Hazard-rate analysis and patterns of recurrence in early stage melanoma: moving towards a rationally designed surveillance strategy.

BACKGROUND: While curable at early stages, few treatment options exist for advanced melanoma. Currently, no consensus exists regarding the optimal surveillance strategy for patients after resection. The objectives of this study were to identify patterns of metastatic recurrence, to determine the influence of metastatic site on survival, and to identify high-risk periods for recurrence. METHODS: A retrospective review of the Duke Melanoma Database from 1970 to 2004 was conducted that focused on patients who were initially diagnosed without metastatic disease. The time to first recurrence was computed from the date of diagnosis, and the associated hazard function was examined to determine the peak risk period of recurrence. Metastatic sites were coded by the American Joint Committee on Cancer (AJCC) system including local skin, distant skin and nodes (M1a), lung (M1b), and other distant (M1c). RESULTS: Of 11,615 patients initially diagnosed without metastatic disease, 4616 (40%) had at least one recurrence. Overall the risk of initial recurrence peaked at 12 months. The risk of initial recurrence at the local skin, distant skin, and nodes peaked at 8 months, and the risk at lung and other distant sites peaked at 24 months. Patients with a cutaneous or nodal recurrence had improved survival compared to other recurrence types. CONCLUSIONS: The risk of developing recurrent melanoma peaked at one year, and the site of first recurrence had a significant impact on survival. Defining the timing and expected patterns of recurrence will be important in creating an optimized surveillance strategy for this patient population.