Abstract

The standard approach of performing a completion axillary lymph node dissection (cALND) after a positive sentinel node for breast cancer patients is no longer generally accepted. This study applied the criterion of a 27% risk of having residual positive lymph nodes calculated by the MD Anderson nomogram to perform a cALND. This 27% cut-off is based on the number of positive non-sentinels in the Z0011 trial. A cohort of 166 cN0, sentinel positive breast cancer patients was used to validate the MD Anderson nomogram. ROC (Receiver Operating Characteristic) analysis shows an AUC (Area Under the Curve) of 0.76 and an optimal cut-off at 34% risk of positive non- SLNs (sensitivity 86%, specificity 57%). The 27% cut-off has a sensitivity of 88% and a specificity of 41% to detect positive non-sentinels. In a second cohort (N= 114) the 27% cut-off criterion was prospectively applied and appeared to be practice changing. Although we take minimal risk to leave disease behind (2/166 patients >3 positive nodes), 30.7 % in the first cohort and 54.4 % of the patients in the second cohort could be spared a cALND. The Z0011 criteria would have had more impact, omitting 90% of the cALND, but leaves more disease behind. The impact of leaving disease behind on survival remains unanswered but is awaited by long term follow up of large prospective cohort studies.