Efficacy of lymphatic mapping, sentinel lymphadenectomy, and selective complete lymph node dissection as a therapeutic procedure for early-stage melanoma

Background: Lymphatic mapping, sentinel lymphadenectomy, and selective comp lete lymph node dissection (LM/SL/SCLND) is an increasingly popular alterna tive to elective lymphadenectomy (ELND) for patients with early-stage melan oma. Although several reports have demonstrated the accuracy of the LM/SL t echnique, there are no data on its therapeutic value. Methods: We performed a matched-pair statistical analysis of 534 patients w ith clinical stage I melanoma; one half of the patients were treated with L M/SL and the other half were treated with ELND. Patients in the two treatme nt groups were matched for age (54% were less than or equal to 50 years of age), gender (63% were male patients), site of the primary melanoma (49% we re on the extremities, 36% on the trunk, and 15% on the head and neck), and thickness of the primary melanoma (7% were <0.75 mm, 42% between 0.75 and 1.5 mm, 43% between 1.51 and 4.0 mm, and 8% >4 mm). Patients in the LM/SL g roup underwent complete regional lymphadenectomy (SCLND) only if the LM/SL specimen contained metastatic melanoma. Results: The overall incidences of nodal metastases were no different (P = .18) between LM/SL (15.7%) and ELND (12%) groups, but the incidence of occu lt nodal disease was significantly (P = .025) higher among patients with in termediate-thickness (1.51-4.0-mm) primary tumors who underwent LM/SL (23.7 %) instead of ELND (12.2%). Survival data were compared by the log-rank sco re test. LM/SL/SCLND and ELND resulted in equivalent 5-year rates of diseas e-free survival (79 +/- 3.3% and 84 +/- 2.2%, respectively; P = .25) and ov erall survival (88 +/- 3.0% and 86 +/- 2.1%, respectively; P = .98). The LM /SL and ELND groups also exhibited similar incidences of same-basin recurre nces (4.8% vs. 2.1%, P = .10, respectively) and in-transit metastases (2.6% vs. 3.8%, P = .48) after tumor-negative dissections. Patients who underwen t ELND showed a higher incidence of distant recurrences (8.9% vs. 4.0%, P = .03), but this may be related to the longer follow-up period for these pat ients (median, 169 months), compared with the LM/SL-treated patients (45 mo nths). Among patients with rumor-positive nodal dissections, the 5-year ove rall survival rates were higher, and approached significance (P = .077) for patients treated by LM/SL/SCLND (64 +/- 12%) compared with ELND (45 +/- 10 %). Conclusions: These findings suggest that LM/SL/SCLND is therapeutically equ ivalent to ELND but may be more effective for identifying nodal metastases in patients with intermediate-thickness primary tumors.