Role of Sentinel Node Biopsy in the Management of Malignant Melanoma

Drs. North and Spellman concisely review the role of sentinel node biopsy in the management of patients with malignant melanoma and provide an excellent summary of the current state of this technique. A number of comments should be made about this review. These comments relate to (1) the technical aspects of the procedure and (2) its clinical indications. Technical Aspects With regard to the former, I disagree with the authors' contention that lymphoscintigraphy is unnecessary in the management of patients with extremity melanoma. In about 5% of patients with extremity melanoma, lymphatic drainage is found to be other than antici-pated based on anatomy alone. Lymphoscintigraphy clearly defines these abnormal drainage patterns. More importantly, lymphoscintigraphy is a prerequisite for the use of a hand-held gamma counter to precisely locate the sentinel node; even in the setting of a predictable nodal drainage pattern, preoperative lymphoscintigraphy with intraoperative use of the hand held gamma counter markedly shortens the length of time required to find the node, lessens the extent of dissection required, and, after removal of the sentinel node, permits counting over the nodal basin to detect relevant residual nodes. Furthermore, as demonstrated by Reintgen's group, the combination of blue dye and radioisotope leads to a higher success rate in the detection of the sentinel node [1]. The timing of the radioisotope injection relative to surgical dissection is important. Most investigators would agree that the ideal interval is between 1 and 4 hours. In the series of Krag et al, a number of patients had the injection up to 24 hours prior to surgery [2]. This long interval may permit "pass through" of radioisotope into second-echelon nonsentinel nodes. If one uses the hand-held gamma counter to ensure that all significant radioactivity has been removed from the nodal drainage basin, this can lead to excision of multiple second-echelon nonsentinel nodes. As the authors note, McCarthy et al have criticized the choice of colloid used. Although it is true that sulfur colloid has particles of varying size, the larger particles can easily be excluded from the injection at the primary site by microfiltration with a 22or 10-mcm filter. We have found this technique to be quite satisfactory, and associated with prompt migration of the isotope. Antimony, advocated as an alternative by McCarthy et al, is not currently available in the United States and is not widely used. The authors describe the amount of blue dye injected at the primary site as 2 to 3 mL. In most instances, very satisfactory staining of the afferent lymphatics can be achieved with a much smaller injection, ie, 0.5 to 1.0 mL. This minimizes tissue staining at the site of the primary, decreasing the likelihood of locally persistent dye after wide excision. The role of immediate "frozen-section" immunostaining of sentinel nodes also is as yet undefined. At centers where this procedure is not done routinely, it is unlikely that pathology departments will commit to the overhead necessary to have this investigation available, particularly as it does not seem to have an impact on long-term regional control or survival. It would appear that, for most centers, immediate frozen section with hematoxylinand eosin (H & E) staining is appro-priate, reserving immunohistochemical evaluation to be done subsequently on the permanent sections. If the sentinel node is found to contain metastatic disease on subsequent evaluation by either permanent section H&E or immunohistochemical staining, there does not appear to be any detriment to the patient to going back and performing a selective lymph node dissection at that time. Indeed, immediate frozen section of the node may be unnecessary. Moreover, there may be