Axillary lymph node metastases in patients with small carcinomas of the breast: is accurate prediction possible?

Objectives: To find out whether macroscopic classification of the tumour ma rgin is predictive of axillary lymph node metastases and to identify a comb ination of clinical and pathological findings by which axillary node status can be predicted accurately in small carcinomas (T1) of the breast. Design: Retrospective study. Setting: Municipal referral centre, Japan. Subjects: All 1003 patients with T1 invasive carcinoma of the breast who ha d axillary lymph node dissection between January 1970 and December 1996 as part of their treatment. Main outcome measures: The association between the incidence of axillary ly mph node metastases and 10 clinical and pathological factors (age, palpabil ity and size of tumour, macroscopic classification of tumour margin, clinic al axillary status, radiating spiculation on a mammogram, histological type , lymphatic invasion, oestrogen and progesterone receptor status) were anal ysed. Results: Clinical axillary node status, macroscopic classification of tumou r margin, lymphatic invasion, and age of the patient were significant predi ctors of axillary lymph node metastases (p < 0.01 in each case). Among 47 p atients aged 65 or more whose tumours had well-defined margins and with a c linical NO status in the axillae, the incidence of histological axillary ly mph node metastasis was only 6% (n = 3) whereas it was 65% in 57 patients w ith tumours of ill-defined margins whose axillae were N1 or N2. Conclusions: Macroscopic classification of tumour margins is an independent predictor of axillary lymph node metastases for patients with small carcin omas of the breast. However, even with combinations of the examined predict ors of axillary node metastases, the subgroup of patients at minimal risk o f metastasis was less than 5% in T1 breast cancer, whereas three-quarters o f the patients had clear axillary lymph nodes. Most patients with T1 boast cancer will need surgical staging of the axillae by methods such as sentine l node biopsy.