Reliability of identification of 655 sentinel lymph nodes in 263 consecutive patients with malignant melanoma

Background The purpose of this study of sentinel lymph node biopsies (SLN) was threefold: to compare the reliability of lymphazurin blue dye to radioactive technetium 99m sulfur colloid (TC); to evaluate the reliability of frozen section examinations of sentinel lymph nodes; and to determine how much SLN dissections prolonged operative time. Study design We evaluated the records of 263 consecutive patients with intermediate and high-risk melanomas (1.0 mm or thicker, or Clark Level IV or greater), who were treated by a single surgeon at the Yale Melanoma Unit between October 1, 1997, and September 30, 2001, and followed for more than 18 months. Results A total of 655 SLN were identified and removed from these 263 consecutive patients. Radioactive colloid was found to be more reliable (100%) in identifying the SLN than lymphazurin blue dye (51%) in the nodes of the patients. Twenty-eight patients (11%) had positive sentinel lymph nodes, and 2 patients (7%) had false-negative frozen sections. Three patients (11%) had false-negative frozen sections; tumor was found subsequently on permanent sections only after special immunohistochemical stains were used. The location or removal of SLN did not prolong the operative procedure unreasonably, requiring an average of 7 to 20 minutes for removal of SLN, and 33 minutes for frozen section reports, during which time the primary tumor resection and wound coverage were performed. Conclusions SLN were found in all 263 patients. All SLN were identified reliably with radioactive colloid. Because blue dye was found in only half of the radioactive nodes, it is not appropriate to use this as the only marker for locating the SLN. This large series of patients attests to the reliability of frozen sections in identifying SLN harboring metastases, with 82% of the patients with nodal metastases identified in this fashion.