Absence of Association Between Transforming Growth Factor B1 Polymorphisms and Gastric Cancer: A Meta-Analysis

Abstract

Transforming growth factor-β1 (TGF-β1) is a ubiquitous and essential cytokine that plays a dual role in the development of cancer: tumor suppressor in the early stage of carcinogenesis and tumor promoter in the later stage of carcinogenesis. To date, several studies have focused on the association between the TGFB1 polymorphisms and risk of gastric cancer. Conflicting results, however, have been reported. We conducted a meta-analysis to analyze more precisely the effects of the TGFB1 C-509T, T869C, and G915C polymorphisms on the risk of gastric cancer. No significant association between the TGFB1 C-509T, T869C, and G915C polymorphisms and risk of gastric cancer was observed in overall analyses and subgroup analyses according to ethnicity.

Introduction

G astric cancer is the fourth most common malignancy worldwide and can develop in any part of the stomach (Parkin et al., 2005). In 2008, approximately 989,600 new cases were diagnosed and 738,000 patients died from gastric cancer (Jemal et al., 2011). Over two-thirds of the cases and deaths occurred in developing countries and 42% in China alone (Parkin et al., 2005; Jemal et al., 2011). Helicobacter pylori infection, tobacco smoking, and consumption of alcohol are risk factors for the development of gastric cancer (Nomura et al., 1990; Tredaniel et al., 1997; Parkin, 2006; Sung et al., 2007; Shimazu et al., 2008). On the other hand, consumption of fresh fruits and vegetables appears to lower the risk of gastric cancer, probably because such food contains antioxidant vitamins (such as A and C) (Palli, 2000). However, not all individuals develop gastric cancer even though they had the same environment, indicating that genetic variations, such as single-nucleotide polymorphism (SNP) in humans may contribute to the risk of gastric cancer (El-Omar et al., 2001).

Transforming growth factor-beta (TGF-β), a pluripotent cytokine, plays a biphasic role in cancer development: tumor suppressor in the early stage of carcinogenesis and tumor promoter in the later stage of carcinogenesis (Blobe et al., 2000; de Caestecker et al., 2000; Massague et al., 2000; Derynck et al., 2001). TGF-β exists in 3 isoforms called TGF-β1, TGF-β2, and TGF-β3. Of the three isoforms, TGF-β1 is most frequently upregulated in tumor cells, and thus most studies focus on the role of TGF-β1 in human cancers. The TGFB1 gene is located on the long arm of chromosome 19 (19q13.1) (Yoshie et al., 2007). TGF-β1 plays an essential role in carcinogenesis, and is involved in various cellular processes including proliferation, differentiation, migration, apoptosis, angiogenesis, and immunosurveillance (Elliott and Blobe, 2005; Mishra et al., 2005). In the early stage of cancer development, TGF-β1 functions as a tumor suppressor. In the later stage of cancer development, however, TGF-β1 acts as a tumor promoter by stimulating tumor invasion or suppressing immune response (Blobe et al., 2000; de Caestecker et al., 2000; Massague et al., 2000; Derynck et al., 2001).

Emerging evidence has shown that enhanced serum concentration of TGF-β1 was detected in a wide range of human cancers, including breast cancer (Gorsch et al., 1992), lung cancer (Barthelemy-Brichant et al., 2002), prostate cancer (Ivanovic et al., 1995), colorectal cancer (Shim et al., 1999; Langenskiold et al., 2008), hepatocellular carcinoma (Song et al., 2002), and gastric cancer (Lin et al., 2006; Li et al., 2008b). The high expression of TGF-β1 has also been demonstrated in gastric cancer and in the gastric mucosa of first-degree relatives of patients with gastric cancer (Ebert et al., 2000). Overexpression of TGF-β1 is associated with the invasion, metastasis, and poor prognosis of gastric carcinoma (Maehara et al., 1999; Saito et al., 1999). Moreover, conditional loss of TGF-β signaling resulted in increased susceptibility to gastrointestinal carcinogenesis in mice (Hahm et al., 2002). These findings indicate that TGFB1 may be a candidate gene for the development of gastric cancer. Previously, three common SNPs were identified in the TGFB1 gene: one in the promoter region (C-509T), and the others in the coding region with nonsynonymous amino acid substitution (Leu10Pro, and Arg25Pro, respectively). Several studies have reported that these variations influence the production of TGF-β1 (Awad et al., 1998; Yamada et al., 1998; Grainger et al., 1999.

Many molecular epidemiological studies have been performed to investigate the association between the three polymorphisms and risk of gastric cancer. However, the results were inconclusive. For example, Jin et al. (2007) reported that the −509 CT genotype was associated with a significantly reduced risk of gastric cancer. In contrast, other studies reported that the −509 CT genotype was associated with a significantly increased risk of gastric cancer (Li et al., 2008a; Guo et al., 2011). Because a single study might have lower power to estimate the overall effects, we conducted a meta-analysis by combining data from published case–control studies to provide more analysis of the effects of the TGFB1 polymorphisms on the risk of gastric cancer.

Materials and Methods

Selection of published studies

We searched for articles using the following terms (“TGF-beta” or “TGF-β” or “transforming growth factor-beta” or “transforming growth factor-β”) and (“polymorphism”) and (“gastric cancer”) in the PubMed, Embase, China National Knowledge Infrastructure (CNKI), and Chinese Biological Medicine (CBM) databases (last search: July 13, 2011). No language restrictions were used. The inclusion criteria were: (1) articles investigating the association between TGFB1 C-509T (rs1800469), T869C (rs1800470; rs1982073; T29C; Leu10Pro), and G915C (R25P; rs1800471) polymorphisms and gastric cancer risk; (2) case–control studies; and (3) providing enough data for computing odds ratios (ORs) with 95% confidence intervals (CIs). We excluded the studies that were repeatedly published in other languages by the same research group (Zhou et al., 2007; Guo et al., 2010). Thus, a total of eight articles were included in this meta-analysis (Fig. 1). Of the eight studies, six studies investigated the association between the TGFB1 C-509T polymorphism and risk of gastric cancer, six studies investigated the association between the TGFB1 T869C polymorphism and risk of gastric cancer, and four studies investigated the association between the TGFB1 G915C polymorphism and risk of gastric cancer.

FIG. 1.

Flow diagram of the selection of eligible studies.

Data extraction

Two investigators worked independently to extract the following information from each study: the first author's name, the year of publication, country of origin, the year of sample collection, ethnicity, the number of cases and controls, genotyping methods, matching criteria, and Hardy–Weinberg equilibrium (HWE; Table 1).

Table 1.

Overview of Literatures Included in the Meta-Analysis

References	Year of publication	Country (region)	Year of sample collection	Ethnicity	Sample size (cases/controls)	Genotyping methods	Matching criteria	Polymorphisms	HWE
Garcia-Gonzalez et al. (2006)	2006	Spain (Zaragoza)	2002–2005	Caucasian	142/342	PCR-RFLP	Age and sex	T869C; G915C	Yes
Chen et al. (2010b)	2010b	China (Shenzhen)	—	Asian	80/102	PCR-RFLP	—	G915C	Yes
Guo et al. (2011)	2011	China (Hebei)	2004–2009	Asian	468/584	PCR-RFLP	Age and sex	C-509T; T869C; G915C	No
Jin et al. 2007	2007	China (Jiangsu)	2003–2005	Asian	636/676	PIRA-PCR	Age and sex	C-509T; T869C	Yes
Li (2008a)	2008a	China (Shandong)	2004–2005	Asian	167/193	PCR-RFLP	—	C-509T; T869C	No
Zhang et al. (2008)	2008	China (Shanghai)	2004–2006	Asian	414/414	PCR-RFLP	Age and sex	C-509T; T869C	Yes
Chen et al. (2010a)	2010a	China (Shenzhen)	—	Asian	80/102	PCR-RFLP	—	C-509T	Yes
Guan et al. (2009)	2009	United States (Texas)	1990–2008	Caucasian	171/353	PCR-RFLP	Age, sex and ethnicity	C-509T;T869C; G915C	No; Yes; Yes

PCR-RFLP, polymerase chain reaction–restriction fragment length polymorphism; HWE, Hardy–Weinberg equilibrium; PIRA-PCR, primer-introduced restriction analysis–polymerase chain reaction.

Statistical analysis

The pooled ORs and corresponding 95% CIs were performed to evaluate the association between the three polymorphisms in TGFB1 gene and gastric cancer risk. Four different ORs were calculated: (a) heterozygote comparison (C-509T: CC vs. CT; T869C: TT vs. TC; and G915C: GG vs. GC); (b) homozygote comparison (C-509T: CC vs. TT; T869C: TT vs. CC; and G915C: GG vs. CC); (c) dominant model (C-509T: CC/CT vs. CC; T869C: TC/CC vs. TT; and G915C: GC/CC vs. GG); and (d) recessive model (C-509T: TT vs. CC/CT; T869C: CC vs. TT/TC; and G915C: CC vs. GG/GC). The Q-test and I ² statistics were used to assess the between-study heterogeneity (Higgins and Thompson, 2002). If the p-value for Q-test ≤0.10 or heterogeneity index I ²>50%, indicating obvious heterogeneity between studies, then a random-effects (the DerSimonian and Laird method) model was used to compute the pooled ORs (DerSimonian and Laird, 1986). In contrast, if the p-value for Q-test > 0.10 and heterogeneity index I ²≤50%, indicating the absence of heterogeneity between studies, then a fixed-effects (the Mantel-Haenszel method) model was used (Mantel and Haenszel, 1959). Logistic meta-regression analysis was conducted to explore potential reason for the between-study heterogeneity. Publication bias was carried out using a Begg's funnel plot (Egger et al., 1997). All the statistical analyses were done using STATA software, version 10.0 (STATA Corp., College Station, TX).

Results

A flow diagram of the selection of eligible studies is shown in Figure 1. In all, 57 records were identified through screening the PubMed, Embase, CNKI, and CBM databases. Forty-eight records remained after removing duplicates. We excluded 37 records because they were review articles (n=3) or where absence of polymorphism (n=29), gastric cancer (n=4), and control (n=1) was noted. Then, full-text articles were assessed for eligibility, and 3 records were excluded for overlapped data (n=2) or lacking available data (n=1). Finally, 8 studies, including 6 studies of the TGFB1 C-509T polymorphism (1936 cases and 2322 controls), 6 studies of the TGFB1 T869C polymorphism (1998 cases and 2562 controls), and 4 studies of the TGFB1 G915C polymorphism (861 cases and 1381 controls) were included in this meta-analysis.

The main characteristics of the eligible studies are shown in Table 1. Of the eight studies, six were of Asians, and two were of Caucasians. All the gastric cancer cases in the included studies were pathologically confirmed. Most of the studies (87.5%) performed polymerase chain reaction–restriction fragment length polymorphism analysis for genotyping. Several quality control methods were used to avoid genotyping error, including positive and negative control, internal control, random repeat, blindness to the case–control status, and different genotyping assays for validation. Three studies were found with significant deviation from HWE (Glinsky, 2008; Guan et al., 2009; Guo et al., 2011).

The main results of this meta-analysis are summarized in Tables 2 –4. No significant associa tion between the TGFB1 C-509T, T869C, and G915C polymorphisms and the risk of gastric cancer was observed in overall analyses and subgroup analyses according to ethnicity. Obvious heterogeneity was found in the effect of TGFB1 C-509T and T869C polymorphisms on the risk of gastric cancer. To explore the source of the heterogeneity, we assessed all comparison models by HWE, ethnicity, and sample size (≤500 and > 500 subjects) with meta-regression analysis. We found that the variable HWE could explain the heterogeneity (TGF-β1 T869C: TT vs. TC, and TT vs. CC). The results of the meta-analysis were not altered after exclusion of studies of Hardy–Weinberg disequilibrium (HWD).

Table 2.

Association Between the Transforming Growth Factor-B1 C-509T Polymorphism and Gastric Cancer Risk

			CC versus CT		CC versus TT		CT/TT versus CC (dominant)		TT versus CC/CT (recessive)
	n^a	Cases/controls	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}
Total	6	1936/2322	0.89 (0.56–1.41)^b	<0.001	0.83 (0.55–1.24)^b	<0.001	1.16 (0.76–1.77)^b	<0.001	1.13 (0.98–1.31)	0.31
Asian	5	1765/1969	0.83 (0.48–1.42)^b	<0.001	0.76 (0.48–1.18)^b	<0.001	1.26 (0.77–2.07)^b	<0.001	1.16 (1.00–1.34)	0.36
HWE	3	1130/1192	1.20 (0.75–1.91)^b	0.009	1.14 (0.91–1.42)	0.11	0.89 (0.59–1.36)^b	0.012	1.07 (0.88–1.30)	0.29

Number of comparisons.

Random-effects model was used when p-value for heterogeneity test <0.10.

OR, odds ratio; CI, confidence interval.

Table 3.

Association Between the Transforming Growth Factor-B1 T869C Polymorphism and Gastric Cancer Risk

			TT versus TC		TT versus CC		TC/CC versus TT (dominant)		CC versus TT/TC (recessive)
	n^a	Cases/controls	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}
Total	6	1998/2562	0.84 (0.61–1.16)^b	<0.001	0.79 (0.55–1.15)^b	<0.001	1.55 (0.99–2.44)^b	<0.001	0.93 (0.63–1.39)^b	<0.001
Asian	4	1685/1867	0.73 (0.47–1.14)^b	<0.001	0.69 (0.42–1.16)^b	<0.001	1.40 (0.89–2.23)^b	<0.001	1.12 (0.96–1.30)	0.12
Caucasian	2	313/695	1.12 (0.82–1.51)	0.68	1.05 (0.72–1.53)	0.50	1.88 (0.45–7.80)^b	<0.001	0.56 (0.12–2.54)^b	<0.001
HWE	4	1363/1785	1.08 (0.91–1.28)	0.87	1.08 (0.88–1.32)	0.92	1.30 (0.73–2.30)^b	<0.001	0.75 (0.44–1.28)^b	<0.001

Number of comparisons.

Random-effects model was used when p-value for heterogeneity test <0.10.

Table 4.

Association Between the Transforming Growth Factor-B1 G915C Polymorphism and Gastric Cancer Risk

			GG versus GC		GG versus CC		GC/CC versus GG (dominant)		CC versus GG/GC (recessive)
	n^a	Cases/controls	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}	OR (95%CI)	p_{heterogeneity}
Total	4	861/1381	0.89 (0.65–1.21)	0.91	0.71 (0.23–2.14)	0.59	1.02 (0.76–1.37)	0.52	1.00 (0.32–3.14)	0.81
Asian	2	548/686	0.88 (0.55–1.40)	0.97	0.79 (0.18–3.49)	0.41	1.15 (0.73–1.80)	0.80	1.26 (0.29–5.54)	0.41
Caucasian	2	313/695	0.89 (0.59–1.35)	0.46	0.61 (0.12–3.18)	0.28	0.93 (0.63–1.38)	0.18	0.72 (0.12–4.40)	0.72
HWE	3	393/797	0.89 (0.61–1.31)	0.76	0.50 (0.12–2.07)	0.52	0.97 (0.67–1.40)	0.35	1.14 (0.26–5.08)	0.64

Number of comparisons.

Begg's funnel plot was used to evaluate the publication bias, and no evidence of publication bias was detected in this meta-analysis.

Discussion

In this study, we failed to find any association between the TGFB1 C-509T, T869C, and G915C polymorphisms and risk of gastric cancer in overall analyses. After subgroup analyses according to ethnicity, we did not find any association between the three polymorphisms in TGFB1 gene and gastric cancer in both Asians and Caucasians. Moreover, three out of eight studies showed significant deviation from HWE. When we excluded these studies of HWD, the results of the meta-analysis were not modified. These findings indicate that the TGFB1 polymorphisms may not be risk factors for the development of gastric cancer.

Several common functional polymorphisms in the TGFB1 gene have been reported. One is the C-509T polymorphism located in the promoter region of the TGFB1 gene. The −509 T allele was reported to be associated with increasing levels of TGF-β1 compared with the −509 C allele (Grainger et al., 1999). Moreover, the variation could alter TGFB1 transcription activity and the interaction with the transcription factor Yin Yang 1 (Silverman et al., 2004). Two additional polymorphisms, located in codon 10 (Leu→Pro) and codon 25 (Arg→Pro) of the TGFB1 gene were reported to be associated with interindividual variation in levels of TGF-β1 production (Awad et al., 1998; Yamada et al., 1998; Grainger et al., 1999). Given the functional relevance, several epidemiological studies have focused on the association between the TGFB1 polymorphisms and risk of gastric cancer.

In the last decades, evidence of the relationship between TGF-β1 and gastric cancer has been provided: in patients with gastric cancer, elevated expression of TGF-β1 was detected (Lin et al., 2006; Li et al., 2008b), and the high expression was related to the progression and invasion of tumor (Maehara et al., 1999; Saito et al., 1999); in mice model, loss of TGF-β activity led to increased susceptibility to gastrointestinal tumorigenesis (Hahm et al., 2002). However, no significant association between the TGFB1 polymorphisms and risk of gastric cancer was observed in this meta-analysis. This finding was confirmed by two recent genome-wide association studies that did not identify TGFB1 as the susceptibility gene of gastric cancer (Sakamoto et al., 2008; Abnet et al., 2010).

Obvious between-study heterogeneity was detected in this meta-analysis. Logistic meta-regression analysis, therefore, was performed to explore possible reasons for the heterogeneity. After taking HWE, ethnicity, and sample size into consideration, we found that the variable HWE but not ethnicity, and sample size could explain the heterogeneity. To check whether HWD could influence the results of this meta-analysis, we recomputed the summary ORs omitting these studies (Glinsky, 2008; Guan et al., 2009; Guo et al., 2011). No significant association was found, suggesting the results were robust.

Several limitations of this study need to be addressed. First, obvious heterogeneity among studies was observed in overall analyses and also subgroup analyses, which is likely to result in misleading results. Second, the relatively small sample size included in this meta-analysis might not provide sufficient power to assess the relationship between the TGFB1 polymorphisms and gastric cancer risk. Finally, further evaluation was prevented due to lack of available data on the effect of gene–gene, gene–environment interactions on gastric cancer.

In conclusion, this meta-analysis provides evidence that TGFB1 C-509T, T869C, and G915C polymorphisms are not associated with the development of gastric cancer. Well-designed studies with thousands of individuals are warranted to confirm these findings. Further studies investigating the effect of gene–gene and gene–environment interactions on gastric cancer risk would also be valuable.

Footnotes

Disclosure Statement

Niu H and Chen ZL designed the study and wrote the article. Niu Z, Zhang XL, and Chen ZL collected the data, performed the statistical analysis, and interpreted the results. All authors read and approved the final article.

References

Abnet

C.C.

, Freedman

N.D.

, Hu

, Wang

, Yu

, Shu

X.O.

, Yuan

J.M.

, Zheng

, Dawsey

S.M.

, Dong

L.M.

, Lee

M.P.

, Ding

, Qiao

Y.L.

, Gao

Y.T.

, Koh

W.P.

, Xiang

Y.B.

, Tang

Z.Z.

, Fan

J.H.

, Wang

, Wheeler

, Gail

M.H.

, Yeager

, Yuenger

, Hutchinson

, Jacobs

K.B.

, Giffen

C.A.

, Burdett

, Fraumeni

J.F.

Jr. , Tucker

M.A.

, Chow

W.H.

, Goldstein

A.M.

, Chanock

S.J.

, Taylor

P.R.

2010. A shared susceptibility locus in PLCE1 at 10q23 for gastric adenocarcinoma and esophageal squamous cell carcinoma. Nat Genet, 42:764–767.

Awad

M.R.

, El-Gamel

, Hasleton

, Turner

D.M.

, Sinnott

P.J.

, Hutchinson

I.V.

1998. Genotypic variation in the transforming growth factor-beta1 gene: association with transforming growth factor-beta1 production, fibrotic lung disease, and graft fibrosis after lung transplantation. Transplantation, 66:1014–1020.

Barthelemy-Brichant

, David

J.L.

, Bosquee

, Bury

, Seidel

, Albert

, Bartsch

, Baugnet-Mahieu

, Deneufbourg

J.M.

2002. Increased TGFbeta1 plasma level in patients with lung cancer: potential mechanisms. Eur J Clin Invest, 32:193–198.

Blobe

G.C.

, Schiemann

W.P.

, Lodish

H.F.

2000. Role of transforming growth factor beta in human disease. N Engl J Med, 342:1350–1358.

Chen

C.X.

, Yu

Y.H.

, Huang

Q.M.

2010a. The relationship between transforming growth factor beta 1 gene polymorphism and gastric cancer with Helicobacter pylori infection. Clin J Gastroenterol Hepatol, 19:202–204.

Chen

C.X.

, Yu

Y.H.

, Zhang

2010b. Association between transforming growth factor-beta 1 gene polymorphism and gastric cancer. Pract Prev Med, 17:1057–1059.

de Caestecker

M.P.

, Piek

, Roberts

A.B.

2000. Role of transforming growth factor-beta signaling in cancer. J Natl Cancer Inst, 92:1388–1402.

DerSimonian

, Laird

1986. Meta-analysis in clinical trials. Control Clin Trials, 7:177–188.

Derynck

, Akhurst

R.J.

, Balmain

2001. TGF-beta signaling in tumor suppression and cancer progression. Nat Genet, 29:117–129.

10.

Ebert

M.P.

, Yu

, Miehlke

, Fei

, Lendeckel

, Ridwelski

, Stolte

, Bayerdorffer

, Malfertheiner

2000. Expression of transforming growth factor beta-1 in gastric cancer and in the gastric mucosa of first-degree relatives of patients with gastric cancer. Br J Cancer, 82:1795–1800.

11.

Egger

, Davey Smith

, Schneider

, Minder

1997. Bias in meta-analysis detected by a simple, graphical test. BMJ, 315:629–634.

12.

El-Omar

E.M.

, Chow

W.H.

, Rabkin

C.S.

2001. Gastric cancer and H. pylori: host genetics open the way. Gastroenterology, 121:1002–1004.

13.

Elliott

R.L.

, Blobe

G.C.

2005. Role of transforming growth factor Beta in human cancer. J Clin Oncol, 23:2078–2093.

14.

Garcia-Gonzalez

M.A.

, Strunk

, Piazuelo

, Benito

, Santolaria

, Jimenez

, Sopena

, Pascual

, Simon

M.A.

, Sainz

, Lanas

2006. TGFB1 gene polymorphisms: their relevance in the susceptibility to Helicobacter pylori-related diseases. Genes Immun, 7:640–646.

15.

Glinsky

G.V.

2008. SNP-guided microRNA maps (MirMaps) of 16 common human disorders identify a clinically accessible therapy reversing transcriptional aberrations of nuclear import and inflammasome pathways. Cell Cycle, 7:3564–3576.

16.

Gorsch

S.M.

, Memoli

V.A.

, Stukel

T.A.

, Gold

L.I.

, Arrick

B.A.

1992. Immunohistochemical staining for transforming growth factor beta 1 associates with disease progression in human breast cancer. Cancer Res, 52:6949–6952.

17.

Grainger

D.J.

, Heathcote

, Chiano

, Snieder

, Kemp

P.R.

, Metcalfe

J.C.

, Carter

N.D.

, Spector

T.D.

1999. Genetic control of the circulating concentration of transforming growth factor type beta1. Hum Mol Genet, 8:93–97.

18.

Guan

, Zhao

, Niu

, Tang

, Ajani

J.A.

, Wei

2009. The VEGF −634G>C promoter polymorphism is associated with risk of gastric cancer. BMC Gastroenterol, 9:77.

19.

Guo

, Dong

, Guo

, Chen

, Yang

, Kuang

, Shan

2011. Polymorphisms of transforming growth factor-beta1 associated with increased risk of gastric cardia adenocarcinoma in north China. Int J Immunogenet, 38:215–224.

20.

Guo

, Shan

B.E.

, Zhang

, Liu

, Dong

Y.R.

, Guo

Y.L.

, Yang

Z.B.

, Dong

Z.M.

2010. Correlation of polymorphisms in the promoter region of transforming growth factor (beta)1 with the risk of developing gastric cardia adenocarcinoma. Tumor, 30:596–602.

21.

Hahm

K.B.

, Lee

K.M.

, Kim

Y.B.

, Hong

W.S.

, Lee

W.H.

, Han

S.U.

, Kim

M.W.

, Ahn

B.O.

, Oh

T.Y.

, Lee

M.H.

, Green

, Kim

S.J.

2002. Conditional loss of TGF-beta signalling leads to increased susceptibility to gastrointestinal carcinogenesis in mice. Aliment Pharmacol Ther, 16,Suppl 2:115–127.

22.

Higgins

J.P.

, Thompson

S.G.

2002. Quantifying heterogeneity in a meta-analysis. Stat Med, 21:1539–1558.

23.

Ivanovic

, Melman

, Davis-Joseph

, Valcic

, Geliebter

1995. Elevated plasma levels of TGF-beta 1 in patients with invasive prostate cancer. Nat Med, 1:282–284.

24.

Jemal

, Bray

, Center

M.M.

, Ferlay

, Ward

, Forman

2011. Global cancer statistics. CA Cancer J Clin, 61:69–90.

25.

Jin

, Wang

, Chen

, Hu

, Zhou

, Tan

, Wang

, Hua

, Ding

, Shen

, Zhang

, Wang

, Xu

, Shen

2007. Variant alleles of TGFB1 and TGFBR2 are associated with a decreased risk of gastric cancer in a Chinese population. Int J Cancer, 120:1330–1335.

26.

Langenskiold

, Holmdahl

, Falk

, Angenete

, Ivarsson

M.L.

2008. Increased TGF-beta 1 protein expression in patients with advanced colorectal cancer. J Surg Oncol, 97:409–415.

27.

, Cao

B.W.

, Dai

, Cui

, Yang

H.L.

, Xu

C.Q.

2008a. Correlation of transforming growth factor beta-1 gene polymorphisms C-509T and T869C and the risk of gastric cancer in China. J Gastroenterol Hepatol, 23:638–642.

28.

, Yue

Z.C.

, Zhang

Y.Y.

, Bai

, Meng

X.N.

, Geng

J.S.

, Fu

S.B.

2008b. Elevated serum level and gene polymorphisms of TGF-beta1 in gastric cancer. J Clin Lab Anal, 22:164–171.

29.

Lin

, Kikuchi

, Obata

, Yagyu

2006. Serum levels of transforming growth factor beta1 are significantly correlated with venous invasion in patients with gastric cancer. J Gastroenterol Hepatol, 21:432–437.

30.

Maehara

, Kakeji

, Kabashima

, Emi

, Watanabe

, Akazawa

, Baba

, Kohnoe

, Sugimachi

1999. Role of transforming growth factor-beta 1 in invasion and metastasis in gastric carcinoma. J Clin Oncol, 17:607–614.

31.

Mantel

, Haenszel

1959. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst, 22:719–748.

32.

Massague

, Blain

S.W.

, Lo

R.S.

2000. TGFbeta signaling in growth control, cancer, and heritable disorders. Cell, 103:295–309.

33.

Mishra

, Derynck

, Mishra

2005. Transforming growth factor-beta signaling in stem cells and cancer. Science, 310:68–71.

34.

Nomura

, Grove

J.S.

, Stemmermann

G.N.

, Severson

R.K.

1990. Cigarette smoking and stomach cancer. Cancer Res, 50:7084.

35.

Palli

2000. Epidemiology of gastric cancer: an evaluation of available evidence. J Gastroenterol, 35,Suppl 12:84–89.

36.

Parkin

D.M.

2006. The global health burden of infection-associated cancers in the year 2002. Int J Cancer, 118:3030–3044.

37.

Parkin

D.M.

, Bray

, Ferlay

, Pisani

2005. Global cancer statistics, 2002. CA Cancer J Clin, 55:74–108.

38.

Saito

, Tsujitani

, Oka

, Kondo

, Ikeguchi

, Maeta

, Kaibara

1999. The expression of transforming growth factor-beta1 is significantly correlated with the expression of vascular endothelial growth factor and poor prognosis of patients with advanced gastric carcinoma. Cancer, 86:1455–1462.

39.

Sakamoto

, Yoshimura

, Saeki

, Katai

, Shimoda

, Matsuno

, Saito

, Sugimura

, Tanioka

, Kato

, Matsukura

, Matsuda

, Nakamura

, Hyodo

, Nishina

, Yasui

, Hirose

, Hayashi

, Toshiro

, Ohnami

, Sekine

, Sato

, Totsuka

, Ando

, Takemura

, Takahashi

, Ohdaira

, Aoki

, Honmyo

, Chiku

, Aoyagi

, Sasaki

, Yanagihara

, Yoon

K.A.

, Kook

M.C.

, Lee

Y.S.

, Park

S.R.

, Kim

C.G.

, Choi

I.J.

, Yoshida

, Nakamura

, Hirohashi

2008. Genetic variation in PSCA is associated with susceptibility to diffuse-type gastric cancer. Nat Genet, 40:730–740.

40.

Shim

K.S.

, Kim

K.H.

, Han

W.S.

, Park

E.B.

1999. Elevated serum levels of transforming growth factor-beta1 in patients with colorectal carcinoma: its association with tumor progression and its significant decrease after curative surgical resection. Cancer, 85:554–561.

41.

Shimazu

, Tsuji

, Inoue

, Wakai

, Nagata

, Mizoue

, Tanaka

, Tsugane

2008. Alcohol drinking and gastric cancer risk: an evaluation based on a systematic review of epidemiologic evidence among the Japanese population. Jpn J Clin Oncol, 38:8–25.

42.

Silverman

E.S.

, Palmer

L.J.

, Subramaniam

, Hallock

, Mathew

, Vallone

, Faffe

D.S.

, Shikanai

, Raby

B.A.

, Weiss

S.T.

, Shore

S.A.

2004. Transforming growth factor-beta1 promoter polymorphism C-509T is associated with asthma. Am J Respir Crit Care Med, 169:214–219.

43.

Song

B.C.

, Chung

Y.H.

, Kim

J.A.

, Choi

W.B.

, Suh

D.D.

, Pyo

S.I.

, Shin

J.W.

, Lee

H.C.

, Lee

Y.S.

, Suh

D.J.

2002. Transforming growth factor-beta1 as a useful serologic marker of small hepatocellular carcinoma. Cancer, 94:175–180.

44.

Sung

N.Y.

, Choi

K.S.

, Park

E.C.

, Park

, Lee

S.Y.

, Lee

A.K.

, Choi

I.J.

, Jung

K.W.

, Won

Y.J.

, Shin

H.R.

2007. Smoking, alcohol and gastric cancer risk in Korean men: the National Health Insurance Corporation Study. Br J Cancer, 97:700–704.

45.

Tredaniel

, Boffetta

, Buiatti

, Saracci

, Hirsch

1997. Tobacco smoking and gastric cancer: review and meta-analysis. Int J Cancer, 72:565–573.

46.

Yamada

, Miyauchi

, Goto

, Takagi

, Okuizumi

, Kanematsu

, Hase

, Takai

, Harada

, Ikeda

1998. Association of a polymorphism of the transforming growth factor-beta1 gene with genetic susceptibility to osteoporosis in postmenopausal Japanese women. J Bone Miner Res, 13:1569–1576.

47.

Yoshie

, Kobayashi

, Tai

, Galicia

J.C.

2007. The role of genetic polymorphisms in periodontitis. Periodontol 2000, 43:102–132.

48.

Zhang

, Di

J.Z.

, Zhu

Z.Z.

, Wu

H.M.

, Wang

, Zhu

, Zheng

, Hou

2008. Association of transforming growth factor-beta 1 polymorphisms with genetic susceptibility to TNM stage I or II gastric cancer. Jpn J Clin Oncol, 38:861–866.

49.

Zhou

, Jin

G.F.

, Jiang

G.J.

, Wang

H.M.

, Tan

Y.F.

, Ding

W.L.

, Wang

L.N.

, Chen

W.S.

, Ke

, Shen

, Xu

Y.C.

, Shen

H.B.

2007. Correlations of polymorphisms of TGFB1 and TGFBR2 genes to genetic susceptibility to gastric cancer. Ai Zheng, 26:581–585.