Loco-regional recurrent melanoma: I Natural history

Abstract

<h2>Summary</h2> Loco-regional recurrent melanoma as a clinical entity includes (a) local recurrence, (b) in-transit metastasis or satellitosis, and (c) regional lymph node involvement. Among patients with recurrent melanomas 10–20% have local recurrence, 12–20% in-transit metastasis, 40–60% nodal involvement (Type C±A/B), and 10–30% distant metastasis (Table 1). For patients with Stage I melanomas, 80% of the local and regional nodal relapses occurred before 2 years, while in-transit and systemic metastasis appeared later. With surgical re-treatment, 40% of patients with local recurrence and 20–30% of those with nodal metastasis will be free of disease 5 years later (Table 2). The survival results of patients with satellitosis are more variable depending on: (i) sex of the patient with females having a better prognosis, (ii) site of primary melanoma with extremities have a better prognosis, (iii) number of satellites and distance from primary site, (iv) presence or absence of metastasis in the regional nodes, either concurrently or at the time of initial lymphadenectomy. For all loco-regional relapses, those that appear soon after initial treatment have a poorer prognosis. The incidence of loco-regional recurrences is higher when melanomas: (i) occur in the head and neck region, (ii) are located distal to elbows and knees, (iii) are of clinical Stage II rather than I, (iv) present with pathological proven metastasis in regional nodes. A common denominator for loco-regional relapse is inadequate treatment of the primary lesion and/or regional nodes. Recently, depth of microinvasion of the primary melanoma has been shown to be the most important prognostic factor. It not only correlated with the survival rates and incidences of occult nodal and systemic metastasis, it also predicts accurately the chance of loco-regional recurrences. For instance, patients with Clark Level II lesions had an overall relapse rate less than 5% at 5 years, while over 70% of those with Level V lesions recurred, including 50% with systemic dissemination, and 20% with locoregional relapses (Table 4). The suggestion that lymphadenectomy and its associated lymphedema <i>per se</i> increased in-transit metastasis or satellitosis is not proven. Studying 2549 patients with clinical Stage I melanoma at initial treatment (Table 5), the relapse rate in regional nodes (Type C) varied from 0 to 5% of those who had prophylactic lymphadenectomy compared with 2–38% of those who were only observed. The incidence of local and intransit metastasis (Types A and B) were 2–5% (median—5%) for those with nodal resections against 4–20% (median—8%) for those without. Since all patients in these 9 series were treated with surgery alone without any adjuvant therapy it can be concluded that prophylactic lymphadenectomy not only decreased regional nodal recurrences, but also local and in-transit metastasis. The indications and selection criteria of high-risk patients who benefited from prophylactic lymphadenectomy are not discussed in this review [interested readers may refer to Lee (21)]. Currently, there is no evidence that delaying node dissection for 4–6 weeks after excision of the primary lesion will decrease loco-regional recurrence or improve survival rates. There is some data showing that in-continuity dissection of the nodes might give better local-regional controls than discontinuous node resections. Therefore, whenever feasible, an in-continuity nodal dissection is preferred. But when the primary melanoma on an extremity is separated from the regional lymphatic drainage by a major joint, excision of large amounts of skin, intervening subcutaneous tissue and fascias seems excessive and is not recommended.