Background and Objectives:
The introduction of multimodal therapy has improved the prognosis in stage III breast cancer. A knowledge of the likely axillary lymph node status at presentation is important, both to plan surgical therapy to the axilla and to establish the effect of induction therapy on the axillary nodes.
Methods:
The study involved a chart review of 114 patients with stage III breast cancer who were treated by modified radical mastectomy without prior systemic therapy. A standard method was used for axillary dissection and numbers and levels of pathologically involved lymph nodes were recorded. The incidence of lymph node metastases was correlated with tumour size, grade, and clinical T stage. The accuracy of clinical axillary staging and the relationship between level III invasion and the number of level I and II nodes invaded was also assessed. Results: Overall, 96 of 114 (84%) patients had axillary nodal metastases, and 37 of 114 (32%) patients had level III metastases. Eighteen of 43 tumours (42%) 30 mm or less had level III metastases and 27% of larger tumours had level III metastases (6/25 31-49-mm tumours, and 12/42 50+-mm tumours). Of 98 gradable cancers, only 1 out of 10 welldifferentiated tumours had level III metastases, but the rate in moderately and poorly differentiated tumours was 36% (19/53) and 37% (13/35), respectively. Clinical node staging was unreliable. A group of patients with a low likelihood of level III metastases who might benefit from an axillary procedure less than level III dissection could not be identified preoperatively. Conclusions: Patients with stage III breast cancer have a high incidence of level III axillary lymph node metastases. A subgroup with a low incidence of level III metastases could not be identified in this study. Until axillary downstaging has been convincingly demonstrated with induction therapy, a less than complete axillary procedure may leave the patient at high risk of axillary relapse.