© 2000 AlphaMed Press MMP-1 is a Prognostic Marker for Hematogenous Metastasis of Colorectal Cancera Department of Surgical Oncology; b Department of Transfusion Medicine, Graduate School of Medical Sciences, Faculty of Medicine, the University of Tokyo, Tokyo, Japan; c Institute of Molecular and Cellular Biosciences, the University of Tokyo, Tokyo, Japan; d Cancer Institute Hospital,Tokyo, Japan Correspondence: Eiji Sunami, M.D., Department of Surgical Oncology, Graduate School of Medical Sciences, Faculty of Medicine, the University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan. Telephone: +81-3-5800-8653; Fax: +81-3-3811-6822.
Background. Degradation of basement membrane and extracellular matrix by matrix metalloproteinases (MMPs) is believed to be an essential step in the complicated process of hematogenous metastasis. MMP-1 is a member of collagenases, a family of MMPs that degrades collagens type I, II, and III, main components of the interstitial stroma. The purpose of this study was to investigate the expression of MMP-1 in colorectal cancer and its correlation with hematogenous metastasis. Patients and Methods. We examined 133 cases of colorectal cancer (Dukes A: 72; Dukes B: 26; Dukes C: 23; Dukes D: 12). Sections were cut from formalin-fixed, paraffin-embedded samples containing the deepest site of cancer invasion and stained immunohistochemically with a monoclonal antibody to MMP-1. According to the area of the tumor that was stained, patients were divided into high- and low-MMP-1 expression groups. Results. MMP-1 expression was observed in the cytoplasm of cancer cells, some stromal cells, and a few normal epithelial cells of colonic mucosa. High MMP-1 expression was found in 47 (35.3%) cases and low in 86 (64.7%). Hematogenous metastasis was identified in 14 (29.8%) of high-MMP-1 groups and 12 (13.9%) of low-MMP-1 groups. MMP-1 expression significantly correlated with hematogenous metastasis of colorectal cancer, but no correlation was found between MMP-1 expression and the other clinicopathological features investigated. Conclusions. MMP-1 expression may be a novel marker for hematogenous metastasis of colorectal cancer, and its inhibition may be a strategy for prevention of metastasis. Key Words. MMP-1 • Colon cancer • Hematogenous metastasis • Immunohistochemistry
Colorectal cancer is one of the cancers for which surgical treatment is rather effective. The five-year survival rate in the cases of curative surgical resection including lymphadenectomy reaches about 80% in cancer of the colon, and about 70% to 75% in that of the rectum. One of the most difficult problems in the treatment of colorectal cancer is hematogenous metastasis; about 25% to 35% of colorectal cancer cases show hematogenous metastasis, including synchronous and metachronous. The most frequent site of hematogenous metastases is the liver, accounting for about 60% to 80% of the cases of hematogenous metastasis of colorectal cancer [1, 2]. The five-year survival rate in the cases of liver metastasis is as low as 25% to 50%, even when the samples are limited to curative resection [3, 4]. Consequently, the detection of cases with high potential for development of hematogenous metastasis and the prevention of its development are important matters when considering the prognosis of colorectal cancer. Hematogenous metastases of cancer develop through a complex process, in which cancer cells detach from the primary site, invade the vasculature by degradation of the surrounding tissue, migrate to the secondary organ, where the cells attach to and transmigrate through the endothelial cell layer and invade and grow to form the metastatic lesion. In this complex process of metastasis, matrix metalloproteinases (MMPs) are reported to play an important role [5, 6]. MMP is a family of enzymes that specifically degrades the extracellular matrix. Extracellular matrix is composed of various kinds of collagens, proteoglycans, laminin, and fibronectin. It plays a role in maintaining tissue compartments as the connective tissue surrounding organs and is one of the greatest obstacles for cancer cells to invade and achieve metastasis [5, 7]. MMP-1 is a kind of collagenase that belongs to the MMP family and specifically degrades collagens I, II, and III. Its expression has been found in various cancer tissues at both protein and mRNA levels, and its overexpression has been demonstrated to be associated with poor prognosis in colorectal and esophageal cancers [8, 9]. However, there are no reports yet on its relationship with hematogenous metastasis. In this study, we investigated the expression of MMP-1 in colorectal cancer and its relation with the clinicopathological features, especially those related to hematogenous metastasis.
Monoclonal Antibody The monoclonal antibody to human MMP-1 was purchased from Oncogene Science (Cambridge, MA). This antibody was confirmed to specifically react with MMP-1 by Western blotting (data not shown).
Samples The average observation period after resection was 56 months (range: two months to seven years) and "no recurrence" here is defined as cases without recurrence at least five years after resection.
Clinicopathological Data
In terms of cancer stage, 72 cases were Dukes A, 26 Dukes B, 23 Dukes C, and 12 Dukes D. Hematogenous metastases were found in 26 cases (12 synchronous and 14 metachronous). Metachronous hematogenous metastasis was observed at an average of 22 months after resection (range: 4-40 months). According to the localization, hematogenous metastases were found in the liver in 16 cases, four in the lung, four in both liver and lung, and two in other sites. The relation between hematogenous metastasis and the other clinicopathological features are shown in Table 1
Immunohistochemistry Samples used were formalin-fixed, paraffin-embedded specimens of primary colorectal cancer. Sections were cut from samples containing the deepest site of cancer invasion, and stained immunohistochemically. The immunohistochemical procedures were performed using the avidin-biotin-peroxidase complex method. Briefly, endogenous peroxidase activity was abolished by treatment with 0.3% H2O2 in methanol, and nonspecific binding of antibodies was blocked by preincubation with 5% normal goat serum in phosphate-buffered saline (PBS). Then, the tissue sections were subjected to antigen retrieval by incubating with 0.1% pepsin in PBS for 6 min and boiling in citric acid buffer for 30 min. Anti-MMP-1 antibody was applied as the first antibody and biotinylated antimouse immunogloblin (DAKO; Glostrup, Denmark) as the second antibody. Vectastain ABC kit (Vector Laboratories; Burlingame, CA) was used according to the manufacturer's recommendation. Finally the sections were incubated with diaminobenzidine and 0.006% H2O2 and counterstained with Mayer's hematoxylin. The slides were evaluated independently by two investigators without any information of the patients' clinicopathological features. According to the tumor area that was stained for MMP-1, the patients were divided into high- and low-MMP-1 groups. The cut-off value was set at 30%, the mean value of staining, and those cases with more than 30% of the tumor area stained for MMP-1 were grouped as high-MMP-1 and those with less than 30%, low-MMP-1.
Statistical Analysis
MMP-1 expression was found predominantly in cancer cells, but also some stromal cells and endothelial cells stained weakly. In the MMP-1-positive cells, it was diffusely observed in the cytoplasm of the cells (Fig. 1
According to MMP-1 immunoreactivity, high MMP-1 expression was observed in 47 (35.6%) cases and low expression in 86 (64.7%). Hematogenous metastases were found in 14 (29.8%) of high-MMP-1 and 12 (13.9%) of low-MMP-1 cases. The correlation between MMP-1 expression and hematogenous metastasis was found to be statistically significant (p < 0. 05).
MMP-1 expression did not correlate with any of the other clinicopathological features evaluated, namely depth of tumor invasion, lymphatic invasion, venous invasion, lymph node metastasis, and Dukes' classification (Table 2
Cumulative patient disease-free survival was assessed by the method of Kaplan-Meier and the disease-free survival curves for high- and low-MMP-1 cases were compared by the log-rank test. As shown in Figure 2
One of the great obstacles that we face when treating colorectal cancer patients is the development of hematogenous metastasis. It is one of the main factors that determine the prognosis of these patients [3, 4, 10]. Consequently, identification of those patients with high risk of metastasis and the development of new strategies for its prevention are crucial events to ameliorate the prognosis of those patients that have already developed colorectal cancer. In the complex process of hematogenous metastasis, cancer cells must degrade the extracellular matrix proteins, main components of basement membrane and interstitial stroma, and gain access to blood vessels [5, 6]. The MMPs are known as a family of enzymes that degrade extracellular matrix proteins such as collagens, fibronectin, and laminin. The MMPs are divided into three groups according to their structure and substrate specificity: A) collagenase group; B) gelatinase/type IV collagenase group; C) stromelysin group, and D) others [5, 11]. Type IV collagen is the main component of the basement membrane, and its degradation is a crucial event in the process of hematogenous metastasis. Type IV collagenase (MMP-2 and MMP-9) and matrilysin (MMP-7) are the main enzymes that degrade type IV collagen, and their expression in cancer tissues has been reported to correlate with the metastatic potential [12-14]. On the other hand, MMP-1, a member of the collagenase group, specifically degrades collagens type I, II, and III, the main components of the interstitial stroma [5]. MMP-1, also known as interstitial collagenase or collagenase-1, is the predominant collagenase of resident cells, and is produced by human endothelial cells [15], tissue macrophages [16], synovial cells [17, 18], and fibroblasts [19, 20]. MMP-1 expression can be induced in monocytic cell lines by bacterial endotoxin and zymosan [21, 22]. It is also produced by keratinocytes, playing a beneficial role in wound healing by facilitating their movement over the collagen-rich dermis during re-epithelialization [23], although MMP-1 overexpression may have a deleterious effect during wound healing [24]. MMP-1 expression is also detected in interstitial lung diseases [25], synovial fluid from patients with osteoarthritis [26], and vulnerable regions of human atherosclerotic plaques [27, 28]. MMP-1 expression is found in basal cell carcinoma, cancer of thyroid, esophagus, stomach, lung, breast, liver, head and neck, and colon [8, 9, 29-35]. In esophageal cancer, Murray et al. reported a particularly poor prognosis associated with MMP-1 expression that was an independent prognostic factor [9]. In colorectal cancer, the same group also clearly demonstrated that the presence of MMP-1 was associated with a poor prognosis and that MMP-1 expression had prognostic value independent of Dukes' stage [8]. However, in this report, it is not clarified which factor worsens the prognosis of those patients with high MMP-1 activity. In the present study, therefore, we aimed to investigate the correlation between MMP-1 expression and hematogenous metastasis, the most probable cause of poor prognosis in colorectal cancer. Using immunohistochemistry, we demonstrated that many colorectal cancers strongly express MMP-1 in the cytoplasm of cancer cells. Consensus has not yet been obtained on exactly which cells produce and express MMP-1. MMP-1 staining is reported mainly in stromal cells of neoplastic tissue [36], or in both cancer cells and stromal cells [8]. We have performed gelatin zymography of colorectal cancer cell lines and have observed that cancer cells also produce MMP-1 (data not shown), therefore, it is not implausible that cancer cells express it. In studies using in situ hybridization, overexpression of MMP-1 mRNA was found in stromal cells [29, 37]. The discrepancy between the immunohistochemical studies, including ours, and those using in situ hybridization may be dependent on the difference in the levels of MMP-1 mRNA and on the different capacity of stromal cells and cancer cells to store MMP-1. Also the threshold of both methods may be relevant. Another possibility is that, similar to MMP-2, cancer cells utilize the enzyme produced by stromal cells [38]. In the present study, we clearly demonstrated that MMP-1 expression correlates with hematogenous metastasis in colon cancer and is independent of Dukes' classification, venous invasion, lymphatic invasion and lymph node metastasis. Patients with low-MMP-1 expression had a significantly lower incidence of hematogenous metastases than those with high-MMP-1 expression. Therefore, MMP-1 expression may be a new prognostic factor for hematogenous metastasis of colorectal cancer. As MMPs seem to play important roles in tumor invasion and metastasis, recently they have gained attention as targets for new anticancer therapy strategies. Inhibitors of MMPs have been shown to prevent tumor spread both in vitro and in vivo [39-41] and to inhibit tumor angiogenesis [42, 43], and some of them are being developed for clinical use. Marimastat, a synthetic low-molecular weight inhibitor of MMPs, is currently in phase I/II and III clinical trials with satisfactory preliminary results, but some side effects are reported [44-50]. One reason for the development of side effects is that almost all the cells in the human milieu produce various types of MMPs, and most of the newly developed drugs are promiscuous for MMPs. The development of more specific drugs and the establishment of more objective markers for identification of those patients who are most likely to benefit from this kind of therapy are necessary. In the present study we demonstrated that MMP-1 is a good prognostic marker for hematogenous metastasis of colorectal cancer, and the anti-MMP therapy specific for MMP-1 may be a possible strategy for its prevention in the near future.
This work was supported partly by a Grant-in-Aid for Scientific Research from the Ministry of Education, Sciences, Sports, and Culture of Japan and partly by a grant from the Ministry of Health and Welfare of Japan.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||