This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

#202 Introduction: Controversy remains over whether to perform sentinel node biopsy (SNB) before or after neoadjuvant chemotherapy (NAC). We examined the practice patterns, feasibility, and accuracy of this procedure in high risk breast cancer patients treated with NAC in a multi-institutional correlative science study.
 Methods: Patients with biopsy-proven breast cancer >3 cm enrolled into the I-SPY TRIAL to undergo 4 weeks of anthracycline-based NAC, 4 weeks of taxane treatment, then surgical intervention. Study protocol did not dictate axillary treatment. Timing of SNB was dictated by the surgeon. Practice patterns, outcome of SNB and axillary lymph node dissection (ALND), locoregional recurrence and distant metastases were recorded with a mean follow-up of 2.9 years.
 Results: 237 patients enrolled, 221 completed the trial, 210 had complete data at the time of analysis; Table 1 shows axillary practice patterns.
 
 Overall, 43% had a positive SNB and/or ALND after NAC. 129 (61% of 210) patients presented with clinically positive nodes, 39 of which had a post-NAC SNB. 5/39 had no ALND (all SNB negative). Table 2 shows results for those who had a post-NAC SNB and ALND. For this subset of patients, sentinel node ID, accuracy, and false negative (FN) rates were 80%, 91% and 15% respectively. If SNB was negative, 20% of patients still had a positive ALND. 81 (39% of 210) patients presented with clinically negative nodes, 22 of which had a post-NAC SNB. 8/22 post-NAC SNB patients had no ALND (6 negative, 2 positive for 1mm disease). Table 2 shows results for those who had a post-NAC SNB and ALND. For this subset of patients, sentinel node ID, accuracy, and FN rates were 100%, 100% and 0%.
 
 Overall, there were 26 deaths; 96% occurred in those who presented with clinically positive nodes, 77% had positive post-NAC nodes. A negative axilla post-NAC was predictive of longer DFS over those with axillary disease post-NAC (p<0.05).
 Conclusions: In clinically node negative patients, post-NAC SNB is feasible and accurate before or after NAC. Our data suggests that a post-NAC SNB is sufficient; this avoids an additional operation and allows us to gain information on post-NAC axillary status which is of prognostic significance. In clinically positive patients, SNB does not adequately reflect axillary disease; even when SNB was negative, 20% still had axillary disease. At this time, we recommend that ALND be performed on all clinically node positive patients. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 202.