Circulating Inflammation Biomarkers and the Risk of Esophageal Adenocarcinoma: A Nested Case-control Study in the Department of Defense Serum Repository

Abstract Background: We previously identified associations of esophageal adenocarcinoma (EA) risk with four inflammation-related candidate biomarkers: TNFR2, IL17A, VEGFR3 and resistin. Methods: We aimed to replicate these candidates and discover novel associations with additional proteins. We conducted a nested case-control study of men with prediagnostic biospecimens stored at the US Department of Defense Serum Repository, including 203 incident EA cases. Controls were matched to cases in an ~2:1 ratio by date of birth, race, service branch, and blood draw date. Multiplex immunoassays (Olink/Proseek panels) measured 254 proteins detected in ≥10% of all samples. Multivariable-adjusted conditional logistic regression models calculated associations between biomarker quantiles and EA. P-values (<0.05) were used to indicate the statistical significance of candidates, and false discovery rate (FDR) was applied to the additional proteins. Odds ratios (ORs) from the current analysis and from previous studies were combined for the candidate markers using fixed effects meta-analysis. Results: Among the four candidates, the highest category of TNFR2 was associated with significantly increased EA risk (ORQ4vsQ1=1.87, 95% confidence interval: 1.02-3.42). In the meta-analysis, associations with EA were positive for TNFR2 (meta-analyzed ORhighest-vs-lowest=2.04, 1.12-2.95) and inverse for IL17A (meta-analyzed ORhighest-vs-lowest=0.53, 0.26-0.80). Of the additional 250 proteins, 45 were associated with EA risk and 6 (MCP3, IL6, TNFR1, HGF, TFF3 and FURIN) remained significant after FDR correction. Conclusions: We confirmed associations of TNFR2 and IL17A with EA risk. Additionally, our study expands the range of proteins associated with EA development. Impact: This is the largest assessment of inflammation-related proteins with EA to date.