Suh M, Harvey S, Wikoff D, Mittal L, Ring C, Goodmanson A, Proctor D. 2017. Meta-analysis of hexavalent chromium and stomach cancer. Poster presented at Society of Toxicology Annual Meeting, March 13, Baltimore, MD.
Abstract
Hexavalent chromium [Cr(VI)] is a known human lung carcinogen by inhalation, but no consistent association has been observed for non-respiratory cancers among workers exposed to Cr(VI). Three published meta-analyses investigated possible association between Cr(VI) exposure and stomach cancer. Two studies reported no significant association between Cr(VI) exposure and stomach cancer; one observed statistically elevated risk. However, the inclusion criteria were not consistent across studies, and some were criticized for including studies of workers with limited or no Cr(VI) exposure, excluding eligible studies, or including environmentally-exposed populations. To evaluate the potential for increased cancer risk among Cr(VI)-exposed workers and the basis for varying conclusions, we assessed the Cr(VI) epidemiology literature for stomach cancer risk measures and previously published meta-analyses, updating with newly published studies and including subgroup analyses to assess the effect of differing inclusion criteria. Sensitivity analyses were conducted to explore the robustness of our findings; heterogeneity of study populations and publication bias were also evaluated. We identified 225 studies for potential inclusion, and 45 studies were found to be eligible. When registry studies and proportionate mortality ratio studies were removed, heterogeneity was substantially minimized (I2 = 18.9%) and stomach cancer risk was not significantly increased (meta-RR = 1.08; 95% CI 0.96 to 1.21, 39 studies). Subgroup analyses based on cohort studies (stomach cancer meta-RR = 1.04, 95% CI 0.92 to 1.17, 34 studies, I2 = 20.8%) or studies of workers in industries with increased lung cancer risk (stomach cancer meta-RR = 1.13, 95% CI 0.93 to 1.38, 19 studies, I2 = 24.5%) showed no significant associations. Compared to the previous analyses, these results highlight the importance of inclusion criteria. Additionally, analyses conducted to date have not consistently evaluated measures of study quality and validity, which could impact the overall conclusions especially considering that most studies did not account for known risk factors, such as smoking and asbestos exposure. As such, it was determined that a systematic assessment of the body of literature with updated inclusion/exclusion criteria relative to previous assessments, and considerations for internal and external validity, would allow for refined characterization and quantification of possible association between Cr(VI) exposure and stomach cancer.