The article titled “Patterns of Recurrence in Patients with Melanoma after Radical Lymph Node Dissection” by Nathonsohn et al1 is highly instructive and educational. This article discussed the patterns of recurrence of melanoma after radical lymph node dissection (RLND), but the authors have identified and have discussed the prognostic factor of recurrence of melanoma after RLND. The authors have identified only 1 prognostic factor in their 141 patients that can predict the recurrence of the disease after RLND. This predictor is Breslow thickness more than 4 mm. The authors did not consider other predictors such as location of the lesion, absence or presence of ulceration of the lesion, whether the lesion was the superficial spreading type or the nodular type, or the cytological analysis of the lesion. The authors have documented that there is a 52% failure rate during the follow-up period of 41 months, of which 70% of patients had a systemic recurrence. So the systemic failure rate is very high, even in experienced hands. It has been documented that in this series of 141 patients, 26% patients had prophylactic RLND. It is a generally accepted fact that currently there is no indication for prophylactic lymph node dissection. There is clear evidence in the literature2 that patients do not do well with prophylactic RLND for melanoma. We know that regional basins act as a filter for systemic spread of the disease. Because of prophylactic RLND, patients lose their immunological mechanisms to fight against the malignant process. Thus, I would like to know what are the indications of prophylactic RLND in 38 basins. It seems to me that the indication for prophylactic RLND was a Breslow thickness of 1.5 to 4.0 mm, which is an intermediate thickness. I do not think such criteria are appropriate indications for RLND in the absence of a clinically palpable regional basin node or positive sentinel node. It is very likely that such high systemic recurrence could be caused by 26% of the patients having had prophylactic RLND. Patients who had RLND with negative sentinel nodes for melanoma had a recurrence rate of 8.6%, with a local recurrence rate of 1.7%.3 Even in this series, the group with negative sentinel nodes and RLND had a higher incidence of systemic disease. Obviously the authors have not analyzed these 38 patients separately to evaluate their actual surgical result. We know that local failure and in-transit failure is caused by surgeon failure, local failure and basin failure is caused by surgeon failure and the biological behavior of the tumor, and systemic failure is caused by biological behavior of the tumor. Because the authors have combined these groups, I believe that the overall results do not really reflect the overall result of treatment of melanoma with RLND. Better results can be achieved if surgeons do not do prophylactic node dissection and do not combine data from patients with prophylactic RLND dissection with those who had therapeutic RLND.