Background: In the era of sentinel node biopsy (SNB), mainly aimed at local control, the axillary dissection (AD) was performed for breast cancer with positive node. It was argued that only level I/II lymph nodes dissection might be inadequate because nodes involvement was found in level III and interpectoral region when axillary node was positive. It is necessary to assess the extent of AD after modern preoperative axillary staging. The aim of this study was to investigate the incidence and associated factors of node involvement in level III/interpectoral region. Methods: A consecutive series of 338 core needle biopsy confirmed T0-2 invasive breast cancer cases were included in this study. Axillary node metastases were proved by ultrasound guided needle biopsy (NB) if ultrasonographic abnormal node was detected prior to SNB or by SNB if no abnormal node was detected. Cases of negative NB but positive SNB with image abnormal node were excluded. Prior to AD included level III and interpectoral lymph nodes, 4 to 8 cycles of neoadjuvant chemotherapy with anthrocyclin and/or taxane regimen were completed for each case. The chi-square test was used to determine the relation between level III/interpectoral lymph nodes metastases and clinicopathological factors. Multivariate logistic regression was analyzed for covariate selection. Results: A median of 19 axillary nodes was harvested per case (range: 5-46, average: 19.2). The pathologic complete response rate of axillary nodes was 35.3% (70/198) in NB positive subgroup. Level III and interpectoral lymph nodes were harvested in 76.9% (260/338) of cases (range: 1-8, average: 1.9, median: 1) and 49.7% (168/338) of cases (range: 1-10, average: 1.6, median: 1), respectively. The incidence of positive level III and interpectoral lymph nodes were 8.9% (30/338) and 8.9% (30/338), respectively. Node involvement of level III/interpectoral region was found in 13.3% (45/338) of these cases. The incidence of node involvement in level III/interpectoral region of NB positive subgroup (14.6%, 29/198) was not significantly higher compared with SNB positive subgroup (11.4%, 16/140, P=0.391). Increasing tumor size was significantly correlated with increasing likelihood of node involvement in level III/interpectoral region (T0-1: 6.3% vs T2: 16.7%, P=0.008). In SNB positive T0-1 subgroup, there was no positive node found in level III/interpectoral region. Multivariate analysis showed that tumor size was the only independent factors predicting node involvement in level III/interpectoral region (OR=3.488, 95%CI:1.427-8.528, P=0.006).

Conclusions: The incidence of node involvement in level III/interpectoral region was 13.3% of nodes positive T0-2 breast cancer treated with neoadjuvant chemotherapy. Tumor size may be the predictor of node involvement in this region.

Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-01-27.