Objective: To examine and quantify the potential dose–response relationship between green tea intake and the risk of gastric cancer.Design: We searched PubMed, EMBASE, Web of Science, CBM, CNKI and VIP up to December 2015 without language restrictions.Setting: A systematic review and dose–response meta-analysis of observational studies.Subjects: Five cohort studies and eight case–control studies.Results: Compared with the lowest level of green tea intake, the pooled relative risk (95 % CI) of gastric cancer was 1·05 (0·90, 1·21, I²=20·3 %) for the cohort studies and the pooled OR (95 % CI) was 0·84 (0·74, 0·95, I²=48·3 %) for the case–control studies. The pooled relative risk of gastric cancer was 0·79 (0·63, 0·97, I²=63·8 %) for intake of 6 cups green tea/d, 0·59 (0·42, 0·82, I²=1·0 %) for 25 years of green tea intake and 7·60 (1·67, 34·60, I²=86·5 %) for drinking very hot green tea.Conclusions: Drinking green tea has a certain preventive effect on reducing the risk of gastric cancer, particularly for long-term and high-dose consumption. Drinking too high-temperature green tea may increase the risk of gastric cancer, but it is still unclear whether high-temperature green tea is a risk factor for gastric cancer. Further studies should be performed to obtain more detailed results, including other gastric cancer risk factors such as smoking and alcohol consumption and the dose of the effective components in green tea, to provide more reliable evidence-based medical references for the relationship between green tea and gastric cancer.

Background: A large number of studies have been conducted to investigate associations between genetic variants and esophageal cancer risk in the past several decades. However, findings from these studies have been generally inconsistent. We aimed to provide a summary of the current understanding of the genetic architecture of esophageal cancer susceptibility. Methods: We performed a comprehensive field synopsis and meta-analysis to evaluate associations between 95 variants in 70 genes or loci and esophageal cancer risk using data from 304 eligible publications, including 104,904 cases and 159,797 controls, through screening a total of 21,328 citations. We graded levels of cumulative epidemiologic evidence of a significant association with esophageal cancer using the Venice criteria and false-positive report probability tests. We constructed functional annotations for these variants using data from the Encyclopedia of DNA Elements Project and other databases. Results: Thirty variants were nominally significantly associated with esophageal cancer risk. Cumulative epidemiologic evidence of a significant association with overall esophageal cancer, esophageal squamous cell carcinoma, or esophageal adenocarcinoma was strong for 13 variants in or near 13 genes (ADH1B, BARX1, CDKN1A, CHEK2, CLPTM1L, CRTC1, CYP1A1, EGF, LTA, MIR34BC, PLCE1, PTEN, and PTGS2). Bioinformatics analysis suggested that these variants and others correlated with them might fall in putative functional regions. Conclusions: Our study summarizes the current literature on the genetic architecture of esophageal cancer susceptibility and identifies several potential polymorphisms that could be involved in esophageal cancer susceptibility. Impact: These findings provide direction for future studies to identify new genetic factors for esophageal cancer.

This study aimed to construct and validate an immunoscore nomogram that may be used to predict the prognosis of oesophageal cancer. With the gene expression data of oesophageal cancer in a public database, we used CIBERSORT to estimate the fractions of 22 infiltrating immune cell types. We then built an immunoscore signature based on 12 types of infiltrating immune cells using the least absolute shrinkage and selection operator (LASSO) model. This immunoscore was used as an independent predictor in the prognostic model (training cohort: [hazard ratio (HR), 4.78; 95% confidence interval (CI), 2.64-8.67; P < 0.001], validation cohort: [HR, 2.15; 95% CI, 1.04-4.45; P = 0.040]). Subgroup analysis by clinical features showed that overall survival was significantly different between the high-immunoscore group and the low-immunoscore group. The predictors that constituted the individualized prediction nomogram were immunoscore, age, and tumour stage. The nomogram had good discrimination and calibration. Decision curve analysis showed that the immunoscore nomogram was clinically useful. Therefore, the novel immunoscore signature based on infiltrating immune cells can be used as a reliable predictor of the prognosis of oesophageal cancer, and the immunoscore nomogram is a convenient tool for predicting the survival of individual patients.