A single institute's experience on melanoma prognosis: a long-term follow-up.

Background: The prognosis of cutaneous melanoma is correlated to histopathologic parameters such as Breslow thickness, the presence of mitosis, ulceration and lymphatic involvement at the moment of the diagnosis. On the other hand, the prognostic value of parameters such as age, sex, and tumor localization are still a matter of debate. We evaluated herein the prognostic factors in melanoma patients during a long-term follow-up (60 months).

Methods: Melanoma patients presenting stage IB-III at diagnosis were included. Breslow thickness, ulceration, lymphatic involvement, patients' age, sex and tumor localization were correlated to patients' prognosis. Univariate Cox regressions and multivariate Cox proportional-hazards regression were performed. Successively, Kaplan-Meier was used for variables significantly associated with overall melanoma survival.

Results: A total of 115 melanoma patients were included in this study. During follow-up 82 (72.17%) patients survived and 33 (28.7%) died. In our dataset, Breslow thickness >2 mm (P=0.0007), patients age >50 years (P=0.005) and positive sentinel lymph node (P=0.0003) seem to be the most important variables correlated with the presence of metastases at 5 years follow-up. However distant metastases were also observed during follow-up in 14/26 patients presenting negative sentinel lymph node at diagnosis.

Conclusions: Given the vital importance of target drugs and the newly introduced immunotherapies in cutaneous melanoma management, we would suggest that mutational analyses should also be extended to the subgroup of patients presenting microstaging parameters related to a poor prognosis in a long-term follow-up of 60 months.