RESUMO
OBJECTIVES: To perform long-term celiac disease (CD) screening in an HLA-DQ2 (+) cohort from the general population and to assess the influence of risk genotypes on its development. METHODS: In 2004, an HLA-DQ2 (+) cohort was selected. After the first CD screening at age 2 to 3 years, we performed a follow-up screening 8 to 10 years later. Antitransglutaminase 2 antibodies were determined using a rapid test kit. Results were confirmed by serum IgA antitransglutaminase 2 and IgA endomysial antibody determination. CD diagnosis was carried out by intestinal biopsies. Four HLA-DQ2 genotypic groups were used: G1: DQ2.5/DQ2.5 (G1A) or DQ2.5/ DQ2.2 (G1B); G2: DQ2.2/DQ7.5 (DQ2.5 trans); G3: DQ2.5/ X; G4: DQ2.2/X. RESULTS: CD prevalence after 10 years of follow-up was 5.8% (95% confidence interval 3.8-8.7). One of every 3 HLA-DQ2(+) children carried at least 1 haplotype DQ2.2 or DQ7. The homozygous genotype DQ2.5/DQ2.5 and the HLA-DQ2.5 trans genotype increased CD risk 4- and 3-fold, respectively. The homozygous genotype DQ2.5/ DQ2.2 did not increase the CD risk. Children carrying G1 or G2 genotypes were diagnosed with CD earlier and more frequently during the follow-up compare with those carrying G3 or G4 genotypes. Approximately 81% of children with spontaneous antibody negativization after the first screening maintained negative antibodies. CONCLUSIONS: A repeated screening of at-risk children during their follow-up allowed us to diagnose new CD cases. In our cohort, HLA- DQ2.5 trans genotype conferred a higher risk in the development of CD than HLA- DQ2.5/DQ2.2. The majority of children with potential CD and CD autoimmunity at 10 years of age remained healthy.