Tumor necrosis factor-α promoter -308 G/A polymorphism and susceptibility to sepsis in very-low-birth-weight infants.

OBJECTIVES: To determine whether the tumor necrosis factor-α -308 G/A polymorphism is associated with blood culture-proven sepsis in two large cohorts of very-low-birth-weight infants. DESIGN: Genetic association studies. SETTING: Prospective, population-based, multicentered cohort of 1944 very-low-birth-weight infants born in 14 German study centers between 2003 and 2008 and 976 mothers, and a second prospective cohort of 926 very-low-birth-weight infants born in 2009 (German Neonatal Network). MEASUREMENTS AND MAIN RESULTS: In cohort I, 344 of 1944 (18.2%) very-low-birth-weight infants had at least one episode of blood culture-proven sepsis develop. The sepsis incidence stratified to genotype was 19.3% for G/G, 15.8% for G/A, 10.0% for A/A genotype (Cochrane-Armitage trend test: G/G vs. G/A: odds ratio, 1.32; 95% confidence interval, 1.03-1.71; G/G vs. A/A: odds ratio, 1.74; 95% confidence interval, 1.06-2.91; p = .03). There was a trend for association of tumor necrosis factor-α -308 A/G genotype with late-onset sepsis episodes (incidence: 17.2% for G/G, 12.5% for G/A, 10.0% for A/A genotype; Cochrane-Armitage trend test: G/G vs. G/A: odds ratio, 1.43; 95% confidence interval, 1.09-1.9; G/G vs. A/A: odds ratio, 2.05; 95% confidence interval, 1.19-3.56; p = .009). However, after adjustment for multiple testing, no significant associations were found. Furthermore, the genotype of the investigated 976 mothers had no impact on sepsis risk for their very-low-birth-weight infants. We additionally studied a second prospective cohort of 926 very-low-birth-weight infants and found no associations with sepsis risk. CONCLUSIONS: No association was found between the tumor necrosis factor-α -308 G/A polymorphism blood culture-proven sepsis in two large cohorts of very-low-birth-weight infants. A recent meta-analysis demonstrated that the tumor necrosis factor-α -308 A allele is associated with higher sepsis risk in adult cohorts. Thus, potential differences between adults and infants need to be incorporated in future study designs evaluating risk profiles for sepsis.