Plasma amino acid levels in a cohort of patients in Turkey with classical phenylketonuria

After approval of the study protocol by the Harran University Ethics Committee for Human Research (certificate of approval No. E40428; approval date October 4, 2018, session 10/40428) following the Declaration of Helsinki and its contemporary amendments, and after obtaining written informed consent to participate in the study from the parents of all patients and controls included, we conducted a study in a community in Sanliurfa province, Turkey. All participants with phenylketonuria included in the study were followed up regularly from October 15, 2017 to October 15, 2018 in the pediatric metabolic diseases clinic. Data were obtained retrospectively from each phenylketonuria patient's medical record at the time of their most recent visit, which included age, sex, height, weight, adherence to dietary treatment, and plasma AA analysis results. Blood samples were collected from age-matched healthy control patient-participants attending the Pediatric Metabolism Disorders clinic but with normal results and volunteered by their parents in the month from 15 October 2018 to 15 November 2018, and data were obtained, which included age, sex, height, weight, and plasma AA analysis results on the same day. The phenylketonurics were divided into 2 groups according to their adherence to dietary treatment: those who adhered [diet compliant (DC)] and who had not adhered to their diet [diet incompliant (DIC)], and plasma AA analysis results were compared with those of healthy children in the control group. The participants included in the control group were selected from among healthy children who did not have any disease or pathological condition. The patients in the DC group were selected among the patients who had the target plasma levels of Phe (<12 years 120–360 μmol/L, ≥ 12 years 120–600 μmol/L) in all 3 consecutive follow-up examinations within the last 6 months. The patients in the DIC group were selected among the patients who exceeded the target plasma levels of Phe (<12 years >360 μmol/L, ≥ 12 years >600 μmol/L) as determined in all 3 consecutive follow-up examinations within the last 6 months. Patients who did not attend the follow-up examinations (3 times in the last 6 months) were excluded from the study. The plasma samples taken for the LC-MS/MS analysis were analyzed immediately after they are collected. Informed consent from the parents of all participants included in the study was obtained and documented before the venous blood samples were taken.

We included 150 patients evaluated by the clinic for pediatric metabolic diseases and met the study criteria. Of the patients, 56% were boys and 44% were girls. At the time of admission to our clinic, the mean age of the patients was 6.55 ± 4.35 years (range 0–16). There were 44 (29%) patients in the DC group and 56 (37%) patients in the DIC group, and 50 (34%) patients in the control group. When the groups were analyzed in terms of sex, weight for height percentile distribution, and mean age values, they were found to be similar (Table 1). In the DIC group, the rate of parental consanguinity was 87.5%, while it was 75% in the DC group. The distribution of parental consanguinity rates between the 2 groups was found to be similar (Pearson chi-square test, 2 × 2 table, P = 0.106). The rate of parental consanguinity for the control group was 22% and significantly lower than that of other groups (Pearson chi-square test, 3 × 2 table, P < 0.001. The proportion of patients with siblings with a phenylketonuria history was 68% (n = 38) in the DIC group, 20% (n = 9) in the DC group, and 4% (n = 2) in the control group. The proportion of patients with siblings with a phenylketonuria history was significantly higher in the DIC group than in the DC group (the Pearson chi-square test, 2 × 2 table, P < 0.001). Only 8 patients had a history of an unexplained death of siblings during birth, all of whom were in the DIC group. When the patients were analyzed in terms of comorbidity and/or pathological condition, we found that 2 patients had epilepsy, 2 had mild mental motor retardation, and 1 had a hearing problem in the DIC group, while 1 patient had epilepsy and 2 had mild mental motor retardation in the DC group. None of the patients in the control group had any comorbidity or pathological condition.

In the analyses conducted, no significant difference was found between the DIC, DC, and control groups in terms of the mean Tau, Trp, Phe, Tyr, Leu, Met, Ile, Pro, Gln, Cit, Hc2, Cys, Cy2, Arg, OH-Lys, Orn, Lys, His, 3-MHIS, 1-MIHS, Val, Thr, Ser, Ala, Asp, or Hyp values (Table 2). When the mean Phe, Tyr, GABA, Gly, Asn, and Glu values of the DIC, DC, and control groups were compared, a significant difference was observed between the groups (Table 2). In post hoc analysis, a significant difference was found between the DIC, DC, and control groups in the mean Phe level (Table 3), but no significant difference was found between the DIC and DC groups in the mean Tyr level, while the mean Tyr level was significantly higher in the control group than both in the DIC and the DC groups (Table 3).

A significant difference was observed between the groups in the mean GABA, Gly, Glu and Asn levels (Table 2). In the post hoc analysis, the mean GABA and Gly values were significantly higher in the DIC group than they were in the DC and control groups, and no significant difference was observed between the DC and control groups in the mean GABA and Gly levels (Table 3). In the post hoc analysis, the mean Glu and Asn levels were significantly higher in the DC and control groups than they were in the DIC group, while no significant difference was observed between the DC and control groups (Table 3).

The Pearson correlation analysis revealed a positive correlation between the GABA and Phe (r = 0.462, P < 0.001) and Gly and Phe (r = 0.511, P < 0.001) levels (Table 4). Moreover, a negative correlation was found between the Glu and Phe (r = −0.475, p < 0.001) and Asp and Phe (r = −0.412, P < 0.001) levels in the Pearson correlation analysis (Table 4).

Numerous congenital metabolic diseases can be diagnosed by using a dried blood spot (heel prick) test in the early neonatal period. Early diagnosis of phenylketonuria is extremely important, especially as it is an important cause of preventable mental retardation. Phenylketonuria is common in geographical regions where consanguineous marriages, such as in Turkey and Iran, are frequent [16, 17]. We found that the rates of consanguineous marriages in the parents and a history of phenylketonuria in siblings were higher in children with classic phenylketonuria than they were in healthy children.

Compliance with dietary treatment is most effective in infancy and childhood [18]. The diet severely interferes with culturally normal eating habits, particularly in older children and adolescents, and this often results in problems with compliance with dietary treatment recommendations [5, 18]. In the present study, the ages of the patients in the DC and DIC groups were similar. We consider that sociocultural factors, such as seasonal agricultural work and refugee life, may affect this situation. Indeed, in the present study, we found that refugees were more common among the DIC group patients.

Especially in regions where phenylketonuria is more common, besides early diagnosis, close follow-up of patients is also necessary. Depending on the age of the patients and their adherence to the diet (blood Phe level), white matter abnormalities (WMA) may arise in the brain tissue of patients with phenylketonuria, and cognitive functions of children may be adversely affected [19,20,21]. Although the pathogenesis of resulting WMA and the severity of adversely affected cognitive functions have not been fully elucidated, it has been shown that WMA may regress within 6 months with decreased blood Phe levels [3, 19, 22]. Various scoring tests and cranial magnetic resonance imaging (MRI) techniques are used in the follow-up of a patient's cognitive functions [2].

While the blood Phe level can be measured precisely from venous blood using LC-MS/MS, many other variables can be measured simultaneously. One of these variables is the blood Tyr level. Current recommendations are to add Tyr to the diet in the treatment of patients with phenylketonuria [2]. In our study, the mean plasma Tyr levels in the patients with phenylketonuria (DIC and DC groups) were found to be lower than that of participants in the control group, suggesting that Tyr supplement added to the diet might be clinically beneficial in patients with phenylketonuria. In the present study, the plasma levels of neurotransmitter AAs, which are considered to play an important role in the CNS functions, have been investigated along with the blood Phe level. Yuan et al. [23] revealed that decreased GABA level and increased Asp, Glu, Tau, and l-Ser levels in plasma could be correlated with Parkinson's disease. Bugajska et al. [24] found a correlation between low plasma exogenous AA levels and autism in a pediatric population, so that there may be a correlation between autism and phenylketonuria [25]. Zaki et al. [26] found that the levels of some AAs in autistic children, other than Phe in plasma, differed from those in healthy children. A study demonstrating a correlation between major depressive disorder (MDD) and plasma AA levels was conducted by Ogawa et al. [27]. They found that plasma Asn levels were lower, whereas Glu levels were higher in children with MDD than they were in healthy children.

As is seen in various studies, plasma AA levels may vary in many diseases affecting the CNS. To our knowledge, the present study is the first to show the relationship between dietary compliance and plasma excitatory and inhibitory AA levels in patients with phenylketonuria. We found that the mean plasma GABA and Gly values were higher, and the mean plasma Glu and Asn values were lower in the DIC group than they were in the DC and control groups. As expected, no significant difference in levels between the DC and control groups was observed, except for the mean plasma levels of Phe and Tyr. A positive correlation was found between plasma Phe level and both GABA and Gly levels, whereas we found a negative correlation between plasma Phe levels and both Glu and Asn levels.

Our present study is limited because no intelligence test was performed on the patients in this situation, so any correlation between plasma Phe level and intelligence test score could not be evaluated. Failure to determine cerebrospinal fluid levels of neurotransmitters and lack of results of neuromotor development tests (e.g., Denver) of patients can be considered as a limitation. Socioeconomic parameters, such as parental education level and monthly income, which might affect dietary compliance, were not evaluated in the present study. Multicenter studies to be conducted with plasma AA analysis results of patients with phenylketonuria in different clinical centers are warranted.

High blood plasma levels of inhibitory AAs, such as GABA and Gly, and low plasma levels of excitatory AAs, such as Asn and Glu, were measured in patients in Sanliurfa province, Turkey, with phenylketonuria who did not follow their diets. The plasma levels of excitatory and inhibitory AAs may be used as auxiliary parameters in the follow-up of cognitive functions of patients with phenylketonuria. More extensive studies are needed to demonstrate any negative association of cognitive function and the plasma levels of these AAs.