Purpose: No primary lesion is identified in 10% to 20% of patients presenting with palpable evidence of regional metastatic melanoma. Because the prognostic significance of unknown primary melanoma (MUP) is unclear, we compared clinical outcomes of patients with MUP and known primary melanoma (MKP) with regional nodal metastases.
Patients and methods: We reviewed our 13,000-patient prospective melanoma database (1971 through 2005) to identify patients managed with regional lymphadenectomy for palpable nodal metastases from MUP or MKP. Multivariate analysis identified prognostic factors significant for survival. MUP and MKP were then matched by significant covariates. Overall survival (OS) was estimated by Kaplan-Meier method and compared by log-rank analysis.
Results: Multivariate analysis of data from 1,571 study patients identified four significant covariates associated with worse prognosis: age >or= 60 years (hazard ratio [HR] = 1.294; P = .0017), male sex (HR = 1.335; P = .0004), nodal tumor burden >or= one (HR = 1.256; P < .0001), and known primary (HR = 1.507; 95% CI, 1.220 to 1.862; P = .0001). Five-year OS was significantly higher for 262 patients with MUP than for 1,309 patients with MKP (55% +/- 6% v 44% +/- 3%; P = .0021). Computerized matching of MUP and MKP by four significant covariates (age, sex, nodal tumor burden, and decade of diagnosis) yielded 221 matched pairs. Median and 5-year OS rates were 165 months and 58% +/- 7%, respectively, for MUP as compared with 34 months and 40% +/- 7%, respectively, for MKP (P = .0006).
Conclusion: Lymphadenectomy is effective for nodal metastasis from MUP. The significantly better postoperative survival for MUP versus MKP suggests a strong endogenous immune response against the primary melanoma. Immunologic studies to identify cell-mediated and antibody components of this response may lead to new approaches for determining melanoma prognosis and treatment.