Meta-analysis On The Association Of CCR5△32, SDF1-3'a And CCR2-64I Gene Polymorphism And HIV-1 Infection

1 BackgroundMany studies suggested that the chemokines receptors CCR5, CCR2 and CXCR4 ligand SDF1 play an important role in HIV-1 infected and disease progression. Several studies have reported the role of CCR5A32, SDF1-3'A and CCR2-64I gene polymorphism in HIV-1 infection risk, However, due to the heterogeneity of population; sample size of a single study were inadequate and other factors, making relevant research there is a big controversy. In order to help resolve this uncertainly, we have performed a meta-analysis to further clarify the association between CCR5Δ32, SDF1-3'A and CCR2-64I polymorphisms and risk of HIV-1 infection.2 Objectives2.1. To explore the association between CCR5A32 gene polymorphism and susceptibility to HIV-1 infection.2.2. To explore the association between SDF1-3'A gene polymorphism and susceptibility to HIV-1 infection.2.3. To explore the association between CCR2-64I gene polymorphism and susceptibility to HIV-1 infection. 3 Methods3.1 Criteria for InclusionThe inclusion criteria were (1) Study object:CCR5A32, SDF1-3'A and CCR2-64I gene polymorphism and risk of HIV-1; (2) independent case-control or cohort studies; (3) Observed measures:gene frequency of CCR5A32, SDF1-3'A and CCR2-64I; (4) genotype distribution of control population must be in Hardy-Weinberg equilibrium (HWE); (5) If the results were same in several studies, it will be treated as a single study in this meta-analysis; and (6) HIV-1 infected patients were not HIV-1 anti-retro viral treatment, if patients be treated, the literature is not included.3.2 Criteria for exclusionThe inclusion exclusion were (1) no available genotype frequency; (2) no control population; and (3) duplication of a previous study.3.3 Search strategyWe searched MEDLINE (U.S National Library of Medicine) for all genetic association studies on the CCR5A32, SDF1-3'A and CCR2-64I polymorphisms and the susceptibility of HIV-1 published from December 1990 to September 2010 using the PUBMED search engine. The search used the keywords and subject terms "SDF1-3'A," "CCR5A32" "CCR2-64I" "polymorphism," and "HIV-1," and the language was not limited.3.4 Statistical analysisCarefully read the included documentation, collect and analyze relevant data, this meta-analysis was performed with STATA version 9.0 (STATA Corporation, College Station, TX).4 Results4.1. Study characteristics 34 studies met the inclusion criteria.28 studies met the inclusion criteria of CCR5Δ32; 16 studies met the inclusion criteria of CCR2-64I; 16 studies met the inclusion criteria of SDF1-3'A.3710 cases and 4655 controls were enrolled in meta-analysis of CCR5Δ32 gene polymorphism; 2317 cases and 2385 controls were enrolled in meta-analysis of CCR2-64I gene polymorphism; and 1953 cases and 2212 controls were enrolled in meta-analysis of SDF1-3'A gene polymorphism.4.2. Meta-analysis results4.2.1 CCR5A32When the homozygote YY and heterozygote CY (YY+CY) were compared with the wild-type genotype (CC), the pooled ORs for the 19 studies were 1.12 (95% CI: 0.87-1.44, P=0.382), the heterozygote CY were compared with the wild-type genotype (CC), the pooled ORs for the 18 studies were 1.12 (95% CI:0.94-1.32), P=0.193), the homozygote YY were compared with the wild-type genotype (CC), the pooled ORs for the 7 studies were 0.81 (95% CI:0.45-1.43, P=0.464). In the subgroup analysis for subset of cases, statistically significantly decreased risk was found in LNTP cases (dominant model:OR=2.09,95% CI=1.42-3.09, P=0.000; recessive model:OR=2.12,95% CI=1.44-3.13, P=0.000).4.2.2 CCR2-64IWhen the heterozygote (CY) was compared with the wild-type genotype (CC), the pooled ORS for the 16 studies were 0.95 (95%CI:0.82-1.09), P-value of heterogeneity test was 0.87,I-squared was 0.0%, p-value of significant test was 0.45; the heterozygote and homozygote (CC+YY) were compared with the wild-type genotype (CC), the pooled ORS for the 16 studies were 0.97 (95%CI:0.84-1.11), P-value of heterogeneity test was 0.92, I-squared was 0.0%, P-value of significant test was 0.66; the homozygote (YY) was compared with the wild-type genotype (CC), the pooled ORS for the studies were 1.21 (95%CI:0.84-1.74), P-value of heterozygote was 0.63, I-squared was 0.0%, p-value of significant test was 0.30.4.2.3 SDF1-3'AWhen the heterozygote (CY) was compared with the wild-type genotype (CC), the pooled ORS for the 16 studies were 0.86 (95%CI:0.70-1.06), P-value of heterogeneity test was 0.02, I-squared was 46.5%, P-value of significant test was 0.15; the heterozygote and homozygote (CC+YY) were compared with the wild-type genotype (CC), the pooled ORS for the 16 studies were 0.83 (95%CI:0.67-1.03), P-value of heterogeneity test was 0.003, I-squared was 56.1%, P-value of significant test was 0.09; the homozygote (YY) was compared with the wild-type genotype (CC), the pooled ORS for the studies were 0.79 (95%CI:0.50-1.26), P-value of heterogeneity test was 0.007, I-squared was 41.0%, P-value of significant test was 0.32. In the subgroup analysis for ethnicity, when the heterozygote (CY) was compared with the wild-type genotype (CC) in Caucasian population, the pooled ORS were 0.87(95%CI:0.60-1.26), P-value of heterogeneity test was 0.46 I-squared was 0.0%, P-value of significant test was 0.99; the heterozygote and homozygote (CC+YY) were compared with the wild-type genotype (CC), the pooled ORS were 0.95 (95%CI:0.66-1.35), P-value of heterogeneity test was 0.98, I-squared was 0.0%, P-value of significant test was 0.98; the homozygote (YY) was compared with the wild-type genotype (CC), the pooled ORS for the studies were 1.87 (0.80-4.38), P-value of heterogeneity test was 0.60, I-squared was 0.0%, P-value of significant test was 0.15. there is a big heterozygote in Asian population, we further stratified analysis for Asian population, In the subgroup analysis for China population, when the heterozygote (CY) was compared with the wild-type genotype (CC), the pooled ORS were 0.96 (95%CI:0.76-1.21), P-value of heterogeneity test was 0.56, I-squared was 0.0%, P-value of significant test was 0.71; the heterozygote and homozygote (CC+YY) were compared with the wild-type genotype (CC), the pooled ORS were 1.00 (95%CI:0.80-1.25), P-value of heterogeneity test was 0.48,I-squared was 0.0%, P-value of significant test was 0.99, the homozygote (YY) was compared with the wild-type genotype (CC), the pooled ORS for the studies were 1.12(0.85-1.74), P-value of heterogeneity test was 0.55, I-squared was 0.0%, p-value of significant test was 0.28.4.3 Publication biasThe egger's regression test and Begg's adjusted rank correlation test indicate no evidence of publication bias in CCR5A32 meta-analysis (dominant model:P=0.922 and 0.484; recessive model:P=0.932 and 0.544); the egger's regression test and Begg's adjusted rank correlation test also suggested the absence of publication bias in SDF1-3'A meta-analysis (P=0.404 and 0.344 for dominant model, P=0.444 and 0.558 for additive model, P=0.404 and 0.344 for recessive model); the egger's regression test and Begg's adjusted rank correlation test also suggested the absence of publication bias in CCR2-64I meta-analysis (P=0.334 and 1.000 for dominant model, P=0.363 and 0.502 for additive model, P=0.431 and 0.964 for recessive model). The shapes of the funnel plot indicate that it did not any evidence of obvious asymmetry in these models. The results indicated that the results of this meta-analysis are relatively stable and that publication bias can not affect the results of our meta-analysis.5 ConclusionsThis meta-analysis indicated that the heterozygote of CCR5delta32 may decrease risk in the Caucasian of HIV-1-infected long-term non-progressors (LNTP); this meta-analysis indicated that the mutation of CCR2-64I can not decrease the risk of HIV-1-infected in the whole population; and this meta-analysis indicated that the mutation of SDF1-3'A can not decrease the risk of HIV-1-infected in Caucasian and China population.