We must continue a critical reexamination of traditional treatment of any disease in the light of current understanding of its natural history. Since the inclusion of a regional lymph node dissection in the treatment of melanoma without palpable regional nodes remains a point of debate,1 it is crucial that any attempt to answer this point be made in accordance with current concepts of melanoma histogenesis and pathological characteristics.
The excellent clinicopathological correlation between morphologic presentation of the melanoma and patient survival, and between depth of invasion of the melanoma and survival, demonstrated first by Clark's group in Boston2 and confirmed by McGovern in Australia,3 cannot be ignored in evaluating efficacy of any therapeutic regimen (Table 1). Clark's terminology is precise (Figure) and suggests that usage of the word "level" instead of "stage" of invasion be adopted to avoid the ambiguity that arises with the usage of "stage,"