[Effects of PNPLA3, TM6SF2 gene polymorphisms and its interactions with smoking and alcohol drinking on hepatitis B virus-associated hepatocellular carcinoma].

Objective: To explore the SNP effects of patatin-like phospholipase domain which containing 3 (PNPLA3), transmembrane 6 superfamily member 2 (TM6SF2) gene, environmental effects of smoking, alcohol drinking and interaction between gene-gene, gene-environment and drinking-smoking on hepatitis B virus-associated hepatocellular carcinoma (HBV-HCC). Methods: We collected anticoagulant peripheral blood from patients of HBV-HCC, chronic hepatitis B (CHB), liver cirrhosis (LC) and from healthy controls to detect the single nucleotide polymorphism (SNP) of patatin-like phospholipase domain containing 3 (PNPLA3) gene loci rs738409 and transmembrane 6 superfamily member 2 (TM6SF2) gene loci rs58542926, using the flight mass spectrometry method. The optimal assignment value of gene polymorphisms was defined by using the online SNP stats. Hardy-Weinberg (H-W) balance was tested for SNP. Effects of the genetic and environmental factors to HBV-HCC were analyzed by using the multiple classification logistic regression method. The gene-gene, gene-smoking and alcohol drinking interaction effects were investigated by Fork-Life analysis and binary logistic regression methods. Results: The frequency distribution of CHB group rs738409 loci seemed not in conformity with the H-W balance (χ(2)=11.980, P<0.005). Two loci frequency distributions in the other groups were all in accordandce with the H-W balance. After adjusting for influences on age and sex and comparing to the healthy group, the rs58542926 mutation appeared as OR=1.659, 95%CI: 1.026-2.684, P=0.039, in the HBV-HCC group. When comparing to CHB group, the HBV-HCC group presented that drinking as OR=1.680, 95%CI: 1.121-2.519, P=0.012. When comparing to the LC group, the ORs of drinking and smoking were 1.539 (1.071-2.213) and 1.453 (1.005-2.099) respectively, in the HBV-HCC group. When comparing to the CHB+LC group, interactions between the HBV-HCC group were found rs738409 and rs58542926 on additive model OR=1.548 (U=1.885, P=0.029) and OR=1.658 (P=0.024) on logistic regression model while drinking was rs738409 on interaction additive model with OR=1.811(U=1.965, P=0.024). As for drinking and mutation of rs738409, the multiplication model of logistic regression showed no statistically significant differences. Interaction between smoking and drinking appeared as OR=1.756 (P<0.001) in the logistics regression multiplication model. Conclusions: Factors as mutation of TM6SF2, smoking and drinking all appeared as risk factors for HBV-HCC. Mutations of both PNPLA3 and TM6SF2, together with smoking and drinking all served as risk factors for HBV-HCC. However, the mutation of single PNPLA3 appeared as a protective factor on HBV-HCC.