Objectives: To retrospectively assess whether completeness of node dis
section has any bearing on regional control in cutaneous melanoma and
to examine the efficacy of a subsequent dissection in patients with is
olated nodal recurrence.Design: Case series, 18-month minimum follow-u
p. Setting: Academic surgical practice. Study Participants: Patients w
ith cutaneous melanoma who had undergone a regional node dissection an
d subsequently developed recurrence in the same nodal basin in which a
lymphadenectomy had been performed with no evidence of distant metast
ases. Of 1030 instances of regional node dissection, 28 met these crit
eria. Main Outcome Measures: Nodal recurrence in the previously dissec
ted lymph node basin as the only site of recurrence and survival follo
wing a subsequent lymph node dissection. Results: The 28 instances of
isolated nodal recurrence represent a regional failure rate of 2.7%. I
n those cases where the first dissection was performed within our divi
sion, the rate is 0.8%. Recurrence for cervical, axillary, or inguinal
sites was similar. In 71% of the cases, more than one node was positi
ve at the time of recurrence. Four patients have shown disease-free su
rvival greater than 3 years following a subsequent lymphadenectomy. Co
nclusion: Node dissection is a therapeutic procedure and, therefore, m
ust consist of complete lymphadenectomy with meticulous attention to s
urgical detail. Approached in this fashion, only a small subgroup of p
atients will show recurrence in a previously dissected nodal basin, a
few of whom can be salvaged by a second dissection.