This month in Gastroenterology

Abstract

Colorectal cancer (CRC) is the second leading cause of death from cancer in the United States and the third leading cause of death from cancer in Taiwan. Currently, pathological staging is the most commonly used prognostic predictor for colorectal cancer. While the 5-year survival is approximately 90% for stage I disease and 10% for stage IV disease, there is less predictability of the 5-year survival in stage II and III disease. The 5-year survival for stage II disease ranges from 60% to 80% and the 5-year survival for stage III disease ranges from 30% to 60%. This study by Lin et al suggests that connective tissue growth factor (CTGF) is an independent prognostic marker for colorectal cancer and may help better predict prognosis in patients with stage II and III CRC. Furthermore, these authors investigated the role of CTGF in the invasion and metastasis of human colorectal cancer. The expression of CTGF, a member of the CCN family, has been shown to positively correlate with tumor development, and progression of some cancers has been shown to suppress growth of oral squamous cell carcinoma cells transplanted into mice. The study by Lin et al examined the level of CTGF expression in normal adjacent epithelium, premalignant lesions, and CRC specimens in 119 patients diagnosed with CRC at the National Taiwan University Hospital. These authors also examined invasion properties of CTGF-transfected cells and the role of CTGF in CRC metastasis. CTGF expression was determined by immunohistological staining (Figure 1A). Stainings were classified as level 0 (negative staining), level 1 (<5% of tumor cells stained), level 2 (<50% of tumor cells stained), and level 3 (>50% of tumor cells stained). These authors did not find a significant relationship between CTGF expression and patient age, sex, tumor stage, grade of differentiation, preoperative CEA level, or invasion depth. However, CTGF expression was associated with disease outcome. Patients with low CTGF expression had shorter survival and a higher recurrence rate (P < .001 and P < .001, respectively). Patients were also stratified by tumor stage. Patients with stage II or III disease whose tumor had a high level of CTGF expression had better disease-free status (HR = 0.143, P < .001) and better overall survival (HR = .186, P < .001) than patients with a low expression of CTGF. The 5-year survival for stage II patients with high CTGF expression was 87.5% compared with 47.5% in patients with low CTGF expression (P = .005). Similarly, the 5-year survival for stage III patients with high CTGF expression was 76.7% compared with 28.1% in those with low CTGF expression (P = .004). CTGF expression was also measured in 4 human cancer cell lines, HCT116, COLO205, HT-29, and Caco-2. HCT116 cells had the lowest expression of CTGF. The study results also demonstrated an inverse relationship between invasion ability of the cancer cells and CTGF expression levels. To explore the relationship between CTGF expression and hepatic metastases in vivo, BALB/c mice were given intrasplenic/portal injections of CTGF transfectants (CT26/AS-CTGF) and Neo control cells. The number of hepatic metastatic nodules was significantly greater in mice injected with CT26/AS-CTGF as compared with Neo control cells (P = .0039), demonstrating a role of CTGF as a negative regulator of hepatic metastasis by CRC in mice. In previous work, β-catenin/Tcf signal pathway has been shown to be involved in CRC invasion and metastasis; therefore, these authors transfected HCT116 cells with CTGF and found that CTGF overexpression suppresses β-catenin/Tcf activity and decreases the expression of downstream effector gene MMP-7. These authors conclude that CTGF is an independent prognostic factor with the potential of providing a means of successfully differentiating stage II and III patients. The study by Lin et al also suggests that CTGF inhibits CRC invasion and metastasis, possibly through the inhibition of the β-catenin/Tcf signal pathway, providing a novel target for the treatment of CRC. See page 9 Several animal studies have suggested a beneficial role of coffee products on the liver. Additionally, previous epidemiologic studies conducted outside of the US have identified an inverse relationship between coffee consumption and lower levels of liver enzymes. While the mechanism of the possible protective effect of coffee or its products is unclear, potential protective effects could include antagonistic effects of the A1 and A2 adenosine receptors, antioxidant effects, or the inhibition of lipid peroxidation. If coffee or one of its components is protective to the liver, patients at high risk for liver disease may benefit from increased coffee or caffeine intake (Figure 2). Ruhl et al examined the relationship of coffee and caffeine intake with serum ALT activity in a large, national, population-based study of persons at high risk for liver disease. Persons who participated in the third US National Health and Nutrition Examination Survey, 1988–1994, and had excessive alcohol consumption, hepatitis B or C, iron overload, impaired glucose metabolism, or an increased BMI ≥26.9 kg/m2 or waist-to-hip ratio ≥0.94 were included in the analysis. An elevated ALT was found in 8.7% of the patients included in the analysis. Multivariate analysis revealed a decreased risk of elevated ALT with increasing intake of coffee (P value for trend = .034) and caffeine (P < .001). The odds ratio of participants who drank more than 2 cups per day compared with persons who did not drink coffee was just below the level of significance (OR = 0.56, 95% CI = 0.31–1.0). The odds ratio of participants in the highest caffeine intake quintile was 0.31 with a 95% confidence interval (CI) = 0.16–0.61. These findings were consistent among subgroups at risk for liver injury. No statistically significant interactions were found for coffee or caffeine intake with the other risk factors for liver disease. These data support a relationship between coffee and, especially, caffeine consumption and a lower risk of ALT elevation. These findings warrant further investigation of the beneficial effects of coffee and caffeine on the liver. See page 24 Metastasis of colorectal cancer involves enhanced motility of cells to spread beyond the primary tumor, although the mechanisms involved are not clearly understood. For cells to be motile, molecular components of the cytoskeleton, cell adhesion, and signaling systems must be simultaneously activated to dynamically reshape the cell and remodel the actin cytoskeleton to form a distinct leading edge known as filopodia and lamellipoida. The acquired properties of this region provide the cell with the necessary mechanical forces for directional movement. Several small GTP-binding proteins, Cdc42, and Rac initiate this process by activation of WASP family proteins and the Arp2/3 complex. A result of these processes may be activation of certain isoforms of α-actinin, which are actin-binding effector proteins that regulate actin cytoskeleton. Of the 4 isoforms of human α-actinin, 2 non-muscle types, actinin-1, and actinin-4 appear to be expressed in intestinal epithelial cells. Actinin-1 is found at points of focal contacts and adherens junctions, whereas actinin-4 is highly concentrated at the leading edge of motile cells and in cytoplasmic regions with sharp cell extensions. In mouse macrophages, actinin-4 is preferentially localized to moving structures, such as the dorsal ruffles of mouse macrophages. The cytoplasmic localization of actinin-4 is also associated with the invasive phenotype of breast cancer and is a prognostic predictor of breast cancer. Thus, actinin-4 appears to be involved in cell movement, cancer invasion, and metastasis. The study by Honda et al examined the possibility that actinin-4 plays a role in colon cancer metastasis to lymph nodes. By quantitative immunofluorescence, the expression level of actinin-4 protein was observed to be significantly increased in 73.1% (19 of 26 cases) of colorectal cancer over the corresponding normal intestinal epithelium. The increased expression of actinin-4 was most significant in dedifferentiated cancer cells at the invasive front. To better define a causal relationship, a colorectal cancer cell clone was established where actinin-4 expression could be specifically induced through a tetracycline-regulated promoter (DLD1 Tet-off ACTN-4 cells). When actinin-4 expression was induced, DLD1 Tet-off ACTN-4 cells spread filopodia and significantly increased their motility and actinin-4 protein was concentrated at the leading edges (Figure 3). When these cells were injected into the mesocecum of severe combined immunodeficient (SCID) mice, they metastasized into regional mesenteric lymph nodes (see right panel), in contrast to control cells that did not. The expression of actinin-4 in focally dedifferentiated cancer cells at the invasive front was significantly correlated with the frequency of lymph node metastasis of colorectal cancer. These findings are quite important and provide rationale for development of therapies specifically targeting actinin-4 to prevent or limit colorectal cancer metastasis. See page 51 Nonsteroidal anti-inflammatory drug (NSAID)-induced gastric and duodenal ulcers are most commonly attributed to the inhibitory effects of these agents on prostaglandin synthesis via blockade of cyclooxygenase (COX) isoenzymes, COX-1 and COX-2. This has been the rationale for development of selective COX-1 (eg, SC-560) and COX-2 inhibitors (eg, celecoxib and NS-398), which may have fewer injurious effects on gastric mucosa than nonselective inhibitors like indomethacin. However, it has been observed that NSAIDs also cause gastrointestinal damage through induction of necrotic and apoptotic cell death, although there is little understanding of how this occurs. One possibility examined by this study is that NSAIDs affect survivin, a 16.5-kilodalton protein member of the inhibitors of apoptosis protein family and a broad spectrum suppressor of cell death. Its action involves binding to caspase-3 and -7, and inhibition of caspase-independent cell death. Survivin is expressed in mucosal neck cells, which are progenitor cells in both rat and human stomach. The study by Chiou et al therefore examined the role of survivin in NSAID-induced gastric mucosal and gastric cell injury in vivo and in vitro. Indomethacin treatment caused a dose-dependent inhibition of survivin protein levels, which was associated with severe injury of gastric mucosa and RGM-1 cells. Inhibition of survivin expression with interference RNA (siRNA) in RGM-1 cells resulted in cell damage and increased susceptibility to injury by indomethacin (Figure 4). In contrast, celecoxib treatment caused exfoliation of the mucosal surface epithelium, but did not cause deep erosions or alter survivin expression. NS-398 and SC-560 treatment also had no effects on survivin levels or on gastric mucosa. Thus, indomethacin, but not selective COX-1 or COX-2 inhibitors alone or in combination, inhibit survivin expression in gastric mucosal cells. This mechanism of gastric injury is distinct from that associated with COX-1 and COX-2 inhibitors, which cause mucosal injury via a nonsurvivin-mediated process. See page 63