Axillary node sample to evaluate the axilla.

Abstract

Lymphatic drainage from the breast is principally to the ipsilateral axilla. In patients with breast cancer the status of the lymph nodes in the axilla is an important prognostic factor and can be used to determine local and systemic therapies. Clinical assessment of the axillary lymph nodes is unreliable, and imaging techniques, although they show some promise, are at present not practical. Standard policy for management of the axilla is clearance of the axillary lymph nodes (either level II or level III), which is justified on the grounds that it both stages and treats the axilla. In those who are axillary node-negative, however, it is an unnecessary operation and is associated with some morbidity. Various methods to obtain lymph nodes for histologic assessment in an attempt to stage the axilla have been tried. The pectoral node biopsy, where a single node is removed from the axillary tail, has been shown to be unreliable. A triple-node biopsy (pectoral node, apical node, internal mammary node) provides excellent prognostic data but is difficult to perform in patients who have been treated by breast conservation. The four-node sampling technique has been evaluated in Edinburgh in two randomized trials comparing node sampling to level III axillary clearance. It was shown to be reliable for staging the axilla; and in those who are node-negative no further treatment is required. Detailed morbidity has been assessed in patients who underwent axillary clearance; and node sampling plus radiotherapy and node sampling without radiotherapy showed that those with node sampling had the least morbidity. Those who had node sampling plus radiotherapy have reduced movement around the shoulder joint, and the axillary clearance group have increased swelling of the upper limb and slightly reduced abduction. The sentinel node biopsy is presently being assessed in several centers by randomized studies. Several large series have shown the technique to be accurate (98%) when the sentinel node is identified (around 90% of cases).