It is known that melanoma can metastasize and recur many years after the first diagnosis. Although predictive and prognostic factors for melanoma are well defined, there is still insufficient information about the factors affecting the recurrence period and the effect of the recurrence time to survival. This study investigates the course of melanoma to show prognostic factors comparing early and late recurrence patients. The main objective is to uncover the effect of the recurrence time on the progression of the disease. In this retrospective study, late recurrence (LR) was defined as melanoma recurrence 10years after the first diagnosis and early recurrence (ER) was defined as recurrence within 10years. Gender, age, localization of primary tumour, time to first metastasis, survival rates, histological subtype, stage, tumour thickness, invasion level, ulceration and regression of the primary melanoma were documented. Survival curves were evaluated using the Kaplan-Meier and compared with the log-rank test. Multivariate Cox proportional hazard models were used to identify significant independent prognostic factors for melanoma-specific survival (MSS). A total of 1537 melanoma patients were analysed. Early metastasis was developed in 1438 patients (93.6%), and 99 patients (6.4%) developed late metastasis. Late recurrence patients were younger (P<0.001) and had fewer ulcerated (P=0.005), fewer head/neck localized (P=0.009) and thinner (P<0.001) melanomas than ER patients. The MSS time (mean±SD) was nearly identical for LR (31±4.4months 95% CI [22.3-39.7]) and ER (32±1.9months [28.3-35.7]). Multivariate regression analysis revealed male gender (hazard ratio [HR=1.4, P<0.001), truncal tumour localization (HR=1.7, P<0.001), tumour thickness (HR=1.4, P<0.045) and ulceration (HR=1.3, P<0.008) as significant independent prognostic factors for MSS. Although ER and LR patients are found to have different clinicopathologic features, the time of the first recurrence after diagnosis do not seem to have an effect on the survival.
Read full abstract