Abstract
This chapter addresses a spectrum of neoplasms that can be found in the vagina, ranging from the more common, the squamous intraepithelial lesions, to the more obscure. Particularly problematic are the adenocarcinomas, which can be primary (originating in endometriosis), metastatic from a range of sites, or emerge in misplaced embryonic remnants (cloacogenic neoplasia). Tumors that can be particularly confusing include vaginal melanomas and spindle cell epitheliomas (mixed tumors).
Keywords
cloacogenic neoplasia, squamous cell carcinoma, vagina, vaginal intraepithelial lesions, vaginal melanoma
Introduction
This chapter focuses principally on epithelial neoplasms of the vagina (soft tissue tumors are discussed in detail in Chapter 9 ). The most common lesions in the epithelial group are squamous, ranging from mild dysplasia to invasive squamous carcinomas. There is ongoing surveillance of women with in utero exposure to diethylstilbestrol (DES) for glandular lesions, ranging from adenosis to clear cell carcinoma.
The most important challenges faced by the pathologist are the classification of precursors, recognition of potentially serious lesions, including the exclusion of invasion, and recognition of rare entities that could cause diagnostic confusion or inappropriate management.
Vaginal Squamous Intraepithelial Lesions
Terminology
In the prior edition, vaginal intraepithelial neoplasia was placed under the squamous intraepithelial lesion terminology. Another term, vaginal intraepithelial lesion (VAIL), has also been used.
The recent Lower Anogenital Squamous Terminology Standardization (LAST) Project is now in line with this. Vaginal precursors fall into two categories—low-grade squamous intraepithelial lesions (LSILs) and high-grade squamous intraepithelial lesions (HSILs), as they do throughout the lower anogenital tract. The LAST consensus recommendations from the College of American Pathologists and American Society for Colposcopy and Cervical Pathology encourage this two-tiered approach throughout the lower anogenital tract (LAT) and standardization of the nomenclature used.
The main goal is to decrease interobserver variability and provide more useful clinical data. In the vagina, LSIL comprises exophytic condylomatous lesions as well as cytologically similar flat lesions. Similarly, HSIL encompasses vaginal intraepithelial neoplasia grades 2 and 3 (VAIN 2 and 3). Needless to say, despite the efforts to compress precursors into two distinct groups, many lesions will fall in the gray zone between the two, and the distinction between LSIL (VAIN 1) and HSIL (VAIN 2) will always be problematic. The reader can take some comfort from the fact that difficulty in separating VAIN 1 and VAIN 2 is universal although of very clinical significance.
Clinical Background
A more recent reference with data from the U.S. Centers for Disease Control and Prevention’s National Program of Cancer Registries and the National Cancer Institute’s Surveillance, Epidemiology and End Results Program shows the incidence of vaginal carcinoma in situ in the United States to be approximately 0.1 cases per 100,000 women. The risk factors are identical to those for cervical and, to a lesser degree, vulvar intraepithelial lesions, specifically age at first intercourse and number of sexual partners, as well as conditions that compromise immune function. The risk of HSIL has been associated with the following:
- 1.
Patients who have undergone a prior hysterectomy for HSIL of the cervix
- 2.
Women who have had radiation therapy to the cervix or vagina
- 3.
Women who have been taking immunosuppressive agents
There are, however, some slight differences between SILs and similar lesions of the vulva and cervix, related to differences in the sites of origin and the HPV types involved. The vagina is similar to the cervix in that lesions analyzed from this area contain a wider range of HPV types than are seen in the vulva.
The vulva is a combination of skin and mucosa and is prone to infections by HPV types 6, 11, and 16. In contrast, the target epithelium in the cervix is the squamocolumnar junction and the area hosting the dynamic remodeling of epithelial phenotypes that characterizes the cervical transformation zone. The relative frequencies of LSIL and HSIL are similar to those in the cervix, approximately 3 : 1. Approximately 65% of women with VASILs have a concomitant or prior squamous intraepithelial lesion of the cervix.
Natural History
A great majority of SILs regress spontaneously. The progression of LSIL to HSIL or cancer is extremely uncommon. Immunosuppressed women are also more likely to have persistent multifocal lesions. HSILs, those associated with high-risk HPV, and those associated with squamous intraepithelial lesions (SILs) in other anogenital sites are also more likely to recur and be multifocal.
An important variable in cancer risk is the age of the patient. The average patient with SIL is between 43 and 60 years of age. This is later than cervical precancers—10 to 20 years later—in one study.
About 5% of HSILs may progress to invasion, despite close follow-up. A recent study of the rate of progression to invasive vaginal cancer and the potential risk factors were evaluated in 205 women with biopsy-proven HSIL. Of these, 12 cases progressed to vaginal squamous cell cancer (5.8%), with a mean time interval from treatment to progression of 54.6 months (range, 4–146 months). The rate of progression was significantly higher in women diagnosed with VAIN 3 compared with those diagnosed with VAIN 2 (15.4% vs. 1.4%; p < 0.0001), which again speaks to a biologic overlap between VAIN 2 and condyloma. A lower risk of progression in VAIN 2 may warrant further subclassification of HSILs in certain clinical situations, such as younger patients. Women with HSIL who had a previous hysterectomy showed a significantly higher rate of progression to invasive vaginal cancer compared to patients who had not previously undergone a hysterectomy (16.7% vs. 1.4%; p < 0.0001). However, the risk is influenced by other cofactors, including age, lack of estrogen, prior irradiation, and vaginal trauma.
Examination of the Patient
Vaginal inspection is a required component of the pelvic examination and usually entails rapid inspection of the mucosal surface following speculum placement. The indications for vaginal colposcopy are summarized in Box 12.1 .
- 1.
An abnormal Papanicolaou (Pap) smear following hysterectomy
- 2.
An abnormal Pap smear after apparently successful treatment of cervical neoplasia
- 3.
Any Pap smear unexplained by cervical colposcopy or sampling of the endocervical canal
- 4.
Any palpable or unexplained grossly visible vaginal lesion
- 5.
All women with cervical, vulvar, or perianal, or anal human papillomavirus disease
- 6.
Confirmed cervical neoplasia in an immunosuppressed patient
- 7.
Monitoring all women with a history of in utero diethylstilbestrol exposure
- 8.
Any woman with abnormal, unexplained, recalcitrant vaginal discharge or bleeding
- 9.
Before cervical conization for noncorrelating cytology, histology, and colposcopic impression
The entire vagina needs to be closely examined because the appearance of the lesions can be subtle. Special attention should be given to the upper third of the vagina, where most disease is present. In DES-exposed women, glandular lesions are most common in the anterior upper third of the vagina.
Acetic acid and Lugol’s solution can be used in colposcopic evaluation of the vagina. Acetic acid will highlight hyperkeratosis, which is often associated with LSILs, and can identify areas of punctation that may be associated with HSILs ( Fig. 12.1 ).
Lugol’s solution should not be used in patients with an allergy to iodine. It stains glycogen-containing tissue a dark mahogany brown color, whereas nonglycogenated epithelium, specifically intraepithelial neoplasia, does not stain, typically appearing mustard yellow in color ( Fig. 12.2 ). HSIL, however, may appear simply as flat white epithelium. Thus, any potential lesion should be biopsied. Abnormal vascular patterns can be seen with colposcopy and are associated with HSIL. Well-developed vascular patterns of punctuation or mosaicism and atypical irregularly branched vessels are highly suspicious for invasive cancer. The chief limitation of the use of Lugol’s solution is lack of specificity—that is, atrophic or inflammatory conditions of the vagina accompanied by immature squamous epithelium (and lacking glycogen), which will also not stain. Thus, the use of Lugol’s solution is not as helpful for women with atrophy or inflammation of the vagina. The yield, however, can be improved considerably with a short course of intravaginal estrogen therapy, when clinically indicated, before colposcopy in postmenopausal women. Furthermore, perioperative use of intravaginal estrogen may directly facilitate healing after vaginal surgery.
In the examination of DES-exposed women, adenosis appears as red granular tissue. Atypical vascular patterns are also often seen early in glandular lesions. If indurations or nodules are palpated, these should be sampled to rule out carcinoma, particularly clear cell carcinoma (CCA; see later in this chapter, “Clear Cell Carcinoma”).
Cytology specimens should be obtained from the cervix by cytobrush and spatula and four quadrants (circumference) of the upper third of the vagina (by spatula). All the specimens can be put in one container.
The lesions are similar to their counterparts in the cervix and fall into a spectrum ranging from LSIL, exophytic, or flat condyloma. HSIL corresponds to VAIN 2 or VAIN 3.
Low-Grade Squamous Intraepithelial Lesions
Diagnostic Criteria
The first category of LSIL includes exophytic condylomas, which are commonly associated with HPV types 6 and 11. These lesions are discussed in Chapters 6 and 13 and consist of exophytic verrucopapillary lesions with superficial cell nuclear enlargement and koilocytotic atypia. They may be subtle histologically; the primary features will be increased nuclear density in the upper epithelial layers and mild anisokaryosis ( Fig. 12.3 ).
The second category is the flat condyloma, which has a similar distribution of cellular atypia but lacks the exophytic architecture ( Fig. 12.4 ). These may or may not be p16-positive, depending on the HPV present. Central to the diagnosis of LSIL is the absence of conspicuous nuclear atypia in the lower epithelial cell layers.
Differential Diagnosis
There are two benign entities most likely to be confused with LSIL. The first is nonspecific acanthotic and hyperkeratotic changes associated with prolapse ( Fig. 12.5A ). The other is small, fibroepithelium-l–like, polyp-like excrescences in the introitus of reproductive age women (see Chapter 11 ; see Fig. 12.5B ). Again, both of these lack cellular atypia and are no longer as likely to be as confused with HPV infections as they were in the 1980s. The diagnosis of low-grade abnormalities relies principally on morphologic criteria; immunostaining or HPV nucleic acid testing are not indicated. Exceptions are those cases in which sexual abuse is a critical variable that must be excluded.
Some lesions might be confused with HSIL. One is a LSIL with some expansion of the immature cell layers ( Fig. 12.6 ). The pathologists should focus on the nuclear features, including chromatin complexity and distribution and nuclear overlap, while considering the overall pattern of growth, polarity, and differentiation. Tangential sectioning of LSIL might also cause some confusion ( Fig. 12.7 ).
High-Grade Squamous Intraepithelial Lesions
Diagnostic Criteria
HSIL, similar to the cervix and vulva (see Chapters 6 and 13 ), displays full-thickness atypia ( Fig. 12.8 ). Lesions may be further subclassified, when clinically indicated, as VAIN 2 or VAIN 3. Care should be taken to recognize papillary architecture, which may be associated with a risk of concomitant invasion ( Fig. 12.9 ) .
Differential Diagnosis
Mimics of HSIL in the vagina include atrophic changes and/or reactive, inflammatory, epithelial changes. Atrophy is associated with an increased nuclear-to-cytoplasm ratio and an increase in cell density due to closely opposed atrophic cell nuclei, which can mimic HSIL in biopsies as well ( Fig. 12.10 ). Inflammation might retard maturation and increase the number of epithelial layers with immature cells, raising a concern for HSIL ( Fig. 12.11 ).
As noted in examination of the patient, if clinically indicated, a short 2-week course of vaginal estrogen improves colposcopic accuracy in postmenopausal patients with vaginal atrophy. In our experience, this helps the pathologic classification in that any reactive atypia due to atrophy can be significantly decreased or removed. Others have described significant decreases in cytologic atypia associated with intravaginal estrogen use, although much of this may represent treated atrophy-induced atypia.
Squamous Intraepithelial Lesions of Uncertain Grade
To best guide patient management, the pathologist is encouraged to try to categorize the lesion as within a general LSIL or HSIL category. Understanding the threshold for high-grade atypias and being aware of mimics of HSIL are essential. However, it is our experience that many VASILs fall in a gray zone between LSIL and HSIL or SIL. The reader should feel free to classify a lesion as VASIL of uncertain or intermediate grade if unsure. This is preferable to forcing a diagnosis of HSIL when the atypia is not decisive, particularly in young women.
H&E morphologic evaluation remains the gold standard. Consulting with a second pathologist is known to improve accuracy, especially in a two-tiered grading system.
Postradiation Atypia: Reactive and Neoplastic
Prior radiation therapy can be associated with benign but exuberant reactive atypia that may also mimic HSIL ( Fig. 12.12 ). Knowledge of a patient’s prior radiation therapy can alert the pathologist to look for associated benign reactive changes, such as increased nuclear size and hyperchromasia. It should be kept in mind, however, that postradiation HSIL will typically follow radiation therapy after an interval of from 1 to 10 years. These lesions are typically associated with HPV nucleic acids. There is no evidence for a separate pathway (non-HPV) for these lesions.
Biomarkers
Immunohistochemistry with p16 is the biomarker of choice for separating LSIL or HSIL from their mimics. Strong and diffuse (block-positive) p16 staining is supportive of a diagnosis of VASIL and reflective of high-risk (HR) HPV; however, HR HPV can be associated with LSIL and HSIL. One should not use p16 as a primary adjudicator to make a diagnosis of HSIL. The most concerning hazard with this practice can be overdiagnosis and overtreatment. It should be reserved for situations in which it needs to be determined whether a lesion is an HSIL or atrophic or reactive change. Staining with p16 is not usually warranted when the differential diagnosis is LSIL versus reactive change because the distinction is not critical in terms of management; HR HPV can be found in LSIL, and a positive p16 might be overinterpreted as indicating HSIL.
Proliferative markers, in particular Ki-67 (MIB-1), are prone to false-positive staining, especially in reactive or inflammatory lesions. We reserve its use for the evaluation of atrophy-related atypia. Uninflamed atrophic epithelium typically exhibits scattered basal epithelial staining.
Management
The histologic diagnosis of squamous intraepithelial lesions, age of the patient, and available tools for treating the patient influence management. In young women with low-grade lesions who have a high rate of spontaneous regression, the therapeutic option recommended is simple observation. If therapy is decided on, treatment of vaginal LSIL should not be aggressive, because the risk of complications may outweigh the risk of the disease.
Decisive treatment measures should be reserved for HSIL of the vagina. HSIL can be treated in many different ways, including topical therapies, surgical therapy, and, on rare occasions, radiation.
Topical agents include imiquimod cream and 5-fluorouracil (5-FU). The topical application of 5% imiquimod cream for the treatment of VAIN has been studied, with relatively good treatment outcomes being found. The regimen used by Haidopoulos et al. was 5% imiquimod cream applied to the vaginal lesions three times a week for 8 weeks. However, more randomized controlled trials on the use of imiquimod in VAIN are necessary. Topical 5-FU is used to treat VAIN much less frequently today than in previous years because it can be quite painful. Patients frequently complain of vaginal irritation or burning, and ulcerations may form. The application of zinc oxide cream or petroleum jelly to the vestibule can be used as a barrier to help protect against tissue damage which may occur when the medication leaks onto the vestibule and perineum. Topical trichloroacetic acid has also been recommended by some health care providers. Other treatment options include surgery via CO 2 laser or sharp excision. If ablative treatments are to be used, the lesion should be fully visible, and a previous biopsy should have been done to exclude invasion.
Treatment failure as well as complications can occur with any of these methods. Complications include extensive vaginal ulcerations and difficulty in their subsequent healing. Loop electroexcision has been used for the treatment of vaginal precancers; however, it can be risky to perform because surrounding tissue damage may occur, including a potential for bowel and bladder injury. Laser therapy is best used for extensive persisting LSIL or HSIL of the vagina ; it is particularly helpful when there is disease extending from the cervix to the superior vagina.
Surgery via sharp excision is an excellent alternative in the proper setting. In one study of posthysterectomy patients with HSIL (VAIN 3) at the vaginal apex in the region of a vaginal cuff scar, upper vaginectomy was found to be the treatment of choice, whereas multifocal HSIL or colposcopic well-defined lesions, involving large areas of vaginal mucosa, were found to be successfully managed by CO 2 laser ablation. Intracavitary radiation therapy is also an effective treatment ; however, it is rarely used because other therapies are usually successful. It is associated with higher rates of morbidity than other forms of treatment. Radiation therapy is reserved for patients who have failed previous treatments or are a poor surgical risk.
To avoid complications following surgery or radiation treatment, such as adhesions or dyspareunia, the clinician should recommend vaginal dilation following surgery if there are concerns for potential vaginal scarring. It is important to advise patients that with any treatment, high-grade VAIN can relapse and has a risk for progression to invasive cancer. Therefore, adequate follow-up examinations are mandatory in this patient population.
Malignant Epithelial Neoplasms of the Vagina
Introduction
Vaginal cancers constitute approximately 1% to 4% of all gynecologic cancers. The most common tissue type of carcinoma is squamous cell carcinoma. Adenocarcinoma in the vagina is rare and often represents a metastasis from another site. Primary vaginal adenocarcinoma can be seen arising in endometriosis. Vaginal clear cell adenocarcinoma (CCA) is associated with in utero DES exposure. These tumors will be addressed in the following discussion.
Squamous Cell Carcinoma
Introduction and Pathogenesis
Although squamous cell carcinoma (SCC) is the most common epithelial malignancy in the vagina, it accounts for less than 2% of all gynecologic malignancies. To some degree, the following characteristics of squamous cell carcinoma of the vagina serve to distinguish it from primary cervical cancer:
- 1.
Most vaginal SCCs occur in the sixth to eighth decades, which is approximately 20 years older than the mean age for squamous carcinoma of the cervix.
- 2.
Vaginal carcinomas in younger women (<50 years) are more closely associated with HPV-related (or cervical) neoplasms, being commonly linked to a site in the upper vagina, vaginal infections, and prior cervical intraepithelial neoplasia.
- 3.
Approximately 20% of patients have had a prior cervical or squamous neoplasm, either preinvasive or invasive.
Staging of Vaginal Cancer
The staging systems of the International Federation of Gynecology and Obstetrics (FIGO) and American Joint Committee on Cancer (AJCC) are similar ( Table 12.1 ). The AJCC staging includes specific details such as tumor size (T), nodal status (N), and metastases (M), which are needed for use in standardized cancer reporting, such as the College of American Pathologists (CAP) Cancer Protocols. A review of five series, which included 1375 cases of vaginal cancer, reported the following distribution of patients by FIGO stage: stage I (26%), stage II (37%), stage III (24%), and stage IV (13%).
| Extent of Carcinoma | FIGO Staging | AJCC Staging (Based on TNM) a |
|---|---|---|
| Carcinoma in situ (HSIL) | Formerly stage 0 | Stage pTis |
| Carcinoma confined to vaginal wall | Stage I | Stage I, pT1 |
| Carcinoma involves paravaginal and subvaginal tissues | Stage II | Stage II, pT2 |
| Carcinoma extends to pelvic wall | Stage III | Stage III, defined as pT3 by extent or pT1-3 with N1, regional lymph nodes |
| Invades bladder or rectal mucosa or extends directly to the pelvis | Stage IVA | Stage IVA |
| Spread to distant organs | Stage IVB | Stage IVB Any T, any N, M1, distant metastasis |
a Amin MB, Edge S, Greene FL, et al: AJCC cancer staging manual, ed 8, New York, Springer, 2017.
Two special circumstances warrant mention in regard to staging vaginal carcinomas—namely, a broad categorization of stage I disease and a potential to misinterpret stage II cervical disease as stage I vaginal disease.
In the vagina, stage I (AJCC and FIGO) includes all carcinomas confined to the vagina, is not further subdivided, as it is elsewhere, and there are no currently recognized criteria to define early, superficially invasive SCC (SISCCA). The most recent LAST guidelines summarize that this is because of the rarity of cases, with a lack of statistical and prognostic significance. Additional studies are in progress and, until further guidelines are established, in cases of suspected SISCCA of the vagina, describe the dimensions in detail, including width and depth (in millimeters) and correlate these findings with the health care provider in case a more conservative excision is indicated. Second, biopsies from the upper two-thirds of the vagina with invasive SCC in patients with concurrent invasive cervical squamous carcinoma are staged as cervical primaries and represent at least AJCC stage II of cervix, extending beyond the uterus, rather than AJCC stage I of the vagina.
Clinical Presentation
The patient with vaginal cancer usually presents with painless bleeding or a vaginal discharge that is malodorous. Urinary symptoms and pain have also been reported. A rectovaginal examination is often helpful in delineating submucosal extension, paravaginal infiltration, and rectal involvement. Tumors may be exophytic, ulcerative, annular, constricting, polypoid, sessile, indurated, or fungating. They may be located anywhere in the vagina. Approximately 40% of squamous cell carcinomas of the vagina occur on the anterior vagina, 30% on the posterior, and 28% on the lateral walls of the vagina.
Diagnostic Criteria
The histopathologic findings in vaginal squamous carcinoma are similar to those in other sites. The CAP tumor reporting protocol follows the World Health Organization (WHO) classification—keratinizing, nonkeratinizing, basaloid, and verrucous or warty. In our experience, two histologic presentations predominate. The first is a typical invasive SCC ( Fig. 12.13 ). The second is a papillary carcinoma ( Fig. 12.14 ). Papillary carcinomas in particular may not demonstrate frank invasion and are hard to discriminate from HSIL, VAIN 3, papillary type, as depicted in Fig. 12.9 . A more detailed histologic examination to confirm stromal invasion is required. A subset of these tumors is often superficial and has a history of prior cervical carcinoma, in which case a secondary implant from the cervix rather than a primary vaginal carcinoma must be considered.
