Key words
vaginal tumour - malignant melanoma - multivisceral resection
Schlüsselwörter
Vaginaltumor - malignes Melanom - multiviszerale Resektion
History and Clinical Findings
A 44-year-old patient presented to the gynaecological outpatient clinic of our hospital
due to dragging pain in the lower abdomen and vaginal bleeding after intercourse.
The patient had undergone a vaginal hysterectomy for multiple uterine myomas 5 years
previously. She had no history of other operations or secondary illnesses. She had
not had regular gynaecological check-ups for the last 5 years. The subsequent clinical
examination found a granulomatous tumour at the cranial end of the vagina.
Diagnostic Investigation
Biopsy samples indicated an infiltrating nodular melanoma. Further diagnostic investigations
were done to determine whether it was a primary tumour or metastasis. Magnetic resonance
tomography (MRT) showed a large nodular lesion at the stump of the vagina with a diameter
of 5 × 4.5 × 9 cm in close proximity to the rectum, sigma and bladder ([Fig. 1]). Several regional lymph nodes were suspicious but there was no infiltration of
the pelvic wall ([Fig. 2]). Additional computer tomography confirmed the findings and additionally demonstrated
non-specific wall thickening in the sigma with a diameter of 23 × 15 mm. Thoracic
X-ray and thoracic CT were unremarkable. Coloscopy showed no pathological findings.
Anorectal endosonography, however, demonstrated a tumour with a diameter of 4 cm,
located directly ventral and adjacent to the wall of the rectum. Dermatological and
ophthalmological examinations were unremarkable. For
the final clinical diagnosis, full-body PET-CT was done and showed, in addition to
the known malignant melanoma in the area at the end of the vagina (diameter in this
case: 9 × 7.2 cm), bilateral pararectal metastases, left-sided para-iliacal metastasis
and para-aortal lymph node metastases. The interdisciplinary tumour board concluded
that multivisceral resection was indicated.
Fig. 1 Vaginal tumour at the cranial end of the vagina (MRT).
Fig. 2 Vaginal tumour with lymph node infiltration (MRT).
Therapy
A double-J catheter was inserted for urological management in preparation for the
resection. Subsequently we performed an en bloc tumour resection with partial resection
of the vagina, deep anterior resection of the sigma and rectum and partial resection
of the bladder roof. In addition, left-sided adnexa extirpation was done, together
with bilateral iliacal and para-aortal lymphadenectomy, appendectomy and creation
of a protective cecostomy. The postoperative course was without complications and
the patient was closely monitored. The final gynaecological examination showed a non-inflamed
6 cm vaginal remnant. The patient refused the recommended adjuvant palliative therapy
with dacarbazine. The patient died only 4 months after the operation.
Histology
The final histological examination showed an ulcerated, nodular, vaginal melanoma
measuring 85 mm at its widest point with a resection margin of at least 35 mm. Examination
of the 22-cm resected tissue adherent to the sigma and rectum showed tumour infiltration
extending into the subserous adipose tissue with a still intact colonic mucosa. Metastases
were noted in the left- and right-sided lymph nodes of the obturator fossa, in the
left para-aortal lymph nodes and in the pericolic and perirectal lymph nodes. In addition,
a small nodular melanoma infiltrating the tip of the mesenteriolum of the appendix
was found.
Microscopically, the tumour consisted of giant, mainly epithelioid cells, distinct
polymorphic hyperchromatic cells, numerous atypical mitoses and a few giant tumour
nuclei. Dark brown pigmentation was noted in the tumour cells in some sections ([Fig. 3] and [4]). Using Chung et al.ʼs classification, the tumour was classified as level IV (invasion
> 3 mm), pN1 (34/61) pM1 L1 V0 R0.
Fig. 3 Immunohistology using melanoma monoclonal antibody, clone HMB45.
Fig. 4 Detection of tumour cells using Masson-Fontana stain.
Discussion
As vaginal melanomas are very rare, most published studies consist of case reports,
with only a few case series with limited numbers of patients. Most (90–95 %) malignant
vaginal lesions are squamous cell carcinomas. Between 5–10 % are clear cell adenocarcinomas,
and less than 3 % are malignant melanomas. In women, 1.6 % of melanomas are located
in the genital tract [3], [16], most of them around the vulva (70 %), followed by vaginal (21 %) and cervical (19 %)
locations. Vaginal melanoma metastases are even rarer with only 5 cases reported in
the literature to date [5]. It has been suggested that vaginal melanomas arise from melanocytes, which are
present in the vagina in around 3 % of women [4]. Most vaginal melanomas are diagnosed in the 6th or 7th decade of life in postmenopausal
women [5]. The most common symptom of vaginal melanoma is
vaginal bleeding, followed by dyspareunia [11], [14]. The 5-year survival rate for this very aggressive, rapidly growing tumour is very
poor, and is reported to be 0–25 % irrespective of the chosen therapy [10], [16]. Vaginal melanomas are usually only detected in advanced stages. Around 50 % of
patients already have lymph node metastasis at diagnosis, and 20 % have distant metastasis
[11].
In the study by Miner et al. (n = 35), typical prognosis-relevant factors such as
age, depth of invasion, pigmentation, ulceration and even adjuvant therapy were not
found to be correlated with patient outcome. Even the microscopic assessment of positive
and negative resection margins did not show any significant difference in recurrence-free
survival times [7].
Meta-analyses by Reid et al. and Buchanan et al. also showed no correlation between
the depth of tumour invasion and patient survival [2], [12]. Tumour size is considered the only prognostic factor. Thus patients with a tumour
size of < 3 cm survived significantly longer than patients with tumour sizes of > 3 cm
(12 months vs. 41 months) [2], [12]. Another series with 14 patients confirmed the correlation between tumour size and
prognosis. Thus, 3 of 7 patients with a tumour < 3 cm survived more than 5 years while
none of the patients with a tumour > 3 cm survived longer than 5 years [10]. The median survival rate after diagnosis is approximately 20 months [7].
Primary surgery is considered the method of choice and appears to be superior to primary
radiation (25 vs. 13 months; p = 0.039). After excluding very advanced, surgically
non-resectable tumours, the survival benefit of surgery was even more pronounced (25
vs. 9 months; p = 0.006) [9]. Nevertheless, due to the limited number of cases in the literature, there are no
standard therapy recommendations. Various surgical procedures have been described,
including “wide local excision”, colpectomy, radical resection with total abdominal
hysterectomy, bilateral salpingo-oophorectomy, and evisceration. Only Van Nordstrand
et al. reported in their study that radical surgery (total colpectomy or evisceration)
offered a superior outcome compared to “wide local excision” or radiation with regard
to the 2-year survival rate [15]. Other authors did not find any significant differences in survival rates with different
surgical
procedures [7], [12]. Due to the poor prognosis of vaginal melanoma, the aim must be to obtain local
excision with a tumour-free resection margin. Sentinel lymph node biopsy, an established
method to determine the lymph node status of cutaneous melanomas, has been reported
as an alternative to complete lymphadenectomy which is associated with a high morbidity
[9]
[8].
There are no consistent recommendations with regard to the benefits of adjuvant chemotherapy.
Pegylated interferon alpha-2b can prolong survival times in patient with cutaneous
melanoma and lymph node metastasis [1]. Analogous to cutaneous melanoma, dacarbazine or interleukin-2 are used in the palliative
treatment of vaginal melanomas. The lack of studies means that it is not clear whether
adjuvant chemotherapy prolongs recurrence-free survival in patients with vaginal melanoma.
In contrast to malignant cutaneous melanomas, c-kit mutations have been detected immunohistologically
in mucosal melanomas in approx. 20 % of cases. There are several case reports of successful
“off label” therapies using c-kit blockers (e.g. imatinib, sunitinib) when c-kit mutations
were present [13].
Radiotherapy is usually used to treat primary non-resectable tumours and when resection
margins are positive [6], [7]. Therapy consists of intracavitary brachytherapy with 2–5 applications of caesium-137
(0.7–20 Gy per dose) [11]. Due to the limited data, which is exclusively retrospective, it is not possible
to make any assertions regarding the value of adjuvant radiotherapy.
Conclusion
Vaginal melanoma is a very rare tumour entity with a poor prognosis. To date, tumour
size is the only prognostic factor, and complete surgical resection is the therapy
of choice. The data is insufficient to assess the value of sentinel lymph node biopsy
and adjuvant radio- or chemotherapy. Because this tumour entity is almost asymptomatic,
it demonstrates the importance of regular preventive medical check-ups as the prognosis
is extremely poor once symptoms become manifest.
Acknowledgement
We would like to thank Prof. Dr. Eckhard Koepcke, Rostock, for his advice and Dr.
Annegret Radtke for providing the histological images.
