Évaluation du risque d'envahissement ganglionnaire axillaire dans les cancers du sein inférieurs à 3 centimètres. Analyse d'un modèle prédictif à partir de 893 cas

Abstract

Study aim The aim of the study was to assess, by clinical and histological predictive factors, the axillary lymph-node involvement (pN +) in early breast cancers. Materials and methods Eight hundred ninety-three patients with unilateral invasive breast cancer were studied. The evaluated parameters included clinical size ( T), pathological size (pT), histological subtype (ductal infiltrating, according to grading 1, 2, 3, lobular infiltrating and others), age (less than 40, 40 to 60 and above 60). Furthermore, a new parameter, the dosimetric breast size, recently described, was included ( Eur J Cancer 1997; 33: 2432–4). Results The global rate of pN + was 25.3%, with respectively, pN 1: 10%, pN 2–3: 8.4% and pN > 3: 6.9%. According to T, the pN + rates were, respectively, 13.8%, 19.8% and 36.2% in the T 0, T 1 and T 2 ≤ 3 cm groups. According to pT, the pN + rates were, respectively, 11.1%, 17.7%, 23.5%, 30.1% and 36% in the following groups: 0–9.9mm, 10–14.9 mm, 15–19.9 mm, 20–24.9 mm and 25–29.9 mm. For the ductal infiltrating carcinoma, according to the gradings 1,2 and 3, we found, respectively, 18.3%, 27.2% and 37.8% of pN +. For the lobular infiltrating carcinoma and the other histological subtypes, the rates were 22.7% and 10%, respectively. For the three age categories cited above the pN + rates were, respectively, 30.3%, 25.8% and 22.4%. According to breast size we found 30.1 % and 24.4% of pN + respectively for small and medium or large dosimetric breast size. After a multivariate analysis, three factors were significant for pN + risk: clinical tumor size ( P = 0.0001), histological subtype ( P = 0.0005) and dosimetric breast size ( P = 0.004). With a combination of these three factors, the pN + rates varied from 5% to 50%. Conclusions The authors conclude that both clinical and pathological characteristics of the primary tumor (specified by previous core biopsy) can indicate the risk for axillary node metastases, and allow selection of candidates for limited axilla surgery (sentinel node).