Sentinel node dissection in the treatment of melanoma: Report of three cases and review of the literature

BACKGROUND. Elective lymph node dissection for treatment of cutaneous malignant melanoma is controversial. Sentinel node dissection involves removing the primary lymph node in a nodal basin that drains a particular cutaneous lesion. Theoretically, this node would collect regional metastases first. Therefore, if this node is negative, the chances are low that the melanoma would have spread either systemically or to other nodes within this basin. Removing one node would decrease the morbidity associated with radical lymph node dissection. However, the actual risk of widespread metastases in sentinel node-negative patients is yet to be determined, pending results of large, multi-center studies currently under investigation. OBJECTIVE. To present three cases of intermediate thickness cutaneous melanoma treated with selective lymph node dissection and to review the techniques of selective lymphadenectomy and lymphoscintigraphy. METHODS. In a nonrandomized prospective evaluation, patients with intermediate depth melanomas or in transit metastases without signs of systemic disease were given the opportunity for further investigation by sentinel node dissection to determine if additional lymph node dissection or adjunctive therapies would be advantageous. RESULTS. All three patients had negative sentinel node examinations. Two are without visceral or nodal metastases 1 year after the procedure. The third had in-transit metastases from the outset, had in- transit metastases on sentinel node/lymphatic examination, and now has systemic cutaneous metastases. Complications of sentinel node dissection were limited to transient postoperative lymphedema of the extremities and transient seroma formation. No postoperative wound infections or permanent nerve damage were noted. CONCLUSIONS. Our preliminary findings in this limited series suggests that sentinel node dissection appears to be a procedure of low morbidity and relatively high predictive value.