Arsenic is a class-I human carcinogen, and exposure through drinking water is a serious public health in many countries, increasing risk for a wide array of diseases, including cancers of the bladder, kidney, lung, and skin. Observational studies suggest that individuals who metabolize arsenic quickly are at lower risk for toxicities such as arsenical skin lesions. Because single nucleotide polymorphisms (SNPs) in the arsenite methyltransferase (AS3MT) gene region are known to effect arsenic metabolism efficiency (AME), Mendelian randomization can be used to determine if the association between AME and arsenic toxicity is causal, or due to unmeasured confounding factors (e.g., nutritional status). Using data on 2,060 randomly-selected participants from an arsenic-exposed Bangladeshi cohort, we estimated associations for two AS3MT SNPs (rs9527 and rs11191527) with relative concentrations of three arsenic metabolites measured in urine that represent AME: inorganic arsenic (iAs), monomethylarsonic acid (MMA), and dimethylarsinic acid (DMA). Using data on 2,483 skin lesion cases and 2,857 controls, we generated SNP-based predictions of “genetically-determined” arsenic metabolite levels and tested these predicted values for association with case-control status. Our results confirmed the causal role of AME in arsenic toxicity, with casual odds ratios of 0.90 (95% CI: 0.86-0.94), 1.20 (CI: 1.12-1.29), and 1.25 (CI: 1.15-1.37) for one standard deviation increases in DMA%, MMA%, and iAs%, respectively. These directions of these associations are consistent with those observed in prior observational studies. Using data on a subset of cases and controls with prospective arsenic exposure measures, we further demonstrate the causal role of AME in arsenic toxicity by providing evidence for interaction between water arsenic and AS3MT genotypes in relation to skin lesion risk (P=0.02). In summary, we have used Mendelian randomization and gene-environment interaction analysis to confirm the causal role of AME in arsenic toxicity. With this relationship firmly established, future studies can utilize AS3MT SNPs to assess the causal role of arsenic exposure in relation to a wide array of health conditions with hypothesized connections to arsenic, even when prospective exposure data is not available.

This proffered talk is also presented as Poster 16.

Citation Format: Brandon L. Pierce, Lin Tong, Maria Argos, Farzana Jasmine, Muhammad Kibriya, Habibul Ahsan. Genetically determined differences in arsenic metabolism efficiency influence risk for premalignant skin lesions in Bangladesh: Mendelian randomization and gene-environment interaction. [abstract]. In: Proceedings of the AACR Special Conference on Post-GWAS Horizons in Molecular Epidemiology: Digging Deeper into the Environment; 2012 Nov 11-14; Hollywood, FL. Philadelphia (PA): AACR; Cancer Epidemiol Biomarkers Prev 2012;21(11 Suppl):Abstract nr PR6.