Few studies have explored the association between stem cell pathways and esophageal cancer and potential interactions between genetic markers and environmental exposures. We hypothesize that genetic susceptibility markers in the stem cell pathway might be associated with esophageal cancer and there might be interaction between these markers and environmental exposures. In a population-based case-control study conducted in Jiangsu, China, we evaluate the associations between esophageal cancer and various environmental factors and genetic markers using logistic regression model. The focus was on independent associations of cancer stem cell-related genes and their potential modification of the associations of known risk factors. In our analysis, we found that environmental tobacco smoking was associated with non-smokers with an adjusted OR of 1.29 (95% CI: 1.04-1.59) with a corresponding PAR of 11.4% (0.90%-21.7%).Garlic consumption was found to be inversely associated with esophageal cancer risk with an adjusted OR of 0.67 (95% CI: 0.53-0.85) for those in the entire study that self-reported often consumption of raw garlic. This inverse association was seen across stratum of drinking status and smoking status with an adjusted OR of 0.67 (95% CI: 0.48-0.96) among the group of frequent drinkers who were ever smokers.

We analyzed 5 SNPs of the Wnt pathway and its association with esophageal cancer. We found rs4730775 (Wnt2) to have a potential weak inverse association of 0.89 (95% CI: 0.75-1.07 T/T+C/T vs C/C) in the study population overall. The association for Wnt2 (rs4730775) varied by smoking status; never smokers had a negative association with esophageal cancer with an aOR of 0.72 (95% CI: 0.53-0.98; T/T+C/T vs. C/C). When stratified by alcohol status, never and infrequent drinkers had inverse of associations with Wnt2 rs4730775 (T/T+C/T vs C/C aOR: 0.79 95% CI: 0.62-1.00) while there was no association for frequent and daily drinkers (T/T+C/T vs C/C aOR: 1.03 95% CI: 0.79-1.34). Likewise, the inverse association of Wnt2 was seen in the non-smoking/non-drinking category for rs4730775 (aOR: 0.67, 95% CI: 0.46-0.98 T/T+C/T vs C/C, respectively). No association was observed in the drinkers/smokers group for rs4730775 (Wnt 2).

Similarly, rs4835761 (Wnt8A) had an inverse association with an aOR of 0.82 (95% CI: 0.64-1.05, G/G+A/G vs A/A) in never and infrequent drinkers while there was no association observed in frequent and daily drinkers (aOR 0.97 95% CI: 0.74-1.28, G/G+A/G vs A/A. For rs222851 (DVL2) we observed a potential weak association for never and infrequent drinkers with an aOR of 1.19 (95% CI: 0.93-1.52, G/G+A/G vs A/A) and no association in frequent and daily drinkers aOR of 0.88 (95% CI: 0.74-1.06, G/G+A/G vs A/A). For non-drinkers and non-smokers, there was an inverse association observed with the rs2241802 (FZD3) with an adjusted OR of 0.68 (95% CI: 0.46-1.00; G/A+A/A vs. G/G) while there was no association observed in the drinkers/smokers group (aOR=1.00, 95% CI: 0.82-1.49; G/A+A/A vs. G/G).

We did not observe an association of rs2953 miRNA589 (CTNNB1 binding site) with esophageal cancer with an adjusted OR of 0.95 (95% CI: 0.80-1.13; G/T+G/G vs. T/T). When stratified by smoking status, we observed a potential inverse association among smokers with an adjusted OR of 0.84 (95% CI: 0.69-1.03; G/T+G/G vs. T/T) while there was no association observed among never smokers. Joint effects with rs2953 miRNA589 and other Wnt pathway SNPs did not appear to have an association when including the entire study population. However, when performing a stratified analysis of smokers and non-smokers the aROR for rs2953 miRNA589 and rs3729629 Wnt2 demonstrated an inverse association in smokers (aROR=0.58, 95% CI=0.35-0.96; G/G+A/G vs A/A for rs2953 and C/C+G/C vsG/G for rs3729629). This association remained when we stratified on both drinking and smoking status. For ever smokers and frequent/daily drinkers, the aROR was 0.48 (95% CI: 0.26-0.885 G/G+A/G vs A/A for rs2953 and C/C+G/C vsG/G for rs3729629).

In summary, we have identified the protective association between garlic intake and esophageal cancer and observed moderate association between environmental tobacco smoking (ETS) and esophageal cancer. The observed associations between genetic markers and esophageal cancer are relatively weak and there are some indications of potential gene-environmental interaction. This is the first time that ETS is associated with esophageal cancer in a large population-based case-control study in a Chinese population and garlic intake is inversely associated with esophageal cancer. Our observations suggest that in addition to tobacco cessation and alcohol control, the prevention strategy for esophageal cancer should include avoidance of ETS and increasing intake of raw garlic.