Meta-analysis of non-sentinel node metastases associated with micrometastatic sentinel nodes in breast cancer

BACKGROUND: The need for further axillary treatment in patients with breast cancer with low-volume sentinel node (SN) involvement (micrometastases or smaller) is controversial. METHODS: Twenty-five studies reporting on non-SN involvement associated with low-volume SN involvement were identified using Medline and a meta-analysis was performed. RESULTS: The weighted mean estimate for the incidence of non-SN metastases after low-volume SN involvement is around 20 per cent, whereas this incidence is around 9 per cent if the SN involvement is detected by immunohistochemistry (IHC) alone. Subset analyses suggest that studies with axillary dissection after any type of SN involvement result in somewhat higher estimates than studies allowing omission of axillary clearance, as do studies with more detailed histological evaluation of the SN compared with those with a less intensive histological protocol. Higher-quality papers yield lower pooled estimates than lower-quality papers. CONCLUSION: The risk of non-SN metastasis with a low-volume metastasis in the SN is around 10-15 per cent, depending on the method of detection of SN involvement. This should be taken into account when assessing the risk of omission of axillary dissection after a positive SN biopsy yielding micrometastatic or immunohistochemically positive SNs. Copyright (c) 2004 British Journal of Surgery Society Ltd