Sentinel lymph nodes in a swine model with melanoma: contrast-enhanced lymphatic US.

Abstract

To determine if lymphatic channels and sentinel lymph nodes (SLNs) with and without metastases can be detected with lymphatic ultrasonography (US) after peritumoral injection of a US contrast agent and to determine if lymphatic US can be used to assess SLNs for the presence of metastatic infiltration. Six swine with 17 melanomas were evaluated. Conventional gray-scale, color flow, and gray-scale phase-inversion harmonic US examinations were performed. A US contrast agent was administered in four sites around each melanoma (1-mL total dose). Lymphoscintigraphy was followed by injection of a blue dye and then dissection. SLNs identified at lymphatic US were characterized by two readers in consensus as normal or as having metastases; results were compared with histologic findings. Statistical analyses included the sign test and the kappa statistic. Lymphatic US depicted 28 SLNs, while lymphoscintigraphy depicted 27 "hot spots" suspected of representing SLNs (including two false-positive findings). Dissection after blue dye injection helped identify 31 SLNs. There were no false-positive US findings for SLN detection. Five of six nodes not seen with lymphoscintigraphy were detected with lymphatic US. The accuracy of SLN detection was 90% (28 of 31) for lymphatic US and 81% (25 of 31) for lymphoscintigraphy (P =.29). Lymphatic US correctly depicted metastases in 19 of 20 SLNs, and five of the eight normal SLNs were correctly characterized, with an accuracy of 86% (kappa = 0.62). Detection of SLNs with lymphatic US compared favorably with that at lymphoscintigraphy. Lymphatic US can depict metastases within the SLN, which was not possible with lymphoscintigraphy.