SPARC Is Associated with Gastric Cancer Progression and Poor Survival of Patients Free

According to global estimates of cancer incidence in 2002, gastric cancer is the most frequent cancer-related cause of death second only to lung cancer. The incidence of gastric cancer is estimated to be 934,000 cases with 56% of the new cases occurring in East Asia, including 41% in China and 11% in Japan (1). Although the global incidence of gastric cancer has decreased in recent years, its mortality rate in China is the highest among all tumors and represents 25% of gastric cancer mortality worldwide. Despite recent advances in chemotherapy and surgical techniques, the overall 5-year survival rate in China is low at 40%. Most gastric cancer is diagnosed at stage III or IV, and the rate of lymph node metastasis is higher (50–75%; ref. 2). The pathogenesis of gastric carcinomas is multifactorial and includes genetic predisposition and environmental factors. Genetic predisposition has been found to be accompanied by several genetic alternations including tumor suppressor genes, oncogenes, cell adhesion molecules, growth factors, and genetic stability (3). Therefore, it is of great clinical value to further understand the molecular mechanisms involved in gastric cancer and to find valuable diagnostic markers as well as novel therapeutic strategies.

Secreted protein, acidic and rich in cysteine (SPARC), is a unique matricellular glycoprotein that is expressed by many different types of cells and is associated with development, remodeling, cell turnover, and tissue repair. Its principal functions in vitro are counteradhesion and antiproliferation, which proceed through different signaling pathways (4). Recent studies have revealed other biological functions including cell proliferation, migration, morphogenesis, deadhesion, antiproliferation, differentiation, and angiogenesis (5). SPARC was reported to be overexpressed in a variety of human malignancies, including glioma, colorectal, breast, esophageal, prostate, bladder, and thyroid carcinomas (6–11). Overexpression of the SPARC gene was observed in human gastric cancer (12, 13). SPARC has adaptors that mediate cell-extracellular matrix (ECM) interactions (14) and are expressed in tissues undergoing repair or remodeling. Consistent with its function as a mediator of tissue remodeling, SPARC regulates the expression of proteins involved in ECM turnover and the formation of collagens and matrix metalloproteinases (MMP; ref. 15). SPARC upregulates MT1-MMP levels and MMP-2 activity, and their upregulation and activation may be a consequence of increased SPARC expression (16). SPARC protects cells from stress-induced apoptosis in vitro through an interaction with integrin β1 heterodimers that enhance ILK activation and prosurvival activity (17).

Here, we analyzed the prognostic significance of SPARC, intergrin β1, and MMP-2. The purposes of the current study were to examine the expression of SPARC in surgical specimens of gastric carcinoma, to explore the possible correlation between SPARC expression and clinicopathologic variables, to correlate expression of SPARC with lymph node metastasis and distant metastasis, and to determine the prognostic value of SPARC expression.

Gastric cancer tissues were collected from gastrectomy specimens of 436 patients (median age, 60.0 y; range, 30–91 y; 311 male, 125 female) from the Department of Surgery, Zhejiang Provincial People's Hospital from January 1998 to January 2004. Tissue had been formalin-fixed, paraffin-embedded, and clinically and histopathologically diagnosed at the Departments of Gastrointestinal Surgery and Pathology. All patients had follow-up records for over 5 y. The follow-up deadline was December 2008. The survival time was counted from the date of surgery to the follow-up deadline or date of death, which was mostly caused by carcinoma recurrence or metastasis. There were 55, 163, and 218 cases from the cardia, body, and antrum, respectively. According to the WHO histologic classification of gastric carcinoma formulated in 2002, there were 326 tubular adenocarcinomas, 16 papillary adenocarcinomas, 29 mucinous adenocarcinomas, 65 signet ring cell carcinomas, and 13 highly differentiated adenocarcinomas; 128 were classified as well or moderately differentiated adenocarcinomas, 293 as poorly differentiated, and 2 as undifferentiated adenocarcinomas. There were 61 cases with distant metastasis. Ninety cases were categorized as stage I, 104 were stage II, 173 were stage III, and 69 were stage IV. Ninety-two noncancerous human gastric tissues were obtained from gastrectomies of adjacent gastric cancer margins greater than 5 cm. Routine chemotherapy was given to the patients with advanced stage disease after operation, but no radiation treatment was administered to any of the patients included in our study.

Blocks containing a total of 528 cases (436 cancer samples + 92 noncancerous tissue samples) were prepared as described previously (18, 19). Core tissue biopsies (2 mm in diameter) were taken from individual paraffin-embedded gastric tumors (donor blocks) and arranged in recipient paraffin blocks (tissue array blocks) using a trephine. Because it has previously been proven that staining results obtained from different intratumoral areas in various tumors correlate well (20), a core was sampled in each case. An adequate case was defined as a tumor occupying >10% of the core area (21). Each block contained more than three internal controls consisting of nonneoplastic gastric mucosa. Four-micrometer-thick sections were cut from each tissue array block, were deparaffinized, and were dehydrated.

Immunohistochemical analysis was done to study altered protein expression in 92 noncancerous human gastric tissue controls and 436 human gastric cancer tissues (22, 23). In brief, slides were baked at 60°C for 2 h followed by deparaffinization with xylene and rehydrated. The sections were submerged into EDTA antigenic retrieval buffer and microwaved for antigenic retrieval, after which they were treated with 3% hydrogen peroxide in methanol to quench endogenous peroxidase activity, followed by incubation with 1% bovine serum albumin to block nonspecific binding. Sections were incubated with mouse anti-SPARC (Santa Cruz), mouse anti–intergrin β1 (ADCAM USA), and with mouse anti-MMP-2 (Santa Cruz) overnight at 4°C. Normal goat serum was used as a negative control. After washing, tissue sections were treated with secondary antibody. Tissue sections were then counterstained with hematoxylin, were dehydrated, and were mounted. The cytoplasm and stroma with SPARC was stained as buffy, whereas intergrin β1 and MMP-2 were stained as buffy in cytoplasm. The degree of immunostaining was reviewed and scored independently by two observers based on the proportion of positively stained tumor cells and intensity of staining (24–26). Tumor cell proportion was scored as follows: 0 (≤5% positive tumor cells), 1 (6–25% positive tumor cells), 2 (26-50% positive tumor cells), and 3 (>51% positive tumor cells). Staining intensity was graded according to the following criteria: 0 (no staining), 1 (weak staining, light yellow), 2 (moderate staining, yellow brown), and 3 (strong staining, brown). Staining index was calculated as the product of staining intensity score and the proportion of positive tumor cells. Using this method of assessment, we evaluated SPARC, intergrin β1, and MMP-2 expression in benign gastric epithelia and malignant lesions by determining the staining index with scores of 0, 1, 2, 3, 4, 6, or 9. The cutoff value for high and low expression level was chosen based on a measure of heterogeneity using the log-rank test with respect to overall survival. An optimal cutoff value was identified as follows: a staining index score of ≥4 was used to define tumors with high SPARC, intergrin β1, and MMP-2 expression, and a staining index score of ≤3 was used to indicate low SPARC, intergrin β1, and MMP-2 expression.

All statistical analyses were done using the SPSS11.0 software. Measurement data were analyzed using the Student's t test, whereas categorical data were studied using χ² or Fisher exact tests. Survival curves were estimated using the Kaplan-Meier method, and the log-rank test was used to compute differences between the curves. Multivariate analysis using the Cox proportional hazards regression model was done to assess the prognostic values of protein expression. Correlation coefficients between protein expression and clinicopathologic findings were estimated using the Pearson correlation method. Statistical significance was set at P < 0.05.

SPARC protein was detected in 67 (72.83%) of 92 human nontumor mucosa, and all samples expressed the protein at low levels. SPARC protein was detected in 334 (76.61%) of 436 human gastric cancer cases. High expression of SPARC protein was detected in 239 (54.82%) tumors, and low expression was detected in 95 (21.79%) tumors. SPARC was mainly localized in the cytoplasm or stroma of primary cancer (Figs. 1 and 2). MMP-2 protein was detected in 74 (80.44%) of 92 human nontumor mucosa. High expression of MMP-2 protein was detected in 4 (4.35%) nontumor mucosa, and low expression was detected in 70 (76.09%) nontumor mucosa. MMP-2 protein was detected in 307 (70.41%) of 436 human gastric cancer cases. High expression of MMP-2 protein was detected in 177 (40.59%) tumors, and low expression was detected in 130 (29.82%) tumors. MMP-2 was mainly localized in the cytoplasm of primary cancer (Fig. 3). Integrin β1 protein was detected in 70 (76.09%) of 92 human nontumor mucosa. High expression of intergrin β1 protein was detected in 2 (2.18%) nontumor mucosa, and low expression was detected in 68 (73.91%) nontumor mucosa. Integrin β1 protein was detected in 347 (79.59%) of 436 human gastric cancer cases. High expression of intergrin β1 protein was detected in 195 (44.73%) tumors, and low expression was detected in 152 (34.86%) tumors. Intergrin β1 was mainly localized in the cytoplasm of primary cancer (Fig. 4).

Positive expression of SPARC correlated with age, size of tumor, differentiation, depth of invasion, vessel invasion, lymph node and distant metastasis, and tumor-node-metastasis (TNM) stage (P < 0.05; Table 1). SPARC expression did not correlate with gender, location of tumor, and histologic type (P > 0.05; Table 1). The expression of MMP-2 correlated with age, location of tumor, size of tumor, depth of invasion, vessel invasion, lymph node and distant metastasis, and TNM stage (P < 0.05; Table 1). MMP-2 did not correlate with gender, differentiation, and histologic type (P > 0.05; Table 1). Integrin β1 correlated with age, location of tumor, size of tumor, depth of invasion, vessel invasion, lymph node and distant metastasis, and TNM stage (P < 0.05; Table 1). Expression of intergrin β1 was not correlated with gender, differentiation, and histologic type (P > 0.05; Table 1). The factors with possible prognostic effects in gastric carcinoma were analyzed by Cox regression analysis. The study revealed that the depth of invasion (P = 0.039); lymph node (P = 0.009) and distant metastasis (P = 0.005); TNM stage (P = 0.000); and the expression of SPARC (P = 0.001), MMP-2 (P = 0.004), and intergrin β1 (P = 0.000) were independent prognostic factors in patients with gastric carcinoma. However, the location of the tumor, tumor size, histologic type, differentiation, and vessel invasion had no prognostic value.

Ninety-seven gastric cancer cases had a low expression of SPARC and MMP-2 simultaneously. Two hundred and seven gastric cancer cases had a high expression of SPARC and MMP-2 at the same time. There was a significant correlation between SPARC and MMP-2 (r = 0.391, P = 0.000; Table 1). One hundred and sixty-three gastric cancer cases had a low expression of SPARC and intergrin β1 simultaneously, and 161 gastric cancer cases had a high expression of SPARC and intergrin β1 at the same time. There was significant correlation between SPARC and intergrin β1 (r = 0.502; P = 0.000; Table 1).

In stages I, II, and III, the 5-year survival rate of patients with a high expression of SPARC were significantly lower than those in patients with low expression. In stage I, the cumulative 5-year survival rate was 91.4% in the low SPARC protein expression group but was only 65.0% in the high expression group (P = 0.002). In stage I, the cumulative 5-year survival rate was 69.1% in the low SPARC protein expression group but was only 50.0% in the high expression group (P = 0.04). In stage III, the cumulative 5-year survival rate was 29.4% in the low SPARC protein expression group but was only 15.6% in the high expression group. In stage IV, the expression of SPARC did not correlate with the 5-year survival rate. The cumulative 5-year survival rate was 0 in the low SPARC protein expression group but was only 3.3% in the high expression group in stage IV (P = 0.82). The cumulative 5-year survival rate of patients with a simultaneously high expression of SPARC, MMP-2, and intergrin β1 were significantly lower than those in patients with simultaneously low expression (P < 0.05; Fig. 5).

Overexpression of the SPARC gene was observed in human gastric cancer. However, high levels of SPARC often correlate with enhanced invasion, metastasis, and poor prognosis (13, 27–29). The possible clinical significance of SPARC has remained unclear in gastric cancer patients. Therefore, we used immunohistochemical techniques to examine the relationships between SPARC expression and the clinicopathologic characteristics of patients with gastric cancer.

The current study revealed that SPARC, intergrin β1, and MMP-2 were upregulated in gastric cancer tissues in comparison with those in normal gastric tissues. SPARC, intergrin β1, and MMP-2 protein levels were found to significantly correlate with the prognosis of gastric cancer. In addition, a high level of SPARC, intergrin β1, and MMP-2 protein expression in gastric cancer lesions is closely associated with the age of the patients, size of tumor, differentiation, depth of invasion, vessel invasion, lymph node and distant metastasis, and TNM stage. In stages I, II, and III, the 5-year survival rate of patients with a high expression of SPARC was significantly lower than those in patients with low expression. In stage IV, the expression of SPARC did not correlate with the 5-year survival rate. Further multivariate analysis suggested that the depth of invasion; lymph node and distant metastasis; TNM stage; and upregulation of SPARC, MMP-2, and intergrin β1 were independent prognostic indicators for the disease.

High levels of SPARC often correlate with enhanced invasion, metastasis, and poor prognosis (27, 28). Higher expression of SPARC was significantly associated with lymph node metastasis, lymphatic invasion, and perineural invasion. Expression of SPARC in patients in stage II and above was significantly higher than those in stage I. The 3-year survival of patients with lower expression of SPARC was significantly better than those with a higher expression (13). The expression of SPARC by peritumoral fibroblasts portends a poorer prognosis for patients with pancreatic cancer (30).

Previous studies have shown that SPARC has adaptors that mediate cell-ECM interactions (14) and are expressed in tissues undergoing repair or remodeling. Consistent with its function as a mediator of tissue remodeling, SPARC regulates the expression of proteins involved in ECM turnover and formation including collagens and MMP (15). A recent study found that SPARC protects cells from stress-induced apoptosis in vitro through an interaction with integrin β1 heterodimers that enhance ILK activation and prosurvival activity (17). We also examined MMP-2 and intergrin β1 expression in gastric cancer specimens and its correlation with SPARC status. A possible correlation between MMP-2 and intergrin β1 expression and lymph node metastasis has previously been described (31–37). We found a positive correlation between SPARC and MMP2, and SPARC and intergrin β1 expression on gastric cancer invasion and metastasis, suggesting that cells with positive SPARC expression may promote gastric cancer cell invasion and metastasis. SPARC upregulates MT1-MMP levels and MMP-2 activity and their upregulation and activation may be a consequence of increased SPARC expression (16). Expression of SPARC mRNA was associated significantly with MMP-2 mRNA expression, and SPARC accumulation may reflect a functional correlation with MMP-2 expression, which could play a key role in the progression of esophageal carcinoma (8).

Our study suggests that overexpression of SPARC, intergrin β1, and MMP-2 is a common feature in gastric cancer that might play an important role in the progression and metastases of gastric cancer. Our study identified upregulation of SPARC, intergrin β1, and MMP-2 in gastric cancer, including an assessment of SPARC, intergrin β1, and MMP-2 expressions in gastric cancer tissues and noncancerous gastric tissues. The importance of SPARC, intergrin β1, and MMP-2 upregulation in gastric cancer are further highlighted by our results that correlate the pathologic parameters of tumors with a poorer patient prognosis. These results suggest that SPARC, intergrin β1, and MMP-2 may be useful as prognostic and survival indicators. Our study has provided a basis for the development of a novel biomarker for the diagnosis and prognosis of gastric cancer.

No potential conflicts of interest were disclosed.

Grant Support: Zhejiang Provincial Department of Science and Technology Research Foundation 2008C33040 and Zhejiang Provincial Medical Science Research Foundation 2007A013.

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 with 18 U.S.C. Section 1734 solely to indicate this fact.