Meta-analysis of the association between HLA-G 14-bp insertion/deletion polymorphism and susceptibility to viral infections and cancer risk

Abstract

BACKGROUND/OBJECTIVE: HLA-G 14-bp insertion/deletion (I/D) polymorphism has been linked to both cancer susceptibility and various viral infections. This meta-analysis examined the relationship between HLA-G 14-bp I/D polymorphism with different viral infections on patients with or without complications of cancer. METHODS: We carried out a meta-analysis according to PRISMA guidelines, pooling data from 26 case-control studies on HLA-G 14-bp I/D polymorphism including association with viral diseases and cancer cases. 5906 cases and 6963 healthy controls were included in the meta-analysis. RESULTS: Analysis under the random model showed lack of association between HLA-G 14-bp I/D polymorphism and viral infection either in the presence or absence of cancer in the overall population. Interestingly, subgroup analysis revealed variable outcomes across the subgroups analysed. Heterogeneous results depended mainly on virus characteristics (type, genome and family) and ethnicity. The HLA-G 14-bp I/D polymorphism was associated with Hepadnaviridae (DNA genome) infection in patients without cancer under the allelic model (D vs. I: OR = 0.801, 95% CI = 0.714-0.898, p < 0.001). Conversely, it was associated with Flaviviridae (RNA genome) infection in the presence of cancer under the same model (D vs. I: OR = 1.972, 95% CI = 1.022-3.806, p = 0.043). Whereas the HLA-G 14-bp I/D polymorphism was not associated with Coronaviridae infection or Retroviridae infection in patients without cancer. The results of analysed subgroups influenced by ethnicity indicated that HLA-G 14-bp I/D polymorphism was associated with viral infection under the allelic model for mixed populations and under the allelic model and the genotypic models for Caucasians. In complicated infections with the presence of cancer, the HLA-G 14-bp I/D polymorphism is associated with viral infection under the allelic model (D vs. I: OR = 1.855, 95% CI = 1.459-2.358, p < 0.001) and genotypic models (DD + DI vs II: OR = 4.410, 95% CI = 2.754-7.064, p < 0.001; and DD vs DI + II : OR = 1.434, 95% CI = 0.984-2.088, p = 0.061) in Caucasians. Particularly, it gives protection against hepadnavirus infection in patients without cancer for Caucasians under the allelic model (D vs. I: OR = 0.752, 95% CI = 0.640-0.883, p = 0.001) and genotypic (DD + DI vs. II: OR = 0.534, 95% CI = 0.422-0.677, p < 0.001 ; DD + II vs. DI: OR = 1.819, 95% CI = 1.449-2.284, p < 0.001; and DI vs. II: OR = 0.509, 95% CI = 0.399-0.650, p < 0.001) models. But it seems to offer a protection against flavivirus infection in Caucasians when expressing cancer under the allelic (D vs. I) and genotypic (DD vs. DI + II; DD + DI vs. II; DD + II vs. DI; DI vs. II; and DD vs. II) models. CONCLUSION: This meta-analysis highlights the potential implication of virus structure, patient history, and ethnicity in HLA-G 14-bp I/D polymorphism regulation. This polymorphism appears to be associated withcancer susceptibility and immune protection during viral infection. It is hoped that further research will give more clinical utility of HLA-G polymorphism as a valuable molecular tool for prognostic assessment and therapy planning.