Reliability of lymphoscintigraphy in indicating the number of sentinel nodes in melanoma patients.

Abstract

This study was undertaken to establish the reliability of lymphoscintigraphy in indicating the number of sentinel nodes in patients with melanoma. Lymphoscintigraphy was performed with dynamic imaging after injection of 60 MBq 99mTc-nanocolloid (1.6 mCi) and static imaging after 2 hours in 200 patients with clinically localized primary melanoma of the skin. The following day, sentinel nodes were retrieved with the blue dye technique and a gamma detection probe (Neoprobe 1000/1500). The discrepancies between the number of sentinel nodes indicated by lymphoscintigraphy and the actual number of sentinel nodes as established by the surgeon were evaluated. Lymphoscintigraphy showed drainage to 393 sentinel nodes in 255 lymphatic fields in 199 patients. In 48 lymphatic fields (19%) in 46 patients (23%), the number of sentinel nodes was different from the number that was visualized with scintigraphy. Additional sentinel nodes were found by the surgeon because a lymphatic vessel was not seen on the lymphoscintigraphy (43%), because a sentinel node was not visualized separately from other hot nodes or vessels or the injection site (36%), or because a sentinel node was blue and not hot (4%). Fewer sentinel nodes were found than suggested by scintigraphy because a lymphangioma was mistaken for a sentinel node (4%) or because a single elongated node was depicted as two hot spots (6%). Although lymphoscintigraphy is indispensable for lymphatic mapping, the predicted number of sentinel nodes is accurate in only 81% of lymph node fields. The limited discriminating power of the gamma camera is an important cause of discrepancies.