Axillary conservation in early breast cancer.

Management of the axilla is controversial. There has never been a widely agreed means of balancing an understanding of nodal status as prognostic information, the need to prevent axillary recurrence and the wish to avoid treatment-related morbidity, particularly arm lymphoedema. Balancing these conflicting requirements is difficult, particularly as the results of previous trials of locoregional therapy in early breast cancer are less relevant in these days of earlier (commonly screen-detected) presentation, more detailed pathological assessment, improved systemic therapies, and better surgery and radiotherapy. Improved outcomes and a reduced risk of locoregional recurrence are welcome, but make it difficult to conduct contemporary clinical trials, particularly those that attempt to reduce the intensity of treatment in an effort to reduce morbidity. Recent developments and clinical research concerning axillary management need to be appreciated in this context. The routine use of axillary node clearance (ANC) or axillary radiotherapy (ART) for the clinically nodenegative axilla was replaced by the use of image-guided biopsy to confirm nodal involvement before axillary treatment, and more recently by sentinel lymph node biopsy (SLNB) to establish nodal status. The avoidance of ANC and the lymphoedema it so commonly causes in node-negative early breast cancer following the introduction of SLNB represents one of the most significant advances in early breast cancer management over the past 15 years. Since the widespread adoption of SLNB and the refinement of histopathological methods to quantify the degree of involvement of positive nodes, there have been further efforts to define the best local treatment for the sentinel node-positive patient. It is now clear that patients with isolated tumour cells (ITCs) or micrometastases (deposits of 2 mm or less in diameter) in these nodes do not require further intervention to the axilla, as the risk of axillary recurrence is very low and not influenced by further surgery or radiotherapy1. Until recently, the majority of patients with macrometastases (deposits larger than 2 mm) in the sentinel lymph node were routinely treated with ANC or ART. The publication of the American College of Surgeons Oncology Group Z0011 (Alliance) trial in 2010 challenged this practice2. The Z0011 trialists planned to recruit 1900 patients with T1–2 breast cancer and one or two involved sentinel lymph nodes treated with breastconserving surgery. Patients were randomized to completion ANC or no further axillary surgery over a 4-year recruitment period. All patients were required to have whole-breast radiotherapy without any additional radiation to the axilla or supraclavicular regions. Systemic therapy was given at the discretion of the investigators. Overall, disease-free and locoregional recurrence-free survival were no different between groups, but surgical morbidities including lymphoedema were significantly lower in the SLNB-only group. The need for routine axillary therapy in this patient group came under question, and a more conservative approach to the axilla was rapidly adopted in the USA following the publication of the results of Z00112. The American Society of Clinical Oncology has advised that ANC/ART can be avoided in patients with nodepositive early breast cancer if their circumstances reflect the Z0011 eligibility criteria3. There are, however, many concerns about this trial that limit the interpretation and application of its results. Recruitment was less than 50 per cent of predicted, and involved 115 centres. Around 20 per cent of patients were lost to follow-up, information on nodal status was missing in 11 per cent of patients, there was no surgical quality assurance, and overall survival rather than axillary recurrence was chosen as the primary endpoint. Additionally, around half of the patients had micrometastatic disease (for which avoidance of further axillary therapy is routine), leaving 430 patients with axillary nodal macrometastases. Perhaps the most significant concern related to the lack of prospective quality control for radiotherapy; this allowed the possibility that radiation oncologists, aware of the type of axillary surgery performed, may have introduced bias by adjusting radiotherapy to treat the axilla in those patients treated by SLNB alone4. The concerns about radiotherapy prompted an evaluation of this treatment within the trial, revealing an inconsistent practice with