Genotype‐phenotype correlations and BH4 estimated responsiveness in patients with phenylketonuria from Rio de Janeiro, Southeast Brazil

Background Genetic heterogeneity and compound heterozygosis give rise to a continuous spectrum of phenylalanine hydroxylase deficiency and metabolic phenotypes in phenylketonuria (PKU). The most used parameters for evaluating phenotype in PKU are pretreatment phenylalanine (Phe) levels, tolerance for dietary Phe, and Phe overloading test. Phenotype can vary from a “classic” (severe) form to mild hyperphenylalaninemia, which does not require dietary treatment. A subset of patients is responsive to treatment by the cofactor tetrahydrobiopterin (BH4). Genotypes of PKU patients from Rio de Janeiro, Brazil, were compared to predicted and observed phenotypes. Genotype‐based estimations of responsiveness to BH4 were also conducted. Methods Phenotype was defined by pretreatment Phe levels. A standard prediction system based on arbitrary assigned values was employed to measure genotype‐phenotype concordance. Patients were also estimated as BH4‐responders according to the responsiveness previously reported for their mutations and genotypes. Results A 48.3% concordance rate between genotype‐predicted and observed phenotypes was found. When the predicted phenotypes included those reported at the BIOPKU database, the concordance rate reached 77%. A total of 18 genotypes from 30 patients (29.4%) were estimated as of potential or probable BH4 responsiveness. Inconsistencies were observed in genotypic combinations including the common “moderate” mutations p.R261Q, p.V388M, and p.I65T and the mild mutations p.L48S, p.R68S, and p.L249F. Conclusion The high discordance rate between genotype‐predicted and observed metabolic phenotypes in this study seems to be due partially to the high frequency of the so‐called “moderate” common mutations, p.R261Q, p.V388M, and p.I65T, which are reported to be associated to erratic or more severe than expected metabolic phenotypes. Although our results of BH4 estimated responsiveness must be regarded as tentative, it should be emphasized that genotyping and genotype‐phenotype association studies are important in selecting patients to be offered a BH4 overload test, especially in low‐resource settings like Brazil. The correlations between phenotypic characteristics and the mutations found in PKU patients from Rio de Janeiro, Brazil, were evaluated. A reasonable relationship between mutation severity and the inverse of pretreatment Phe levels was observed. The correspondence of the genotype‐predicted and the observed phenotypes reached 48.3%.