Locoregional mucosal melanoma: Epidemiology, clinical diagnosis, and treatment

INTRODUCTION — Mucosal melanomas generally arise from the mucosal epithelium lining the respiratory, alimentary, and genitourinary tracts, all of which contain melanocytes. Melanoma can also arise from melanocytes within the skin, which is classified as cutaneous melanoma.

Mucosal melanomas generally carry a worse prognosis than those arising from cutaneous sites. Our understanding of the optimal management for mucosal melanoma is limited because of the rarity of this disease, its unique biology, and clinical challenges of mucosal melanoma arising from each anatomic location.

The epidemiology, clinical presentation, and management of patients with locoregional mucosal melanoma arising in the head and neck, vulvovaginal, and anorectal regions are discussed here. The management of patients with metastatic mucosal melanoma, cutaneous melanoma, and uveal and conjunctival melanoma is discussed separately.

●(See "Treatment of metastatic mucosal melanoma".)

●(See "Overview of the management of advanced cutaneous melanoma".)

●(See "Adjuvant and neoadjuvant therapy for cutaneous melanoma".)

●(See "Systemic treatment of metastatic melanoma lacking a BRAF mutation".)

●(See "Systemic treatment of metastatic melanoma with BRAF and other molecular alterations".)

●(See "Initial management of uveal and conjunctival melanomas".)

EPIDEMIOLOGY — Mucosal melanoma has distinct epidemiologic and clinical features that differentiate it from cutaneous melanoma and have important implications for patient management.

●Incidence – Mucosal melanoma represents approximately 1.3 percent of the melanomas in the United States and accounts for an estimated incidence of 1400 patients [1,2]. Although the incidence of cutaneous melanoma has been increasing rapidly in the United States, the incidence of mucosal melanoma has generally remained stable over time [3,4].

●Anatomic sites – Mucosal melanomas arise primarily in the head and neck, anorectal, and vulvovaginal regions (55, 24, and 18 percent of cases, respectively). Rarer sites of origin include the urinary tract, gallbladder, and small intestine.

●Age and sex – In general, patients diagnosed with mucosal melanomas are older, with a median age of 70 years. However, mucosal melanoma of the oral cavity presents in younger patients (mean age of 55 years) [5-8]. Mucosal melanoma is more common in females than males, primarily due to the development of disease in the genital tract [1,9,10].

●Risk factors – Cutaneous melanoma is associated with exposure to ultraviolet light, but the anatomic location of mucosal melanoma precludes ultraviolet exposure as a risk factor, with the exception of conjunctival melanoma [11]. No clear environmental or other predisposing risk factors have been identified in this disease. (See "Melanoma: Epidemiology and risk factors".)

●Multifocal, amelanotic tumors – An estimated 20 percent of mucosal melanomas are multifocal [12], compared with less than 5 percent of those arising in the skin [13]. Approximately 40 percent of mucosal melanomas are amelanotic, compared with less than 10 percent of cutaneous melanomas [13].

●Metastatic disease – Mucosal melanoma patients are more likely to present with metastatic disease at the time of diagnosis compared with cutaneous melanoma (23 versus 5 percent) [1,14]. This is primarily due to lack of visibility of the primary site and absence of clinical symptoms, which frequently leads to a delay in diagnosis.

Mucosal melanoma has a more aggressive underlying biology compared with other melanoma subtypes, which may also influence disease stage at presentation. Patients with mucosal melanoma have a five-year overall survival (OS) rate of only 25 percent [1], and they also have inferior stage-matched survival compared with other melanoma subtypes, both at the time of diagnosis and at the development of metastatic disease [15,16].

●Ethnicity – Mucosal melanomas make up a greater proportion of all melanomas diagnosed in Black, Asian, and Hispanic individuals when compared with melanomas occurring in White individuals, reflecting the much lower incidence of cutaneous melanoma in these populations [17]. Melanomas of mucosal origin comprise up to 25 percent of all melanomas diagnosed in Chinese patients (most prospective trials are derived from this population), 9 percent of all melanomas diagnosed in Black or Hispanic patients, and less than 2 percent of all melanomas diagnosed in non-Hispanic White patients [1,18]. Despite a greater proportion of mucosal melanomas in Black, Asian, and Hispanic populations, the absolute incidence of mucosal melanoma remains greater in White populations.

DIAGNOSIS — The definitive diagnosis of mucosal melanoma of any primary site is made using biopsy and histopathology, which is discussed separately. (See "Melanoma: Clinical features and diagnosis", section on 'Diagnosis confirmation'.)

MANAGEMENT CONSIDERATIONS

Surgical considerations — Regardless of the primary site, initial management for locoregionally confined mucosal melanoma is complete wide local surgical resection. Complete surgical resection, if feasible, offers the best chance for prolonged disease-free survival. However, complete resection with optimal margins is frequently difficult because of the lentiginous growth pattern frequently associated with mucosal melanoma, disease multifocality, and anatomic constraints. Patient preferences and quality of life considerations are critical in determining the extent of surgery.

The morbidity of an aggressive resection must also be balanced against the risk of metastatic disease. Most patients will ultimately develop distant metastases despite adequate local disease control of the primary disease with surgery. (See 'Epidemiology' above.)

Is there a role for neoadjuvant therapy? — The role of neoadjuvant therapy in patients with mucosal melanoma is evolving. Further studies in this area may demonstrate the benefits of initial immunotherapy and lead to the development of predictive and prognostic biomarkers. Patients who are interested in this approach should be offered enrollment in clinical trials, where available. Further details on the role of neoadjuvant immunotherapy in advanced cutaneous melanoma are discussed separately. (See "Adjuvant and neoadjuvant therapy for cutaneous melanoma".)

In preliminary results from a phase II trial of 21 patients with resectable mucosal melanoma, at median follow-up of 59 weeks, neoadjuvant toripalimab and axitinib demonstrated a complete pathologic response rate of 29 percent [19].

MUCOSAL MELANOMA OF THE HEAD AND NECK

Clinical presentation — Mucosal melanomas account for less than 10 percent of all melanomas arising in the head and neck region. Mucosal melanomas of the head and neck commonly occur in the following locations [20-25]:

●Nasal cavity (most commonly involving the turbinates and nasal wall) – 55 percent

●Paranasal sinuses (most commonly involving the maxillary and ethmoid sinuses) – 15 percent

●Oral cavity (most commonly involving the hard palate and upper alveolus) – 25 percent

Much less commonly, mucosal melanomas arise in the pharynx [26], larynx [27], or esophagus [28,29].

Patients with sinonasal mucosal melanomas frequently present with nasal obstruction, epistaxis, or loss of smell [30-32]. Mucosal melanoma of the oral cavity typically presents as a painless bleeding mass, as an area of ulceration, as a region of mucosal discoloration, or with ill-fitting dentures [22].

Approximately 60 percent of mucosal melanomas arising from the oral cavity are identified incidentally during routine dental examination. Irregular isolated areas of pigmentation within the oral mucosa must be differentiated from melanosis, a benign condition which is common in Black patients and generally is symmetrical. Such lesions should be biopsied.

Staging evaluation — Upon diagnostic confirmation, the initial staging evaluation for patients with mucosal melanoma arising from the head and neck should include clinical examination with endoscopic inspection for paranasal disease, computed tomography (CT) and/or magnetic resonance imaging (MRI) of the primary site of disease, and CT of the neck, chest, abdomen, and pelvis and/or positron emission tomography (PET)-CT imaging to assess for lymph node involvement or distant metastases [33]. Regional lymph node involvement at presentation has been estimated in approximately 25 percent of those with oral cavity lesions and 6 percent of those with sinonasal mucosal melanoma [1]. Central nervous system (CNS) imaging is also obtained at diagnosis, as the brain can be observed as an initial site of metastasis in 7 percent of cases of mucosal melanoma [34].

Staging system — The American Joint Committee on Cancer (AJCC) staging system for mucosal melanoma arising from the head and neck is generally used for this subset of disease and reflects the overwhelmingly poor prognosis of this disease (table 1) [35].

The AJCC head and neck stage system differs from that for cutaneous melanoma, in which prognosis depends upon the local, regional, and distant burden of disease. Staging begins at stage III as the most limited form of disease, to reflect the poor prognosis in these patients, and includes three subcategories of stage IV disease (IVA, IVB, and IVC), depending upon the local extent and presence of regional and distant disease [36-38].

Management of locoregional disease

Surgical candidates — Complete surgical excision is the primary treatment for AJCC stage III and IVA mucosal melanomas of the head and neck if the lesion can be resected with negative margins. Surgery is not typically used as primary therapy for patients with lymph node involvement (IVB) or distant metastases (IVC), unless there is a need for local control of disease. Patients with complete resection may be evaluated for adjuvant radiation therapy (RT), although treatment is individualized. (See 'Adjuvant radiation' below.)

Management of the primary site — Mucosal melanomas of the oral cavity are approached surgically in the same manner as squamous cell carcinomas at the same site. Local resection should be accompanied by appropriate reconstruction using free flaps for mandibular, soft tissue, or structural defects. Similarly, mucosal melanoma of the larynx and pharynx should be managed by the appropriate surgical approach to achieve local control.

Craniofacial resection has been used when tumor invades the skull base, either as part of the initial presentation or as an isolated local recurrence [39,40]. Endoscopic resections are being performed more commonly and may be accomplished with less morbidity and equivalent local control [41].

Despite aggressive surgical management, local recurrence is common [5,22,25,31]. Factors associated with local recurrence include tumor size, incomplete resection, and vascular invasion [22,42]. The median time to first recurrence in the head and neck is 6 to 12 months [20,31]. (See 'Prognosis' below.)

Although the initial site of relapse is frequently local or regional, such recurrence usually is a harbinger of disseminated disease, with most patients dying of distant rather than uncontrolled local disease [5,20,30,43,44]. Additional surgery can be considered for locally recurrent disease; however, the decision to proceed with re-resection for local or regional recurrence should be evaluated on a case-by-case basis.

Management of the neck — Surgical management of the neck depends upon the presence or absence of clinical nodal disease and disease location (sinonasal versus oral cavity). General principles are as follows:

●Clinically negative nodes

•For patients with sinonasal mucosal melanoma and clinically negative neck nodes, we suggest surveillance alone rather than elective lymph node dissection. In this population, the incidence of nodal disease is relatively low (between 10 and 17 percent) [22,45]. (See 'Surveillance' below.)

•For patients with mucosal melanoma arising from the oral cavity and clinically negative neck nodes, we suggest elective lymph node dissection rather than observation for control of nodal disease [46]. In this population, the incidence of nodal disease is estimated at up to 25 percent [22,25]. As in squamous cell carcinoma of the head and neck, the regional lymph nodes can be addressed at the same time as the primary resection, clearing the first and second echelons of nodes. Furthermore, when reconstruction using free flaps is required after primary resection, the neck must be entered and the nodal basin exposed, which allows for evaluation and clearance of nodal disease at that time.

•Sentinel lymph node biopsy does not have an established role in patients with mucosal melanoma of the head and neck, although successful identification of sentinel lymph nodes in patients with mucosal melanoma is feasible [47,48].

●Clinically evident nodal disease – For patients with clinically evident nodal disease, therapeutic neck dissection can be offered based on the need for local control and symptoms [45]. Therapeutic neck dissection can be accomplished with limited morbidity and may be performed for optimal regional control. However, no significant difference in five-year overall survival (OS) has been observed in patients with and without lymph node recurrence [20].

Adjuvant radiation — For patients with high-risk disease (palpable lymph nodes or extracapsular extension), we suggest adjuvant RT. Adjuvant RT may reduce local recurrences following wide local excision and neck dissection in patients with high-risk features such as palpable lymph nodes or extracapsular extension [38], although no OS benefit has been demonstrated. Treatment decisions should be individualized, as there are limited prospective clinical trials and optimal dosing and fractionation schedules have not been established. (See "Adjuvant radiation therapy or chemoradiation in the management of head and neck cancer" and "Radiation therapy in the management of melanoma".)

Local recurrence occurs in 29 to 79 percent of cases despite aggressive surgery and can be clinically devastating [5,25,43,49-51]. Observational data and one meta-analysis evaluating adjuvant RT have reported an improvement in locoregional control but no impact on survival [32,43,51-55]. Such benefit, however, has not been observed in all series [22,56].

Nonsurgical candidates — Definitive, or primary, RT may be an alternative option for patients who are not surgical candidates or in whom a complete resection is not feasible (eg, inability to obtain adequate resection margins or anatomic constraints) [43]. However, such treatment decisions are individualized, as there are limited prospective clinical trials and optimal dosing and fractionation schedules have not been established. (See "Radiation therapy in the management of melanoma".)

Tumor control rates ranging from 61 to 85 percent have been reported with definitive RT in patients with mucosal melanomas arising from the head and neck [57,58]. Analysis of these series is complicated by the selection bias toward the use of RT in patients with more extensive tumors and narrow or involved surgical margins.

Prognosis — The prognosis in patients with mucosal melanoma arising in the head and neck is generally poor. The five-year survival rates range between 12 to 30 percent for primary lesions of the nasal cavity, oral cavity, and paranasal sinuses. Better outcomes are observed for mucosal melanoma arising from the nasal cavity and inferior outcomes for those with disease arising from the paranasal sinuses [22,38,59-61].

VULVOVAGINAL MELANOMA

Clinical presentation — Melanomas arising from the female urogenital tract occur primarily in the vulva and vagina (95 and 3 percent, respectively) [1]. Vulvar melanomas represent 10 percent of all malignant tumors involving the vulva, but less than 1 percent of all melanomas [62,63]. Melanomas arising elsewhere in the urogenital tract (ie, urinary bladder, urethra, or cervix) are rarer [64-68]. (See 'Epidemiology' above.)

Patients with vulvovaginal melanoma commonly present with pruritus, vaginal bleeding, a vaginal discharge, dyspareunia, or a mass [69-74].

Risk factors — A subset of vulvar melanomas may share common risk factors with cutaneous melanoma, particularly those arising from the outer, nonglabrous hair-bearing portion of the labia majora, Some vulvar melanomas may develop from contiguous nevi [75]. Chronic inflammatory disease, viral infections, chemical irritants, and genetic factors have also been implicated as risk factors [76]. (See "Vulvar cancer: Epidemiology, diagnosis, histopathology, and treatment".)

Staging evaluation — Upon diagnostic confirmation, the initial staging evaluation for patients with vulvovaginal melanoma should include clinical assessment with pelvic examination, CT and/or MRI of the primary site of disease, and CT of the chest, abdomen, and pelvis and/or positron emission tomography (PET)-CT imaging to assess for the presence of distant disease [77]. Central nervous system (CNS) imaging with brain MRI is also obtained at diagnosis [34].

Staging system

Vulvar melanoma — Vulvar melanomas should be staged according to the tumor, node, metastasis (TNM) system for cutaneous melanoma [77]. (See "Tumor, node, metastasis (TNM) staging system and other prognostic factors in cutaneous melanoma", section on 'Eighth edition AJCC TNM staging'.)

Outcomes based on stage of disease at presentation are illustrated by the five-year relative survival rates in 223 women diagnosed between 1985 and 1989 in the United States based upon data in the National Cancer Database [62]:

●Stage 0 – 77 percent

●Stage I – 70 percent

●Stage II – 50 percent

●Stage III – 48 percent

●Stage IV – 24 percent

The Gynecologic Oncology Group performed a prospective clinical-pathologic study of primary vulvar melanomas evaluating the predictive value of the microscopic staging systems described by Clark [78], Breslow [79], and Chung [80], as well as the International Federation of Gynecology and Obstetrics (FIGO) system and the 1992 American Joint Committee on Cancer (AJCC) melanoma staging system. In this study, the AJCC cutaneous melanoma staging system was found to be the most significant predictor of recurrence-free survival in females with this disease [81]. Subsequent studies have provided additional evidence that using the modified AJCC staging system for cutaneous melanoma has prognostic utility for vulvar melanoma [82,83].

Vaginal melanoma — Vaginal melanoma are classified using a simplified three-stage clinical system [18,84]. In this system, stages I, II, and III correspond to clinically localized disease, regional lymph node involvement, and distant metastases, respectively.

No particular staging system is a useful predictor of prognosis for vaginal melanoma, although tumor size (<3 cm versus ≥3 cm), has been shown to predict survival in some series [85,86]. This three-stage system, which was initially developed for mucosal melanomas of the head and neck, has limited prognostic utility for vaginal mucosal melanoma because a majority of patients present with localized (stage I) disease. However since it is easy to use and broadly applicable, it remains a reasonable staging system for vaginal melanoma, in the absence of a predictive alternative.

Management of locoregional disease — For patients with vulvovaginal melanoma, we suggest wide local excision of the primary tumor rather than more radical surgical approaches, as data have not demonstrated an overall survival (OS) benefit with more morbid surgery. Additionally, most patients ultimately develop distant metastatic disease regardless of the primary surgical procedure [87]. Therefore, patient preference and quality-of-life considerations are critical factors in determining the extent of surgery for the initial management.

Vulvar melanoma — Wide local excision has generally replaced pelvic exenteration for vulvovaginal melanoma for patients without evidence of distant metastases. Surgical management of vulvar melanomas should include resection of the primary tumor by wide local excision. Radical vulvectomy is reserved for large tumors primarily to obtain local disease control. Melanomas <1 mm thick should be treated with 1 cm skin margins; if anatomically feasible, margins can be extended to 2 cm for thicker melanomas [88,89]. The excision contains all layers of skin and subcutaneous tissues, extending to the muscular fascia below.

More radical procedures are associated with significant morbidity. Although some series suggested improved local control with more aggressive surgical procedures, the impact of improved local control upon survival is limited by the concomitant development of distant metastases in the majority of cases. Retrospective data also suggest no difference in OS with more conservative wide local excision [82,85,90-96].

Vaginal melanoma — Standard management of patients with vaginal melanoma involves wide local excision if possible. Achieving negative margins in these cases can be difficult without pelvic exenteration given the high frequency of multifocality and anatomic constraints. However, the impact of such radical surgery upon survival when compared with more conservative procedures (with or without radiation therapy [RT]) is unclear. Thus, surgery may be combined with RT in select cases. (See 'Adjuvant radiation' above.) [97].

Regional lymph nodes — Evaluation of sentinel nodes is feasible in patients with vulvovaginal melanoma, but the role of this approach is uncertain [76,98,99]. A positive sentinel lymph node biopsy can identify patients with locoregionally advanced nodal disease who may qualify for additional adjuvant systemic therapy. However, elective regional lymphadenectomy does not provide a survival benefit. (See 'Resected locoregional lymph nodes (adjuvant therapy)' below.)

While small series suggest a survival benefit for elective regional lymphadenectomy in patients with vulvar melanoma, a more contemporary study performed by the Gynecologic Oncology Group identified positive nodes as merely a prognostic factor along with tumor thickness. In this series, only 7 out of 56 patients undergoing a lymphadenectomy had positive nodes [81]. In addition, several studies have demonstrated that radical vulvectomy with bilateral inguinofemoral lymphadenectomy does not confer a survival benefit when compared with wide local excision alone for vulvar melanoma [95,100].

Prognosis — Despite aggressive therapy, the prognosis remains poor for most patients with mucosal melanoma arising in the vulva or vagina. While patients with vulvar melanoma have five-year survival rates of 24 to 77 percent, those with vaginal melanoma have five-year survival rates of 5 to 25 percent [71,88,101-105].

ANORECTAL MELANOMA

Clinical presentation — Anorectal mucosal melanoma accounts for approximately 0.05 percent of all colorectal malignancies and 1 percent of all anal canal cancers [106]. The site of origin is the rectum or anal canal in 42 and 33 percent of cases, respectively, while the primary site cannot be determined in the remainder. Although the risk factors for anorectal mucosal melanoma are not known, epidemiologic data suggest that there is an increased risk associated with human immunodeficiency virus infection [106-108]. (See "HIV infection and malignancy: Management considerations", section on 'Melanoma'.)

The majority of cases arise from the mucocutaneous junction; however, they can also arise from the skin of the anal verge, the transitional epithelium of the anal canal, or the rectal mucosa (figure 1). While primary site of disease in vulvovaginal melanoma has prognostic relevance, this is not the case for anorectal melanoma [109].

Patients typically present with bleeding, a mass, anorectal pain, or a change in bowel habits. Occasionally, melanoma is an incidental finding on pathologic evaluation of a hemorrhoidectomy or anal polyp specimen. Anorectal melanoma is pigmented in only one-third of cases. Most patients present with lesions that are >2 mm thick. Regional lymph node involvement is found in approximately 60 percent of patients at presentation [1], and distant metastases are present at diagnosis in approximately 30 percent of cases [110-115].

Staging evaluation — Upon diagnostic confirmation, the initial staging evaluation of patients with anorectal melanoma should include a rectal examination, possible rectal ultrasound, and CT of the chest, abdomen, and pelvis and/or positron emission tomography (PET)-CT imaging to assess for distant metastases. Central nervous system (CNS) imaging with brain MRI is also obtained at the time of diagnosis [34].

Staging system — Retrospective series have used a simple system in which localized disease only, regional lymph node involvement, and distant metastases are classified as stages I, II, and III, respectively [116].

Factors that may adversely affect prognosis in patients with localized disease include the presence of perineural invasion, tumor size and thickness, and the presence of amelanotic melanoma [111,112,117-119].

Patients with lymph node metastases or distant metastases at presentation have an especially poor prognosis. A review of 183 patients with anorectal mucosal melanoma identified in the Surveillance, Epidemiology, and End Results (SEER) database found the following [110]:

●Stage I – median survival, 24 months; five-year survival, 26.7 percent

●Stage II – median survival, 17 months; five-year survival, 9.8 percent

●Stage III – median survival, 8 months; five-year survival, 0 percent

Although anorectal melanoma is excluded from the American Joint Committee on Cancer (AJCC) staging system for anal cancers, a modification of the AJCC eighth edition staging system for cutaneous melanoma where T stage is classified into three groups: thin (T1), intermediate (T2 to T3), and thick (T4), can be used to stratify patients for survival [120].

Management of locoregional disease

Surgery (primary tumor) — For most patients, we suggest wide local, sphincter-sparing excision rather than abdominoperineal resection, as morbidity is higher with abdominoperineal resection and the benefit of local control with more aggressive surgery is limited by the risk of distant metastases regardless of the primary surgical procedure.

The primary goal of surgery is to perform a negative margin, sphincter-sparing excision. Abdominoperineal resection is reserved for patients with bulky local disease and for carefully selected patients with local recurrence [121].

Multiple observational studies have looked at the impact of the extent of surgery on long-term outcomes. Some series suggested improved local control with abdominoperineal resection, a procedure accompanied by a high morbidity rate and functional limitations [115,117,122]; however, retrospective data suggest that there is no difference in overall survival (OS) with more conservative wide local excision [110,111,118,123,124].

The benefit of improved local control upon survival is limited. Most patients ultimately develop distant disease regardless of the primary surgical procedure; thus patient preference and quality-of-life considerations are critical in determining the extent of surgery.

The most important factor in surgery appears to be the ability to achieve negative (R0) surgical resection margins [118,123]. In a series of 251 cases from the Swedish National Cancer Registry, the five-year survival rates following surgical excision for those in whom an R0 resection was achieved was 19 percent, compared with 6 percent in those in whom a complete local excision was not possible [123]. On multivariate analysis, resection status and tumor stage were significantly associated with prognosis, but the type of resection (abdominoperineal resection or local excision) was not significant.

Indications for inguinal lymphadenectomy — For patients with clinically node-negative disease, we suggest against elective inguinal lymphadenectomy, given the morbidity of bilateral inguinal node dissection and lack of survival benefit from an elective dissection. Inguinal lymphadenectomy is reserved for patients with clinically apparent nodal disease.

Mesorectal, pelvic sidewall, and inguinal lymph nodes are at risk for involvement in anorectal mucosal melanoma. However, available data suggest that lymph node metastasis does not predict outcome in patients undergoing radical resection, and sentinel lymph node biopsy does not have an established role in these patients [118,125,126].

Adjuvant radiation therapy — Adjuvant radiation therapy (RT) for anorectal melanoma may improve locoregional control despite a limited impact on OS [127]. Patients who are being evaluated for adjuvant RT should be offered a balanced discussion about these clinical benefits versus the potential side effects of RT.

Sphincter-sparing local excision followed by hypofractionated RT has been used as an alternative to abdominoperineal resection to prevent local recurrence. However, there are limited data on preserving quality of life with this approach in mucosal melanoma. In an observational series of 54 patients from MD Anderson Cancer Center treated over a 20-year period, this approach achieved local control in 82 percent of cases; however, the five-year OS rate was only 30 percent [128].

Prognosis — The reported five-year survival rates for patients with anorectal melanoma are approximately 20 percent [117,118,123,129].

RESECTED LOCOREGIONAL LYMPH NODES (ADJUVANT THERAPY)

Treatment approach — For patients with mucosal melanoma at high risk of recurrence, there are limited data for adjuvant systemic therapy, particularly in Western patient populations. Management decisions should be made on an individual basis. Patients should also be encouraged to enroll on clinical trials, where available. Our treatment approach is as follows:

●BRAF wild-type tumors – For patients with tumors lacking a BRAF V600 mutation, options include adjuvant chemotherapy (temozolomide plus cisplatin) or adjuvant immunotherapy; either treatment is reasonable in the absence of comparative data. Expectant observation may be offered to those who are ineligible for or decline adjuvant systemic therapy. (See 'Surveillance' below.)

●BRAF mutant tumors – For the uncommon patient with a BRAF V600 mutation-positive tumor (less than ten percent of all mucosal melanomas [130]), targeted therapy with dabrafenib plus trametinib is also an option, in addition to adjuvant approaches available for BRAF wild-type mucosal melanoma. Management of these patients is extrapolated from those with BRAF mutant cutaneous melanoma who receive adjuvant systemic therapy, and is discussed separately. (See "Adjuvant and neoadjuvant therapy for cutaneous melanoma", section on 'BRAF-mutated tumors'.)

Adjuvant chemotherapy — Adjuvant chemotherapy (temozolomide plus cisplatin) improved relapse-free survival benefit in randomized trials of patients with mucosal melanoma and resected regional lymph nodes [131,132]. However, most studies were conducted in Asian populations and require further study in a broader patient population before adjuvant chemotherapy can be offered as a standard of care for Western patients with mucosal melanoma. This approach might be an option in carefully selected patients who are not candidates for clinical trials, have undergone a potentially curative resection, have no serious coexisting illness, and have the potential for long-term survival excluding their risk of melanoma related mortality.

Based on initial data from a phase II randomized trial [131], a subsequent phase III trial was conducted in 204 with completely resected Stage I to III mucosal melanoma [132]. Patients were randomly assigned to either adjuvant chemotherapy (temozolomide 200 mg/m² days 1 to 5 plus cisplatin 25 mg/m2 days 1 to 3, every three weeks for six cycles) or one year of adjuvant interferon-alfa-2b. In preliminary results, at median follow-up of 24 months, compared to interferon-alfa-2b, chemotherapy improved relapse-free survival (median 15.5 versus 9.5 months; hazard ratio [HR] 0.56; 95% CI 0.4-0.77) and distant metastasis-free survival (median 16.8 months versus 9.6 months, HR 0.53; 95% CI 0.38-0.74) [132]. Median OS was approximately 41 and 36 months, respectively.

Adjuvant immunotherapy — Adjuvant checkpoint inhibitor immunotherapy with nivolumab (table 2) or pembrolizumab (table 3) is an established treatment option for patients with locoregionally advanced cutaneous melanoma, but data are limited for those with mucosal melanoma. Patients with mucosal melanoma can be evaluated for this approach, preferably in the context of clinical trial [38]. (See "Adjuvant and neoadjuvant therapy for cutaneous melanoma".)

As an example, only 29 patients with mucosal melanoma were included in a phase III trial evaluating adjuvant nivolumab in advanced melanoma (CheckMate 238). No patients with mucosal melanoma enrolled on a separate phase III trial evaluating adjuvant pembrolizumab (KEYNOTE-054). Further details on these trials are discussed separately. (See "Adjuvant and neoadjuvant therapy for cutaneous melanoma", section on 'High-risk node-positive disease (stage IIIB, IIIC, IIID)'.)

For patients with resected locoregional mucosal melanoma and no distant metastases, limited data are also available for adjuvant nivolumab plus ipilimumab. The use of adjuvant nivolumab plus ipilimumab for this population is best evaluated in the context of a clinical trial. In a single-arm, open-label clinical trial (SALVO), 35 patients with resected mucosal melanoma were treated with four cycles of adjuvant nivolumab (3 mg/kg) plus ipilimumab (1 mg/kg) every three weeks followed by maintenance nivolumab [133]. In this study, two-year recurrence-free and overall survival were 37 and 68 percent, respectively.

The use of adjuvant nivolumab plus ipilimumab for metastatic cutaneous melanoma after definitive surgery or radiation therapy is discussed separately. (See "Adjuvant and neoadjuvant therapy for cutaneous melanoma", section on 'Adjuvant nivolumab plus ipilimumab'.)

We do not offer toripalimab as adjuvant therapy for resected locoregional mucosal melanoma, as this approach failed to improve overall survival (OS) over adjuvant interferon alfa-2b in a randomized phase II trial [134].

SURVEILLANCE — Surveillance for patients with mucosal melanoma is extrapolated from that used for patients with cutaneous melanoma. (See "Staging work-up and surveillance of cutaneous melanoma", section on 'Surveillance'.)

While there is clinical variability in the approach to surveillance, imaging strategies for mucosal melanoma should include imaging of the chest, abdomen and pelvis, and brain [34]. For those with disease involving the head and neck, surveillance neck imaging should be included.

Surveillance should also include routine examination of the oral mucosa as well as mirror and fiberoptic examination for mucosal melanomas arising from the head and neck, proctoscopy or sigmoidoscopy for anorectal mucosal melanoma, and speculum examination for vulvovaginal melanoma. These examinations are warranted because of the risk of locally recurrent disease and the relatively high incidence of multifocal disease in mucosal melanoma.

In a retrospective study of 81 patients with mucosal melanoma identified the most common initial sites of metastasis as the lungs, liver, and nonregional lymph nodes, regardless of the primary site of disease. The rate of brain involvement as an initial site of metastasis was 7 percent, and no primary site was free of risk.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Melanoma screening, prevention, diagnosis, and management".)

INFORMATION FOR PATIENTS

UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5^th to 6^th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10^th to 12^th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.

Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)

●Basics topic (see "Patient education: Melanoma skin cancer (The Basics)")

●Beyond the Basics topics (see "Patient education: Melanoma treatment; advanced or metastatic melanoma (Beyond the Basics)" and "Patient education: Melanoma treatment; localized melanoma (Beyond the Basics)")

SUMMARY AND RECOMMENDATIONS

●Clinical presentation – Mucosal melanoma is a rare disease that arises from the mucosal epithelium lining the respiratory, alimentary, and genitourinary tracts. (See 'Introduction' above.)

•Epidemiology – It is more common in older adults and females, more likely to be multifocal and amelanotic, and generally has a worse prognosis than cutaneous melanoma. (See 'Epidemiology' above.)

•Sites of disease – The most common sites of involvement are the head and neck, vulvovaginal, and anorectal regions. (See 'Mucosal melanoma of the head and neck' above and 'Vulvovaginal melanoma' above and 'Anorectal melanoma' above.)

●Treatment of localized disease – For patients with locoregional disease of any primary site, initial management is complete wide local surgical resection. However, many still develop distant metastases despite adequate local primary disease control. (See 'Surgical considerations' above.)

●Head and neck melanoma – For patients with mucosal melanomas of the head and neck who are surgical candidates, we suggest surgery rather than radiation (Grade 2C). The surgical approach is the same as those with squamous cell carcinomas at the same anatomic site. (See 'Management of locoregional disease' above.)

•Surgical management of the neck – The approach to management of the neck is as follows (see 'Management of the neck' above):

-Sinonasal disease and clinically negative nodes – We suggest surveillance alone rather than elective lymph node dissection (Grade 2C).

-Oral cavity disease and clinically negative nodes – We suggest elective lymph node dissection rather than observation (Grade 2C).

-Clinically evident nodes – Therapeutic neck dissection can be offered based on the need for local control and symptoms, although no overall survival (OS) benefit has been demonstrated.

•Indications for adjuvant radiation therapy (RT) – For patients with resected high-risk disease (ie, palpable lymph nodes or extracapsular extension), we suggest adjuvant RT (Grade 2C). Adjuvant RT in such patients may decrease local recurrences, although no OS benefit has been demonstrated. (See 'Adjuvant radiation' above.)

•Nonsurgical candidates – For those who are not surgical candidates or in whom complete resection is not feasible, definitive RT is an alternative option for control of local disease. (See 'Nonsurgical candidates' above.)

●Vulvovaginal melanoma – For patients with vulvovaginal melanoma, we suggest wide local excision of the primary tumor rather than more radical surgeries (Grade 2C). (See 'Management of locoregional disease' above.)

●Anorectal melanoma – For patients with anorectal melanoma, we suggest wide local excision with sphincter-sparing surgery rather than abdominoperineal resection (Grade 2C). (See 'Management of locoregional disease' above.)

For patients with clinically node-negative disease, we suggest against elective inguinal lymphadenectomy (Grade 2C), given the morbidity of the surgery and lack of survival benefit. We reserve inguinal lymphadenectomy for those with clinically apparent nodal disease. (See 'Indications for inguinal lymphadenectomy' above.)

●Resected locoregional lymph nodes – For patients with positive locoregional lymph nodes treated with surgical resection, data on adjuvant therapy are limited, particularly in a Western patient population. Management decisions should be made on an individual patient basis. (See 'Resected locoregional lymph nodes (adjuvant therapy)' above.)

•BRAF wild-type tumors – For those with BRAF wild-type mucosal melanoma, options include adjuvant chemotherapy (temozolomide plus cisplatin) or adjuvant immunotherapy with nivolumab (table 2) or pembrolizumab (table 3); either treatment is reasonable in the absence of comparative data. (See 'Adjuvant chemotherapy' above and 'Adjuvant immunotherapy' above.)

Expectant observation may be offered to those who are ineligible for or decline adjuvant systemic therapy.

•BRAF mutant tumors – Patients with a BRAF V600 mutation also have the option of adjuvant targeted therapy with dabrafenib plus trametinib. (See "Adjuvant and neoadjuvant therapy for cutaneous melanoma", section on 'Adjuvant dabrafenib plus trametinib'.)