MANCHESTER, UK—Certain polymorphisms of the GM-CSF gene appear to be useful in predicting an individual’s predisposition to atopic dermatitis (AD). A recent study demonstrates that having two copies of the so-called C allele at the upstream promotor region –677 (GM-CSF –677*CC) is protective against AD.[1] Of the study participants with AD, none with this genotype had severe disease. No significant genetic associations were found between AD and polymorphisms of genes IL-1ß and TNF-alpha.

GM-CSF, IL1ß, and TNF-alpha have all been recognized as important modulators of Langerhans’ cells, a type of antigen-presenting dendritic cell localized to the skin and functioning as a primary trigger of immune responses. The study objective was to determine which, if any, of the polymorphisms of these factors were associated with a greater frequency or severity of AD.

The researchers performed DNA studies in 113 children with AD and 114 control children without AD, asthma, or allergic rhinitis. The children with AD (median age, 4) were recruited from a referral hospital known for its treatment of unusually severe or persistent AD. Children in the control group (median age, 11) were recruited from an adjoining fracture clinic; infants and young children were intentionally excluded from this group to avoid including anyone who might later develop AD. Most of the participants in both groups were white.

Significant differences were identified between patients and controls in the frequency of polymorphisms of GM-CSF. At the –677 locus, 5% of the AD patients had the genotype CC, 47% had AC, and 48% had AA. The comparative frequencies in the control group were 16%, 68%, and 16%. At the –1916 locus, characterized by the alleles T and C, the frequency distributions (cases versus controls) were: 9% versus 16% for CC; 44% versus 65% for TC; and 47% versus 19% for TT.

Using regression analysis, the researchers showed that children who had not inherited a C allele had twice the risk of AD as did other children. In genotypic terms, children who were either AA (at the –677 locus) or TT ( at the –1916 locus) had odds ratios for AD of 7.5 and 3.3, respectively, compared with children who were homozygous CC.

In a separate analysis, the researchers stratified the children with AD into two subgroups: 58 had mild to moderate disease (5% to 27% of total body surface affected) and 55 had severe AD (27% to 100% of the surface affected). Interestingly, the children with severe AD had an unusual distribution of the –677 GM-CSF genotypes: None had CC, and 50% each had AA or AC. In contrast, 16% of the control subjects had the CC genotype, 16% had AA, and 68% had AC. The children with mild to moderate AD had genotype frequencies between these two extremes.

—Verna L. Schwartz, MS