No evidence for individual blood–brain barrier phenylalanine transport to influence clinical outcome in typical phenylketonuria patients

Brain tissue concentrations of phenylalanine (Phe) can be measured by proton magnetic resonance spectroscopy in patients with phenylketonuria (PKU). In combination with oral Phe challenges, the kinetics of Phe transport at the blood– brain barrier can be characterized. Weglage and colleagues reported interindividual differences and an association of such kinetic parameters with IQ and magnetic resonance imaging (MRI) visible white matter changes in a group of 15 PKU patients. It was interpreted to indicate that “interindividually different blood–brain barrier transport characteristics [. . .] are major causative factors for clinical outcome in PKU.” The results of Weglage and colleagues contradict a recent study by Rupp and colleagues, which yielded a close linear regression for blood and brain Phe concentrations and no correlation between blood and brain ratios (mean, 4.0 0.4) and IQ (r 0.05, not significant). Reanalysis of the data published by Weglage and colleagues and comparison with earlier results of this group were performed to clarify this discrepancy. There is ample evidence that strictness of early dietary treatment, documented, for example, as blood Phe levels during the first 10 years of life ([Phe]blood 10y), determines IQ. As expected, Weglage and colleagues found a strong negative correlation between [Phe]blood 10y and IQ (r 0.68; p 0.005). The major conclusion of this study is based on associations between kinetic parameters and IQ (Kt,app: r 0.45; p 0.09; Tmax/Vmet: r 0.50, p 0.06). However, completing the correlational matrix in Table 3 in the study by Weglage and colleagues, it can be found that the kinetic parameters also are correlated with [Phe]blood 10y (Kt,app: r 0.54, p 0.04; Tmax/Vmet : r 0.41, p 0.13; Spearman rank correlations based on the data presented in Tables 1 and 2, but not mentioned in Weglage and colleagues). If the influence of [Phe]blood 10y on IQ is statistically controlled for, the correlation coefficients of IQ with Kt,app (r 0.13, p 0.65) and with Tmax/Vmet (r 0.32, p 0.26) substantially decrease and become statistically nonsignificant. Numerous studies have shown that the severity of MRI changes is related to blood Phe levels during several weeks to months before investigation. The correlation between MRI changes and preload brain Phe (r 0.77, p 0.001) therefore is not unexpected. In contrast with the literature, the association between MRI changes in adult life is substantially higher with [Phe]blood 10y (r 0.75, p 0.005) than with [Phe]blood 10y (r 0.31, p 0.25) or with preload [Phe]blood (r 0.02, p 0.94). Moller and colleagues described saturation kinetics for Phe transport based on blood–brain Phe ratios (mean 2.1 0.5 for blood Phe values 1.5mM, in which a significant linear correlation was found by the researchers). In their recent work, the corresponding preload ratio was 3.8 1.1. This suggests a systematic difference between the present and earlier studies (Fig). The data of the two patients with lowest Kt,app and highest Tmax/Vmet (no.1 and 2), and hence largest influence on the apparent “individuality” of the kinetic parameters, already were included in earlier studies (nos. 8 and 9 in Moller and colleagues). These two patients, who previously have been characterized as “typical” PKU patients, now seem to represent “particular” patients compared with the remaining 13 patients (eg, Tmax/Vmet is 6 standard deviations above the mean of the 13 new patients), whereas the kinetic parameters of three previously “atypical” patients fit well with the new data set. Thus, the statistics in the recent article appear to be based on an inhomogeneous sample (Fig) composed of 13 recently investigated patients and 2 patients added from an earlier study. If the latter are excluded from analysis, the correlations of IQ with Kt,app (r 0.23, p 0.46) and with Tmax/Vmet (r 0.31, p 0.31) decrease, corroborating the results of Rupp and colleagues. In summary, clinically significant interindividual differences in blood–brain barrier Phe transport and an influence of kinetic characteristics determined in adult life on outcome parameters is not confirmed for most “typical” PKU patients. The conclusion that the reported observations ultimately could lead to individual dietary recommendations is currently not tenable.