Introduction
Mucosal malignant melanoma of the sinonasal tract is very rare, representing less than 4 per cent of nasal fossa neoplasms.Reference Enée, Houliat and Truilhé1, Reference Huang, Liao, Kan and Chen2 This tumour's nonspecific clinical presentation often delays diagnosis.Reference Ganly, Patel, Singh, Kraus, Bridger and Cantu3 The prognosis for patients with this tumour is dismal, with reported five-year overall survival rates ranging from 19 to 31 per cent.Reference Bhattacharyya4 The high incidence of sinonasal melanoma local recurrence can be attributed to submucosal lymphatic spread of the disease.Reference Ganly, Patel, Singh, Kraus, Bridger and Cantu3 In the seventh edition of the seventh edition Union for International Cancer Control (7th UICC) tumour-node-metastasis (TNM) staging system, a specific classification was established for mucosal melanoma of the upper respiratory tract, comprising only stages T3 and T4; the tumour is designated as T3 even when localised to the mucosa, to emphasise its aggressive behaviour.
Most studies of mucosal malignant melanoma are case reports. No consensus has yet been established on the optimal management of mucosal melanomas of the sinonasal tract. Factors affecting local control of the disease have not yet been clearly delineated.
In this paper, we report a retrospective series of 25 cases of sinonasal tract mucosal malignant melanoma managed in the same head and neck department. Our study aimed to describe the treatment of this disease, and to assess potential prognosis factors affecting local control and survival.
Materials and methods
We performed a retrospective analysis of patients with sinonasal mucosal malignant melanoma managed in the otorhinolaryngology and head and neck surgery department of the University Hospital of Lille between January 1991 and July 2008. Twenty-five consecutive patients with mucosal malignant melanoma of the sinonasal tract were identified. The following clinical data were recorded: gender, age at diagnosis, functional symptoms at presentation, and delay in aetiological diagnosis. The tumour localisation (i.e. nasal fossa, septum, lateral nasal wall or maxillary sinus) and extension (i.e. intra- or extra-nasal spread) were established from nasal endoscopy and imaging (i.e. tomodensitometry and/or magnetic resonance imaging (MRI)). The seventh edition of the 7th UICC TNM staging system for sinonasal cancers was used for tumour staging.
Histological diagnosis was established from biopsy specimens obtained via nasal endoscopy performed under local or general anaesthesia. Each specimen underwent specific immunostaining for vimentin, S100 protein and HMB 45.
Treatment of the tumour and any lymph nodes was recorded (i.e. surgery with or without radiotherapy, radiotherapy alone, or chemotherapy). The surgical approach was defined as endonasal or external. The surgical margins were analysed on the definitive specimen. No frozen section analyses were performed.
Patient follow up was based on repeated clinical examination (including nasoendoscopy) and imaging (i.e. sinonasal MRI) every six months. Distant metastases were investigated based on clinical signs (i.e. neurological or hepatic signs or osseous pain) and chest X-ray.
Data were recorded using Microsoft® Excel software, and statistical analysis performed using Medcalc® software. The Kaplan–Meier method was used for actuarial survival analysis. End-points included time to local relapse, regional recurrence, distant metastasis and death. Survival curves were compared using the log rank test. The accepted degree of significance was p < 0.05.
Results
Epidemiological data
From 1991 to 2008, 25 patients were referred to our institution for mucosal malignant melanoma of the sinonasal tract. The sex ratio was 0.9:1 (12 men and 13 women). The median age was 68 years (range, 29 to 90 years). We detected no recurrent environmental or occupational factors. At presentation, the main symptoms were unilateral nasal obstruction (77 per cent) and intermittent epistaxis (68 per cent). The mean delay to diagnosis was 3.5 months (range, one to 17 months). The nasal endoscopy was nonspecific in all patients. More than half of the mucosal melanomas were exophytic, polypoidal lesions; the remaining were sessile, pigmented lesions with or without ulceration.
Local extension
The tumour was located in the lateral nasal wall (i.e. nasal turbinates, intersinonasal septum or ethmoid sinus) in 11 cases (44 per cent), in the maxillary sinus in four cases (16 per cent), in the nasal septum in five cases (20 per cent) and in the floor of the nasal fossa in one case (4 per cent). It was located outside the nasal fossa and sinuses in four cases (16 per cent).
Patients' tumour staging was T3 in six patients (24 per cent), T4a in eight (32 per cent) and T4b in 11 (44 per cent). One patient was N1. No patients had distant metastasis.
Treatment
A surgical procedure with curative intent was performed in 23 patients (92 per cent). An external approach was used in 12 patients (52.2 per cent) and an endonasal endoscopic approach in 11 patients (47.8 per cent). Two of these 23 patients had neck dissection performed as part of the initial treatment (Figure 1).
Flow diagram showing treatment and follow up of the 25 patients. Surgery was the first line of treatment. Local recurrence occurred regardless of external radiotherapy (RT) of the tumour site. T = tumour; N = node
Of the 23 patients who received surgery, 15 received post-operative radiation therapy (to the tumour site in 13 cases, and to both the tumour site and the neck in two cases due to positive lymph nodes on dissection). Adjuvant chemotherapy was never performed. One patient was treated with radiotherapy only, and one patient was managed symptomatically for advanced local tumour (Figure 1).
Follow up and survival analysis
The mean patient follow up was 31.3 months (range, two to 99 months). The three-year overall patient survival was 100 per cent for T3 patients and 52.1 per cent for T4 patients; this difference was not statistically significant (p = 0.14). Disease-free survival was 75 and 27.5 per cent for T3 and T4 patients, respectively; this difference was not statistically significant (p = 0.15).
Recurrence was observed in 17 of the 23 patients who received curative treatment (73.9 per cent), with a median delay of 24.6 months (range, 1.5 to 97 months). Local recurrence occurred in 14 cases. Of these, one patient had concomitant lymph node metastases and one patient was found to have distant metastasis on follow up. There was one case of regional recurrence with hepatic metastasis but without local relapse. There was one case each of nodal recurrence and distant metastasis (Figure 1).
Of the six T3 patients, five had melanoma of the nasal fossa (either lateral nasal wall, nasal floor or septum) treated via an endoscopic approach, while one had melanoma of the maxillary sinus treated with a sublabial transfacial procedure. The small size of the T3 group did not allow survival analysis based on tumour location, treatment or surgical margins.
Of the 17 T4 patients treated with curative intent, data were available on 16 for actuarial survival analysis. Three-year disease-free survival was significantly better for patients with tumours of the nasal fossa, versus tumours elsewhere (p = 0.012), and for patients treated with surgery and adjuvant radiotherapy, versus other treatments (p = 0.05), without impacting on three-year overall survival. The surgical approach and the quality of the surgical margins did not affect the three-year disease-free and overall survival rates (Table I).
3-Year survival analysis for T4 sinonasal melanoma patients
Data for 17 patients were analysed. *Log rank test. T = tumour stage; pts = patients; OS = overall survival; DFS = disease-free survival; RT = external radiotherapy
Discussion
The aetiopathology of mucosal malignant melanoma of the sinonasal tract is not yet well understood. In contrast to cutaneous melanoma, no risk factors have been identified.Reference Owens, Roberts and Myers5, Reference Prasad, Busam, Patel, Hoshaw-Woodard, Shah and Huvos6 Our study identified no likely environmental or occupational factors. This tumour generally occurs after 50 years of age, and has no sex predominance.Reference Enée, Houliat and Truilhé1, Reference Huang, Liao, Kan and Chen2, Reference Lee, Shimizu, Tran, Juillard and Calcaterra7
Macroscopic examination of mucosal melanoma is not very helpful for aetiological diagnosis. Furthermore, histopathological diagnosis is difficult, not only because of the low prevalence of the tumour but also because the tumour lesions can be amelanotic, with negative Fontana staining.Reference Prasad, Busam, Patel, Hoshaw-Woodard, Shah and Huvos6 Thus, immunostaining with specific markers is widely used.Reference Huang, Liao, Kan and Chen2, Reference Pomar Blanco, San Román Carbajo, Bouso Montero, Martín Villares, Fernández Pello and Tapia Risueño8 Usually, melanoma reacts positively to antivimentin antibodies, and strongly positively to antibodies to S100, a calcium-binding protein found in neural tissue. It also reacts strongly to HBM 45 and Melan-A, which are specific monoclonal antibodies prepared using malignant melanoma extract.Reference Owens, Roberts and Myers5, Reference Lee, Shimizu, Tran, Juillard and Calcaterra7, Reference Trapp, Fu and Calcaterra9
Regardless of sinonasal mucosal melanoma treatment, studies have reported high rates of local recurrence (31–85 per cent) and distant metastasis (25–50 per cent), and poor three- and five-year survival rates (19–31 per cent).Reference Enée, Houliat and Truilhé1, Reference Huang, Liao, Kan and Chen2, Reference Dauer, Lewis, Rohlinger, Weaver and Olsen10 The limited size of most reported series and the heterogeneity of available data make it difficult to assess the effectiveness of different treatment modalities, as regards tumour location and survival rate.Reference Meleti, Leemans, de Bree, Vescovi, Sesenna and van der Waal11 Our study, performed over 17 years, identified only 25 mucosal melanoma cases with available data.
Complete surgical resection is generally accepted to be the best treatment approach for sinonasal mucosal malignant melanoma, in order to achieve prolonged survival and cure.Reference Enée, Houliat and Truilhé1, Reference Huang, Liao, Kan and Chen2, Reference Trapp, Fu and Calcaterra9 Precise delineation of the tumour extent is required, using nasal endoscopy and imaging, in order to ensure appropriate surgical intervention.
Tumour stage classification is frequently cited as a factor associated with poor prognosis.Reference Bhattacharyya4 In a retrospective study of 46 sinonasal mucosal melanoma cases managed between 1979 and 1997, Temam et al. identified T classification as an independent factor associated with local control.Reference Temam, Mamelle, Marandas, Wibault, Avril and Janot12 In our study, patient survival tended to be better for T3 cases compared with T4 cases, although this difference was not statistically significant.
Tumour location has also been cited as a factor affecting disease control. In our study, five of our six T3 melanomas were localised; in our T4 patients, local control was better in patients with tumours involving only the nasal fossa, compared with other sites. Tumours occurring in the sinuses may grow asymptomatically and be detected late in the disease course, while tumours arising in the nasal cavity are usually diagnosed at an early stage, with less local infiltration. The latter tumours are more easily accessible, theoretically resulting in better outcomes. Our findings agree with those of Dauer et al., obtained from a review of 61 cases.Reference Dauer, Lewis, Rohlinger, Weaver and Olsen10
When planning surgical treatment of sinonasal mucosal malignant melanoma, the exact tumour size and location must be established in order to ensure complete resection. The transfacial approach is widely used for oncological resection. Due to recent technical advances allowing improved surgical view, the endoscopic endonasal approach is also relevant in skilled hands and for appropriate indications. In our study, this method was used for T3 melanoma of the nasal fossa, and for T4 melanoma accessible for resection and for endoscopic drilling of bony borders. We found no difference in local control of T4 lesions, comparing external versus endoscopic approaches. Endoscopic resection may be used in lesions without extension to the glabella area and the lateral wall of the maxillary sinus.
The quality of the surgical specimen resection margins is thought by many to be associated with local control. However, in our study we observed no impact of this parameter on survival. However, given the small number of patients with positive margins in our study (five), this result should be taken with caution. Moreover, the intricate, three-dimensional nature of the nasal cavity and adjoining structures makes margin interpretation difficult.Reference Ganly, Patel, Singh, Kraus, Bridger and Cantu3, Reference Bhattacharyya4
The effect of neck lymph nodes is not clear, particularly in N0 patients. The reported incidence of positive lymph nodes in sinonasal mucosal malignant melanoma patients is 10 to 30 per cent.Reference Enée, Houliat and Truilhé1 Some authors have recommended systematic selective neck dissection (e.g. supraomohyoid dissection), on the basis of evidence showing poorer overall survival and distant metastasis free survival in the presence of pathological neck lymph nodes.Reference Huang, Liao, Kan and Chen2, Reference Temam, Mamelle, Marandas, Wibault, Avril and Janot12 Our study included only two cases involving neck dissection, so the impact of this on survival was not measurable. Other authors have reported that systematic neck dissection in N0 patients is not significantly associated with survival.Reference Stárek, Koranda and Benes13 The role of sentinel lymph node biopsy has not been evaluated.Reference Stárek, Koranda and Benes13
• Sinonasal mucosal malignant melanoma is rare
• Diagnosis may be delayed by presentation with nonspecific nasal symptoms mimicking chronic rhinosinusitis
• Treatment consists of radical excision (endoscopic or external approach) guided by magnetic resonance imaging and tomodensitometry
• Adjuvant radiotherapy seems to improve local control, but does not affect overall survival
External radiotherapy has been reported as an adjuvant treatment for sinonasal mucosal malignant melanoma, but its effect has not been systematically investigated. Mucosal malignant melanomas were previously thought to be radioresistant.Reference Enée, Houliat and Truilhé1 However, recent radiobiological studies have demonstrated wide variation in the radiosensitivity of melanoma cell lines.Reference Huang, Liao, Kan and Chen2, Reference Lund, Howard, Harding and Wei14 Temam et al. found that radiotherapy appeared to increase the local control rate of mucosal melanoma of the head and neck, independent of primary tumour stage, albeit with no impact on overall survival.Reference Temam, Mamelle, Marandas, Wibault, Avril and Janot12 However, distant metastasis occurred earlier and more frequently in patients with locally advanced tumours, and this could explain why the authors observed no effect of post-operative radiotherapy on overall survival.Reference Meleti, Leemans, de Bree, Vescovi, Sesenna and van der Waal11, Reference Temam, Mamelle, Marandas, Wibault, Avril and Janot12 The same result was described by Owens et al., who studied 48 head and neck mucosal melanoma cases and observed improved locoregional tumour control after the addition of radiotherapy to surgical treatment.Reference Owens, Roberts and Myers5 In our study, the same result was observed for patients with T4 lesions. In other studies, the impact of external radiotherapy on local control has not been established.Reference Ganly, Patel, Singh, Kraus, Bridger and Cantu3, Reference Lund, Howard, Harding and Wei14–Reference Wu, Gomez, Zhung, Chan, Gomez and Wolden16 Because disease is usually localised at presentation, and local disease recurrence often precedes distant metastasis, many centres have introduced adjuvant radiotherapy into their mucosal melanoma treatment protocol, even in the absence of a demonstrated survival benefit.Reference Lee, Shimizu, Tran, Juillard and Calcaterra7, Reference Dauer, Lewis, Rohlinger, Weaver and Olsen10, Reference Temam, Mamelle, Marandas, Wibault, Avril and Janot12 Adjuvant radiotherapy has been used more frequently when the margins of resection are invaded, or when neck dissection lymph nodes are histologically positive.Reference Huang, Liao, Kan and Chen2, Reference Wagner, Morris, Werning and Mendenhall17
Conclusion
Mucosal malignant melanoma is a rare but aggressive type of sinonasal tumour. Treatment aims to achieve radical resection with safe margins; however, this is often difficult due to the anatomical constraints of the nasal cavity. Thus, overall survival and local control rates are still poor. No adjuvant therapy has yet been shown to improve prognosis, although radiotherapy appears to enhance local control. Patients with sinonasal mucosal malignant melanoma are at high risk of local recurrence, and require frequent and regular follow up.
