Vaginal Cancer
Primary vaginal cancer is an uncommon gynecologic malignancy and constitutes only 1% to 2% of gynecologic malignancies. Because of its rarity, there are specific guidelines for the diagnosis of primary vaginal cancer. A malignancy located in the vagina and not involving any adjacent pelvic organs is considered to be a primary vaginal cancer. If the malignancy extends to the cervix or vulva, it is considered a primary lesion of the nonvaginal site. Based on the observation that 95% of patients with recurrent cervical cancer will experience relapse within 5 years, all squamous cell carcinomas identified in the vagina within this period are defined as recurrences. Only those squamous cell carcinomas found more than 5 years after the diagnosis of cervical cancer are defined as primary vaginal cancers.1 If there is a history of endometrial cancer, a diagnosis of adenocarcinoma in the vagina is usually considered a recurrence regardless of the time from primary treatment.
Primary vaginal malignancies are most commonly located in the upper third of the vagina and on the posterior wall (Figure 10-1). The most common histo-logic type is squamous cell carcinoma, which accounts for approximately 80% of all primary vaginal malignancies. Less commonly encountered are adenocarcinomas (10%), melanomas (3%), sarcomas (3%), and other rare tumors. Eighty-four percent of malignancies identified in the vagina are secondary, most commonly originating from the cervix (32%) and endometrium (18%); less common primary sites include the vulva, gastrointestinal tract, and ovary, as well as gestational trophoblastic disease.2
FIGURE 10-1. Squamous cell carcinoma of the vagina. (Reprinted with permission of Robboy SJ and Robboy Associates LLC [Chapel Hill, NC], copyright holder of Robboy’s Pathology of the Female Reproductive Tract, Elsevier Pub, London.)
Prior to the mid-20th century, vaginal cancer was generally considered to be an incurable disease. Advances in radiation therapy and surgical techniques have been associated with significant improvement in cure rates, even in women with advanced disease.
EPIDEMIOLOGY
Key Points
1. Risk factors for vaginal cancer include human papillomavirus infection, smoking, in utero diethylstilbestrol exposure, and a prior history of cervical cancer.
2. Vaginal intraepithelial neoplasia (VAIN) is a potential precursor lesion to invasive squamous cell carcinoma of the vagina.
3. A history of pelvic radiation is a risk factor for vaginal sarcoma.
Primary vaginal cancer is a rare entity and constitutes only 1% to 2% of gynecologic malignancies. It has an incidence of approximately 1 in 100,000 women per year. The American Cancer Society estimates that there will be 2300 new vaginal cancers diagnosed in 2010 with approximately 780 deaths.3
The peak incidence of vaginal cancer is in the sixth and seventh decades. Risk factors for the disease closely mirror those for cervical cancer and include an increased number of lifetime sexual partners, younger age at first intercourse, human papillomavirus (HPV) infection, history of smoking, in utero diethylstilbestrol exposure, and a prior history of cervical cancer.4,5 There is also an association between vaginal cancer and chronic vaginal irritation, as with long-term pessary use. In a 10-year review of the National Cancer Data Base, Creasman and colleagues6 observed that survival was related to the stage of disease and the histologic type, with melanoma having the poorest prognosis. In a later review of the Surveillance, Epidemiology, and End Result (SEER) program by Shah and colleagues,7 stage, tumor size, histology, and treatment modality were the most significant prognostic factors.
Although VAIN is recognized as a premalignant condition, its progression rate to invasive cancer is not well understood. Aho and colleagues8 followed 23 patients with untreated VAIN in an attempt to define the natural history of the disease and reported a 9% progression rate from high-grade VAIN to vaginal cancer among women who were followed for at least 3 years. A subsequent retrospective study by Schookaert and colleagues9 examining the incidence of high-grade VAIN after hysterectomy for high-grade cervical intraepithelial neoplasia also suggested progression from VAIN to invasive vaginal cancer. A total of 7.6% of patients developed VAIN 2, VAIN 3, or invasive carcinoma with a mean interval between hysterectomy and the first vaginal biopsy of 45 months, with 3% of these high-grade vaginal lesions subsequently progressing to invasive vaginal cancer. All of the patients were reported to have had a negative vaginal colposcopy before their hysterectomy for cervical dysplasia, suggesting that the subsequent cancers were of vaginal origin.
One recognized risk factor for the development of vaginal cancer is a history of treatment for an anogenital tumor. In a large, population-based, case control study, Daling and colleagues5 reported that 30% of women with vaginal cancer had a history of treatment for an anogenital tumor. This is most likely secondary to the shared pathophysiology of HPV infections. Up to 90% of anal cancers in women harbor HPV; previous treatment for a genital cancer confers a 10-fold increased risk of anal cancer.10 Other risk factors for squamous cell carcinoma of the vagina include pelvic radiation, tobacco abuse, and other factors associated with HPV, such as an increased number of sexual partners and younger age at first intercourse. Finally, 10% to 20% of women with a primary vaginal cancer have a history of pelvic irradiation, most commonly as treatment for cervical cancer. A prior history of pelvic radiotherapy is more frequent in women with vaginal sarcomas than in women with vaginal cancers of other histologic subtypes.
DIAGNOSIS
Key Points
1. Vaginal cancer may present with an abnormal Pap smear or with vaginal bleeding.
2. Biopsy of suspicious vaginal lesions is critical to diagnosis of this disease.
Vaginal cancer is often identified at the time of screening for cervical cancer. Although there is ample literature addressing optimal strategies to screen for cervical cancer, the rarity of vaginal cancer makes it difficult to assess any screening strategy. It is generally accepted that women who have had a hysterectomy for benign disease do not need continued vaginal cytologic screening if there is no evidence of HPV-related cervical lesions on pathology. There is also agreement that it is likely safe for older women to discontinue screening if there is a history of negative cytology and an absence of risk factors. Women who have been treated for cervical cancer should continue to undergo routine surveillance with vaginal cytologic testing. It is also reasonable to continue to screen women who have undergone a hysterectomy and who have a history of cervical dysplasia.11
The most common presenting symptom of vaginal cancer is abnormal vaginal bleeding. Other presenting symptoms include abnormal discharge and pelvic pain or dysuria, although these often present later in the disease course secondary to tumor spread outside of the vagina. A careful vaginal examination is required in these patients, as a small lesion can be difficult to visualize in the folds of the vagina or can be covered by blades of the speculum. The differential diagnosis for vaginal lesions is broad (Table 10-1), and biopsy should never be delayed when a suspicious lesion is encountered.
Table 10-1 Differential Diagnosis for Vaginal Lesions
The most common symptom at diagnosis of a vaginal melanoma is vaginal bleeding,12,13 but up to one-third of these lesions present as a mass at the time of routine pelvic examination.13 The majority of vaginal melanomas are pigmented or ulcerated at diagnosis, but up to one-fourth have been reported as grossly amelanotic.12,13 The most common location for vaginal melanoma is the lower anterior vaginal wall.
Sarcomas are often bulky lesions that appear most commonly in the upper vagina. Sarcoma botryoides is a subtype of rhabdomyosarcoma and primarily affects children, predominantly arising from the anterior wall of the vagina. The name derives from the Greek word for grapes and grossly has a grapelike cluster appearance.
In general, computed tomography (CT) and, more recently, positron emission tomography (PET) or single-photon emission computed tomography (SPECT) have been shown to be a valuable adjunct in the identification of lymph node metastasis. Although not part of the standard staging criteria, these imaging techniques may have a role in surgical and/or radiation therapy planning.
PATHOLOGY
Key Points
1. Squamous cell carcinomas comprise the majority of vaginal cancers, and up to 70% are associated with infection with high-risk HPV subtypes.
2. Adenocarcinomas of the vagina are commonly of clear cell histology and are attributed to in utero exposure to diethylstilbestrol.
3. Vaginal cancers are clinically staged using International Federation of Gynecology and Obstetrics (FIGO) criteria; this system is limited by the accuracy of determining extension into subvaginal tissues.
Squamous cell carcinomas (Figure 10-2) comprise approximately 80% of primary vaginal malignancies and histologically resemble squamous cell carcinomas of the cervix. There is no established or consistent precursor lesion to vaginal cancer. HPV infection appears to be associated with the development of vaginal cancer, similar to cervical cancer. Daling and colleagues5 detected HPV in the tissue of more than 80% of patients with carcinoma in situ and more than 60% of patients with invasive squamous cell cancer. More recent studies have suggested that approximately 70% of squamous cell vaginal cancers and greater than 90% of VAIN 2 and VAIN 3 can be attributed to high-risk HPV subtypes.14,15
FIGURE 10-2. Squamous cell carcinoma. Hematoxylin and eosin stain of squamous cell carcinoma of the vagina showing malignant invasion of the dermal layer with associated inflammatory reaction. (Image contributed by Dr. Rex Bentley.)
Adenocarcinomas of the vagina (Figure 10-3) account for approximately 10% of primary vaginal carcinomas. The origin of these lesions may include embryonic elements and normal glandular tissue residing in the vagina. The majority of reported cases are clear cell adenocarcinomas in individuals exposed in utero to diethylstilbestrol (DES). These DES-associated adenocarcinomas are histologically identical to the clear cell adenocarcinomas of the ovary and endometrium. DES is a synthetic nonsteroidal estrogen that was used from the 1930s until 1971 primarily to prevent spontaneous miscarriage in patients considered as high risk. Herbst and Scully16 described 7 women aged 15 to 22 years with clear cell adenocarcinoma of the vagina and noted that 6 of these patients had been exposed to DES in utero. DES was subsequently removed from the market in the United States. Herbst and Scully16 subsequently established a registry to track cases of clear cell adenocarcinoma. As of December 2007, the registry has accessioned approximately 760 cases of clear cell adenocarcinoma of the vagina and cervix, two-thirds of which were associated with DES exposure. Hatch and colleagues17 examined the cancer risk in daughters exposed to DES and found that the standardized incidence ratio for clear cell adenocarcinoma was 40 times that expected in the general population. The age-incidence curve peaks between the ages of 15 and 25 years, with an estimated risk of development of 1 per 1000 of these individuals exposed, but cases up to the age of 55 have been accessioned to the registry.
FIGURE 10-3. Clear cell adenocarcinoma. Hematoxylin and eosin stain of clear cell carcinoma of the vagina showing clear cells secondary to a large quantity of glycogen in the cytoplasm. (Image contributed by Dr. Rex Bentley.)
The number of DES-associated adenocarcinomas has sharply declined over the last several decades after the removal of DES from the market. Frank and colleagues18 described a series of patients with primary non–DES-associated adenocarcinoma of the vagina. The patients presented at an older age and, stage for stage, carried a worse prognosis than DES-associated adenocarcinoma. Local and distant recurrence rates were also higher in non–DES-associated malignancies.18 The role of HPV in the pathogenesis of non–clear cell adenocarcinomas of the vagina is unclear.
There are several vaginal cancers of less common histology; these include melanomas, sarcomas, verrucous carcinomas, and endodermal sinus tumors. Vaginal melanoma (Figure 10-4) constitutes 3% to 5% of all vaginal malignancies19,20 and less than 1% of all melanomas19,21 and occurs at a mean age of 60 to 65 years. Unlike many cutaneous melanomas, exposure to ultraviolet radiation is not a risk factor for mucosal melanomas. Melanoma is thought to arise from melanocytes, which are present as a normal variant in the vaginal epithelium in approximately 3% of women.22
FIGURE 10-4. Melanoma. Hematoxylin and eosin stain of melanoma demonstrating melanocytes proliferating along the dermal/epidermal juction in a continuous band with scattered single melanocytes extending into the upper layers of the dermis. (Image contributed by Dr. Angelica Selim.)
Sarcomas account for approximately 3% to 4% of all primary malignancies of the vagina.6,23 Prior radiotherapy is a risk factor, with 35% of cases having a history of pelvic radiotherapy, usually for cervical cancer, 8 to 26 years prior.24 Grade is an important prognostic factor; a series of 24 vulvar and vaginal sarcomas demonstrated no deaths in patients with low-grade tumors, whereas the 5-year overall survival rate was approximately 50% for high-grade lesions.25 There was no survival difference between leiomyosarcomas and other histologic types in this series, although numbers were limited.
Embryonal rhabdomyosarcoma is the most common malignant neoplasm of the vagina in infants and children and accounts for 75% of the vaginal rhabdomyosarcomas. Although a majority of these cases are diagnosed before age 5 years, these malignancies can be found elsewhere in the female genital tract later in life.
Verrucous carcinoma is a rare malignancy of the female genital tract, with only a few cases reported in the vagina. It is primarily found in postmenopausal women. HPV is likely a causative agent or factor in its development, but the pathogenesis is not completely understood. It is a slowly growing tumor that generally “pushes” into adjacent tissue rather than invades it, and accompanying inflammatory enlargement of the regional lymph nodes is encountered as opposed to metastasis.26
Endodermal sinus tumors, or yolk sac tumors, are germ cell tumors that are rarely found in the vagina. They are primarily diagnosed before the age of 3 years. The most common presenting symptom is vaginal bleeding. In children, the majority of germ cell tumors are extragonadal, with thoracic and central nervous system being most common. Only a small proportion of these are primary vaginal tumors.27 The most common presenting symptom is vaginal bleeding. α-Fetoprotein is generally elevated in these patients and can be used as a marker of response to treatment.
Metastatic Spread Patterns
The vagina is bordered laterally by the levator ani muscles, pelvic fascia, and ureters. Deep invasion anteriorly or posteriorly can result in bladder or rectal involvement, which may lead to fistula formation during or after radiotherapy. Disease extending to the sidewall may result in ureteral obstruction and renal failure. The lymphatic drainage of vaginal cancers is complex and depends on the anatomic level of the lesion. Much like cervical cancer, the lymphatic drainage can follow multiple pathways.28 The upper two-thirds of the vagina drain into the internal and external iliac lymph nodes, whereas the lower one-third drains into the sacral, common iliac, and superficial inguinal lymph nodes.
FIGO Staging
Vaginal cancer is clinically staged using the criteria established by the International Federation of Gynecology and Obstetrics (FIGO) (Table 10-2), which uses history and physical examination, routine laboratory evaluation, bimanual rectovaginal examination, cystos-copy, proctoscopy, intravenous pyelogram, and plain film radiographs if there is a concern for bony metastasis. One of the limitations of this staging system is the inability to accurately determine spread into subvaginal tissues. This is reflected in a wide range of survival data within a given stage. Perez and colleagues1 have suggested that stage II disease be subdivided into 2 groups: (1) lesions that involve the submucosa but do not extend into the parametria, and (2) lesions that involve the parametria without reaching the pelvic sidewall, but this has not consistently been demonstrated to have prognostic significance.7,29
Table 10-2 Staging of Vaginal Cancer Adapted From the International Federation of Gynecology and Obstetrics
Because it does not account for tumor size or regional lymph node status, the FIGO staging system is not generally used for melanomas. Likewise, the Clark and Chung methods are not applicable because of the difference in histologic landmarks between cutaneous structures and the vagina. Vaginal melanomas may best be classified using the Breslow thickness,30 which describes the tumor thickness, or the American Joint Committee on Cancer (AJCC) staging system (AJCC 2001), which takes into account thickness, presence or absence of ulceration, lymph node metastasis, and other metastatic sites of disease.19 The size of the lesion is an important prognostic factor as well, with a size of greater than 3 cm generally indicating worse outcomes.31–33 Similarly, the staging of sarcoma botryoides follows that of pediatric rhabdomyosarcoma.
TREATMENT
Key Points
1. Stage I squamous cell carcinomas of the vagina may be treated primarily with surgical excision or primary radiotherapy. Advanced-stage lesions likely requiring postsurgical radiation are typically managed with primary radiotherapy alone.
2. Lymph node dissection is recommended with surgical excision of early-stage vaginal cancers.
3. Radiotherapy is the treatment of choice for a majority of vaginal cancers because of its effectiveness in achieving long-term survival and preservation of adjacent organs; chemosensitization with platinum regimens may improve disease control.
Surgery
Stage I squamous cell carcinoma of the vagina may be treated either surgically or with primary radiotherapy, and this decision should be tailored not only to the size of the lesion, but also to the patient’s desire to maintain optimal sexual function (Figure 10-5). Minimally invasive lesions discovered incidentally at the time of surgical excision of presumed VAIN have been managed with surveillance with acceptable outcomes.34,35 Women who have small, known invasive lesions and who have previously undergone a hysterectomy are usually treated by partial or total vaginectomy, parametrectomy, and lymph node dissection. For women who have an intact uterus, radical hysterectomy is also indicated.36–38 A significant portion of patients who undergo primary surgical therapy also receive adjuvant radiotherapy, with lymph node metastasis or surgical margin involvement being the most common indications.36–39 Those at high risk for requiring postoperative radiation, as when lymph node metastasis is suspected, are generally not also subjected to primary surgical excision, as this combination is associated with higher rates of adverse events.
FIGURE 10-5. Treatment of vaginal cancer by stage algorithm. Vaginal cancer treatment decision tree.
Adenocarcinoma of the vagina is treated similarly to squamous cell carcinoma. In the past, consideration was given to conservation of ovarian and sexual function, as patients with early-stage DES-associated adenocarcinoma were usually younger, but DES-associated malignancies are now a rare entity. Simple excision is not recommended for adenocarcinomas regardless of size because of their more aggressive nature and propensity for lymphatic spread, and as with squamous cell carcinomas, options include radical surgery or primary radiotherapy with or without chemotherapy.18
Although the role of lymph node metastasis is not well defined in the FIGO staging system, it is a poor prognostic indicator. As outlined earlier, the lymphatic drainage of the vagina is somewhat complicated and depends on the level of the lesion. In patients undergoing surgical management of early-stage vaginal cancer, a lymph node dissection is routinely performed.36,38 Because the majority of patients diagnosed with vaginal cancer undergo radiation therapy, the true incidence of lymph node metastasis is unknown. Identification of sentinel nodes has the potential to aid in radiation treatment planning. Frumovitz and colleagues40 were able to identify a sentinel node in 79% of the patients with vaginal cancer evaluated with pretreatment lymphoscintigraphy, and van Dam and colleagues41 were able to detect a sentinel lymph node in 2 of 3 patients with vaginal cancer laparoscopically.
Radiotherapy
Radiotherapy (RT) is an effective, curative treatment modality for vaginal cancer and is often the treatment of choice for many reasons. First, radical surgery to remove large, invasive primary vaginal lesions with negative margins often carries risks of significant morbidity because of the close proximity of the vagina to critical structures; RT affords preservation of adjacent organs. Second, the vagina is an ideal site for brachy-therapy (BT) for conformal treatment to curative doses. Finally, given the broad lymph drainage patterns of vaginal cancer, prophylactic nodal treatment is often more appropriately accomplished with whole-pelvic RT than with multiple large dissections.
Combined Chemotherapy and Radiotherapy for Vaginal Cancer
Interest in combining chemotherapy and RT for treatment of vaginal cancer has increased in the last decade as a result of the tremendous success with concurrent cisplatin-based regimens in cervical cancer.42–45 Because of the small numbers and stage heterogeneity, direct comparison of combined chemotherapy plus RT and RT alone is difficult. Most series include a majority of advanced-stage disease, and in this unfavorable cohort, disease control is achievable using combined chemotherapy and RT in approximately 50% to 60% of patients. Concurrent chemotherapy is generally well tolerated in appropriately selected patients, and there is no evidence of significantly worse late toxicity as compared with RT alone. As such, it has been our approach to include concurrent platinum-based chemotherapy in the treatment of appropriate patients with advanced-stage disease.