Purpose: The aim of the present investigation was to evaluate the prognostic significance of urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitor-1 (PAI-1) tissue contents in primary breast adenocarcinoma.
Patients and methods: Patients from 3 medical centers were included between 1994 and 2001. Biologic factors in tumor extracts were all assayed in the same laboratory. PAI-1 and uPA were treated as continuous or dichotomized variables. Metastasis-free survival analyses were performed using the Cox model and the classification algorithm and regression tree (CART) in the whole population and in the subsets of node-negative (pN0) or positive (pN+) cases. Kaplan curves of metastasis-free survival were designed for different groups in relation to uPA/PAI-1 combinations. Urokinase-type plasminogen activator tumor level appears related to the anatomic degree of extension; conversely, PAI-1 tumor content is independent of tumor size and nodal extension.
Results: In univariate analysis, adjusted on usual prognostic factors, the metastasis risk increased with PAI-1 level in all patients (HR [hazard ratio], 3.1; P < .001), in pN0 (HR, 4.3; P < .001), and pN+ patients (HR, 2.3; P = .019). It increased also with uPA in all patients (HR, 2.6; P = 0.006). In multivariate analysis, when both variables were included, PAI-1 remained significant in all patients (HR, 2.4; P = .002) and pN0 patients (HR, 3.2; P = .003) but not uPA. CART analyses confirmed that the best partitioning factor was PAI-1.
Conclusion: There was no evidence for any additional impact of uPA. PAI-1 is an independent prognostic factor in particular in pN0 breast ductal carcinoma.