Primary mucosal melanoma of the head and neck: a proposal for microstaging localized, Stage I (lymph node-negative) tumors.

BACKGROUND

The current study was conducted to identify histologic predictors of survival in patients with localized, lymph node-negative (Stage I, N0M0) primary mucosal melanomas of the head and neck (HNMM).

METHODS

The histology of 39 sinonasal, 20 oral, 1 pharyngeal, and 1 laryngeal Stage I HNMM was reviewed by 2 pathologists without knowledge of patient outcome. The invasion was evaluated as Level I: melanoma in situ (without invasion or with microinvasion only); Level II: invasion into the lamina propria only; and Level III: invasion into deep tissue (e.g., skeletal muscle, bone, or cartilage). The tumor architecture was defined as pseudopapillary when tumor cells clustered around blood vessels resembling papillae and sarcomatoid when they resembled high-grade pleomorphic sarcoma. Survival analysis was performed with Kaplan-Meier survival curves using disease-specific survival (DSS) as the endpoint.

RESULTS

The 5-year DSS rate was 43% (median, 41.5 months). The median survival was found to decrease significantly with increasing level of invasion: Level I (n = 4): 138 months; Level II (n = 29): 69 months; and Level III (n = 28): 17 months (P = 0.003). The presence of pseudopapillary and sarcomatoid architecture (n = 20) and undifferentiated cells (n = 16) were found to be associated with a significantly poor DSS (P < 0.05). However, on multivariate analysis, only the level of invasion remained an independent predictor of survival (P = 0.03). Tumor thickness, vascular invasion, and necrosis were found to have no significant influence on survival.

CONCLUSIONS

Microstaging according to invasion into three tissue compartments was found to be a significant and independent predictor of poor survival in patients with localized, lymph node-negative, Stage I HNMM. The presence of sarcomatoid and pseudopapillary architecture and undifferentiated cells also appear to be associated with significantly poor DSS.