Sentinel node biopsy for malignant melanoma. Technical details and clinical results on 390 patients

BACKGROUND: The purpose of this paper is to present personal experience with sentinel node biopsy for the treatment of malignant melanoma. Technical details influencing the efficacy of the procedure are presented and the clinical, therapeutic and prognostic advantages of this technique discussed. METHODS: A total of 390 consecutive patients with primary skin melanoma (T2-3,N0,M0) underwent sentinel node biopsy between March 1996 and May 2001. All patients underwent previous excisional biopsy of the primary lesion and clinical and radiographic examination to exclude lymphatic or systemic macroscopic spreading of the disease. Preoperative lymphoscintigraphy (99mTc nanocoll) was routinely performed in the last 315 patients. Intraoperative detection of the sentinel nodes was performed by perilesional, intradermical, injection of blue dye associated with a g probe (Neoprobe(R) 2000) in the last 315 patients. Sentinel nodes, serially sectioned, were all Haematoxylin-Eosin and immunohistochemically stained. All patients positive for micro-metastasis underwent radical lymphadenectomy. Comparative analysis between the incidence of metastasis in sentinel and non-sentinel nodes, according to the clinical stage of the disease, was done. RESULTS: The overall detection rate of sentinel nodes was 97.4%. Relevant differences were found according to the site of dissection and the use of a g probe. The g-probe makes the procedure more effective, less invasive and less expensive. Timing and accuracy of the preoperative lymphoscintigraphy is a basic step of the procedure. The overall incidence of positive sentinel nodes was 14.7% with differences correlated with thickness of primary lesion (0.75-1.5 mm: 5,8%; 1.5-3 mm:18%; 3-4 mm: 24.6%). Metastasis in other non-sentinel nodes was found only with primary tumour thickness exceeding 2 mm. CONCLUSIONS: Sentinel node biopsy is a procedure requiring a multidisciplinary approach (surgery, nuclear medicine and pathology). A specific learning phase (>30 patients) is recommended to obtain reliable results.