Reliability and accuracy of sentinel node biopsy in cutaneous malignant melanoma

AIMS AND BACKGROUND: The aims of this study were 1) to investigate whether sentinel lymph node (SLN) biopsy could become the method of choice for the early detection of metastatic disease in patients with malignant melanoma and 2) to identify those patients with lymph node metastases who could benefit from regional lymphadenectomy. METHODS AND STUDY DESIGN: Our study started in March 1998 and involved 110 patients with primary cutaneous malignant melanoma stage I or II (AJCC) in whom the primary lesion had been surgically removed no more than 90 days previously. On the day of lymph node dissection patients were given an intradermal injection of colloid particles of human serum albumin labeled with technetium-99m and an injection of isosulfan blue. The surgical procedure was usually performed with local anesthesia but in some cases locoregional or general anesthesia was preferred. Contralateral and ipsilateral lymphatic areas were scanned with a hand-held gamma camera (Scintiprobe MR 100) to measure the background and identify the hot point indicating the location of the sentinel node to direct the incision. RESULTS: The combined use of lymphoscintigraphy, isosulfan blue and gamma probe allowed us to identify sentinel nodes in 108 of 110 patients (98.18%) while the SLN was blue in only 90 cases (81.81%). The SLN was positive for metastases in 13 of the 108 patients (12.03%) and regional and distal lymphadenectomy was performed in all of them. The distribution of positive SLNs by primary lesion thickness was as follows: 0.76-1.5 mm: one positive SLN/44 patients (2.27%); 1.51-4 mm: six positive SLNs/51 patients (11.7%); > 4 mm: six positive SLNs/15 patients (40%). Only four of 12 patients with ulcerated cutaneous melanoma had positive SLNs. The patients in our study underwent follow-up visits every four months. The median follow-up was 481 days (range, 97-1271 days). CONCLUSIONS: In patients with primary cutaneous melanoma the histological status of the SLN accurately reflects the presence or absence of metastatic disease in the relevant regional lymph node basin. Complete lymph node dissection should only be performed in patients with positive SLNs. Patients with lesions > 4 mm are likely to develop recurrences and to die of systemic disease, so in these patients the usefulness of SLN biopsy is questionable. In conclusion, sentinel node mapping is a rational approach for the selection of patients who might benefit from early lymph node dissection of the affected basin.