Lymph node tumor volumes in patients undergoing sentinel lymph node biopsy for cutaneous melanoma.

BACKGROUND: Regional lymph node tumor volumes in patients undergoing sentinel lymph node (SN) biopsy (SNB) for treatment of cutaneous melanoma have not been described. The objectives of this study were to describe the lymph node tumor volumes typically seen in this population and to correlate tumor volumes with tumor thickness and positive SN characteristics.

METHODS: Review of a consecutive series of patients with clinically localized cutaneous melanoma who underwent SNB of nonpalpable regional lymph node basins followed by complete lymphadenectomy (LND) was performed. Multiple lymph node sections from positive SNs and nonsentinel nodes (NSNs) in LND specimens were examined microscopically. Individual tumor deposit diameters were measured using an ocular micrometer. Aggregate tumor volumes were calculated for SN and LND specimens. Tumor volumes and SN and LND positivity rates were correlated with tumor thickness, the number of positive SNs, and the presence of multiple SN tumor deposits.

RESULTS: SNB procedures were performed for 149 melanomas in 189 regional nodal basins. The mean tumor depth was 2.48 mm. The mean number of SNs/basin was 2.1. Thirty-two of 149 SNB procedures (21.5%) revealed a total of 34 nodal basins with at least one positive SN. The median tumor volume in positive SNs was 4.7 mm3 (range, 0.1-3618 mm3; mean, 209 mm3). The median aggregate tumor volume in positive LND specimens was 4.9 mm3 (range, 0.1-3618 mm3; mean, 224 mm3). Six basins (17.6%) contained at least one positive NSN. The regional node aggregate tumor volume correlated weakly with tumor thickness (Pearson's correlation coefficient = .302, P = .0934). NSN positivity was not predicted by tumor thickness, American Joint Committee on Cancer tumor stage, number of positive SNs, or number of metastatic deposits within SNs.

CONCLUSIONS: Most melanoma-positive SNs contain minute tumor volumes. Tumor thickness and patterns of SN metastases may not be predictive of tumor burden or the presence of positive NSNs.