Membrane-type-1 matrix metalloproteinase, matrix metalloproteinase 2, and tissue inhibitor of matrix proteinase 2 in prostate cancer: identification of patients with poor prognosis by immunohistochemistry

Summary

Overexpression of the matrix metalloproteinase (MMP) 2 is associated with poor prognosis in many tumor types. Membrane-type-1 MMP (MMP14) activates MMP2 using pro-MMP2 specific inhibitor, tissue inhibitor of matrix proteinase 2 (TIMP2), as a receptor. We evaluated, by immunohistochemistry on 189 T3N0-2M0 prostate cancer (Pca) cases, the influence of MMP2, MMP14, and TIMP2 expression, individually and in association, on Pca disease-free survival (DFS). We evaluated marker expression separately in cancer, stromal, and benign epithelial (BE) cells according to a percentage scale (0%, <10%, 10%-50%, and >50%). Median follow-up was 4.61 years. In BE cells, there was an inverse relationship between initial prostate-specific antigen serum level and T3 stage with MMP14 expression (P = .003) and between pN stage and TIMP2 expression (P = .04). The most significant results with survival were obtained by dichotomizing the cases between those with less than 10% and at least 10% of cells expressing the marker, the latter category representing overexpression. TIMP2 overexpression in stromal cells was associated with a longer DFS with a hazard ratio of 0.573 (P = .02) for time to recurrence. MMP2 overexpression by BE cells correlated with a shorter DFS using a multivariate trend test (hazard ratio = 1.46, P = .02). Stromal cells expressing less than 10% TIMP2 and MMP2 overexpression was the only combination that was significantly associated with a shorter DFS (log-rank test, P = .0001). This study suggests that MMP14 is involved mostly in Pca implantation and that MMP2 and TIMP2 expression by reactive stromal cells might be used as predictors of DFS in T3N0-2M0 Pca.

Introduction

With 300 000 new cases every year, prostate cancer (Pca) is the most common cancer in North American men [1]. However, only 2.9% die of their disease [1]. Therefore, more conservative therapies instead of standard radical treatments, such as radical radiotherapy and prostatectomy, may represent an alternative for elderly patients with high-risk surgical conditions. Tumor behavior cannot be accurately and reliably predicted by our current prognostic factors. The identification of new prognostic factors may allow the development of new treatment strategies with less morbidity.

Matrix metalloproteinase (MMP) 2 is part of the MMP family, which includes more than 20 zinc-dependant endopeptidases that degrade various components of the extracellular matrix [2]. MMPs have also been shown to be involved in the release of growth factors that enhance tumor growth and aggressiveness [3]. Because of their proteolytic and growth-factor–releasing activities, MMPs are thought to be involved in cancer progression.

Pro-MMP2 is a 72-kd protein [2] that is cleaved into an active form of 62 kd. Active MMP2 is mostly involved in type IV collagen and gelatin proteolysis [2]. MMP2 has been demonstrated to be up-regulated in many tumor types and is associated with prognosis in most of them [4]. Whereas most MMPs are transcriptionally regulated, MMP2 is constitutively expressed by stromal cells [2]. MMP2 is mainly regulated by its zymogen inhibitor, tissue inhibitor of matrix proteinase 2 (TIMP2), and by its major activator, membrane-type-1 MMP (MMP14) [2]. MMP14 is a membrane-type nonsoluble MMP. According to Strongin et al [5], who showed in 1995 that MMP2 was activated by MMP14, TIMP2 inhibits successively MMP14 and pro-MMP2 molecules, TIMP2 being the link between MMP14 and MMP2. A free adjacent MMP14 then cleaves pro-MMP2 (Asn³⁷-Leu³⁸), leading to an intermediate molecule. A nearby active MMP2 can finally cleave the intermediate form of MMP2 (Asn⁸⁰-Tyr⁸¹), leading to another active MMP. Because of the unique MMP2 activation mode, TIMP2 is involved in MMP14 inhibition when expressed at high levels and in MMP2 activation when expressed at lower levels [5].

We have previously shown that MMP2 overexpression by cancer cells was associated with a poor prognosis in T3N0-2M0 Pca [6]. Our current project was aimed at testing the hypothesis that the balance between MMP2 and the main molecules involved in its activation and/or in its inhibition—MMP14 and TIMP2—are related to Pca disease-free survival (DFS). We hoped that this study might help to identify new potential targets to treat a subset of patients, such as those with a high surgical risk, with a more conservative approach.