
In the United States, there has been rapid adoption of LBC for cervical screening over the past decade. Most cervical smears in the United States are taken into solution despite the lack of large randomized studies (303). Recent technological improvements in automation of processing and assessment of LBC specimens, particularly the use of an automated imager, are likely to increase this utilization (304). The ability to test LBC specimens for HPV DNA and other sexually transmitted organisms further enhances the clinical appeal of this technology.
The challenging questions asked by screening programs around the world are (i) Is LBC more sensitive and specific than conventional cytology; and (ii) Is implementation of LBC cost effective? Screening programs rely on analyses that adopt a population-based model as opposed to limited laboratory and clinical studies of specific clinical populations. The results of these analyses have been as varied as the screening programs (305,306,307,308,309,310,311,312).
Liquid-Based Cytology: Clinical Efficacy and Cost Effectiveness
A 1999 analysis conducted for the U.S. Agency for Health Care Policy and Research (AHCPR) assessed the efficiency and cost effectiveness of new cervical cytologic screening technologies based on a metaanalysis of published research (287). This study reported that the ThinPrep test was the most cost effective new cervical cytologic screening technology. The report indicated an improved sensitivity for LBC as opposed to conventional cytology but found no precise estimates for the effect on specificity, particularly with reported increased detection of atypical squamous cells of undetermined significance (ASC-US) and low-grade squamous intraepithelial lesions (LSIL). It reported there was insufficient evidence to make recommendations for adopting LBC and suggested further study.
Following an extensive literature review and assessment, the American Cancer Society reported in 2002 that LBC (ThinPrep) was more sensitive but possibly less specific than conventional cytology for the detection of HSIL (313). The revised guidelines of the ACS recommended liquid-based cytology as an alternative to conventional smears with screening to be performed every two years.
The United Kingdom experience lent strong support to the claim that LBC offers improved clinical performance in cervical screening. The 2001 Scottish Cervical Screening Programme pilot study reported that use of LBC increased detection of HSIL and significantly reduced unsatisfactory smear rates when compared with the conventional smear (311).LBC was subsequently introduced into the Scottish Cervical Screening Programme. Similar findings were reported in England after interim clinical assessment of the pilot conducted by the U.K. National Screening Program (308). This led the National Institute for Clinical Excellence, U.K. (NICE) in 2003 to recommend LBC be used as the primary cervical cancer screening tool in England and Wales (309).
The U.K. experience argued strongly in favor of improved clinical effectiveness of LBC (314). Introduction of LBC will be complete in the United Kingdom in 2008 following an implementation pilot (315). The desired end points are to reduce the rate of inadequate samples and increase screening capacity. This has been achieved and specificity has been maintained (316). There has been some increase in detection rates (317) but increased sensitivity was not a specific aim of implementation. Cost effectiveness will be influenced by a number of additional variables, such as availability and implementation of automated screening devices, increased screening intervals with increased screening sensitivity, reduction in unsatisfactory smear rates, colposcopic referral guidelines, introduction of HPV DNA testing in older age groups to further increase safe screening intervals, use of reflex HPV DNA testing for triage of equivocal LBC results, and use of LBC specimens for broader testing for other sexually transmitted diseases.
More recent systematic reviews have concluded that the evidence about liquid-based cytology was not good enough to judge LBC performance relative to conventional cytology. The U.S. Preventive Services Task Force, reporting in 2003, concluded there was insufficient evidence to make a recommendation to adopt LBC (312). Recent metaanalyses of the performance of LBC, based on nonrandomized trials, have reached conflicting conclusions (303,318,319,320).
In 2007, Ronco and colleagues reported the first large randomized, controlled comparison of conventional cytology with liquid based cytology in over 45,000 women presenting for primary screening in 9 centers in Europe (321). The same cytologists read both types of slides and abnormal results were checked by a panel of cytologists. Histologic confirmation of CIN or worse in a colposcopically directed biopsy was the principle end point with pathologists blinded to cytologic results. Sensitivity and specificity could not be calculated as patients with normal cytology did not have colposcopy. The frequency of abnormal findings was greater with liquid-based cytology than conventional cytology (6.3% vs. 3.8%).Liquid-based cytology showed significantly increased sensitivity for CIN1 but not for CIN3 or invasive cancer. LBC was slightly more sensitive in detection of CIN2 lesions but this was not significant. Detection of high-grade disease and invasive cancer was similar in both groups suggesting sensitivity and frequency of false-negative results were also similar. The probability of histologically confirmed CIN after a positive result was lower in the LBC group suggesting lower specificity, lower positive-predictive value and higher false-positive rates. LBC also reported proportionately more slides as atypical squamous cells of undetermined significance.
The finding of lower sensitivity of LBC for high-grade CIN and worse in this large, well designed randomized study was surprising and at odds with manufacturer claims that LBC is “significantly more sensitive” than conventional cytology for detecting cervical neoplasia.
A further recent large, Dutch, prospective randomized comparison of LBC against conventional cytology showed no statistically significant difference in cytologic test positivity rates but significantly fewer unsatisfactory tests (322). Lower unsatisfactory test reporting, the ability to perform additional testing on the liquid-based sample and laboratory benefits such as automation and time savings will need to be weighed against the potential physical and psychological trauma to women associated with false positive results (318,320,323).
The 2001 Bethesda System
The Bethesda System for reporting cervical/vaginal cytological diagnoses was originally developed in 1988 at a United States National Cancer Institute (Bethesda, MD) workshop (41). The recommendations of the 1988 workshop rapidly gained widespread acceptance in laboratory practice in the United States and beyond. In 1991, a second NCI-sponsored workshop reviewed and modified the Bethesda System on the basis of laboratory and clinical experience (43) (Fig. 8.10).
A cervical-vaginal smear report using the revised 1991 Bethesda system had three components: (i) a description of smear adequacy; (ii) a general categorization (i.e., “within normal limits” or “not within normal limits”); and (iii) description of the cytologic abnormality, specifying whether squamous or glandular. Abnormal morphology that may represent preinvasive squamous disease falls into three descriptive categories: ASC-US, LSIL, and HSIL.
With the increased utilization of new cervical cancer screening technologies and in response to recent research findings, in 2001, the NCI sponsored a further multidisciplinary workshop to reevaluate and update the Bethesda System (45). The most clinically relevant changes are described in Chapter 6.
ASCUS-LSIL Triage Study (The ALTS Trial)
In the United States, more than 55 million Papanicolaou tests are performed annually. Five percent are reported as ASC-US and 2% as LSIL. The clinical dilemma is the need to identify the small number of women with CIN 3 and invasive cancer weighed against the high prevalence of ASC-US and LSIL reporting (324,325). Effective colposcopic triage strategies are needed to identify the minority of women with clinically significant disease while avoiding excessive intervention for others.
Figure 8.10 Comparison of terminolgy for classification of HPV-associated squamous disease. Top: The Bethesa System. Equivocal interpretations of ASC-US (atypical squamous cells of undetermined significance) and ASC-H (atypical squamous cells, cannot rule out highgrade squamous intraeptithelial lesions) are noted with stippling, the amount and colour of which suggests the expected frequencies within the differential diagnosis. 2. The CIN Classification 3. The Dysplasia Classsification and Bottom the old Papanicolaou system. (Reproduced with permission from Schiffman M, Castle PE, Jeronimo J, Rodriguez AC, Wacholder S. Human papillomavirus and cervical cancer. The Lancet 2007;370;890-907.)
The ASC-US/LSIL Triage Study (ALTS) (48,54,55,326,327,328,329,330,331) was a multicenter, randomized trial comparing the sensitivity and specificity of the following management strategies to detect CIN 3 among women referred with ASC-US and LSIL smear reports: (i) immediate colposcopy (considered to be the reference standard); (ii) triage to colposcopy based on enrollment HPV DNA testing results from Hybrid Capture 2 (HC 2) and thin-layer LBC results with a colposcopy referral threshold of HSIL; or (iii) conservative management with triage based on repeat cytology results alone at a referral threshold of HSIL. The trial had a majority of young women with a mean age of 29 years and included 2-year follow-up with exit colposcopy. Loop electrosurgical excision procedure (LEEP) was offered to women with histologic diagnoses of CIN 2 or CIN 3 at any visit or persistent CIN 1 at exit.
ALTS Trial: ASCUS Results
There were 3,488 women recruited with a community-based referral smear report of ASC-US. ASC-US interpretation was not highly reproducible, with only 32.4% of women having repeat ASC-US reported on enrollment LBC (54,326). The overall percentage of CIN 2-3 in the ASC-US study population was 15.4%. The 2-year cumulative diagnosis of CIN 3 was 8% to 9% in all study arms (329,330). A single enrollment HPV DNA test identified 92% of women ultimately diagnosed with CIN 3 with 53% of women with an ASC-US smear report being referred for colposcopy. Only 1.4% of women who were HPV negative at enrollment were ultimately found to have CIN 3 over 2 years. Serial cytology with a repeat ASC-US result as threshold for referral required two visits to achieve similar sensitivity (95%) and would have referred 67% of women to colposcopy. HPV triage is at least as sensitive as immediate colposcopy for detecting CIN 3 but refers approximately half as many women to colposcopy. Repeat cytology is sensitive at an ASC-US referral threshold, but requires two follow-up visits and more colposcopic examinations than HPV triage.
The ALTS trial demonstrates that testing for cancer-associated HPV-DNA is a viable option in the management of women with ASC-US. The American Society for Colposcopy and Cervical Pathology (ASCCP) 2001 Consensus Management Guidelines state that reflex HPV testing is the preferred triage for an ASC-US smear result when LBC methods are used (332).
The 2001 Bethesda System did not specifically endorse HPV DNA testing. Simultaneous reporting of the cytology and HPV DNA results is recommended but not always feasible.
ALTS Trial: LSIL Results
There were 1,572 women recruited with a community-based referral smear report of LSIL (331). A cytologic interpretation of LSIL was more reproducible and was associated with a 25% risk of CIN 2-3 within 2 years. There were five invasive cancers and one case of adenocarcinoma in situ detected. No intermediate triage strategy significantly decreased the need for colposcopic referral. Most (more than 85%) LSIL cases are oncogenic HPV DNA positive, and the use of HPV DNA testing for the initial triage of LSIL is discouraged.
Repeat cytology with an HSIL threshold for referral for colposcopy required only 19% of women to be referred for colposcopy but detected only 48% of cumulative CIN 3 cases. This strategy is not sufficiently sensitive for the timely detection of CIN 3. If the referral threshold was decreased to a single ASCUS result or above on repeat cytology, more than 80% of women were referred, achieving 90% sensitivity in detection of CIN 3. Three sequential cytologic examinations with a referral threshold of LSIL demonstrated acceptable sensitivity (93%) and referral rate (69%). High patient retention is critical to this strategy, requiring a commitment from patient and provider to obtain quality cervical samples every 6 months.Two-thirds of women would still be eventually referred for colposcopy.
If compliance can be achieved, cytologic follow-up of LSIL might be considered in selected populations such as adolescents who are at high risk of HPV infection and abnormal smear results but at low risk of cancer. Loss to follow-up of patients remains a major concern. The ALTS data suggest that there is currently no efficient triage for LSIL. In general, the level of risk of CIN 2-3 warrants colposcopic evaluation.
In both the ASC-US and LSIL study groups, the cumulative 2-year detection rates of CIN 3 did not vary significantly by study arm. However, in both study groups, the cumulative detection of CIN 2 was significantly reduced among women who were followed by 6-monthly conventional cytology compared with those referred for immediate colposcopy. This implies significant regression of missed prevalent cases of CIN 2 over the 2 years. CIN 2 may represent a heterogeneous group of lesions, only some of which are incipient CIN 3. CIN 2 lesions are currently treated. To avoid possible overtreatment, it would be useful to determine those lesions likely to regress, possibly through identification of biomarkers of cancer risk among CIN 2 cases.
ALTS Trial: Performance of Colposcopy
The strategy of immediate colposcopy was included as an arm in the ALTS study as the reference standard of optimal sensitivity and safety. Of significant concern was the finding that immediate colposcopy in response to either an ASC-US or LSIL smear report was only 56% sensitive for the cumulative CIN 3 detected during the trial (330). Some of the CIN 3 lesions may have developed after enrollment to be appropriately detected at follow-up visits. Review of CIN 3 cases suggested that many represented prevalent cases missed at initial colposcopic examination. Some of the CIN 3 lesions were detected in LEEP specimens at exit from the study.
Prospective follow-up of CIN 3 lesions is not ethically justifiable. If the goal of cervical screening is the timely detection of CIN 3, the ALTS trial suggests that colposcopically directed biopsy is not a gold standard of absolute sensitivity.
ALTS Trial: Risk of CIN 2-3
The 2-year cumulative risk of CIN 2-3 was virtually the same for women with LSIL and HPV-positive ASC-US (27.6% and 26.7%, respectively) (332). Women with HPV-positive ASC-US were more likely to have negative colposcopy or negative histology on reporting of colposcopically directed biopsies. Both LSIL and HPV-positive ASC-US had an 18% risk of detection of CIN 2-3 at initial colposcopically directed biopsy, which underlies the need for identical initial colposcopic management as indicated in the 2001 ASCCP Management Consensus Guidelines.
Women with documented CIN 1 are most commonly managed expectantly by intermittent colposcopy and repeat cytology. Women with negative findings at colposcopy and biopsy tend to be followed by cytology alone. Women with CIN 1 detected at initial colposcopic workup are presumed to be at higher risk of subsequent CIN 2-3 than women who have no pathology confirmed, but the 2-year follow-up of women in the ALTS study revealed no difference in the subsequent risk of CIN 2-3 between women with no disease documented at initial colposcopy and women with documented CIN 1. This reinforces the need for follow-up of the majority of women (82%) with LSIL or HPV-positive ASC-US smear reports and who have CIN 1 or less diagnosed from the initial colposcopic assessment.
The ALTS longitudinal data were reviewed to determine the most efficient follow-up strategy for detection of prevalent high-grade disease in women referred with LSIL or HPV-positive ASC-US and who had CIN 1 or less at initial colposcopy. An HPV DNA test at 12 months was the single test with the highest sensitivity and lowest referral to repeat colposcopy. HPV DNA testing at the 6-month follow-up examination was equally sensitive, but referred 13% more patients for repeat colposcopy (62.4% vs. 55.0%). Addition of cytology to the HPV test only marginally increased sensitivity but significantly increased referral. Three repeat cytologic tests at a threshold of ASC-US and without HPV testing achieved marginally higher sensitivity but referred a much higher percentage of women for colposcopy and required multiple office visits.
ALTS Trial: Cost-Effectiveness Model
A comprehensive cost-effectiveness analysis of alternative strategies based on the ALTS trial was conducted (333). A policy of ignoring equivocal smear results significantly reduced the effectiveness of cervical cancer screening and was not considered a viable alternative. The three management strategies of repeat cytology, immediate colposcopy, and oncogenic HPV DNA testing produce extremely small differences in cancer incidence reduction, although repeat cytology was less effective than both alternative strategies under all model conditions. Costs associated with each management strategy differed substantially. Reflex HPV DNA testing was always less costly than repeat cytology, as it eliminates the need for repeat clinical examination to obtain a further cervical specimen and reduces the number of colposcopic examinations by 40% to 60%. Referral of all women with ASC-US smears for colposcopy was always more costly than repeat cytology or HPV DNA testing.
ALTS Trial: Implications for Screening
The model then assessed the most efficient screening options by considering alternative strategies to manage equivocal smear results while simultaneously varying screening frequencies and types of cytological tests (334). Triennial screening using liquid-based cytology with reflex HPV DNA testing for ASC-US smear results appeared the most cost-effective model.
In part because of these findings and citing lack of direct evidence that annual screening leads to better outcomes than wider interval screening, the U.S. Preventive Services Task Force recommended that screening for cervical cancer be performed “at least every three years” rather than every year (335). The American Cancer Society Guidelines suggest 3-yearly screening interval for women 30 years of age and older who have had negative results on three or more consecutive cervical smears (313). A combined cytologic and HPV DNA test was also recommended by the American Cancer Society as a reasonable alternative to cytologic testing alone for women 30 years of age or older, with an explicit recommendation that the testing not be performed more often than every three years.
An analysis of data from the National Breast and Cervical Cancer Early Detection Program administered by the Centers for Disease Control and Prevention (CDC) concluded that “women 30 to 64 years of age with three or more previous negative smears who are screened three-yearly (once every three years) after the last negative test rather than annually have an excess risk of cancer of no more than 3 in 100,000” (335). Swedish investigators reported an annual incidence of squamous cervical cancer of 0.8 per 100,000 women with at least one previous negative test (336).
In part because of these findings, the CDC program changed its screening policy, in line with the other national organizations, increasing the interval between screenings to 3 years after three consecutive negative tests. More than 80% of women in the United States have undergone cervical screening in the past three years.
Human Papillomavirus Testing in Primary Screening
Large HPV DNA screening studies, employing both cross-sectional and longitudinal study designs, provide compelling evidence supporting the adjunctive use of HPV DNA testing in routine screening in women older than 30 years and for younger women in certain settings (337,338,339,340,341,342,343,344,345,346,347,348,349,350). These studies demonstrate that HPV DNA testing using clinically available detection tools can identify almost all patients with CIN 3, high-grade glandular neoplasia, and invasive cancer. Addition of LBC to HPV DNA testing increases sensitivity by 5%. Combination HPV DNA testing and LBC has 90% to 100% sensitivity for CIN 3 and invasive cancer. The negative predictive value of such combined testing is above 99% in all studies and approaches 100% in most.
A negative HPV DNA test virtually excludes any risk of underlying disease even in the presence of a reported cytologic abnormality. In response to this collective experience, HPV DNA testing was approved as a primary screening tool in conjunction with cytology in women aged more than 30 years by the FDA in 2003.
Although the high sensitivity and the high negative predictive value are obvious advantages of the HPV test, its use as a primary screening tool is potentially hampered by low specificity and risk of overtreatment of HPV DNA-positive women (351). Specificity and positive predictive value is further decreased when HPV DNA testing is combined with LBC, reflecting the inverse relationship between sensitivity and specificity in screening methodology.
HPV DNA testing identifies many transient HPV infections not associated with highgrade CIN, particularly in younger women. This is the main contributing factor to reduced specificity. Women who are HPV DNA positive but who have a normal cervical smear or have no clinical evidence of HPV-related disease are at greatest risk of developing cervical neoplasia prospectively (352,353). Such women should not be viewed as having “falsepositive” tests but require close follow-up and repeat testing. For HPV testing to be cost effective in primary screening, an efficient policy is required for the management of women who test positive for high-risk HPV-DNA but who have negative or equivocal cytologic reporting (354). This strategy would use HPV DNA testing alone as primary screening and cytology would be used to triage HPV DNA-positive women.
The HART Study
The HART (HPV in Addition to Routine Testing) study from the United Kingdom recruited 11,085 women attending 161 U.K. family practices for routine cervical screening (343). This study confirmed HPV DNA testing to be a more sensitive primary screening technique than conventional (not liquid-based) cytology for detecting high-grade CIN (97.1% vs. 76.6%).High-risk HPV DNA detection combined with conventional cytology had 100% sensitivity for high-grade CIN lesions and above. HPV DNA testing was significantly less specific than conventional cytology (93.3% vs. 95.8%) and had a positive predictive value of 12.8%, less than that of an equivocal or worse conventional smear (15.8%). The high rate of positive HPV DNA testing in women with no significant histologic abnormality in this study appears to be an impediment to the use of HPV DNA testing in primary screening.
The authors estimated that referral rates for colposcopy would be reduced if (i) women with equivocal (and possibly with LSIL) smear reports but who were HPV DNA negative were returned to routine screening; (ii) HPV DNA-positive women with negative or equivocal smears were retested at 12 months; and (iii) the screening interval was extended to 5 years.
Three recent large European population-based, randomized, controlled trials of HPV DNA testing for primary cervical cancer screening have been reported from the Netherlands (355), Sweden (356), and Italy (357,358). In each study, the combination of HPV DNA testing and cytology was compared with conventional cytology alone for the detection of cervical neoplasia.
In the Dutch study (355), implementation of HPV DNA testing in addition to conventional cytology led to a 70% increase in the number of histologically confirmed lesions reported as CIN 3 or worse in the baseline screening round. At the second round of screening, 55% fewer women in the combined screening group had CIN 3 or worse detected compared to the cytology alone group. Similar results were reported for CIN 2 lesions which are usually treated when identified. The decreased detection of high-grade precursor lesions in the second round of screening suggest earlier detection of high-grade lesions which would have persisted rather than a subset of regressive lesions. This is supported by the finding that the decrease in the number of lesions detected at the second round was almost the same as the increase seen at baseline.
This early detection of a substantial proportion of persistent CIN 3 lesions leads to greater protection against invasive cervical cancer from a single round of screening. This would permit safe extension of the screening interval with probable decreased costs and improved participation. Nonparticipation in screening is still the greatest risk factor for future development of cervical cancer. Women who are high-risk HPV DNA negative but have an abnormal smear had a negligible risk of CIN 3 or worse, supporting the model of primary screening with HPV DNA testing and cytology as triage of HPV DNA positive women. Conservative follow-up of HPV DNA positive women with normal cytology led to few referrals for colposcopy and may have avoided detection of rapidly regressive low-grade lesions. Women in this trial were recruited from age 30 years onwards as is the norm in the Netherlands.
The large Italian randomized, controlled trial of HPV DNA testing versus conventional cytology also demonstrated a substantial gain in sensitivity of HPV DNA testing over cytology in detection of CIN 2-3 with only a small reduction in positive predictive value for women aged 35 to 60 years (357,358). Among younger women, 25 to 34 years of age, HPV DNA testing had an even higher sensitivity for detection of CIN 2-3 compared to cytology than with the older age group. Intensive screening and referral for colposcopy of HPV DNA positive young women and teenagers led to the detection of a high incidence of CIN 2-3 lesions soon after initial HPV infection. The difference in detection rates of high-grade CIN among young women when all HPV DNA positive women were referred for colposcopy versus only those who also had an associated cytological abnormality of ASC-US or worse was interpreted by the authors to suggest that most abnormalities, including CIN 2-3 lesions, in young women are destined to regress. Confirmation will require further screening cycles. This raises concerns in relation to increased referral of young women for colposcopy and treatment of high-grade CIN with associated increased risk of pregnancy related morbidity.The recommendation from the study is that young women under 35 years who are HPV DNA positive should only be referred to colposcopy if cytology is abnormal or if HPV positivity persists after 1 year.
The Swedish population based randomized controlled trial of HPV DNA testing versus conventional cytology among women aged 32 to 38 years demonstrated a 51% greater detection of CIN 2-3 or cancer among those screened using HPV DNA testing (356). At subsequent screening rounds, the HPV DNA screened group had a 42% lower detection of CIN 2 or worse and a 47% decrease in CIN 3 or worse. Women with persistent HPV DNA detection in later screening rounds remained at high risk of CIN 2-3 or worse after referral for colposcopy. The authors concluded that the improved sensitivity of HPV-based cervical cancer screening was not overdiagnosis, but attributable to earlier detection of high-grade lesions which do not regress.
Based on currently available data, the combination of HPV DNA testing with cytology significantly increases screening sensitivity. Women with “double-negative” results can be screened safely at longer intervals, offsetting the increased cost of the initial screen (350). The duration of low-risk after a negative HPV DNA test is unknown. Patients identified as being at increased risk on the basis of a positive HPV DNA test but who do not have identifiable disease should be monitored more closely.
There is no consensus regarding the appropriate age for cessation of screening. Lifetime screening continues to save lives but the benefits for regularly screened women beyond age 65 years are minimal. The probability of a false-positive cervical smear result is much greater than a true-positive result after age 65 years for regularly screened women. The high negative predictive value of HPV DNA testing could be used as an additional means of safely exiting women from routine screening beyond a certain age.
The wider use of HPV DNA testing in the primary screening setting will inevitably hinge on the acceptability of testing for a sexually transmitted virus to the target population and the provision of HPV DNA testing at low cost (359). The anticipated lower incidence of disease developing among a younger generation who have received the cervical cancer vaccine as they enter the screening age groups will require the most sensitive screening tests. Highrisk HPV DNA testing may become the preferred primary screening test in this population due to the greater sensitivity and high negative predictive value.
HPV testing may have an important role in primary cervical screening in developing countries (360,361). The crude incidence rate of cervical cancer in some developing countries is as high as 100 cases per 100,000 women, compared with approximately 4 to 10 per 100,000 in developed countries. The simplicity of taking the sample, the stability of the transport medium, and the ability to automate the processing of the specimens are all advantages in developing countries.
Systematic Approach to Colposcopy
Colposcopy is the examination of the epithelia of the cervix, lower genital tract, and anogenital area using magnified illumination after the application of specific solutions to detect abnormal appearances consistent with neoplasia or to affirm normality. Integral to the procedure is targeting biopsies to areas of greatest abnormality.
Indications for Colposcopy
Colposcopy is most frequently performed in response to an abnormal cervical smear. Abnormal findings on adjunctive screening tests, such as HPV testing, can also be the indication for colposcopy. If the cervix is clinically abnormal or suspicious on naked-eye examination, colposcopy is also indicated. Abnormal and unexplained intermenstrual or postcoital bleeding and unexplained, persistent vaginal discharge may also be assessed by colposcopy to exclude a neoplastic cause. Other indications include a personal history of in utero diethylstilbestrol (DES) exposure, vulvar or vaginal neoplasia, or condylomata acuminata, and possibly sexual partners of patients with genital tract neoplasia or condylomata acuminata.
There are no absolute contraindications to colposcopy. The examination may be deferred until after bleeding ceases for women who are menstruating. Acute cervicitis or vulvovaginitis should be evaluated and treated before colposcopy is performed unless poor patient compliance is anticipated. The colposcopic procedure is modified in pregnancy, with a less liberal use of biopsy in the absence of warning signs of high-grade disease or cancer and avoidance of endocervical curettage. Postmenopausal women who are not taking hormone replacement may benefit from a 3-week course of topical or oral estrogen before colposcopy. Patients should avoid use of all intravaginal products for 24 hours before the examination.
Initial Clinical Workup
The patient should be prepared for the examination by a comprehensive explanation of the indication for colposcopy and a thorough verbal description of the procedure.
A complete medical history and general examination should be obtained. A history of previous premalignant cervical disease or cervical treatment should be determined. A history of endogenous or exogenous immune suppression is relevant. A social history of smoking or other “recreational” drug use should be obtained.
A clinical and speculum examination of the cervix, vagina, vulva, and perianal areas should be performed before the colposcopic examination. Squamous neoplasia may be multicentric (involving more than one genital tract site, i.e., cervix, vagina, or vulva) or multifocal (involving several areas at one site).
A bimanual pelvic and rectal examination should be performed, usually on completion of the colposcopy, to exclude clinically apparent coexistent gynecologic or pelvic disease. Uncommonly, abnormal cervical smears are caused by palpable malignancies of the endocervix, uterine body, adnexae, or bowel.
Locating the Source of Abnormal Cells
Colposcopy is performed in the dorsal lithotomy position with a drape covering the patient's legs. The cervix is visualized using a standard speculum. The colposcopic examination involves the application of three standard solutions to the cervix to determine the source of the abnormal cells in the cervical smear.
Delineating the Margins of the Lesion
Once the source of abnormal cells in a cervical smear is located, the peripheral and distal margins of the lesion should be determined.
The distal or peripheral margin of the lesion is usually readily identified. Occasionally, the lesion may extend onto the vaginal fornices, especially in the DES-exposed patient.
Delineation of the proximal or upper margin of the lesion requires the colposcopic visualization of the new squamocolumnar junction, which establishes the colposcopy as satisfactory or unsatisfactory. Failure accurately to delineate the position of the new squamocolumnar junction represents one of the most common colposcopic triage errors. An endocervical speculum may be helpful if the proximal margin is within the canal.
Endocervical curettage is performed to exclude an occult cancer in the canal (362). With the increasing incidence of cervical adenocarcinoma in situ (AIS) and invasive adenocarcinoma, many of which are associated with squamous CIN lesions, endocervical curettage may provide a safeguard against missing such lesions.
Routine performance of endocervical curettage is controversial. When the entire new squamocolumnar junction can be visualized, it is reasonable to omit the routine endocervical curettage. A negative endocervical curettage from a patient with an abnormal high-grade cytology and an unsatisfactory colposcopy does not exclude occult endocervical cancer, arid excisional cone biopsy remains mandatory. When specifically indicated, collection of an endocervical sample using a cytobrush has been shown to be a more sensitive sampling device than endocervical curettage for endocervical squamous and glandular disease. Specificity, however, is decreased (364).
Colposcopically Directed Cervical Biopsy
Cervical biopsies should be directed to the most significant lesions. Multiquadrant lesions may require multiple biopsies. Any area suspicious for occult invasion must be carefully sampled. The most reliable method of ensuring the accuracy of targeted biopsies is to grade lesions by deriving a colposcopic score. Cervical biopsies should be taken through the colposcope. The colposcopic grading score of the lesions and the biopsy sites should be carefully recorded.
Documentation of Colposcopic Findings
The findings of the colposcopic examination should be carefully documented. Photodocumentation can be extremely valuable. A system for recording patient information, laboratory results, management plan, and tracking log should be established and maintained to ensure appropriate patient care and follow-up. Modern computerized systems provide for many of these needs in a most effective manner.
The Abnormal Transformation Zone
If the transformation zone is deviated along a neoplastic pathway, epithelial and vascular alterations produce the characteristic morphologic appearances of the abnormal transformation zone (365). The colposcopic signs of the abnormal transformation zone are described in Table 8.1.
Squamous metaplasia, repair and regeneration, inflammation, and infection may all produce abnormal colposcopic transformation zone findings, such as acetowhite epithelium and abnormal vessels. Significant changes in the hormonal milieu such as accompany pregnancy, oral contraceptive pill use, estrogen withdrawal, and estrogen replacement can produce abnormal colposcopic signs in the absence of cervical disease. Atypical vessels, considered one of the colposcopic hallmarks of invasive cancer, can also occur in association with benign conditions, including immature metaplasia, nabothian follicles, inflammation, radiation treatment, and granulation tissue.
Table 8.1 Colposcopic Signs of the Abnormal Transformation Zone
Colposcopic Grading Systems
The basis of colposcopic management decision making is the process of cytologic-colposcopic-histologic correlation, with each component affording certain safeguards. There are four basic colposcopic diagnoses: (i) normal; (ii) low-grade disease (HPV infection/CIN 1)i (iii) high-grade disease (CIN 2 or 3)i and (iv) invasive cancer. Colposcopic grading systems have been developed to provide an objective, accurate, reproducible, and clinically meaningful prediction of the severity of CIN lesions based on discriminatory analysis of specific colposcopic signs (366,367,368,369).
Routine determination of a colposcopic diagnosis permits quality-control measures to be implemented in colposcopy (370,371). In colposcopic quality control programs, the colposcopist is required to achieve at least an 80% accuracy rate in colposcopic-histologic correlation or receive remedial training in colposcopic assessment of cervical lesions (372).
In current colposcopic experience, the Reid Colposcopic Index represents the most reproducible and clinically valid means of standardizing the evaluation of cervical lesions (368,369) (Table 8.2).
Reid Colposcopic Index The Reid Colposcopic Index uses four colposcopic features of premalignant cervical lesions to achieve predictive accuracy. The colposcopic index permits accurate differentiation of low-grade from high-grade disease. It is not designed to differentiate premalignant from malignant cervical neoplasia. The four colposcopic criteria usedare (i) the margin of the lesion; (ii) the color of the acetowhitening; (iii) the type of vascular pattern; and (iv) the iodine staining reaction.
The four colposcopic signs are scored individually and sequentially. The value of these colposcopic signs is maximized by combining them into a weighted scoring system. Scores of 0, 1, or 2 are assigned for each criterion, as described in Table 8.2. The total score is then reported as a ratio, the denominator of which is constant at 8. The numerator is the score derived from adding the four scores derived from evaluation of the four colposcopic signs and fluctuates as the predictor of disease severity. Scores of 0 to 2 are predictive of low-grade lesions (HPV infection/CIN 1; Fig. 8.11). Scores of 6 to 8 usually denote high-grade lesions (CIN 2 to 3; Fig. 8.12). Scores of 3 to 5 represent an area of overlap between low-grade and high-grade lesions.
Table 8.2 Scoring System for Deriving the Colposcopic Index
Figure 8.11 Colposcopy of low-grade cervical lesions showing acetowhite epithelium with fine abnormal vascular pattern.
Figure 8.12 Colposcopy of high-grade cervical lesion showing dense acetowhite epithelium and coarse abnormal vascular pattern.
The overall predictive accuracy of the index exceeds 90% after a short training period. The colposcopic index permits a significantly more accurate colposcopic-histologic agreement than can be achieved by less systematic approaches to colposcopic diagnosis.
Table 8.3 Colposcopic Warning Signs of Invasive Cancer
Colposcopic Warning Signs of Invasive Cancer
Although a rare event in many colposcopic settings, invasive cancer must not be missed and still remains the major challenge to the colposcopist. Clinicians working in oncologic settings may over time acquire significant experience in the colposcopy of occult and overt cervical cancer. Although most invasive cancers are clinically apparent and do not require colposcopy for identification, early invasive lesions may be clinically occult. Exclusion of invasive cancer demands both a high index of suspicion and knowledge of warning signs. Colposcopic warning signs are shown in Table 8.3.
Other warning signs for invasive cancer include:
ASCCP 2006 Consensus Guidelines
In 2001, the American Society for Colposcopy and Cervical Pathology (ASCCP) Consensus Guidelines were developed to assist in the management of women with cytological abnormalities (332) and in the management of cervical cancer precursors (373), in part in response to the 2001 Bethesda System.
Since 2001, improved understanding of the pathogenesis and natural history of cervical HPV infection and cervical cancer precursors, of the future pregnancy implications of treatment for CIN among young women, and of the management of adenocarcinoma in situ has led to a critical review of the earlier Guidelines (374,375). The ASCCP 2006 Consensus Guidelines are presented in Figs. 8.13 and 8.14. Although the Guidelines have been developed for the U.S. setting, most of the recommendations are relevant internationally.
Summary of Key Aspects of the ASCCP 2006 Consensus Guidelines
Management of Women with CIN 1 Preceded by ASC-US, ASC-H, or LSIL Cytology
Women with a histological diagnosis of CIN I (usually implying previous referral for colposcopy and colposcopically directed biopsy), preceded by ASC-US, ASC-H, or LSIL cytology should be followed by yearly HPV-DNA testing or 6- to 12-month Pap smears (Fig. 8.14A). If HPV DNA testing remains positive or if repeat cytology is reported as ASCUS or greater, repeat colposcopy is recommended. If the HPV DNA test is negative or two consecutive Pap smears are reported as negative, return to routine screening is recommended. If CIN 1 persists for 2 years or more, continued follow-up or treatment is appropriate. Treatment can be ablative or excisional. If colposcopy is unsatisfactory, the endocervical sample is positive for CIN or the patient has been previously treated, a diagnostic excisional procedure is recommended.
Figure 8.13 A-J Algorithms from the 2006 Consensus Guidelines for the Management of Women with Cervical Cytological Abnormalities. (Reprinted from The Journal of Lower Genital Tract Disease Vol. 11, Issue 4, with the permission of ASCCP © American Society for Colposcopy and Cervical Pathology 2007. No copies of the algorithms may be made without the prior consent of ASCCP.)
Figure 8.13 A-J Continued
Figure 8.13 A-J Continued
Figure 8.13 A-J Continued
Two consecutive negative Pap smear reports in prospective follow-up of low-grade lesions, whilst statistically reassuring, does not necessarily represent disease regression and is no guarantee against disease progression.
Management of Women with CIN 1 Preceded by HSIL or AGC-NOS Cytology
Women with a histological diagnosis of CIN 1 which has been diagnosed in the assessment of abnormal Pap smears reported as HSIL (CIN 2-3) or atypical glandular cells not otherwise specified (AGC-NOS) can be managed by either an excisional diagnostic procedure or 6-monthly colposcopy and cytology for 1 year (Fig. 8.14B). This is provided women with a cytological finding of AGC-NOS have a satisfactory colposcopic examination and negative endocervical sampling. A full review of the cytological, colposcopic, and histological results can also be undertaken with management according to the guidelines for any revised diagnosis.
Figure 8.14 A-F Algorithms from the 2006 Consensus Guidelines for the Management of Women with Cervical Histological Abnormalities. (Reprinted from The Journal of Lower Genital Tract Disease Vol. 11, Issue 4, with the permission of ASCCP © American Society for Colposcopy and Cervical Pathology 2007. No copies of the algorithms may be made without the prior consent of ASCCP.)
Figure 8.14 A-F Continued
Figure 8.14 A-F Continued
If prospective observation is selected, a diagnostic excisional procedure is recommended if the repeat cytology at either 6 or 12 months is reported as HSIL or AGC-NOS. If histology of a colposcopically directed biopsy taken in prospective follow-up confirms high-grade CIN, management is according to guidelines irrespective of the cytology result. After two consecutive “negative for intraepithelial neoplasia or malignancy” results in prospective follow-up, woman can return to routine cytological screening. The same warning applies in relation to the implications of negative Pap smears in the follow-up of histologically proven CIN although, in this situation, colpsocopy has been included in the follow-up regimen, affording a higher level of reassurance. If CIN 1 is preceded by HSIL or AGC-NOS cytology and colposcopy is unsatisfactory, a diagnostic excisional procedure is recommended except in special populations such as pregnant women.
CIN 1 in Adolescence and Pregnancy
For adolescents with CNI 1, follow-up with annual cytology is recommended (Fig. 8.14C). Only those with HSIL or greater at 12 months should be referred for colposcopy. At 24 months, those with ASC-US or greater should be referred. Prospective follow-up by HPV DNA testing in this age group is of no value due to the frequency of positive results.
Management of Women with CIN 2, 3
The heterogeneity of CIN 2 lesions is significant and regression rates are higher than for CIN 3. The histological distinction between CIN 2 and CIN 3 remains subjective and these diagnoses remain combined in the 2006 Consensus Guidelines to define the threshold for treatment for squamous intraepithelial lesions.
Both excisional and ablative procedures are acceptable treatment modalities for women with histologically proven CIN 2-3 with satisfactory colposcopy (Fig. 8.14D). An excisonal procedure is recommended for residual/recurrent CIN 2-3. Ablation is unacceptable for women with a histological diagnosis of CIN 2-3 and unsatisfactory colposcopy. Cytologic and colposcopic follow-up of CIN 2-3 is unacceptable except in specific circumstances.
Acceptable posttreatment follow-up options include 6-monthly cytology alone, 6 monthly combined cytology and coloposcopy, and HPV DNA testing at 6 to 12 months. If HPV DNA testing is positive or if the repeat cytology is ASC-US or greater, very common situations particularly at a 6-month posttreatment examination, referral for colposcopy and endocervical sampling is recommended. If HPV DNA testing is negative or if two consecutive posttreatment cytology results are “negative for intraepithelial lesion or malignancy,” routine screening for at least 20 years is recommended and should be annual for at least 5 years. The power of a negative HPV DNA test as a predictor of normality posttreatment for CIN 2-3 should be emphasized.
If CIN 2-3 is reported histologically at the margins of an excised specimen or in an endocervical sample obtained immediately after the procedure, 4 to 6 monthly cytologic follow-up with endocervical sampling is preferred. The role of colposcopy in this follow-up option is not clearly defined in the Guidelines. In practice, cytology and colposcopy will be performed in most clinical settings. A repeat diagnostic excisional procedure is acceptable. The Guidelines allow for hysterectomy if a repeat diagnostic, excisional procedure is not feasible although great care should be taken to exclude an undisclosed invasive cancer within the endocervical canal prior to hysterectomy.
For women with histologically proven residual/recurrent CIN 2-3, the Guidelines permit a repeat excisional procedure or hysterectomy. If a repeat excisional procedure is not feasible, our practice is to perform a modified radical hysterectomy.The specimen is evaluated intraoperatively, if necessary by frozen section, to determine any need for lymphadenectomy.
CIN 2, 3 in Adolescence and Pregnancy
For adolescents with a histological diagnosis of CIN 2-3 not otherwise specified, the Guidelines state that either treatment or 6-monthly observation by cytology and colposcopy for up to 24 months is acceptable provided colposcopy is satisfactory (Fig. 8.14E). Allowing for the subjectivity in this histological distinction, observation is preferred for a diagnosis of CIN 2 alone but treatment is acceptable. Treatment is recommended for a specified histological diagnosis of CIN 3 or if colposcopy is unsatisfactory. Although invasive cervical cancer is very rare in this age group, prospective follow-up of a histological diagnosis of CIN 2-3, not otherwise specified, in young women should be limited to those women likely to be compliant with the recommendations. After two consecutive “negative for intraepithelial lesion or malignancy” results, implying negative cytology and colposcopy with satisfactory colposcopic examinations, these adolescents and young women can return to routine cytologic screening. An annual screening interval should be recommended.Treatment is recommended if CIN 3 is diagnosed histologically or if CIN 2-3 persists for 24 months.
Management of Women with Cervical Adenocarcinoma In Situ (AIS)
Hysterectomy remains the preferred management recommendation for women with a histological diagnosis of AIS on a specimen from a diagnostic excisional procedure (Fig. 8.14F). A histological diagnosis of AIS from a punch biopsy or a cytologicaldiagnosis of AIS is not sufficient to justify hysterectomy without a diagnostic excisional procedure. The difficulty in defining colposcopic limits of AIS lesions, the frequent extension of disease to within the endocervical canal and the presence of multifocal, “skip lesions” (i.e., lesions which are not contiguous) compromise conservative excisional procedures.
Negative margins in an excisional specimen do not mean the lesion is completely excised. If future fertility is desired, conservative excisional management is acceptable. The overall failure rate of excision is less than 10%. Margin status is a useful clinical predictor of residual disease as is endocervical sampling at the time of excision. If a conservative excisional procedure is performed and margins are involved or the endocervical sample at the time of excision shows AIS or CIN, reexcision is recommended. A reassessment at six months using a combination of cytology, colposcopy, HPV DNA testing, and endocervical sampling is acceptable. Long-term follow-up is recommended for women who do not undergo hysterectomy for AIS.
Table 8.4 Triage Rules for Ablative Therapy for Cervical Intraepithelial Neoplasia
Treatment of Cervical Intraepithelial Neoplasia
Historically, Anderson (376) from the United Kingdom in 1965 and Kolstad and Klem (377) from the Norwegian Radium Institute in 1969, demonstrated that cone biopsy was as effective as hysterectomy in preventing the progression of carcinoma in situ to invasive cervical cancer. In 1973, Stafl and Mattingly (378) from Wisconsin demonstrated thatcolposcopically directed punch biopsies, taken by an experienced colposcopist, were as accurate as cone biopsy in obtaining a histologic diagnosis in women with an abnormal cervical smear. This facilitated the use of physical modalities to destroy the abnormal transformation zone in selected patients.
Subsequently, high primary cure rates with minimal morbidity were reported for ablative techniques such as cryosurgery (379), electrocoagulation diathermy (380), and the carbon dioxide laser (381). Patient selection is based on a set of triage rules (Table 8.4). Diagnostic conization is now performed only for specific indications in which there remains a genuine risk of undisclosed invasive cancer.
In the 1990s, loop electrosurgical excision procedures (LEEP) gained in popularity because of concerns regarding the occurrence of invasive cervical cancer in patients who had undergone ablative treatment (382). Invasive cancer has been reported after each of the ablative modalities (383). When cancer occurred after ablative therapy, it occurred within 12 months in 66% of cases and within 2 years in 90%. This suggested that a triage error was made in the initial assessment and invasive cancer was missed. Reports of a low incidence of misclassification of invasive cervical cancer or high-grade glandular neoplasia as squamous intraepithelial disease have raised concerns about the safety of ablation of high-grade squamous lesions (384,385,386).
LEEP allows for excision of the transformation zone with removal of a volume of tissue similar to that removed by ablative procedures, and with no greater morbidity. When the procedure is performed by an inexperienced operator, adequate histologic evaluation can be difficult because of diathermy artifact and orientation difficulties.
Pregnancy-Related Morbidity Associated with Treatment for CIN
All excisional procedures used in the treatment of cervical intraepithelial neoplasia are associated with adverse obstetric morbidity. Recent research and metaanalyses have confirmed that excisional procedures for CIN are associated with an increased risk of preterm delivery and low birth weight babies in future pregnancies(387,388,389,390,391,392,393,394).
The greatest increased risk is associated with cold-knife conization where there are significantly increased risks of perinatal mortality, extreme preterm delivery and very low birth weight infants (388,392). A recent Norwegian population based cohort study reported a 17.2% incidence of preterm delivery among women who gave birth after cervical conisation versus 6.7% among women who gave birth before cervical conisation and 6.2% in women who had never undergone cervical conisation (392). The excess risk was highest for late abortion and for preterm delivery prior to 33 weeks.
Large loop excision is also associated with an increased risk of preterm delivery and low birth weight babies, but is not associated with increased perinatal mortality, or with extreme preterm delivery (reviewed 388,393). It is probable that loop excision procedures that remove large amounts of cervical tissue would have the same negative impact as cold-knife conization. Clinicians should inform women of the risks of adverse pregnancy related outcomes associated with excisional treatment procedures.
Ablative procedures are associated with fewer adverse pregnancy outcomes than excisional procedures (388), although radical diathermy treatment is associated with significant pregnancy related morbidity. Treatment should be avoided where possible for young women and adolescents, particularly with low-grade lesions, which have a high rate of spontaneous regression. The “see-and-treat” approach, combining diagnosis and treatment in one visit, is now not generally recommended due to the potential for unnecessary treatment (394).
The treatment modalities for preinvasive cervical disease are ablative procedures and include cryosurgery, electrocoagulation diathermy, and CO2 laser; and excisional procedures, including LEEP, cervical (excisional) conization, CO2 laser excision, and hysterectomy. A Cochrane Database of Systematic Review (2003) examined surgical treatment modalities for cervical intraepithelial neoplasia (395). The evidence from 28 randomized controlled trials suggested that there is no overwhelmingly superior technique for eradicating CIN. Cryotherapy is an effective treatment of LSIL but not HSIL.
Cryosurgery is a simple, effective, inexpensive, and relatively easy therapeutic option for treatment of selected patients with CIN. Cervical cryosurgery, which was first introduced in 1968, involves the destruction by cryonecrosis of the lesion, including the entire transformation zone. Hypothermia is produced by the evaporation of liquid refrigerants. Compressed nitrous oxide (N2O) is allowed to expand through a small jet, producing an iceball at the surface of a metal probe placed in contact with the surface of the tissues to be frozen. Crystallization of intracellular water results in cell death.
The most appropriate cryoprobe tips are the 19-mm and 25-mm minicone. A water-soluble gel is used to coat the probe tip before the procedure. Temperatures achieved at the cryotip using N2O are recorded at -65°C to -85°C. Cell death occurs in the range of -20°C to -30°C. The lethal zone during cryosurgery begins 2 mm proximal to the iceball margin,with the temperature at the margin of the iceball equal to 0°C. To ensure a 5-mm depth of freezing, a total lateral spread of freeze of 7 mm is required. For cervical cryosurgery,the probe must cover the lesion and the entire transformation zone.
If the transformation zone is large, successive overlapping treatments are required, increasing the duration and discomfort of the procedure. Cryosurgery is therefore used mainly for smaller, ectocervical lesions. It is usually used for low-grade lesions (LGLs) without extension to within the endocervical canal.
Technique The procedure is performed under colposcopic supervision without anesthesia. Prophylactic premedication with nonsteroidal antiinflammatory drugs 30 to 60 minutes before the procedure may reduce pain and cramping associated with prostaglandin release from dying cells. The procedure should not be performed in pregnancy or during the menstrual period. The procedure is performed as follows:
Table 8.5 Comparison of Therapeutic Modalities for Cervical Intraepithelial Neoplasia
A freeze-thaw-freeze technique is commonly used. This technique was reported by Creasman et al. (396) to reduce the failure rate from 29% to 7%, although others claimed similar results from a single freeze (397,398). The second freeze is not commenced until the tissues have visibly thawed from the initial treatment.
Patients experience a watery, malodorous, blood-tinged discharge for 2 to 3 weeks after the procedure. This can be decreased by debridement of the bullous, necrotic tissue using a ring forceps and gauze 48 hours after the procedure. The patient should abstain from vaginal intercourse and tampon use for 4 weeks after the procedure.
Primary cure rates in excess of 90% have been reported for cryosurgical management of CIN lesions. The larger the lesion, the lower the primary cure rate. Cryosurgery for large ectocervical lesions covering the ectocervix is associated with failure rates as high as 42%. Endocervical glandular involvement increases the failure rate from 9% to 27%. Decreasing cure rates with increasing severity of disease, specifically 94% for CIN 1, 93% for CIN 2, and 84% for CIN 3, have also been reported (399). This may in part reflect the increased size of HGLs, which more frequently occupy two or more quadrants of the cervix (400) (Table 8.5).
Loop Electrosurgical Excision Procedure
To minimize the risk of failed detection of early invasive cancer and high-grade glandular neoplasia at the time of colposcopic triage, LEEP of the transformation zone has become a widely used and valuable therapeutic option. The equipment is relatively inexpensive, and the surgical skills are readily acquired. The procedure combines the advantages of conservative ablative procedures in preserving cervical tissue with the safety of histologic assessment of the entire lesion.
Cartier originally developed an electrosurgical method for management of CIN using 5-mm rectangular, thin wire loops to sample and treat the cervix by removing the epithelium and underlying stroma in multiple 5-mm strips (401). The process was time consuming, and thermal injury at the edge of the strips frequently compromised the specimen.
Prendiville et al. (402,403) introduced larger loop electrodes, 1 to 2 cm in width and 0.7 to 1.5 cm in depth, for excision of the entire transformation zone, usually in a single pass. The combination of very thin wire loops and modern electrosurgical generators capable of delivering high powers (35 to 55 W) has allowed electrosurgical cutting with little associated thermal injury.
The technique for electrosurgical loop excision is as follows:
Figure 8.15 “Cowboy-hat” configuration for LEEP.
Complications are minimal, comparing favorably with those associated with CO2 laser procedures (395). Postoperative bleeding occurs in 2% to 5% of patients. Postoperative infection is uncommon. Clinically significant cervical stenosis and cervical incompetence are rare complications, but the patient must be made aware of the possibility of such adverse reproductive sequelae. Cure rates are comparable with those achieved with CO2 laser procedures and with “cold-knife” conization, often in excess of 95% (395).
Electrosurgical loop excision offers several advantages over CO2 laser ablation. The procedure is quicker and easier. However, ease of use carries an attendant risk of overuse. Patient acceptance is improved and intraoperative pain is decreased. The submission of the entire specimen for histologic study increases the probability that unsuspected cancer will be detected and not ablated. In many large studies of LEEP, the unsuspected invasive cancer and high-grade glandular disease rate has been as high as 1% to 2% (404,405,406,407).
Another potential advantage of LEEP is the ability to “see-and-treat” at one visit. However, histologic study of loop-excised specimens removed at a “see-and-treat” approach revealed no disease in 5% to 40% of specimens, particularly in young women referred with minor cytologic abnormalities (408).
Carbon Dioxide Laser Ablation of the Transformation Zone
The CO2 laser is the ideal choice for vaporizing sharply defined tissue volumes to a precisely determined depth (409,410,411). To achieve optimal vaporization with minimal lateral thermal injury, the CO2 laser should be used at the highest power output with which the surgeon is comfortable. This should be a minimum of 25 W but preferably above 60 W. The cautious use of low-power outputs is one of the most common CO2 laser surgery errors causing thermal injury.
Most gynecologists treating cervical preinvasive lesions with the CO2 laser have access to a laser providing 50 W maximum power or less. As such, the laser is best used in continuousmode for the ablation of cervical lesions. However, in addition to using the highest controllable power output, thermal conduction can be further minimized by the choice of rapid superpulse as the temporal mode. The main advantage of higher-powered lasers is the higher average power achieved in rapid superpulse settings. This offers a distinct advantage in situations where control of thermal injury is critical, such as the DES-exposed, breastfeeding, postmenopausal, or postirradiated patient. For transformation zone ablative procedures, the average power density must be kept within the range of 750 to 2,000 W/cm2 (412).
Surgical Control of the CO2 Laser
The CO2 laser can be readily controlled by the surgeon at any power setting provided the appropriate Y-beam geometry is selected (see later in this chapter). For cervical transformation zone ablation, a high-power laser setting is selected. The spot diameter is progressively enlarged by defocusing the beam using the micromanipulator until a point is found where the impact crater (tested on a moistened wooden tongue blade) is hemispherical. This permits controlled tissue vaporization with minimal lateral heat conduction. The higher the average power setting on the laser, the larger the spot size or impact crater in a Y-beam geometry. For transformation zone ablation, this affords the advantage of minimizing thermal conduction and associated tissue damage. The laser energy is delivered in short bursts either by use of the mechanical timer in the laser console or, preferably, by gating the laser pulses using the foot pedal. This permits precise, controlled tissue vaporization.
CO2 laser ablation of the transformation zone should always be performed using a micromanipulator attached to a colposcope or operating microscope. Careful colposcopy is required at the time of transformation zone ablation to determine the lateral extent of disease. The entire transformation zone must be treated, as selective ablation of areas of disease results in much lower primary cure rates.
A major advantage of CO2 laser ablation is the ability to destroy tissue to a precisely controlled depth. The maximum depth of gland involvement with CIN is 5.2 mm, whereas the maximum depth of uninvolved glands is 7.9 mm. Frequency and depth of gland crypt involvement increase with the grade of CIN. The transformation zone is usually destroyed to a depth of 7 to 10 mm.
Control of Intraoperative Pain and Bleeding
CO2 laser ablation of the transformation zone is usually performed under local anesthesia. The cervix is infiltrated with 4 to 6 mL of local anesthetic such as 1% lidocaine, with or without a vasospastic agent. The infiltration is best performed as a slow, subepithelial infiltrate using a dental syringe with a 27-gauge dental needle. A deeper, paracervical block does not always provide adequate anesthesia and its administration is associated with more discomfort and bleeding.
Laser ablation is usually associated with minimal bleeding. Should a small vessel such as an arteriole be encountered, hemostasis is readily achieved by using direct pressure from a moistened cotton-tipped applicator and lasing immediately onto the applicator tip at the site of the vessel. Monsel's solution is applied to the cervix on completion of the procedure to minimize the risk of postoperative bleeding. CO2 laser ablation of the transformation zone is an excellent treatment for selected patients with CIN, achieving primary cure rates of up to 95%, with minimal morbidity (395).
Sequelae of Conservative Treatment Procedures
Patients can expect a vaginal discharge for up to 3 weeks after the procedure. Infection is rare, but persistence of an offensive discharge or development of postoperative pelvic pain warrants assessment. Minor spotting may occur in the first two postoperative weeks but usually settles promptly. If bleeding is heavier and does not settle quickly, the patient should be examined and hemostasis secured using Monsel's solution. Rarely, sutures may be required to secure hemostasis with secondary bleeding. The patient should refrain from tampon use, douching, and vaginal intercourse for 3 to 4 weeks after surgery.
Repeat Pap smears and colposcopy should be performed at 6 and 12 months posttreatment. If these assessments are normal, the patient may return to annual screening, particularly if a highrisk HPV DNA test is negative. A significant proportion of patients continue to show minor abnormalities on cervical smears in the first 12 months after treatment, reflecting reparative changes or continued expression of minimally developed HPV-induced changes. These patients rarely require further treatment.
Excisional Cervical Conization
Excisional conization performed with a scalpel, sometimes referred to as a “cold-knife conization,” was traditionally the standard response to cytologic abnormalities (413) and remains an important therapeutic option in the management of CIN. It is both diagnostic and therapeutic. The geometry of the conization should adapt to the size and shape of the lesion as well as the geometry of the cervix (Fig. 8.16). The procedure is performed in the following manner:
Cervical conization achieves cure rates for high-grade CIN in excess of 95%. The risk of cervical stenosis and cervical incompetence is higher for cervical conization performed with a scalpel than for CO2 laser and electrosurgical excisional conization. This in part reflects the fact that cervical conization performed with a scalpel has been traditionally used for the most severe lesions, when invasive cancer has not been excluded or when colposcopy has been unsatisfactory, often with significant disease extension to within the endocervical canal (414).
Hysterectomy is rarely indicated in the primary management of CIN. The most common indication is coexistence of a gynecologic condition that warrants hysterectomy. Such conditions include dysfunctional uterine bleeding, fibroids, uterovaginal prolapse, or patient request for sterilization.
Before any hysterectomy, colposcopic assessment is important. If the entire lesion and transformation zone is not seen, if there is any cytologic, colposcopic, or histologic suspicion of invasive cancer, if an endocervical specimen is positive for high-grade neoplasia, or if there is any evidence of high-grade glandular neoplasia, an excisional conization must be performed to exclude invasive cancer before hysterectomy is performed.
Figure 8.16 Tissue excised for cervical conization procedures depending on the extent of disease and the anatomy and shape of the cervix.
In 2% to 3% of patients with high-grade CIN, the disease extends to the vaginal vault (415). If the vaginal cuff is not carefully fashioned in these patients, preferably using a vaginal approach, neoplastic epithelium may be sutured into the vaginal vault. High-grade vaginal intraepithelial neoplasia (VAIN) occurs in the vaginal vault in 1% to 7% of patients who have undergone hysterectomy to treat CIN. Coppleson and Reid (416) reported 38 cases of invasive cancer occurring in the vaginal vault after hysterectomy among 8,998 women (0.4%).
If hysterectomy is performed for the management of CIN, the patient should have vault cytologic testing and colposcopy performed on two occasions in the 18 months after surgery.She should be screened by vaginal vault smears on an annual basis thereafter.
Cervical Adenocarcinoma In Situ
The reported incidence of cervical glandular neoplasia has increased, predominantly among nulliparous and primiparous women in the reproductive age groups with up to 30% of cases occurring in women younger than 35 years of age (417,418,419,420,421).
These changes in the clinical profile of cervical cancer have focused much attention toward AIS. There is convincing evidence that AIS is a precursor lesion (422,423,424,425,426).The mean age of diagnosis of AIS is 15 years younger than that for invasive adenocarcinoma (427,428). AIS frequently coexists with invasive adenocarcinoma in histologic specimens.Patients who have a cone biopsy performed in response to cytologic evidence of AIS already have invasive cancer in up to one-third of cases. Women diagnosed with cervical adenocarcinoma frequently have had previous cytologic evidence of endocervical atypia on smears for intervals of 2 to 10 years.
Specific HPV types, in particular HPV 18, are strongly implicated in the etiology of highgrade glandular neoplasia (131). Persistent high-risk HPV infection, particularly with HPV 18, is the greatest risk factor for developing AIS. Prolonged oral contraceptive usage, beyond 5 years, may be a cofactor in the development of glandular neoplasia, particularly in young women (131,429). Widespread vaccination against HPV 16 and 18 which are associated with 90% of glandular neoplasms, should significantly decrease the incidence of both AIS and adenocarcinoma (15,212,213).
The relationship between AIS and lesser degrees of cervical glandular neoplasia is more controversial. No prospective study of glandular dysplasia has been undertaken, and the neoplastic potential of these lesions remains uncertain. Glandular dysplasia, less than adenocarcinoma in situ, represents a heterogeneous group of lesions with variable progressive potential. Glandular dysplasia is much less predictably associated with high-risk HPV types when compared to AIS, further confusing understanding of the significance of such lesions.
Adenocarcinoma in situ is usually diagnosed after an abnormal Pap test result. The abnormal smear may predict the presence of high-grade glandular disease. Only 38% to 69% of AIS cases are detected by cytology prior to conization. The diagnosis increases to 85% following colposcopy, biopsy, and endocervical sampling. Because AIS coexists with high-grade squamous CIN in 50% of cases, the abnormal smear will frequently predict only the squamous lesion (430). The diagnosis of AIS is often a coincidental finding in the histologic assessment of an excised specimen taken in the management of high-grade CIN, which is a compelling argument for the routine excision of high-grade CIN.
The 2001 Bethesda system includes a category for atypical glandular cells (AGC). Patients with AGC smear reports have a 30% to 50% risk of having high-grade cervical disease and are at much higher risk of significant disease than those with ASC smear reports (431,432,433,434,435). An AGC smear report is an indication for referral for colposcopy and careful endocervical assessment. The underlying lesion is most frequently high-grade squamous CIN, which occurs in up to 25% of patients. AIS, cervical adenocarcinoma, and endometrial disease, including hyperplasia and carcinoma, occur in up to 20% of patients (431,432,433,434,435).
The colposcopic features of AIS and early adenocarcinoma are widely seen as nonspecific. A minority view is that most high-grade glandular lesions do have specific colposcopic features. Discrete or extensive stark acetowhitening of individual or fused columnar villi may be seen surrounded by normal villiform structures (Fig. 8.17). Prominent atypical vessels may also be seen, particularly in association with early invasion. Although colposcopy should be performed in response to cytologic or clinical suspicion of glandular neoplasia, excisional conization is mandatory for definitive diagnosis.
The potential for AIS to involve the entire endocervical canal and the frequent association with invasive carcinoma demands formal excisional conization in the management of AIS (436). AIS is often unifocal and is located at the transformation zone (437). The cone biopsy should be fashioned as a cylinder of at least 2.5 cm in depth and be performed with a cold knife to avoid thermal injury to the specimen. Care must be taken in particular with excision of the apex of the cone to avoid traumatic or thermal distortion. If the conization margins are clear of disease, more than 80% of patients have negative cytologic and colposcopic follow-up beyond 12 months from treatment (438,439,440,441). However, even when excision margins are clear, 6% to 25% of women who proceed to hysterectomy are reported to have persistent disease in the specimen (442). For this reason, hysterectomy is still considered the gold standard for women with AIS who do not desire fertility preservation (443,444).
Younger women may be managed by excisional conization alone if margins of excision are clear. Careful cytologic and colposcopic follow-up is important, with recurrent AIS being detected in 7% to 10% and squamous and glandular preinvasive disease in as many as 33% of such patients (445,446). Invasive cancer has been reported as late as 7 years postconization (436,446,447,448,449).
If the margins of excision are positive, more than 50% of patients have residual disease at hysterectomy (443,446). Patient with positive conization margins and/or a positive endocervical curettage require repeat excisional conization with further endocervical sampling to exclude invasive cancer. If the repeat cone biopsy is negative for invasive cancer, hysterectomy is indicated in the older patient and should be seriously considered in the younger patient.
Figure 8.17 Colposcopic appearance of adenocarcinoma in situ lesion showing prominent atypical vessels.
In modern clinical practice, up to 50% of patients with AIS are diagnosed following the histologic assessment of a loop excision specimen. The appropriate management of such cases is controversial. The standard approach has been to proceed to cold-knife conization regardless of margin status. Recent studies have reported follow-up of these women if the margins are clear (440,450,451). A recent metaanalysis comparing cold-knife conization and loop excision showed no difference in the probability of obtaining negative margins, and no difference in the probability of residual disease or subsequent development of cervical adenocarcinoma (452). Further prospective studies are required to determine the safety of conservative management of young women desiring to preserve fertility, who have had AIS diagnosed following a loop excision procedure. HPV DNA testing is a valuable test of cure following conservative treatment of cervical AIS (453).
Classification of Vaginal Intraepithelial Neoplasia
Vaginal intraepithelial neoplasia is classified similarly to cervical lesions: VAIN 1 (mild dysplasia), VAIN 2 (moderate dysplasia), and VAIN 3 (severe dysplasia/carcinoma in situ). VAIN 3 is a premalignant lesion, but the natural history of the lesser degrees of VAIN has not been submitted to prospective study. VAIN 1 is an HPV-induced change without an established progressive potential but is associated with high-risk HPV types in 64% to 84% of cases (454). Management should be conservative, usually by observation. Most of high-grade VAIN and vaginal cancer is caused by high-risk HPV types.
Since the 1970s, the diagnosis of high-grade VAIN has been made with increasing frequency. The mean age at diagnosis has decreased to 35 years, but patients with high-grade VAIN are older on average than those diagnosed with high-grade CIN. The increased rate of diagnosis of high-grade VAIN is due to increased clinical awareness, improved screening, and an absolute increase in incidence. The rarity of primary vaginal squamous cancer, accounting for 1% to 2% of female genital tract cancers, suggests that the malignant potential of VAIN is low, but progression to invasive cancer does occur (455).
High-grade VAIN usually occurs in association with high-grade CIN, which extends onto the vaginal fornices in approximately 3% of cases, but primary foci of high-grade VAIN also occur (456,457). VAIN 2-3 involves the upper third of the vagina in more than 70% of cases and less commonly the lower third, with the middle third infrequently involved. Occasionally, multifocal disease can extend throughout the vagina, particularly in the presence of extensive multicentric intraepithelial neoplasia. This reflects the “field effect” of squamous carcinogenesis in the lower genital tract related to HPV 16 in particular (458). This is further supported by the more rapid development of VAIN among women with a prior history of anogenital neoplasia. A prior history of treatment for cervical cancer and cigarette smoking are major risk factors for development of high grade VAIN (455,459), as isimmunosuppression, e.g., following organ transplantation or due to concurrent HIV infection (455).
High-grade VAIN lesions are asymptomatic and are usually diagnosed by colposcopy and biopsy following abnormal cytologic screening (460). Among women who have not had a hysterectomy, concomitant or antecedent high-grade CIN is detected in over two-thirds of patients with high-grade VAIN. Cervical cytologic testing is therefore usually positive in the presence of VAIN. The vaginal vault, in particular, and the vaginal walls should be inspected at the time of colposcopy. In addition, certain specific indications require careful vaginal colposcopy (Table 8.6).
Patients with VAIN have a 10% incidence of coexisting high-grade VIN. Careful vulvar colposcopy should be performed for all women with VAIN.
High-grade VAIN is generally diagnosed by histology of colposcopically directed biopsies. Lesions are usually flat and inconspicuous before application of acetic acid, although occasionally raised pink, red, or white lesions may be seen. Clinically apparent hyperkeratosis or leukoplakia may represent an underlying VAIN lesion. Peeling or ulceration of the vaginal epithelium, particularly in the perimenopausal and postmenopausal patient, may be an indicator of underlying high-grade VAIN. Occasionally, recalcitrant condylomatous lesions of the vagina reveal a high-grade dysplastic morphology.
VAIN 2 to 3 has a colposcopic appearance similar to that of high-grade CIN (Figs. 8.18, 8.19) after application of 5% acetic acid (461). The reaction takes longer to develop than for CIN, and the rugosity of the vagina further impairs detection. Vascular patterns are usually indistinct or absent. A fine capillary punctation is often seen with high-grade VAIN as the acetic acid reaction fades. Prominent abnormal vascular patterns develop late in the neoplastic process. Widely spaced, varicose punctation and, less frequently, mosaicism occurring in an area of high-grade VAIN are highly suspicious for invasive cancer.
Table 8.6 Indications for Vaginal Colposcopy
Figure 8.18 Colposcopic appearance of high-grade vaginal intraepithelial neoplasia with acetic acid.
Figure 8.19 Colposcopic appearance after staining with Lugol's Iodine.
The ability reliably to predict the probable histologic status of vaginal colposcopic lesions is a challenge for the most experienced colposcopist. A lesion may appear inconspicuous and trivial but reveal high-grade dysplasia on biopsy. Examination under general or regional anesthesia may be required, particularly in the presence of extensive disease, to permit accurate diagnosis.
The difficulty in colposcopic assessment of the vagina renders examination after application of aqueous iodine solution invaluable. Poorly differentiated vaginal epithelium is unglycogenated and rejects iodine staining. High-grade VAIN lesions appear mustard yellow against the mahogany-brown staining of normal surrounding mucosa. This assists in the mapping of significant lesions and in obtaining accurate biopsies. The application of aqueous iodine is mandatory for delineation of treatment margins.
Treatment of High-Grade Vaginal Intraepithelial Neoplasia
Vaginal intraepithelial neoplasia can be very difficult to treat, particularly in the presence of extensive, multifocal disease or when the vaginal vault is involved posthysterectomy.Surgical excision, often requiring partial vaginectomy, or vaginal irradiation were historically used as the main treatment modalities. Significant morbidity is associated with both approaches. The CO2 laser is regarded as the treatment of choice for most VAIN cases (462,463). The vaginal wall is relatively thin compared with other genital tract sites, with vital organs in close proximity. Surgical access is, at times, difficult. The CO2 laser provides the surgeon with the ability to treat to a precisely controlled depth and achieve very high cure rates for selected patients (464).
Topical 5-fluorouracil (5-FU) cream can also be used with good effect for carefully selected patients (465,466). 5-FU cream produces chemoinflammation and chemoulceration that often adequately treats VAIN lesions. Care is required to protect the vulvar skin and to avoid persistent denudation of the vaginal mucosa, particularly in the posterior fornix.
Application of 5% imiquimod cream may be considered as an alternative treatment for high-grade VAIN where excision is not indicated (467,468,469). Although the data are limited,imiquimod cream has been demonstrated to achieve high clearance and response rates for VAIN. Recurrence rates are no greater than for other topical treatments and it appears to be safe and well tolerated by most patients. Imiquimod has been used for management of extensive, multifocal VAIN lesions, both as primary treatment, as well as with the aim of decreasing the extent of disease before ablation (469). Conservative ablative therapy requires expert colposcopy, liberal use of directed biopsies, and no cytologic, colposcopic, or histologic evidence of invasive cancer.
Carbon dioxide laser treatment for high-grade VAIN is best performed using a highpowered superpulse or ultrapulse laser. The beam is defocused to an appropriate beam geometry (see vulvar section) and controlled by a micromanipulator attached to a colposcope or operating microscope. The vaginal mucosa is destroyed to the depth of the lamina propria, which is at most 2 to 3 mm. Because the vaginal mucosa contains no gland crypts or skin appendages, superficial treatment only is required. Conservatism is important because delayed healing and scarring occur after overenthusiastic destruction of vaginal mucosa.
Treatment of high-grade VAIN in the vaginal vault represents a particular surgical challenge. Woodman et al. (470) reported results of vaginal vault laser surgery for VAIN following hysterectomy in 23 patients who were followed for a mean period of 30 months. Only six patients remained disease free, and invasive cancer developed in two patients. Hoffman et al. (471) reported 32 patients who underwent upper vaginectomy for VAIN 3. Occult invasive cancer was found in nine patients (28%). This very difficult problem is increasingly viewed as an indication for surgical excision, although CO2 laser ablation may have a role if the patient is young and reliable.
Vulva and Perianal Area
Since approximately 1970, there has been a marked increase in the incidence of highgrade preinvasive vulvar disease and a decrease in the modal age of diagnosis (473). There has been a much smaller increase in the incidence of invasive vulvar cancer, presumably because the preinvasive disease is actively treated (474,475,476). Although more than 95% of cervical malignancies (15) and up to 90% of high grade VIN lesions are HPV-associated (454), HPV DNA is detected only in approximately 50% of vulvar cancers (477). Many of the HPV-negative cancers, particularly in older women, are associated with lichen sclerosus (478,479,480,481,482,483,484).
Preinvasive neoplasia of the vulva has been recognized for more than 75 years, but the descriptive terminology has been confusing. Vulvar carcinoma in situ has been described as Bowen's disease, erythroplasia of Queyrat, carcinoma in situ simplex, bowenoid papulosis, kraurosis vulvae, and leukoplakia. This confusion was compounded by the use of similar terms to describe a group of nonneoplastic vulvar diseases to which Jeffcoate (485) in 1966 assigned the term chronic vulvar dystrophy. In 1986, the International Society for the Study of Vulvar Disease (ISSVD) agreed on a new classification of vulvar epithelial disorders (Table 8.7) (486).
Although this classification represented a significant advance in rationalizing previously confusing terminology, significant shortcomings existed. The vulvar intraepithelial neoplasia (VIN) terminology was introduced for uniformity and consistency with the grade classification for CIN. Although this seemed logical, there existed an established biologic continuum from CIN 1 to CIN 3. The neoplastic biologic continuum from VIN 1 through VIN 3 to invasive cancer has not been established (454).
Although the progression rate of VIN 3 to invasive cancer remains controversial, the malignant potential is undisputed (487,488,489,490,491,492). By contrast, there is no direct evidence that VIN 1 has any malignant potential. The detection of high risk HPV-DNA types in 30% to 40% of VIN 1 lesions, a considerably higher rate than in condylomata acuminata, has been advanced as an argument for retaining low-grade lesions within the VIN diagnosis (454), but the inclusion of such lesions in the neoplastic continuum creates pressure for a more aggressive therapeutic approach and there is a compelling argument for excluding low-grade VIN from the “intraepithelial neoplasia” category (493). When mild squamous atypia is seen in vulvar skin, usually limited to the lower epidermis, the lesion is more likely to be nonneoplastic reactive atypia. Considerable inter- and intraobserver variation occurs with the VIN 1 diagnosis and this diagnostic category is not reproducible (494).
Table 8.7 Classification of Epithelial Vulvar Disorders
Most histologic VIN lesions are categorized as VIN 2-3 and good histologic agreement is obtained when VIN 2-3 are combined as a single high-grade VIN diagnosis (494). Careful histologic and molecular review in the 1990s, particularly by Kurman and associates, led to a reclassification of VIN 3 into three histologic subtypes, namely basaloid, warty (or bowenoid) and differentiated (or carcinoma simplex) (495,496). Further clinical correlation refined the high-grade VIN diagnosis to include two distinct lesions which have different morphology, biology and clinical features (494).
VIN, usual type, is seen adjacent to 30% of both warty (condylomatous) and basaloid types of invasive squamous vulvar cancers. It is mostly high-risk HPV related, as are the associated cancers. It has a definite invasive potential, particularly in women over 30 years of age. A variant of VIN, usual type, is the multifocal, pigmented, papular lesion seen in younger women, often of non-European background, associated with genital warts and sometimes seen in pregnancy. These lesions can regress spontaneously and although there is a definite potential to progress, close prospective follow-up is justified (495,496).
The less-common VIN, differentiated type, is seen in older women, often adjacent to invasive keratinizing squamous cell carcinoma of the vulva (494,497,498). It may occur with chronic vulvar dermatoses, particularly lichen sclerosus, but also with squamous cell hyperplasia, lichen simplex chronicus and erosive lichen planus. VIN, differentiated type, is not HPV associated nor is the associated keratinizing squamous cell cancer. Clinically, these lesions are difficult to distinguish against a dystrophic background. A keratotic nodule or shallow ulcer may be the only clinical indicator.
In 2004, the ISSVD recommended the following modifications to the terminology for squamous vulvar intraepithelial neoplasia (494):
o The more common VIN, usual type encompassing VIN 2, VIN 3 and the older clinical and histologic terms: Bowen's disease, bowenoid papulosis, dysplasia, and carcinoma in situ. These lesions are associated with high-risk HPV types, particularly HPV 16. VIN, usual type, is subcategorized histologically as warty(condylomatous), basaloid, or mixed.
o The less-common VIN, differentiated type. These lesions are not associated with HPV but frequently occur against a background of a vulvar dermatosis, particularly lichen sclerosus.
The 2004 ISSVD classification of VIN is shown in Table 8.7.
Paget's disease of the vulva is an uncommon intraepithelial lesion. It is sometimes associated with underlying invasive carcinoma. These conditions are discussed in Chapter 13.
Clinical Profile of High-Grade Vulvar Intraepithelial Neoplasia
The increased incidence of VIN 3 in recent decades reflects increased clinical awareness, improved diagnostic accuracy, and an absolute increase in disease incidence. Specific genital HPV types, in particular HPV 16, are strongly implicated in the causation of high-grade VIN (454,499,500,501,502). Other vulvar HPV-induced lesions, including condylomata acuminata and subclinical HPV infection, frequently either coexist with or predate the diagnosis of VIN 3. Cigarette smoking, nutritional deficiency, poor personal hygiene, granulomatous vulvar diseases, endogenous and exogenous systemic immune suppression including HIV infection, previous radiation therapy, and pregnancy have been implicated as cofactors in the pathogenesis of VIN 3 (474,489). There is a strong association between VIN 3 and sexually transmitted disease, with rates varying from 20% to 60%.
High-grade VIN lesions tend to be localized and unifocal in the older patient. A higher malignant potential is presumed for such lesions because invasive vulvar cancer occurs predominantly in the older age groups. However, many of the invasive cancers in elderly women occur against a background of lichen sclerosis and without a prior history of VIN 3 or coexisting histologic evidence of VIN 3.
In younger patients, high-grade VIN lesions are frequently multifocal and extensive. Lesions may remain discrete or coalesce to develop a large field of disease. Lesions may extend laterally from the inner aspect of the mucous membranes of the labia minora to the hair-bearing skin of the labia majora and from the clitoris, periclitoral area, and mons pubis anteriorly to the perineum and perianal area posteriorly. Difficult-to-access sanctuary sites, such as the urethra, clitoris, vagina, and anal canal, need to be carefully inspected.
More than 30% of women with VIN 3 experience vulvar symptomatology. The most common symptoms are pruritus, burning, pain, and dysuria (503). Vulvar symptoms are often exacerbated by voiding. Patients may present reporting a localized lump or thickening in the vulvar skin, or they may notice an area of increased or decreased pigmentation. The patient may present with a history of recalcitrant vulvar condylomata acuminata.
Delay in diagnosis of high-grade VIN, even in symptomatic patients, is common (473,489,503). Opportunistic inspection of the vulva, particularly at the time of colposcopy for abnormal cervical cytology, is recommended.
The clinical appearance of VIN 3 lesions varies according to patient age and skin color, as well as the location of the lesions in the vulva and perianal region (Figs. 8.20, 8.21). In both the hairbearing and non-hair-bearing keratinized vulvar skin, lesions tend to be raised or papular. They may be white, red, or brown in color. White lesions are due to hyperkeratosis or dehydration of the outer keratinized layer. Red lesions result from increased vascularity, reflecting either an inflammatory response or increased blood vessel formation secondary to angiogenic factors of neoplasia. Brown or pigmented lesions, which occur in more than 10% of patients, result from melanin incontinence, usually in the keratinized squamous epithelium. On the mucosal surfaces and less frequently on the keratinized surfaces, VIN 3 lesions may be flat or macular. Occasionally, such macular lesions are evident through associated erythema or pigmentation. Usually, macular lesions are subclinical and are detected on colposcopic examination after application of 5% acetic acid solution.
The clinical appearance of VIN 3 in dark-skinned women is similar when detected on mucosal surfaces but may differ in keratinized and hair-bearing areas. Relative hypopigmentation may occur, producing pink or erythematous plaques. Such lesions may blanch densely acetowhite after application of acetic acid solution. Unifocal, localized lesions in older women less frequently involve the mucous membranes. Care must be taken in the assessment of suspicious vulvar lesions in older women because of the increased risk of undisclosed invasive cancer. Warning signs of an occult invasive lesion include yellow discoloration, nodularity, ulceration, thick scale, and abnormal vascularity.
Vulvar intraepithelial neoplasia grade 3 is often found on biopsy of recalcitrant and abnormal appearing condylomata acuminata. VIN 3 is reported in biopsies from 30% of patients with large, persistent condylomatous lesions, particularly if the lesions are pigmented or coalescent and sessile with a micropapilliferous surface. Condylomatous lesions exhibiting a severely dysplastic morphology on biopsy frequently harbor high-risk HPV types, with HPV 16 and 18 detected in more than 70% of such lesions (500).
Colposcopy is now an accepted standard in the diagnostic assessment of preinvasive vulvar disease. After application of 5% acetic acid solution and colposcopic assessment using a magnification of at least 7, lesions appear as clearly demarcated, dense acetowhite areas. The multifocal distribution is usually evident. The acetic acid reaction is best seen in lesions that are nonpigmented or red. Pigmented lesions often develop an acetowhite hue or a rim of acetowhitening. Initial clinical examination may identify clinically apparent lesions.
Figure 8.20 Clinical appearance of vulvar intraepithelial neoplasia showing hyperkeratotic papular lesions.
Figure 8.21 Left: Multifocal VIN 3 lesion with multiple hyperpigmented and hyperkeratotic lesions. Right: Multifocal VIN 3 with confluent hyperpigmented areas on the vulva extending to the perineum and perianal areas.
Colposcopy may permit identification of previously unidentified, subclinical lesions and better define the distribution of clinically evident disease.
In high-grade vulvar preinvasive lesions, vascular patterns are often inconspicuous or absent, particularly in the presence of hyperkeratosis. Macular lesions on the mucous membranes may reveal a capillary punctation pattern, and a fine punctation is sometimes observed in papular lesions. Marked vascular abnormalities characterized by a varicose, widely spaced punctation and, rarely, mosaicism represent a definite warning sign of invasive cancer, and the lesion must be excised. Colposcopic warning signs of vulvar cancer occur late in the neoplastic process, limiting the sensitivity of colposcopy for the identification of early invasive cancer. Histologic evidence of VIN 3 may be seen outside colposcopically identified margins of disease, particularly laterally in the hair-bearing areas.
Diagnosis ultimately depends on liberal use of directed biopsy. This is particularly the case if ablative treatment is being considered, either alone or in combination with excisional procedures. Biopsies are best taken with a Keyes biopsy instrument under local anesthesia in the office setting.
Natural History of High-Grade Vulvar Intraepithelial Neoplasia
Vulvar intraepithelial neoplasia grade 3 coexists with invasive cancer in 30% to 50% of cases. Vulvar dystrophy occurs in up to 50% of specimens, with lichen sclerosus and squamous hyperplasia equally represented. There is no coexistent disease in 10% to 15% of specimens (473,489,503).
Few studies have examined the natural history of untreated VIN. Jones and Rowan from New Zealand (488) reported in 1994 on the follow-up of 113 women with VIN 3 diagnosed between 1961 and 1993. Of 105 women whose disease was treated, 4 (3.8%) developed invasive cancer 7 to 18 years after treatment. Of eight untreated cases of VIN 3, progression to invasive cancer was reported in seven patients (87.5%) within 8 years, and the disease regressed spontaneously in the remaining patient. This very high incidence of progression within a reasonably short time is troubling. More recent studies, with much shorter follow-up of untreated women, suggest a lower progression rate (487,491).
A more recent study from New Zealand reported women who experienced spontaneous regression of VIN 2-3 (495). These women had a median age of 19 years, an initial presentation through a sexual health clinic, and a previous history of condylomata acuminata. Most had multifocal, pigmented lesions. Median time to regression was 9.5 months.
The occurrence, usually in younger women, of multifocal, pigmented, papular vulvar lesions reported histologically as VIN 3 is well recognized and has been described as “bowenoid papulosis” (504). Reports of spontaneous regression, especially associated with pregnancy, suggested distinctive epidemiologic features for bowenoid papulosis, but the term has been abandoned by the International Society for the Study of Vulvar Disease and the International Society for Gynecologic Pathologists. High-grade VIN is a disease with a varied and individual clinical profile and histologic appearance. This range encompasses the entity previously described as “bowenoid papulosis.”
Treatment of High-Grade Vulvar Intraepithelial Neoplasia
Treatment is aimed at control of symptoms and prevention of progression to invasive cancer. Many treatment modalities have been used and, historically, vulvar carcinoma in situwas managed by simple vulvectomy (505). Such a radical approach is unjustified and is associated with significant morbidity, particularly for young women, including scarring, dyspareunia, urinary stream difficulties, loss of elasticity for vaginal delivery, and a “castration-like” self-image.
Since the 1970s, there has been a trend toward more conservative therapy, initially using excisional approaches and more recently, ablative modalities (506,507).
The risk of occult malignancy occurring in association with VIN 3 is too low to mandate complete excision of disease in all patients but too high to allow routine ablation. Women undergoing excisional treatment for VIN 3 are reported to have a 15% to 23% incidence of unsuspected invasive squamous cell carcinoma on histologic examination of the excised specimen(s) (503,508,509).
The clinical profile of VIN, including a broad age range and marked variability in extent, distribution, and symptomatology, demands individualization of the therapeutic approach for each patient. A period of close prospective follow-up without treatment may be appropriate for young, immunocompetent women with multifocal disease, particularly if they are pregnant. The patient must comply with close follow-up and understand and accept the risks of treatment delay.
Wide Local Excision, Skin Flap Procedures, and Superficial (Skinning) Vulvectomy
Localized high-grade VIN lesions are best managed by local, superficial excision. The lesion should be excised with a disease-free margin of at least 5 mm. Wide, local excision is ideal for unifocal and lateral lesions or for hemorrhoids involved with high-grade intraepithelial neoplasia. It is mandatory if a lesion has warning signs of possible invasive cancer.Primary closure of the defect usually achieves uncomplicated healing and a very satisfactory cosmetic and functional outcome. The elasticity of the vulvar skin permits preservation of sexual and reproductive functions, of particular importance in the young patient.
The surgical specimen should be submitted to careful histologic evaluation to exclude invasive disease and to ensure clear margins of excision. Surgical excision with disease-free surgical margins achieves a 90% cure rate for localized disease. If the margins of excision are involved with disease, the cure rate falls to 50%, demanding very close follow-up. As long as all macroscopic disease has been removed, reexcision is not justified for positive margins. Most recurrences occur within three years of treatment, although late recurrence and progression to cancer can occur. Development of symptoms should prompt urgent review.
Large, confluent lesions or extensive multifocal disease, particularly in the presence of colposcopic warning signs of early invasion, require more extensive excisional procedures with rotational flaps to fill the defect, or superficial (skinning) vulvectomy with a splitthickness skin graft (510,511). Primary closure of a large vulvar wound may not be possible without undue tension leading to wound breakdown or excessive scarring. Cutaneous flaps with no muscle component have been used for cases of extensive excision of VIN (512,513,514). Perineal flaps of skin and fascial tissue rotated around a perforating branch of the internal iliac artery have been used to close larger vulvo-vaginal defects (512). Thin skin flaps with less than 1cm of underlying fat can be raised from the buttock and rotated medially to close vulvar defects (513). Good postoperative healing is usually achieved without significant morbidity or long-term negative impact on sexual functioning.
“Skinning” vulvectomy was introduced by Rutledge and Sinclair (510) for extensive VIN lesions, particularly in the hair-bearing skin where the skin appendages may be involved. Lesions are carefully mapped and a shallow layer of vulvar skin is excised, preserving the subcutaneous tissues. The vulvar skin at risk is replaced with epidermis from a donor site on the inner aspect of the thigh or buttock. The clitoris is preserved, with lesions on the prepuce or glans being superficially excised or laser ablated. The epithelium regenerates without loss of sensation.
DiSaia (511) reported a 39% recurrence rate in patients with VIN 3 treated by skinning vulvectomy with split-thickness skin grafting. There were no recurrences in grafted areas, although such recurrence has been reported. Although this procedure has been largely outmoded by CO2 laser treatment for many patients with extensive disease, it remains an important therapeutic option when there is an increased risk of occult invasive cancer.
CO2 Laser Surgery
Vulvar intraepithelial neoplasia is occurring more frequently in young women, and the disease may be very extensive, involving the hair-bearing area of the labia majora in more than 30% of cases. Excision of such wide areas, even with skin grafting, can cause significant scarring and anatomic distortion. With careful, expert colposcopy and liberal use of directed biopsy, the undisclosed cancer risk in selected patients is low.
Some consider an ablative procedure in these young patients using the CO2 laser to be the treatment of choice (515,516,517,518,519). The morbidity associated with ablation of large areas of VIN 3 has been found to be unacceptable in some studies. The initial 2 weeks following more extensive laser ablative procedures will be associated with significant pain, particularly with micturition. The use of appropriate laser technology and settings, advanced surgical expertise with careful control of depth of ablation, and appropriate postoperative care will mitigate much of the potential morbidity. CO2 laser ablation is particularly useful in patients with periclitoral and perianal disease, where the preservation of anatomy and function is a substantial benefit.
Physical Principles Governing Vulvar Laser Surgery
Figure 8.22 Diagrammatic representation of CO2 laser beam geometry.
For ablative treatment of VIN, a high-power laser setting is selected. The spot diameter is progressively enlarged by defocusing the beam using the micromanipulator until a point is found where the impact crater is hemispherical. This permits controlled tissue vaporization with minimal lateral heat conduction. The laser should be first tested on a moistened tongue blade to defocus the beam to the hemispherical Y-beam geometry before use on the skin.
Surgical Strategies Governing Vulvar CO2 Laser Surgery
First surgical plane Destruction to the first surgical plane removes the surface epithelium to the level of the basement membrane. The laser beam is rapidly oscillated across the target tissue with the spot describing a series of roughly parallel lines. When the impact debris is wiped away with a moistened swab, the moist “sand-grain” appearance of the papillary dermis will be evident.
Second surgical plane Ablation to the second plane removes the epidermis and the superficial papillary dermis. This plane is achieved by a slightly slower oscillation of the laser beam across the first surgical plane, scorching but not penetrating the papillary dermis. The visual effect is a shrinking of the target tissue because of dehydration, and a finely roughened, yellowish surface similar in appearance to chamois cloth is produced. Ablation extends to the deep papillary dermis with minimal thermal injury to the underlying reticular dermis. The second surgical plane is the preferred depth of ablation for condylomata acuminata treated with the CO2 laser.
Third surgical plane Destruction to the third surgical plane removes the epidermis, papillary dermis, and superficial reticular dermis containing the upper portions of the skin appendages, specifically the pilosebaceous ducts and hair follicles. This is achieved by a slower, purposeful movement of the laser beam across the second surgical plane. The tissue is seen to relax and separate as the midreticular dermis is exposed as moistened gray-white fibers representing coarse collagen bundles. Healing occurs from the base of the skin appendages, and scarring is absent or minimal. Ablative procedures for VIN 3 should be carried to the depth of the third surgical plane (Fig. 8.23).
The skin appendages are involved with the VIN process in more than 50% of cases (517). Depth of hair follicle involvement is usually less than 1 mm but may extend to 2 mm. Measured sweat gland involvement has been more than 3 mm in depth. Beyond 3 mm, the equivalent of a third-degree thermal defect is created, resulting in delayed healing, scarring, and alopecia. The implications of residual disease after treatment of VIN are different from those of residual CIN, which may be buried and escape detection. Although the surgeon should be aware of vulvar skin appendage involvement, this is not an indication to destroy beyond the midreticular dermis.
Fourth surgical plane Destruction of the reticular dermis creates a thermal injury extending to the subcutaneous tissues and must be avoided.
Narcotic analgesia is usually required in the immediate postoperative period or, alternatively, prolonged epidural analgesia can be used. Regular sitz baths followed by topical application of a mixture of equal parts 1% lidocaine and 2% silver sulfadiazine creams to the surgical site aid in pain relief. The postoperative discomfort is often most severe on the third to the sixth postoperative days. Patients should have available appropriate oral narcotic analgesics to provide relief after discharge from hospital.
Figure 8.23 Diagrammatic representation of three surgical planes.
Regardless of treatment modality, recurrence of VIN is common. Even with modern treatment and management of VIN, invasive cancer will still develop in 3% to 5% of women, considerably higher than the risk of cervical cancer post treatment of CIN. Recurrent VIN is a significant problem and represents both incomplete primary treatment and disease recurrence. Lifelong vigilance is an important component of the management of high-grade VIN.
Immune Response Modifiers: Imiquimod
Surgery is the treatment of choice for high-grade VIN, to remove visible lesions, relieve symptoms, and prevent invasive cancer. However, the margins of excised specimens are positive for disease in 24% to 68% of cases and disease recurrence is common. Surgery does not eradicate HPV, the primary cause of most cases of VIN. The discovery that VIN 2-3 lesions, particularly multifocal lesions in younger women, are strongly HPV associated, argues for the possible efficacy of immune response modifiers in treatment of high-grade VIN (375,376).
Imiquimod (Aldara) is an imidazoquinoline, a novel synthetic compound that is a topical immune response stimulator. It enhances both the innate and acquired immune pathways, particularly the T helper cell type 1-mediated immune response, resulting in antiviral, antitumor, and immunoregulatory activities (520,521). Imiquimod causes cytokine induction in the skin, which up-regulates the host immune system to recognize the presence of a viral infection or tumor, theoretically leading to eradication of the lesion. It also stimulates activation, maturation, and migration of Langerhans cells, the major antigen-presenting cells of the skin, which are depleted by HPV infection (521,522).
A patient-applied topical 5% imiquimod cream is clinically efficacious and safe in the management of condylomata acuminata (523,524,525,526). It was licensed in 1997 for the treatment of anogenital condylomata acuminata and is recommended for this application in sexually transmitted disease guidelines from the U.S. Centers for Disease Control and Prevention, as well as guidelines from Europe, Latin America, and Australia.
The beneficial effects, patient acceptability, and low morbidity of imiquimod in the treatment of genital condylomata acuminata have led to its recent evaluation in the treatment of VIN 2-3. Case reports demonstrated efficacy against VIN 2-3, including in an immunesuppressed lung transplant patient (527). Pilot studies, applying imiquimod one to three times a week at night, have reported a 30% complete response rate and 60% partial response after 6 to 30 weeks of treatment (469,528). In contrast to surgical treatment, imiquimod focuses on the cause of many VIN cases, and preserves the anatomy and function of the vulva.Exclusion of invasive cancer is an important aspect of pretreatment assessment.
In a recent placebo controlled, randomized trial of imiquimod in the treatment of high-grade VIN, 26 women were treated with the active cream (529). Over an observation period of one year, 9 women (35%) showed complete response and 12 (46%) showed a partial response. Regression from VIN 2-3 to low grade disease was seen in 18 of 26 patients treated (69%) and 15 of these tested negative for HPV DNA after treatment. Three patients developed early invasive cancer to a depth less than 1 mm, two after placebo treatment and one after imiquimod, reinforcing the need for close follow-up. Patients reported good symptom relief and there was no reported adverse influence on health-related quality of life, body images, or sexuality. Imiquimod is well tolerated and less invasive than surgery.
No treatment modality is ideal for every woman. Treatment should be individualized according to age, distribution, severity, associated disease, and previous treatment.
Multicentric Lower Genital Tract Neoplasia
The concept of multicentricity of lower genital tract neoplasia is well established (530) reflecting the “field effect” of high-risk HPV types (531,532). Multiple primary preinvasive or invasive lesions involving the cervix, vagina, vulva, and/or anus can occur synchronously or metachronously in this region. Multicentric preinvasive disease has a higher recurrence rate after treatment than unicentric disease (533). Continued detection of high-risk HPV DNA results in a risk of recurrence of 45%. Other risk factors for recurrence include age, immunosuppression, smoking, choice of treatment modality, and positive surgical margins.
High-grade perianal intraepithelial neoplasia (PAIN) occurs in more than 30% of patients with VIN 3 or multicentric squamous neoplasia. High-grade PAIN may occur in recalcitrant perianal condylomata acuminata or as thickened, hyperkeratotic, often pigmented papular lesions usually visible to the naked eye. Proctoscopic examination using the colposcope after application of acetic acid may reveal high-grade squamous preinvasive disease extending to above the dentate line. Squamous cancer of the anus remains an uncommon disease (534), although its incidence has increased significantly in homosexual men (535,536). Viral analysis confirms a strong association with HPV 16 (537).
High-grade PAIN is managed similarly to VIN 3. Conservation of normal tissues by careful colposcopic delineation of diseased areas is important. The CO2 laser may afford some therapeutic advantage in this area because disruption of nerve fibers with full-thickness excision can lead to diminished ability to differentiate feces and flatus, leading to a degree of anal incontinence. Disease may also extend posteriorly to the anus and onto the natal cleft. Although considerable postoperative care is required for pain control and wound care, modern CO2 laser surgery is usually the treatment of choice in this difficult situation after exclusion of invasive cancer. Topical imiquimod cream may be a useful primary treatment or adjuvant in difficult to access sites.