Aims Sentinel lymph node biopsy (SLNB) has been widely accepted as the lymph node sampling procedure of choice for melanoma patients. Completion lymph node dissection (CLND) is the current standards of practice suggest for patients with a positive SLNB result. The purpose of this study is to compare the demographic characteristics, melanoma features and parameters of the sentinel lymph node (SLN) to identify which are the most important for the prediction of non sentinel lymph node (NSLN) involvement and to correlate with the clinical course in terms of disease-free survival (DFS) and overall survival (OS).
Patients and methods SLN involvement was found in 100 patients treated in our department since 2000. Univariate and multivariate logistic regression were used to test the significance of the predictor variables’ associations with both SLN and NSLN metastasis. Cox proportional multivariate analyses were carried out to evaluate the independent prognostic variables on OS and DFS. The variables included were: demographic data (sex, age); melanoma features (Breslow thickness, histologic subtype, ulceration); histological SLN data included: number of positive SLNs, localisation of SLN, pattern of invasion (micro/macro metastases), depth invasion of metastatic cells in the SLN (Starz Classification 2004). The three SLN variables were grouped together in a new variable (SLN index risk) ranging from 0 to 3, which takes into account the number of adverse prognostic indicators.
Results A significantly higher risk of NSLN involvement was associated in univariate analysis to the presence of SLN macro-metastases and Starz stage III. Multivariate analysis confirmed the independent role of the SLN index risk together with Breslow thickness and ulceration. Multivariate OS analysis identified the SLN index risk as the most significant adverse prognostic indicator, followed by Breslow thickness. The same parameters were found to be independent prognostic factors on DFS.
Conclusion This study provides a straightforward and easily applicable means of identifying melanoma patients with a very high likelihood of harbouring NSLN metastases. Future studies on larger patient cohorts and longer follow-up time are needed to confirm these data, which could help, whenever confirmed, to identify patients who will benefit from CLND and others who on the other hand could be spared by a procedure with a relevant degree of potential co-morbidities.