Proposed quality standards for regional lymph node dissections in patients with melanoma.

Abstract

The experience of the Sydney Melanoma Unit (SMU) is documented to offer quality assurance (QA) standards and an acceptable range for lymph node yield for regional lymph node dissection (RLND) in melanoma patients. Surgery is the most effective treatment for melanoma involving lymph nodes (LN). QA for RLND procedures, including adequacy of surgery and histopathology, is not well developed. The number of LN removed is one auditable measurement, known as a reliable predictor of surgical quality in other tumors. Data were retrieved from the SMU prospective database for patients treated from 1993 to 2006. There were 2039 RLND by SMU surgeons and 324 by non-SMU surgeons. The axilla, groin, cervical dissections of < or = 3 levels (CD < or = 3) and cervical dissections > or = 4 levels (CD > or = 4) were assessed. At axillary dissection the mean number of LN resected by SMU surgeons was 21.9 (median 21; range 1-83; 90% of cases > or = 10 LN), groin dissection mean 14.5 LN (median 13; range 1-54; 90% of cases > or = 7 LN), CD < or = 3 dissection mean 19.5 LN (median 18.5; range 1-52; 90% of cases > or = 6 LN) and CD > or = 4 dissection mean 38.9 LN (median 36; range 5-103; 90% of cases > or = 20 LN). SMU surgeons retrieved significantly more LN than non-SMU surgeons for axillary and groin dissections (P < 0.0005). These data benchmark performance for melanoma RLND. Cases with a low node count (below the 90 percentile) should be assessed critically. Standard RLND operations should have a reproducible mean and predictable distribution of LN retrieved.