Urokinase-type plasminogen activator in epithelial ovarian cancer: A poor prognostic factor, associated with advanced stage

Objective: Expression of urokinase-type plasminogen activator (uPA), a CSF-1 (macrophage colony stimulating factor) inducible gene, has been associated with invasion and metastatic properties, as well as with poor prognosis in a variety of epithelial cancers. In ovarian cancer metastases, co-expression of CSF-1 and CSF-1 receptor has been shown to be an independent poor prognostic factor for clinical outcome. Moreover, CSF-1 stimulated invasion of ovarian cancer cells in vitro is mediated by uPA activity. Ovarian cancers contain a higher concentration of uPA and its receptor, uPAR, than benign ovarian tumors. Reports, however, on the prognostic value of uPA content in ovarian cancer tissue extracts, which contain tumor epithelium, stroma, and infiltrating inflammatory cells, have been conflicting. The objectives of this study are to determine: (1) the prognostic value of uPA or uPAR expression in ovarian cancer epithelium, (2) the level of uPA or uPAR staining in the stromal cells, and (3) if there is an association between expression of uPA/uPAR and CSF-1/CSF-1 receptor in ovarian cancers. Methods: Immunohistochemistry was used to study the primary and metastatic tissues from 131 epithelial ovarian carcinomas, which included all stages of disease. The intensity and extent of staining for uPA and uPAR in the tumor epithelium was scored. Kaplan-Meier curves of survival were compared with the log-rank test. The Cox regression model was used for multivariate analysis. The Chi square test was used to study the association between uPA/uPAR immunohistochemical positivity and clinicopathologic parameters, as well as with CSF-1 and CSF-1 receptor expression. Results: For the invasive cases, 66% of the primary tumors and 57% of the metastases expressed uPA, while only 32% of the primary tumors and 23% of the metastases expressed uPAR. There was low level staining for uPA or uPAR in the stromal cells surrounding the tumor. Tumor cell expression of uPA in the invasive primary tumors was strongly predictive of a shorter overall survival; the median survival of 29 months for the 76 patients whose primary tumors expressed uPA was extended to 53 months for the 40 patients whose primary tumors did not express uPA. The prognostic value of uPA expression was retained in the subgroup of patients with residual disease greater than or equal to 2 cm, but not among patients with minimal or no residual disease. Urokinase expression was, however, significantly associated with advanced stage and older age, and was not an independent factor for survival on multivariate analysis. Expression of uPAR alone, or in combination with uPA, had no prognostic value in this study. Co-expression of uPA and uPAR was strongly associated with co-expression of CSF-1 and CSF-1 receptor. Conclusions: The current study is the first to describe the prognostic significance of uPA/uPAR expression localized to ovarian cancer epithelium. Two previous studies on ovarian cancer, as well as studies on breast, lung, gastric, colon, and cervical cancers, examined the prognostic value of uPA content in tissue extracts. We demonstrate that uPA expression in the primary tumor epithelium is a common finding which is significantly predictive of a shorter overall survival, but is not an independent factor, most likely due to its close association with advanced stage. On the other hand, ovarian cancer stromal staining for uPA was of a low level and was not a prominent feature. In this study, a significant association was found between co-expression of uPA/uPAR and of CSF-1/CSF-1 receptor; thus components of the uPA system may act as a mediator of the invasive, poor prognosis phenotype conferred by CSF-1 in epithelial ovarian cancers.