[PS2 as a prognostic factor in 1065 cases of human breast cancer. A multicenter study]

pS2 protein assay was performed with Elsa-pS2 kit (CIS-Biointernational) on a group of 1,065 patients with operable breast cancer who underwent breast surgery in the years 1982 through 1990. The median follow-up was 57 months. This group included exclusively infiltrating ductal carcinoma with primary surgery. Age mean was 58 yr; T0-T1, 33.6%; T2-T4, 66.4%; Differentiation grade I, 29%; node negative, 53%; estrogen receptor (ER) positive, 62.4%; progesterone receptor (PR) positive, 55.2%; mean tumor size, 2.4 cm; local recurrence, 5.2%; metastasis, 17.5%. pS2 values varied from 0.1 to 707 ng/mg of cytosol protein (median, 5.6; mean 24.5; 95th percentile 112 ng/mg p). There was no significant relationship between the mean level of pS2 and age, tumor size, nodal status, whereas pS2 was related to histological grade (P < 10(-3)), ER (P < 10(-5)), and PR (P < 10(-5)). By using 2 ng/mg p as pS2 cutoff, 77/391 (19.7%) of ER+PR+ tumors were pS2-, and 122/345 (35.4%) of ER-PR-tumors were pS2+; with this cutoff, a strong relationship existed between pS2 and overall survival, but not between pS2 and relapse-free survival. With Cox multivariate analysis, pS2 protein was classified after lymph node status, histological size, ER, differentiation grade, age, clinical stage, PR. In patients with axillary lymph node involvement (N+), pS2 status could discriminate between good and bad prognosis, specially for patients with small tumors (< 2 cm) and with less than seven invaded nodes. This study showed that pS2 protein was a poor prognostic factor in comparison with classical factors.