Objective: The purposes of this study were to determine which classification best predicts additional lymph node disease and survival, and to suggest a threshold below which a completion dissection may be omitted.
Summary background data: Three micromorphometric parameters of melanoma sentinel node metastases were compared: invasion depth from the capsule (Starz-classification), maximum diameter (Rotterdam-criteria), and location within the node (Dewar-classification).
Methods: The pathology slides of 116 patients with tumor-positive sentinel nodes were reviewed. The follow-up data were obtained from the prospectively kept database. The median follow-up duration was 53 months.
Results: Metastases with an invasion depth under 0.3 mm or diameter less than 0.1 mm were not associated with additional involved nodes. Four percent of the patients with metastases with an invasion depth of 0.3 to 1.0 mm had other involved nodes and 3% of the patients with metastases with a diameter of 0.1 to 1.0 mm. Other nodes were involved in 3% of subcapsular metastases, 9% of both subcapsular and parenchymal metastases, and 33% in case of multifocal or extensive disease. The smallest tumor invasion depth and diameter associated with additional involved nodes was 0.4 mm. Only 5-year overall survival in the 3 successive invasion depth categories were statistically significant: 92%, 83%, and 68%. Five-year overall survival was 81% in patients with one involved sentinel node and 60% if there were more.
Conclusions: Invasion depth and diameter of the metastasis correlate best with the presence of additional nodal disease. Invasion depth best predicts overall survival. It seems justified to refrain from completion dissection in patients with a sentinel node tumor invasion depth up to 0.4 mm.