Can sentinel node biopsy be safely omitted in thin melanoma? Risk factor analysis of 1272 multicenter prospective cases

Background: The indication to sentinel node biopsy (SNB) for thin melanomas (Breslow <1 mm) is still subject to controversies. The aim of this paper is to review all SNB performed for thin melanoma and to analyze factors related to lymphatic metastasis. Moreover, the diagnostic performance of the 5th, 6th, 7th and 8th AJCC classifications for cutaneous melanoma were investigated. Methods: All sentinel node biopsies performed for thin melanomas were selected from a multicentre prospectively-collected database. For each patient the following was collected: age, sex, date of treatment, site of primary melanoma, histopathologic features (Breslow, Clark, number of mitoses/mm 2 , presence of ulceration) and the results of the sentinel node biopsy. Results: From 1998 to 2017 were performed a total of 1272 SNB for thin melanoma. Mean age was 51years with 48.7% of male patients. Overall, 5.6% positive SNB were found. At univariate and multivariate analyses, Breslow thickness and ulceration were related to the presence of lymphatic metastasis. We compared the four versions of the AJCC classification: among pT1a patients there were respectively 5.32%, 5.63%, 3.72% and 3.49% of positive SNB. Conclusions: in thin melanoma Breslow thickness and ulceration were the only factors related to a positive SNB. Although convincing improvements resulted from the implementation of AJCC classifications with a reduction of positive biopsies among pT1a, a 10.71% rate among all positive nodes remains in the low-risk group. No recommendations can be drawn from this research and adjunctive evidences are needed to better identify patients at risk of nodal metastasis.