| 1. BackgroundThe predominant histological type for esophageal cancer (EC) in China is esophageal aqueous cell carcinoma (ESCC,> 95% of total EC). In contrast, esophageal adenocarcinoma (EAC) is the most common type of EC (ECA,>80% of total EC) in western countries. The mechanism for this ethnic difference in terms of histological pattern for EC is not clear. The pathogenesis of EAC has been thought to be related with obesity reflux esophagitis and Barrett's esophagus (BE), BE is a metaplastic chang in the gastoesophageal linging from the usual squamous mucosa to columnar epithelium. Recently, the incidence of ECA has been increasing dramatically in western countries. Reflux esophagitis and BE have been considered as preasor lesions for ECA, compared to the general population, the patients with BE may have 30-50 fold-increased risk of developing EAC. The prevalence of reflux esophagitis and BE in Chinese population is much lower than in westren counties. Historically, reflux esophagitis and BE have been seen as uncommon disease in Asia, especially in China, which is consistent with the observation of very low incidence of EAC in chinese population. However, recent studies with endoscopic examiination and symptom-based mass survey have shown that the prevalance of reflux esophagitis is increasing in Asia, including in china. In case studies in China, the prevalence of reflux esophagitis rangs from 3% to 16%. The reasons for this prevalence increasing are largely unknown. The dramatic socio-economic changes in China in terms of diet and eating habits and other "modern" lifestyle changes may contribute to the increasing prevalance of reflux esophagitis. It is noteworting that reflux esophagitis and related BE are only present in a subset of patients in the high incidence area for esophageal cancer in China, indicating that an altered susceptibility to reflux esophagitis and BE may exist in these individedes. Recent genome-wide association (GWAS) on EC by us, incidate that the interacting by environment and genetic factors may play an important role in esophageal carcinogenesis, and 18 single nucleiotide polymorplisms (SNP) may dramatically increase the risk for EC.Thus, the present study was undertaken in a large population-based, case-control study in Chinese population and to determine the prevalence of BE, to validate the 18 susceptibility SNP loci and related genes identified from our GWAS study in these patients with BE. Further analyses were performed to correlate these SNPs with phenotype, such as body mass index (BMI).2 Materials and Methods2.1 Prevalence of Barrett's esophagus2.1.1 Study populationBased on case-control design,199 cases with Barrett's esophagus (123 male with a mean age of 48±12 and 76 female with a mean age of 50±8.7) and 761 controls (456 male a mean age of 46±10and 305 female a mean age 48±9.8) were enrolled in this study. The data for cases were collected from 30850 cases of 8 province-city hospitals in China2.1.2 Questionnaire and StatisticsThe questionnaire was performed on all the cases and controls by either interview at hospital for clinical information collection, including age, gender, symptoms of reflux, body weight and weight. All the data were inputed into computer using the EXCEL software. The data were analyzed by SPSS 17.0 software. The chi-square test was also applied for comparisons among different groups (a=0.05 was considered as the test standard).2.2 Validate of 18 SNPs in Barrett's esophagus and Control2.2.1 Study Population, Blood sample collection and DNA extraction75 cases with BE and 260 controls from Chinese population were enrolled in the 18 SNPs validation. Peripheral venous blood samples (3-5ml) were collected from each patient and control. Genomic DNA was isolated from hymphocytes of peripheral blood using Flexi Gene DNA (QIAGEN, Germany) kits.2.2.2 SNPs AnalysisAll the data are derived from the Illumina 610-Quad BeadChips and the Sequenom MassArray System of was used to analyze the 18 SNPs as identified from our ESCC GWAS.2.3 Statistical AnalysisSPSS software (version 17.0) was used for statistical analysis. Two-sided P values of less than 0.01 were considered to indicate statistical significance, Odds ratio (OR) and 95% confidence interval (95%CI) were determined to evaluate relative risk, Stratified analysis by age, gender and family history was performed. The genotype and allele frequencies were further compared between each group.3 Results3.1 Analysis of clinical phenotype for BEThe rate of BE detection was 0.65%,199 BE with a mean age of 49.3±12.7 (in male 48±12 and in female 50±8.7). The occurrence of BE in males was more frequently than in females (62%vs38%). Interesting, the detection rate of BE in high risk area for EC was much lower than in low risk area (77%vs33%). The occurrence of BE in no family history was more frequently than in family history (90%vsl 0%).Furthermore, the patients with reflux esophagitis had a higher risk for development of BE (X2=9.433 P=0.025 OR= 1.524 95%CI=1.051-2.208). Similarly the patients with hight BMI (overweight) had a higher risk for BE (X2=4.991 P=0.025 OR=1.524 95%CI=1.051-2.208).3.2 Genetic susceptibility and BEWe selected 18 promising single nucleotide polymorphism (SNPs) from GWAS scanning, after validateion, three SNPs show independent association evidence with BE:rs8102476 on 19q13.1 (genotype frequency X2＝25 df=2 allelic frequency P=3.73×10-6 X2=11.158 df=1 P=0.001 OR=0.550 95%CI= 0.386-0.783),rs6098279 on 20q13.2 (genotype frequency X2=5.239 df=1P=0.022 OR=0.45 95%CI= 1.056-3.916 allelic frequency X2=4.894 df=1 P=0.027 OR=0.48 95%CI= 0.247-0.931)and rs6023640 on 20q13.2 (genotype frequency X2=5.239 df=1 P=0.022 OR=0.45 95% CI=1.056-3.916 allelic frequency X2=4.258 df=1 P=0.039 OR=1.901 95% CI=1.024-3.530). rs5753220 on 22q12.1 (genotype frequency X2=4.746 df=2 P=0.048 allelic frequency X2=3.901 df=1 P=0.048 OR=0.693 95% CI=0.481-0.998)After investigating the patterns of the recombination and linkage disequilibrium (LD) around the risk-associated SNPs and the genes, one gene at each locus was implicated:PKC-potentiated protein phosphatase inhibitory protein-17 (PPP1R14A) at 19q13.1, Downstream of kinases 5 (DOK5) at 20q13.2 and Pescadillo(PESl)at 22q12.1.PPP1R14A gene at rs8102476 within was further analyzed in the different groups of BMI to correlate this SNP with BE, in the normal weight of BMI, the chi-square test showed that distribution of alleles had no significant differences with this SNP between BE and controls (X2=2.69 df=1 P=0.101 OR=1.604 95%CI=0.909-2.831). In the overweight of BMI, the chi-square test showed that distribution of alleles had significant differences with this SNP between BE and controls (X2=4.538 df=1 P=0.033 OR=0.37795%CI=0.150-0.949).DOK5 at rs6098279 within gene was further analyzed in the different groups of BMI to correlate this SNP with BE. In the normal weight of BMI, the chi-square test showed that distribution of alleles had significant differences with this SNP between BE and controls (X2=18.52 df=1 P=1.68E-05 OR=4.829 95%CI=2.225-10.482). In the overweight of BMI, the chi-square test showed that distribution of alleles had significant differences with this SNP between BE and controls (X2=12.166 df=1 P=4.87E-04 OR=5.747 95% CI=1.958-16.872). PES1 gene at rs5753220 within was further analyzed in the different groups of BMI to correlate this SNP with BE. In the normal weight of BMI, the chi-square test showed that distribution of alleles had significant differences with this SNP between BE and controls (χ2=3.134 df=1 P=0.077 OR=1.893 95%CI=0.928-3.863). In the overweight of BMI, the chi-square test showed that distribution of alleles had significant differences with this SNP between BE and controls (χ2=0.062 df=1 P=0.803 OR=0.923 95%CI=0.491-1.734).Rs6023640 within DOK5 gene was further analyzed in the different groups of BMI to correlated this SNP with BE, in the normal weight of BMI, the chi-square test showed that distribution of alleles had significant differences with this SNP between BE and controls (χ2=6.864 df=1 P=0.009, OR=2.671 95%CI=1.248-5.718). In the overweight of BMI, the chi-square test showed that distribution of alleles had no significant differences with this SNP between BE and controls (χ2=2.486 df=1 P=0.092 OR=2.04 95% CI=0.878-4.737).4. Conclusions4.1 The prevalence of BE in Chinese population is very low (0.6%), with a predominance in male; and overall, the detection rate of BE is higher in low risk area than in high risk area for EC.4.2 Overweight apparently increase the risk for development of BE.4.3 The PPP1R14A at rs8102476 genotype CT and TT shows a higher risk of BE than genotype CC in Chinese population, the alleles C of PPP1R14A increases the risk of BE(χ2=11.158 df=1 P=0.001 OR=0.550 95%CI= 0.386-0.783).4.4 The DOK5 at rs6098279 genotype CT shows a higher risk of BE in Chinese population, the alleles C of DOK5 increases the risk of BE (χ2=4.894 df=1 P=0.027 OR=0.48 95%CI= 0.247-0.931).4.5 The DOK5 at rs6023640 genotype GA on shows a higher risk of Barrett's esophagus in Chinese population, the alleles A of DOK5 increases the risk of BE(χ2=4.258 df=1 P=0.039 OR=1.901 95% CI=1.024-3.530)4.6 The PES1 at rs5753220 genotype CT on shows a higher risk of Barrett's esophagus in Chinese population, the alleles T of PES1 increases the risk of BE (X2=3.901 df=1 P=0.048 OR=0.693 95% CI=0.481-0.998 X2=3.924 df=1 P=0.048 OR=0.52895% CI=0.279-0.999)4.7 BE is correlation with rs8102476 in overweight, rs6098279 is correlation with BE in overweight and normal weight, rs6023640 is correlation with BE.4.8 BE is correlation with rs5753220 in overweight. |
PDF Full Text Request