Aims of dietary treatment

The primary aim of dietary treatment is to prevent adverse neurocognitive and psychological outcomes by restricting and titrating the intake of protein from natural foods to maintain blood Phe levels within the guideline-established range that prevents adverse outcomes while providing enough Phe to support protein synthesis and avoid catabolism⁽ Reference van Spronsen, van Wegberg and Ahring ^{3 ,} Reference Vockley, Andersson and Antshel ^{9 )}. There is general consensus on the importance of meticulous control of blood Phe levels within a narrow target range from birth to 12 years of age (120–360 μmol/l)⁽ Reference van Spronsen, van Wegberg and Ahring ^{3 ,} Reference Vockley, Andersson and Antshel ^{9 ,} Reference Camp, Parisi and Acosta ^{10 )}. For patients above the age of 12 years (exception pre-conception and pregnancy), the European guidelines consider 120–600 μmol/l to be safe throughout adolescence and adulthood⁽ Reference van Spronsen, van Wegberg and Ahring ^{3 )}, while US guidelines advocate 120–360 μmol/l throughout life⁽ Reference Camp, Parisi and Acosta ^{10 )}. Protein substitutes have an important role in helping achieve optimal metabolic control.

The second aim of dietary management is to maintain a healthy nutritional status by providing sufficient non-Phe amino acids, energy and other nutrients to support physiological protein synthesis and counterbalance catabolism, without providing excess energy. N and essential amino acid requirements are met with a Phe-free amino acid mixture containing vitamins and other essential nutrients and the addition of special low-protein modified foods provides a necessary energy source⁽ Reference Pena, de Almeida and van Dam ^{11 ,} Reference Pena, Almeida and van Dam ^{12 )}. The specialist dietitian and metabolic physician manage dietary treatment to maintain appropriate blood Phe levels in response to metabolic changes, while ensuring that other nutritional requirements are being satisfied to support normal growth and development in children and a healthy body composition in all patients⁽ Reference Singh, Rohr and Frazier ^{13 )}.

Notwithstanding the theoretical effectiveness of dietary treatment, its actual success depends on providing a diet that is acceptable and practical to the patient. Subjective factors that can hinder acceptance include poor palatability, disagreeable smell, or textures, lack of dietary variety and food neophobia⁽ Reference Evans, Daly and Chahal ^{14 )}. Adherence with protein substitutes that must be consumed at least three times a day throughout life is a particular challenge. Socially acceptable, appropriate food choice is particularly important for improving adherence as patients move from childhood to adolescence, as they have more responsibility for managing their treatment and are exposed to new food choices and peer pressure outside the home⁽ Reference Walter, White and Hall ^{15 )}.

Protein requirement

In PKU, when the protein requirement is met primarily with a Phe-free amino acid mixture, the protein equivalent amount of amino acids should be increased because of poor retention⁽ Reference Walter, White and Hall ^{15 )}. This higher requirement may have both a physiological cause, related to the oxidation of excess free amino acids due to their rapid, non-physiological absorption⁽ Reference Acosta, Yannicelli and Singh ^{16 –} Reference Mönch, Herrmann and Brösicke ^{18 )}, and perhaps a pathological cause, related to higher needs imposed by the condition itself, in the form of increased catabolism. Few studies have addressed this latter point. Van Rijn et al. ⁽ Reference van Rijn, Hoeksma and Sauer ^{19 )} found no difference in protein metabolism between adults with PKU and matched healthy volunteers using an l-[1-¹³C]valine tracer method. In this study, six patients with PKU received amino acids equivalent to 0·96 g protein/kg per d (120 % of the WHO recommended daily allowance)^{( 20 )}, while the six healthy volunteers received 0·8 g intact protein/kg per d (100 % of the recommended daily allowance). The 20 % increase in patients with PKU was administered to account for inefficient absorption/utilisation. The results suggest that, at least in adults with PKU, the increased requirement for amino acids results from reduced retention rather than an increase in metabolic requirement. In children, however, Turki et al. ⁽ Reference Turki, Ueda and Cheng ^{21 )} used the indicator amino acid oxidation technique to study four patients with mild PKU ranging in age from 9 to 18 years, finding that their protein requirement was 1·85 g/kg per d, considerably higher than the recommended amount for healthy children⁽ Reference Turki, Ueda and Cheng ^{21 )} and higher than the recommendation for children with PKU⁽ Reference van Spronsen, van Wegberg and Ahring ^{3 ,} Reference Vockley, Andersson and Antshel ^{9 ,} Reference Singh, Cunningham and Mofidi ^{22 ,} Reference Camp, Lloyd-Puryear and Huntington ^{23 )}. This may reflect methodological differences between older N balance studies that inform guidelines and the more sensitive isotope methods. Moreover, in a 6-week randomised, cross-over study conducted in twenty-five children aged 2–10 years (median 6 years), administering a higher dosage of protein substitute was associated with lower blood Phe levels⁽ Reference MacDonald, Chakrapani and Hendriksz ^{24 )}. Compared with a PKU diet containing the recommended protein requirement for children (1·2 g/kg per d), a similar diet with higher protein content (2·0 g/kg per d) resulted in median blood Phe levels that were 301 μmol/l lower when measured before breakfast (95 % CI 215, 386 μmol/l; P<0·001) and 337 μmol/l lower when measured in the evening (95 % CI 248, 431 μmol/l; P<0·001)⁽ Reference MacDonald, Chakrapani and Hendriksz ^{24 )}.

Either of these mechanisms (reduced retention or increased requirement) justifies the administration of increased amino acid intake or the use of an amino acid mixture formulated to achieve physiological absorption and avoid net protein catabolism, which causes the release of Phe and increases blood Phe concentrations. The current European guidelines suggest administering 140 % of the recommended daily requirement established for healthy individuals⁽ Reference van Spronsen, van Wegberg and Ahring ^{3 )}, and the most recent US guidelines suggest administering 150 % until age 4 years and 120–140 % thereafter⁽ Reference Singh, Cunningham and Mofidi ^{22 )}.

Amino acid absorption

Free amino acid mixtures are formulated to fulfil the specific requirements of patients with PKU. Administering free amino acids bypasses the digestive phase, and the absorption profile is different from that of intact proteins, in that plasma levels of total and essential amino acids are higher, peak faster and decrease more quickly (Fig. 1)⁽ Reference Gropper and Acosta ^{17 )}. Administering a free amino acid mixture with a prolonged release mimicking physiological absorption kinetics is expected to improve the rate of assimilation into protein and prevent fluctuations in plasma amino acid levels.

Fig. 1 Mean change from baseline in total plasma amino acids after feeding whole protein (cottage cheese, ––) or an equivalent amount of a free amino acid mixture with identical amino acid composition (---) in fasting healthy volunteers (n 10). Adapted from Gropper & Acosta⁽ Reference Gropper and Acosta ^{17 )}.

Metabolic effects of amino acid absorption rate

Compared with physiological administration of free amino acid mixtures distributed through the day, bolus administration increases the amount of N excreted in urine⁽ Reference Mönch, Herrmann and Brösicke ^{18 )}, when the rapid increase in blood amino acids exceeds the capacity of anabolic processes to incorporate them into nascent proteins. Dangin et al. ⁽ Reference Dangin, Boirie and Garcia-Rodenas ^{33 )} examined the effect of protein digestion rate on protein accretion using ¹³C-labelled Leu and various fast v. slow protein meals. They compared casein v. free amino acids mimicking casein, and whey protein bolus (fast protein) v. repeated dosing (simulated slow protein)⁽ Reference Dangin, Boirie and Garcia-Rodenas ^{33 )}. Fast uptake was associated with a rapid increase in blood amino acids and more oxidation. Protein retention was greater with slow protein sources (casein v. free amino acids: 38 (se 13) v. –12 (se 11) µmol Leu/kg; P<0·01) and with repeated dosing v. whey bolus (87 (se 25) v. 6 (se 19) µmol Leu/kg; P<0·05). Jones et al. ⁽ Reference Jones, Lees and Andrews ^{34 )} compared continuous enteral nutrition with protein v. free amino acid mixtures in isonitrogenous, isoenergetic diets, and found that patients randomly assigned to receive the free amino acid mixture had significantly higher N loss⁽ Reference Jones, Lees and Andrews ^{34 )}. However, a metabolic study comparing adult patients with PKU fed a protein-restricted, free amino acid mixture-supplemented diet with matched healthy subjects receiving a normal diet revealed no differences in whole-body protein metabolism at a protein intake of 0·8 g/kg per d⁽ Reference van Rijn, Hoeksma and Sauer ^{19 )}.

Infusion of increasing amounts of a balanced mixture of free amino acids (from 0·5 up to 6 mg/kg per min) in fasted healthy young adults resulted in inhibition of protein catabolism, stimulation of protein synthesis and an increase in amino acid oxidation, depending on plasma amino acid levels attained (Fig. 2)⁽ Reference Giordano, Castellino and DeFronzo ^{35 )}. Even small increases in amino acid concentrations compared with fasting concentrations inhibited protein catabolism and, in parallel, stimulated protein synthesis as well as amino acid oxidation. The decrease in protein catabolism was overruled by the increase in amino acid oxidation and most notably protein synthesis⁽ Reference Giordano, Castellino and DeFronzo ^{35 )}.

Fig. 2 Effects of graded hyperaminoacidaemia on protein metabolism: a balanced amino acid solution was infused at 0·5, 1, 2, 4 and 6 mg/kg per min for 180 min in eight fasting healthy volunteers, and amino acid oxidation (–♦–), protein synthesis (–■–) and proteolysis (–▲–) were estimated with [1-¹⁴C]leucine infusion and indirect calorimetry. Redrawn from Giordano et al. ⁽ Reference Giordano, Castellino and DeFronzo ^{35 )}.

Protein synthesis is stimulated both by amino acids and by insulin. When plasma amino acids levels are artificially maintained at post-absorptive levels (i.e. physiological hyperaminoacidaemia), administration of insulin further increases net protein synthesis and decreases protein catabolism, with no significant change in amino acid oxidation⁽ Reference Chevalier, Gougeon and Kreisman ^{36 )}.

Protein synthesis increases in the postprandial state because of anabolic signals from the insulin pathway and the increase in plasma amino acids, especially Leu, that work through the mammalian target of rapamycin (mTOR)⁽ Reference Dickinson, Fry and Drummond ^{37 )}. Insulin and insulin-like growth factor 1 increase protein synthesis at the translation level via mTOR-mediated effects on components of the protein synthetic pathway, including the eukaryotic initiation factor 4E-binding protein 1 (4E-BP1) and the ribosomal protein S6 kinase (p70S6K), both key components of the translation initiation complex⁽ Reference Liu and Barrett ^{38 )}. Thus, the anabolic signal from insulin, corresponding to abundant energy (glucose), is an important factor controlling protein metabolism⁽ Reference Campbell, Wang and Proud ^{39 )}. This underscores the need for sufficient energy intake to promote anabolism and prevent an increase in N excretion when free amino acid mixtures are administered⁽ Reference Mönch, Herrmann and Brösicke ^{18 ,} Reference Weigel, Rauh and Kiener ^{40 ,} Reference MacLeod, Clayton and van Calcar ^{41 )}. An optimal protein:energy ratio has been proposed as 3·0–4·5 g protein/100 kcal (418 kJ), corresponding to approximately 12–18 % of energy from protein⁽ Reference Evans, Truby and Boneh ^{42 )}.

Amino acids themselves stimulate insulin secretion⁽ Reference Floyd, Fajans and Conn ^{43 )}. Branched-chain amino acids, especially Leu, have an insulinogenic effect when administered together with a glucose solution, compared with glucose solution alone⁽ Reference Nilsson, Holst and Björck ^{44 )}. Amino acids also stimulate the release of the incretins glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP)⁽ Reference Chang, Wu and Greenfield ^{45 ,} Reference Lindgren, Pacini and Tura ^{46 )}, which in turn increase the anabolic response by promoting insulin secretion⁽ Reference Campbell and Drucker ^{47 )}. The amino acid mixtures used in the dietary treatment of PKU contain a high level of Leu to promote anabolism. In healthy subjects, preloading with 9 g of either whey or soya protein, or a combination of the insulinogenic amino acids (isoleucine, Leu, lysine, threonine and valine) before a reference meal significantly reduced postprandial glycaemia. This correlated with increased incretin and insulin responses⁽ Reference Gunnerud, Heinzle and Holst ^{48 )}. Compared with administration of free amino acid mixtures, physiological absorption of amino acids from intact protein represses ghrelin levels to a greater extent, providing a sense of satiety⁽ Reference Weigel, Rauh and Kiener ^{40 ,} Reference MacLeod, Clayton and van Calcar ^{41 )}.

Thus, in addition to favouring a positive N balance, more physiological absorption of amino acids in the dietary treatment of PKU may be expected to help maintain normal glycaemia and normal sensation of satiety. However, simulating normal absorption kinetics with existing free amino acid mixtures is problematic because it would complicate the treatment with more frequent dosing. When possible, administration of free amino acid mixtures should coincide with a postprandial state, in order to favour retention of dietary N.

Role of the blood–brain barrier

The most severe pathological effects of PKU are in the brain. The blood–brain barrier (BBB) isolates the brain from fluctuations in blood levels of hormones, nutrients and metabolites, and provides the stable chemical environment required for delicate neural processes. The BBB is the result of tight junctions between the endothelial cells of brain capillaries and a multitude of transporters that regulate the passage of specific molecules. The concentration of essential amino acids inside the brain is approximately ten times lower than in the general circulation and is determined by the controlled transport of LNAA across the BBB⁽ Reference Fernstrom and Wurtman ^{49 )}. This is accomplished through LAT1 (SLC7A5/SLC3A2 heterodimer), which has high affinity for the LNAA precursors important for neurotransmitter and protein synthesis⁽ Reference Matsuo, Tsukada and Nakata ^{50 ,} Reference Kido, Tamai and Uchino ^{51 )}. LAT1 is a member of transporter system L, ubiquitous bidirectional Na-independent exchangers of branched and aromatic neutral amino acids⁽ Reference Verrey ^{52 )}. In BBB endothelial cells, LAT1 is expressed on both the luminal (blood) and abluminal (brain) membranes and facilitates the transendothelial transport of LNAA (uptake substrates) in exchange for non-essential amino acid substrates (antiporter)⁽ Reference Meier, Ristic and Klauser ^{53 )}. Cells inside the brain have a three-fold higher affinity for essential amino acids compared with the BBB endothelia⁽ Reference Pardridge and Oldendorf ^{54 )}, making the BBB the limiting transfer step. This system is nearly saturated at physiological blood amino acid concentrations, and is subject to substrate competition depending on LNAA abundance and affinity. Phe is the amino acid with the highest affinity for LAT1 and the high Phe level in PKU essentially blocks the uptake of other LNAA, reducing the synthesis of proteins and neurotransmitters in the brain⁽ Reference de Groot, Hoeksma and Reijngoud ^{55 ,} Reference de Groot, Sijens and Reijngoud ^{56 )}. The importance of the BBB and amino acid disposition as a mechanism in the pathophysiology of PKU is highlighted by the finding that a minority of patients with pathologically high blood Phe concentrations have lower than expected brain Phe levels and normal intelligence quotient (IQ) without treatment⁽ Reference Möller, Weglage and Wiedermann ^{57 –} Reference Weglage, Wiedermann and Denecke ^{59 )}; however, the mechanism for this rare and irregular outcome remains unidentified⁽ Reference Møller, Paulsen and Koch ^{60 )}.

Large neutral amino acids

One approach to treating PKU involves administering higher than normal amounts of non-Phe LNAA in order to overcome this competition and allow other essential LNAA to enter the brain⁽ Reference Matalon, Michals-Matalon and Bhatia ^{61 )}. Decreasing Phe entrance into the brain by another LNAA would be accomplished most effectively with Leu (and isoleucine)⁽ Reference van Vliet, Bruinenberg and Mazzola ^{62 )}, but this does not address the entire pathogenesis of PKU brain dysfunction, and it might also result in increased brain Leu or isoleucine concentrations, which are both associated with inborn errors of metabolism with brain pathophysiology. Thus, the composition of LNAA in amino acid mixtures must be carefully balanced according to amino acid transporter affinities and importance for brain function. In particular, Tyr and tryptophan should be included at high levels to promote normal brain monoaminergic neurotransmitter concentrations⁽ Reference van Vliet, Bruinenberg and Mazzola ^{62 )}. Three mechanisms driving the improvement observed with LNAA administration have been confirmed in a murine PKU model: LNAA supplementation reduced brain Phe levels, increased brain levels of non-Phe LNAA and increased neurotransmitter levels⁽ Reference van Vliet, Bruinenberg and Mazzola ^{63 )}. However, dopamine levels were not normalised in this study, once again indicating the necessity to carefully balance the composition of the LNAA administered.

Tyrosine

The product of PAH is Tyr, which becomes a conditionally essential amino acid in patients with PKU⁽ Reference Benkovic, Wallick and Bloom ^{64 )}. Tyr is essential for protein and dopamine synthesis in the brain. Patients with uncontrolled PKU have reduced levels of dopamine because Tyr is not produced from Phe and high blood Phe levels prevent exogenous Tyr from crossing the BBB. Tyr supplementation may cause blood Tyr levels to fluctuate greatly during the day in treated patients with PKU⁽ Reference van Spronsen, van Rijn and van Dijk ^{65 –} Reference Grünert, Brichta and Krebs ^{69 )}. CV for blood Tyr concentrations are about ten-fold higher than those of Phe⁽ Reference Grünert, Brichta and Krebs ^{69 )}. This may be due to the small pool of Tyr in patients with PKU and the relatively high level of Tyr in the free amino acid mixtures administered in the PKU diet⁽ Reference Grünert, Brichta and Krebs ^{69 )}. It should be noted that, although the amino acid mixtures contain high levels of Tyr, it is not highly soluble, and solutions containing Tyr must be mixed well before administration. A slower release of free amino acids allowing a more physiological absorption may also reduce the amplitude of blood Tyr peaks and overall degree of fluctuation.

Fluctuations in blood phenylalanine levels over 24 h

In healthy subjects, the blood concentrations of individual amino acids undergo circadian variation, with peak levels occurring in the evening⁽ Reference Feigin, Klainer and Beisel ^{70 )}. This rhythm is stable and not influenced significantly by diet⁽ Reference Feigin, Klainer and Beisel ^{71 )}. Much of the ingested Phe is converted to Tyr in the liver, stabilising blood Phe levels, and maintaining the amplitude of variation generally < 50 %⁽ Reference Cleary, Trefz and Muntau ^{72 )}. Greater circadian fluctuations in blood Phe levels have been reported in patients with PKU⁽ Reference van Spronsen, van Rijn and van Dijk ^{65 ,} Reference MacDonald, Rylance and Hall ^{73 ,} Reference MacDonald, Rylance and Asplin ^{74 )}. These fluctuations may be more apparent with PAH genotypes that have low/no residual enzyme activity⁽ Reference Burgard, Rupp and Konecki ^{75 )}. In addition, it is suggested that BH₄ in pharmacological doses reduces Phe fluctuations in BH₄-responsive patients; however, responsive patients have residual PAH activity with less severe PKU, and generally have lower and less variable blood Phe⁽ Reference Levy, Milanowski and Chakrapani ^{76 ,} Reference Burton, Bausell and Katz ^{77 )}.

The normal pattern of circadian fluctuations in blood Phe is inverted in patients with PKU, with the highest levels occurring in the morning instead because of catabolism triggered by overnight fasting⁽ Reference van Spronsen, van Rijn and van Dijk ^{65 ,} Reference MacDonald, Rylance and Asplin ^{74 )}. Diurnal fluctuations may also result from inappropriate distribution of free amino acid mixtures during the day, with even distribution throughout the day resulting in more stable Phe levels⁽ Reference MacDonald, Rylance and Hall ^{73 ,} Reference MacDonald, Rylance and Davies ^{78 )}; however, administration during waking hours only is not sufficient to prevent nocturnal catabolism. Fluctuations in blood Phe from day to day, or over longer time-frames may result from inconsistent adherence to dietary treatment, failure to adjust treatment to changes in growth rate and catabolism associated with illness.

Although it has been suggested that Phe fluctuations may be associated with poorer outcomes, the mechanism by which fluctuations may affect outcomes is not known⁽ Reference Cleary, Trefz and Muntau ^{72 )}. Fluctuations in Phe levels are implicated as a factor influencing IQ and cognitive performance in PKU⁽ Reference Romani, Palermo and MacDonald ^{6 ,} Reference Arnold, Kramer and Kirby ^{79 –} Reference Hood, Grange and Christ ^{81 )}. In adults, stability of Phe levels may be as important as absolute Phe levels for cognitive outcomes⁽ Reference Romani, Palermo and MacDonald ^{6 ,} Reference Hood, Grange and Christ ^{81 )}. International guidelines do not specifically address Phe fluctuations; however, consistently following management plans, attention to dietary intake during intercurrent illness and measuring Phe levels at the recommended intervals may reduce the frequency and magnitude of Phe fluctuations.

Improving compliance to amino acid feeding

Adherence with diet is fundamental to successful treatment of PKU, and non-adherence represents a major cause of blood Phe levels outside of recommended ranges⁽ Reference MacDonald, Rylance and Hall ^{73 )}. Poor adherence can also have negative effects on overall health status if there is low and inconsistent intake of a macro- and micronutrient-supplemented amino acid mixture.

PKU represents a substantial burden⁽ Reference Bosch, Burlina and Cunningham ^{82 )}. The need to severely reduce the intake of protein-rich foods, access special low-protein modified foods, prepare special meals, and monitor and calculate the dietary protein/Phe intake can be particularly onerous for both carers and patients⁽ Reference Walter, White and Hall ^{15 ,} Reference MacDonald, Harris and Rylance ^{83 –} Reference MacDonald, Smith and de Silva ^{85 )}. This is further compounded by the need to administer multiple daily doses of free amino acid mixtures that some patients find distasteful. Despite efforts to improve palatability, the development of an amino acid mixture with organoleptic properties acceptable to all patients remains a challenge. This is particularly important among older patients who have low adherence rates⁽ Reference Walter, White and Hall ^{15 )}, and may have stopped diet treatment because of difficulties in accepting and tolerating amino acid supplements.

Phenylalanine tolerance

Depending on the amount of residual PAH activity associated with a patient’s genotype, there is considerable variation in the amount of natural protein (Phe) that patients can tolerate while maintaining blood Phe levels in the recommended range. Increasing Phe tolerance may improve adherence to treatment and quality of life by allowing the intake of more natural foods. A subgroup of patients with PKU, comprising mainly those with mild/moderate disorder, is responsive to pharmacological treatment with BH₄. This can minimise dietary restrictions by allowing higher intake of natural protein⁽ Reference Hennermann, Bührer and Blau ^{86 ,} Reference Trefz, Burton and Longo ^{87 )}; however, there is a risk of micronutrient deficiencies due to the decrease in the intake of vitamin and mineral-supplemented Phe-free l-amino acid mixtures⁽ Reference Thiele, Weigel and Ziesch ^{88 )}. Careful monitoring is still required to ensure that the diet is nutritionally adequate, and long-term longitudinal studies are needed to understand the real impact of BH₄ on general nutritional status.