sentinel node ratio as a predictor of non-sentinel lymph node involvement

background: sentinel lymph node biopsy (slnb) has replaced axillary lymph node dissection (alnd) in early breast cancer patients as the first line surgical approach to axillary nodes. further dissection is performed only when sln is involved by tumor cells. however, in a significant proportion of patients, non-sentinel nodes are still not involved and axillary dissection has no additional therapeutic benefits. selective axillary clearance has been considered to prevent unnecessary dissection. the purpose of this study was to define predictors of non-sentinel lymph node involvement in patients with positive slnb. methods: patients with early stage breast cancer and positive slnb who underwent alndin a referral hospital in tehran, iran between2010 and 2012 were recruited into the study. relations between different clinico-pathological characteristics and involvement of non-sentinel nodes were investigated. results: from 139 patients who had positive slnb and underwent alnd, only in 71 cases (51%) positive non-sentinel lymph nodes (nslns) were detected. in univariate analysis, there was no association between tumor size, lymphovascular invasion, er, pr and her-2 expression and nsln metastasis. in contrast, presence of more than one sln (p = 0.016) and a sentinel node ratio (snr) more than 0.5 showed a significant association (p< 0.001). only the latter remained as the significant predictor of nsln involvement in mutltivariate analysis (p < 0.001, or = 3.706). conclusions: based on our results, patients with a snr more than 0.5 were more commonly diagnosed with nsln metastasis. thus, it is recommended that surgeons think twice before skipping alnd in this subgroup of patients.