Epstein Barr virus antibody reactivity and gastric cancer: A population-based case-control study
Nuria Aragonés,Nerea Fernández de Larrea,Roberto Pastor-Barriuso,Angelika Michel,Beatriz Romero,Michael Pawlita,Sara Mayorgas-Torralba,Vicente Martín,Victor Moreno,Delphine Casabonne,Jesús Castilla,Guillermo Fernandez-Tardón,Trinidad Dierssen-Sotos,Rocío Capelo,Dolores Salas,Inmaculada Salcedo-Bellido,María Dolores Chirlaque,Nicole Brenner,Manuela Pedraza,Xavier Bessa,Beatriz Pérez-Gómez,Julia Butt,Manolis Kogevinas,Rosa Del Campo,Silvia de Sanjosé,Tim Waterboer,Marina Pollán
DOI: https://doi.org/10.1016/j.canep.2019.05.008
Abstract:Background: In contrast to the recognized role of Helicobacter pylori in the etiology of non-cardia gastric cancer (GC), there is still insufficient epidemiological evidence for the involvement of Epstein-Barr virus (EBV) in gastric carcinogenesis. We aimed to evaluate the relation of antibody profile and antibody reactivity intensity against four individual EBV proteins to GC risk. Methods: We used information from 281 GC cases and 2071 age and sex frequency matched controls recruited in the frame of the MCC-Spain multicase-control study, between 2008 and 2013. Sociodemographic, lifestyle and environmental factors were assessed in face-to-face interviews. Antibody responses to four EBV proteins (EBNA-1, ZEBRA, EA-D, and VCA-p18) were analyzed by multiplex serology. Odds ratios (OR) and 95% confidence intervals were calculated by using logistic regression mixed models to evaluate the association of seropositivity and antibody reactivity against EBV proteins with GC, adjusting for GC risk factors. Stratified analyses by tumor location (cardia vs. non-cardia) and morphology (intestinal vs. diffuse) were done. Results: Among controls, seropositivity for EA-D, ZEBRA, EBNA-1 and VCA-p18 was 85%, 91%, 97% and 99%, respectively. Even though seropositivity for none of the studied proteins was associated with a higher GC risk, increasing antibody reactivity against EBNA-1 and VCA-p18 was associated with higher OR of GC. This association was present for cardia and non-cardia cancer cases, and for intestinal and diffuse types. Conclusion: Our results support the hypothesis that EBV may play a role in GC etiology, and highlight the importance of evaluating specific antibodies and the dose-response relations when studying widespread infections.