PrimaryTumor Levels ofTissue Inhibitor of Metalloproteinases-1 Are Predictive of Resistance to Chemotherapy in Patients with Metastatic Breast Cancer

Purpose: Only about 50% of metastatic breast cancer patients benefit from cytotoxic chemotherapy. Today, no validated markers exist for prediction of chemotherapy sensitivity/resistance in this patient group.Tissue inhibitor of metalloproteinases-1 (TIMP-1) has been shown to protect against apoptosis, and the purpose of the present study was to test the hypothesis that tumors expressing high levels of TIMP-1are protected against apoptosis-inducing agents and thus less sensitive to apoptosis-inducing chemotherapeutic drugs. Experimental Design: We investigated the association between primary tumor expression levels of TIMP-1protein and objective response to first-line chemotherapy in 173 patients with metastatic breast cancer. Results:When analyzed as a continuous log-transformed variable, increasingTIMP-1levels were significantly associated with lack of response to cyclophosphamide/methotrexate/5-fluorouracil and anthracycline-based chemotherapy (P = 0.01; odds ratio, 2.0; 95% confidence interval, 1.1-3.3). In a multivariate model, including lymph node status, steroid hormone receptor status, menopausal status, dominant metastases site, type of chemotherapy, and disease-free interval,TIMP-1was significantly associated with resistance to treatment (P = 0.03; odds ratio, 1.7; 95% confidence interval, 1.1-3.3). Conclusions: In the present exploratory study, we showed that elevated tumor tissueTIMP-1 levels were significantly associated with a poor response to chemotherapy. By usingTIMP-1, we identified a group of patients withmetastatic breast cancer, whichhardly respond to themost frequently used chemotherapy regimes (i.e., cyclophosphamide/methotrexate/5-fluorouracil and anthracyclines). The use of chemotherapy for treatment of patients with metastatic breast cancer is routine. A large proportion of the patients, however, does not benefit from the treatment and may unnecessarily suffer from substantial side effects. The only biomarkers for use as predictors of response to specific therapy regimens are the estrogen and progesterone receptors (ER and PR) for predicting response to hormonal treatment and HER-2/neu expression for predicting response to trastuzumab (1). The need for markers used to individualize treatment will continue to increase as new therapy regimens are introduced. Tissue inhibitor of metalloproteinases-1 (TIMP-1) is one of four endogenous protease inhibitors belonging to the matrix metalloproteinase proteolytic system (reviewed in ref. 2). In recent years, an increasingly complex role of TIMPs in cancer disease has emerged, and, today, TIMPs are recognized as multifunctional molecules with complicated effect on tumor development and growth (for a review, see ref. 3). TIMP-1 inhibits matrix metalloproteinase–mediated proteolytic degradation of extracellular matrix, and, besides this, TIMP-1 is capable of stimulating cell growth (4, 5) and of inhibiting apoptosis (6–11). Clinically, in primary breast cancer tissue, an association between high levels of TIMP-1 mRNA or protein and a poor patient prognosis has been established (12–16). Furthermore, a recent study including 251 patients indicated that high plasma levels of TIMP-1 are associated with a poor response to hormone therapy in patients with metastatic breast cancer (17). The purpose of the present study was to test the hypothesis that tumors expressing high levels of TIMP-1 are protected against apoptosis-inducing agents and thus less sensitive to chemotherapeutic drugs that work through induction of apoptosis. To test Cancer Therapy: Clinical Authors’ Affiliations: Department of Veterinary Pathobiology, The Royal Veterinary and Agricultural University, Frederiksberg C, Denmark; Department of Medical Oncology, Erasmus MC, Josephine Nefkens Institute, Rotterdam, the Netherlands; Hvidovre Hospital, Hvidovre, Denmark ; and Department of Oncology, Rigshospitalet, Copenhagen, Denmark Received 5/3/06; revised 8/18/06; accepted 9/13/06. Grant support: Danish Cancer Society, Danish Medical Research Counsel, Ingeborg og Leo Dannin Fonden, and Fonden til fremme af klinisk eksperimentel cancerforskning specielt vedrDrende cancer mammae (A-S. Schrohl). The costs of publication of this article were defrayed in part by the payment of page charges.This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section1734 solely to indicate this fact. Note: This work was done partly as a Danish Center for Translational Research effort. This work was done inmemory of Benedicte Kampmann. Requests for reprints: Nils Bru« nner, Department ofVeterinary Pathobiology,The RoyalVeterinary and Agricultural University, Ridebanevej 9, DK-1870 Frederiksberg C, Denmark. Phone: 45-35283130; Fax: 45-35353514; E-mail: nbr@kvl.dk. F2006 American Association for Cancer Research. doi:10.1158/1078-0432.CCR-06-0950 www.aacrjournals.org Clin Cancer Res 2006;12(23) December1, 2006 7054 Research. on April 20, 2017. © 2006 American Association for Cancer clincancerres.aacrjournals.org Downloaded from this hypothesis, we investigated the association between primary tumor expression levels of TIMP-1 protein and objective response to chemotherapy in patients with metastatic breast cancer. In 173 patients with recurrent breast cancer, who were all treated with cyclophosphamide/methotrexate/5-fluorouracil (CMF) or an anthracycline-containing regimen (cyclophosphamide/ epirubicin/5-fluorouracil or cyclophosphamide/Adriamycin/5fluorouracil or Adriamycin only) as first-line systemic treatment, we evaluated objective responses to chemotherapy and then analyzed whether these were related to primary tumor tissue levels of TIMP-1, as determined by ELISA. In addition, we analyzed whether the response to chemotherapy was associated with other clinicopathologic variables. Patients and Methods Patients. The present study, in which coded tumor tissues collected between 1979 and 1993 were used, was done in accordance to the Code of Conduct of the Federation of Medical Scientific Societies in the Netherlands. The study design was approved by the institutional Medical Ethical Committee (MEC no. 02.953). Originally, samples were selected according to the criteria described previously (18, 19) and included in prognostic and predictive studies based on the availability of stored cytosol extracts (in liquid nitrogen), which remained after routine ER and PR analysis. For the present retrospective study, samples were selected from those patients who received first-line chemotherapy (CMF in 94 patients and an anthracycline-containing regimen, cyclophosphamide/epirubicin/5-fluorouracil, cyclophosphamide/ Adriamycin/5-fluorouracil or single agent Adriamycin, in 79 patients) for metastatic disease. At the primary diagnosis, 16 patients presented with distant metastasis (M1 patients), and 157 patients were M0. At primary surgery, 58 patients were lymph node negative. Eighty-four tumors were ER and/or PR positive, and for two tumors, receptor status