Hepatorenal Tyrosinaemia: Impact of a Simplified Diet on Metabolic Control and Clinical Outcome

Background: Tyrosinaemia type 1 is a rare inherited metabolic disease caused by an enzyme defect in the tyrosine degradation pathway. It is treated using nitisinone and a low-protein diet. In a workshop in 2013, a group of nutritional specialists from Germany, Switzerland and Austria agreed to advocate a simplified low-protein diet and to allow more natural protein intake in patients with tyrosinaemia type 1. This retrospective study evaluates the recommendations made at different treatment centers and their impact on clinical symptoms and metabolic control. Methods: For this multicenter study, questionnaires were sent to nine participating treatment centers to collect data on the general therapeutic approach and data of 47 individual patients treated by those centers. Results: Dietary simplification allocating food to 3 categories led to increased tyrosine and phenylalanine blood concentrations without weighing food. Phenylalanine levels were significantly higher in comparison to a strict dietary regimen whereas tyrosine levels in plasma did not change. Non-inferiority was shown for the simplification and liberalization of the diet. Compliance with dietary recommendations was higher using the simplified diet in comparison to the stricter approach. Age correlates negatively with compliance. Conclusions: Simplification of the diet with increased natural protein intake based on three categories of food may be implemented in the diet of patients with tyrosinaemia type 1 without significantly altering metabolic control. Patient compliance is strongly influencing tyrosine blood concentrations. A subsequent prospective study with a larger sample size is necessary to get a better insight into the effect of dietary recommendations on metabolic control.


Introduction
Tyrosinaemia type 1 (Tyr 1) is a rare inherited metabolic disease. With an estimated incidence of 1:120,000 in Europe it belongs to the group of orphan diseases [1]. Tyr 1 is caused by an enzyme defect of fumarylacetoacetate hydrolase (FAH) in the tyrosine degradation pathway. The defect of FAH leads to the accumulation of tyrosine and its toxic metabolites such as succinylacetone (SA). SA serves as a surrogate parameter of toxicity. Toxic metabolites cause organ damage especially to the kidneys, liver and nervous system [1,2]. Affected patients show an increased incidence of impaired cognitive function and compromised intelligence [3][4][5]. The therapy of Tyr 1 consists of nitisinone intake (2-(2-nitro-4-trifluormethylbenzoyl)-1,3-cyclohexanedione, NTBC), low-protein diet and a supplemental amino acid mixture [6][7][8]. Nitisinone is an inhibitor of the tyrosine degradation pathway and therefore leads to an increase of tyrosine blood concentrations [9]. Phenylalanine may be supplemented if needed. Van Ginkel et al. suggested that high phenylalanine concentrations might protect from neurocognitive impairments, higher concentrations than generally assumed may be needed to normalize cerebral phenylalanine levels [10].  2017)) in an attempt to standardize the treatment of Tyr 1. However, different approaches are still followed regarding dietary therapy and monitoring of Tyr 1 at the different metabolic centers due to a lack of scientifically based guidelines [11]. In some centers, a very restrictive diet is recommended, whereas in other centers food with a higher protein content is allowed. Furthermore, it is unclear, whether a single daily dose of nitisinone is sufficient and what plasma levels of nitisinone are required to suppress toxic metabolites [12,13]. The aim of our study was to assess the impact of a liberalized and simplified diet on metabolic control.

Materials and Methods
In 2013, data from 42 patients were analyzed and discussed during a workshop by pediatricians and nutritionists from pediatric units in Germany (n = 8), Switzerland (n = 1) and Austria (n = 2) (see Appendix A). Participants had identified a major problem with dietary compliance in Tyr 1 patients, particularly those prescribed a strict diet. In 2013, a considerable number of 17 out of 42 patients/families (40%) were weighing foods to calculate protein intake (see Appendix A). Many patients did not consume low-protein food (40%) but inappropriate items like regular pastries and pasta (43%), milk and nuts ( Figures A2 and A3). Compliance was much better in younger children whose diet was controlled by the parents than in older children over 7 years of age ( Figure A4).
The workshop's aim was to improve dietary compliance by encouraging families and patients not to calculate tyrosine, phenylalanine or protein intake, not to weigh foods and allow moderate protein foods according to appetite. Food was divided into three groups according to protein content: low-protein food (unrestricted), moderate-protein food (occasionally appropriate) and protein-rich food (inappropriate). There was consensus that meat, sausages and eggs were inappropriate, whereas fruits, vegetables, potatoes, rice, butter, oil, cream, crème fraiche, sour cream, low-protein food (e.g., bread, pasta, pastries, milk), rice and oat drinks were regarded as appropriate. There was no general consensus on normal milk, yoghurt, cheese, normal bread, pasta and pastries, legumes, nuts and seeds, so it was agreed that these may be taken occasionally, i.e., not more than twice a week in low to moderate amounts (Table 1). More moderate-protein food could be allowed on an individual basis if blood levels of tyrosine were in the therapeutic range (<400 µmol). It was assumed, that these relaxed recommendations result in higher dietary compliance and quality of life without hampering metabolic control. So far, it has not been assessed which metabolic centers in Switzerland, Austria and Germany (DACH region) have changed their recommendations in response to the workshop and how these changes affected metabolic control. The present survey compared different treatment strategies and evaluated their impact on the patient's clinical condition and metabolic control in terms of tyrosine and phenylalanine concentrations in blood. The aim was to verify the workshop's hypothesis that not calculating natural protein intake and a relaxed, simplified diet results in increased compliance and is not inferior to a strict diet in terms of metabolic control.
The participating centers were asked to complete questionnaires for the 2011-2013 (prior to the workshop) and 2016-2019 periods (post-workshop) including the following items: (1) general dietary recommendations in the center, (2) long term tyrosine and phenylalanine blood concentrations in individual patients, (3) nitisinone dose and compliance to medication and (4) individual patient compliance in terms of dietary restrictions (simplified versus strict) and medication intake.
The primary outcome parameters were mean blood concentrations of phenylalanine and tyrosine taken every 2-3 months in younger patients and about every 3-6 months in older patients. In some older patients with good metabolic control, blood samples were only taken once per year. The influence of the dietary recommendations on the outcome measures was analyzed using linear regression. Other possible confounding variables such as nitisinone intake, compliance, date of birth and age at diagnosis were analyzed separately. A t-test for dependent samples was performed if patient data were available for both 2011-2013 and 2016-2019 in order to analyze the impact of altered recommendations on the outcome measures.
The maximum and minimum tyrosine and phenylalanine concentrations in the respective time periods were recorded. Data were retrospectively collected by treating physicians and nutritionists from medical records. Compliance was assessed retrospectively by dieticians, nutritionists or the treating physician using a 10-point Likert scale. Patients were grouped by their type of dietary recommendations (strict or simplified). The significance level was set at alpha = 5%.
A positive ethical vote was obtained from the ethical review board of Hanover Medical School as the master commission (EC no. 8264_BO_K_2019). Ethical votes were sought from local ethical bodies of the other participating centers as required.

Results
Nine centers from Germany, Austria and Switzerland completed the questionnaires for a total number of 47 patients diagnosed with Tyr 1. Each center included an average of 5.2 patients (min: 2, max: 10 patients). 64% of the patients (n = 30) were male, 36% (n = 17) female; mean year of birth: 2005 (min: 1988, max: 2018, std.-dev.: 7.4). The mean age at diagnosis was 5.7 months (min: 0, max: 36 months, std.-dev.: 7.7). Three of the patients received liver transplantation, so they were excluded from the evaluation of tyrosine and phenylalanine blood concentrations. The following sections give an overview of the most important findings.

Dietary Recommendations and Target Parameters
The target values for tyrosine ranged from <400 to <800 µmol/L. Six of the centers recommended tyrosine levels of <400, two <500 µmol/L and one center aimed for values <800 µmol/L, phenylalanine values should be in the reference range (30-240 µmol/L, depending on age and laboratory). These recommendations have not changed compared to the period before 2013. All centers recommended a division of the amino acid mixture into three doses per day.
The diet in an individual center was simplified when foods were classified in three categories: unrestricted, occasionally appropriate and inappropriate. All participating centers were asked to categorize the following 21 food items: vegetables, fruits, cream, yoghurt/milk, cream cheese, regular pastries, potato/rice, butter/oil/mayonnaise, egg in food, meat/sausage, cheese, fish/seafood, nuts, rice drink/oat drink, egg, poultry, low protein special foods, low protein milk, sugar-rich drinks, food with <2 g protein/100 g and food with 2-6 g protein/100 g. The more foods were categorized as unrestricted, the more relaxed the diet.
Only four of the nine participating centers changed their food recommendation after the workshop in 2013, three markedly relaxed them, one center did no longer recommend protein calculation. Protein in meals was either estimated or food was weighed and protein content calculated by patients and families. None of the centers recommended calculation of phenylalanine in meals. Table 2 gives an overview of protein calculation, phenylalanine and tyrosine target values and diet for each center. Center 9 clearly stands out from the others as meat, sausage, poultry, fish and cheese were classified as occasionally appropriate while the other centers classified these as inappropriate. Interestingly, Center 9 tolerated tyrosine levels up to 800 µmol/L, much higher than in the other centers, both patients treated by Center 9 had tyrosine levels in their respective therapeutic range.
From 2016-2019 centers 2, 6 and 8 achieved the lowest mean tyrosine blood concentrations (447, 476 and 464 µmol/L, respectively). Based on these low levels, centers 6 and 8, relaxed the diet by allowing regular pastries. This shows that the recommendations given in Table 1 can be relaxed on an individual basis without losing metabolic control. Protein intake via amino acid mixture was recommended by dieticians according to DACH 2015 (33%) or DGE 1985 (44%). All centers suggested to divide amino acid mixture into three doses per day.

Medical Formula Intake and Compliance
Compliance for each patient was judged by the treating physician or nutritionist for diet and medication intake on a 10-point Likert scale, respectively. Compliance with medication was significantly higher (average of 9.0 points) compared to dietary compliance (6.5 points) in the 2016-2019 period. There were considerable differences between centers ranging from 4 points to 10 points. The dietary compliance was at all times higher in the group receiving simplified dietary recommendations (diet 2011-2013: 7.9 vs. 5.9 points with strict diet, 2016-2019: 6.9 vs. 6.1 points with strict diet). Using linear regression, age was found to have a significant impact on compliance (r = 0.48; p = 0.001).

Patient Data Laboratory Findings
After the newborn period, patients below the age of 12 months were seen twice a year by all centers for blood collection and dietary counseling. One center reduced these appointments to one per year for patients > 1 year, two more centers for patients > 7 years. 5 centers controlled their patients twice a year even above the age of 18 years. The mean tyrosine and phenylalanine blood concentrations are shown in Tables 3 and 4. Mean plasma tyrosine concentrations were above the target values of the respective center (Table 2).   In the 2016-2019 period mean phenylalanine concentrations were significantly higher in the group treated with a simplified diet (std. dev.: 9.1; CI: [2.35-39.08]; p: 0.028) indicating that the amount of natural protein was increased compared to a strict diet ( Table 4). The mean tyrosine blood concentrations in the period 2016-2019 did not differ significantly between the groups with a strict or simplified/relaxed diet (std. dev.: 46.32; CI: [−93.17-93.79]; p = 0.995) ( Table 4).
Patients who constantly received strict dietary recommendations (n = 14) and those who constantly received simplified dietary recommendations (n = 6) during the whole observation period did not show significant changes of tyrosine and phenylalanine blood concentrations, compliance or nitisinone intake.  Neurocognitive impairment was observed in 10 patients though no formal testing was done. This clinical feature has previously been linked to elevated tyrosine and reduced phenylalanine levels [4,10]. In our study, there was no correlation of high tyrosine and low phenylalanine levels with poor neurocognitive outcome as judged by the health care providers.

Discussion
Non-inferiority was shown for dietary treatment of Tyr 1-patients with a simplified diet based on three categories of food. There were no significant differences between mean tyrosine concentrations in 2016-2019 for strict (mean: 5551 µmol/L) and relaxed diet (mean: 552 µmol/L). For the centers that did change from a strict to a more relaxed diet there was a significant increase in tyrosine concentrations but to a level of 580 µmol/L in line with both the strict and relaxed diets. Strict dietary recommendations did not lead to significantly lower tyrosine levels compared to the group receiving simplified recommendations. This could be due to lower compliance in the group under strict diet. The group receiving strict dietary recommendations were prescribed the highest average dose of NTBC (0.92 mg/kg body weight per day in comparison to 0.63 mg/kg body weight per day in the group receiving simplified recommendations).
Not all centers which participated in the workshop in 2013 switched from a strict diet to a relaxed diet. The treating physicians and nutritionists may have been reluctant to change the diet in an attempt to maintain optimal metabolic control. In the period 2011-2013 tyrosine levels were even lower under a simplified diet (Figure 1a). It is conceivable that only patients with low mean tyrosine plasma levels chose to relax their dietary recommendations after the workshop in 2013. This may explain the rise in tyrosine levels after the switch.
Overall, a simplification of the dietary recommendations was only implemented by five centers. Centers sticking to a stricter diet also used higher amounts of nitisinone which may reflect a more cautious approach in these centers. In addition to the medical aspects, the quality of life should also be considered. A greater choice of food facilitates social participation. Furthermore, fewer special low-protein foods have to be purchased, which is a financial relief.
Phenylalanine levels in the 2016-2019 period were higher under a simplified diet, showing again non-inferiority of a relaxed diet. Increased phenylalanine levels are thought to be protective in terms of neurocognitive function [10]. After the change of diet, phenylalanine blood concentrations were predominantly within the normal range.
A higher compliance to diet was found in the group receiving a simplified/relaxed diet compared to patients on a strict diet in the 2011-2013 period. In the 2016-2019 period the compliance decreased in the group of patients that switched to a relaxed diet after 2013. We expected the compliance to increase after relaxation of the diet. Since age is a confounding factor for compliance this might be one explanation for this surprising result. On the other hand, compliance was subjectively judged by healthcare professionals who may interpret simplified diet as poor compliance. As could be expected, pharmacological compliance was better compared to dietary compliance. A strong correlation was found between patients' adherence to diet and tyrosine blood concentrations. Already in the workshop in 2013, poor dietary compliance was identified by the participating dieticians and nutritionists as a major unmet need in the treatment of tyrosinaemia patients. Therefore, the main focus should be on increasing compliance to achieve improved metabolic control rather than further restricting food choices. Protein tolerance has been shown to increase with age which argues in favor of a simplified and relaxed diet especially in adolescents and adults [14].
There were some limitations in our study. Since the information on dietary compliance was collected by a questionnaire, it is possible that the results are biased. It can be assumed, that the medical personnel answering the questionnaire did not apply the 10-point Likert scale, used to asses compliance, uniformly. Compliance might have been interpreted differently by the centers. Data were collected for the last 9 years, therefore not all details may have been recalled correctly. Another difficulty was caused by the frequency of clinical appointments. Older patients were seen less frequently than younger patients (once or twice per year). The laboratory findings recorded at the outpatient visit do not necessarily reflect the situation of the last 6-12 months at home. It was not recorded how much amino acid mixture the patients took per kg bodyweight. Accordingly, this aspect was not considered in the evaluation, but could have a relevant influence on metabolic control and outcome. Another confounding factor was age. In this retrospective study, dietary assessments were not performed at standardized intervals. Therefore, the exact amount of appropriate, inappropriate and occasionally appropriate foods could not be quantified.
The study population of 44 patients included in the calculations is relatively large for a study on an ultrarare orphan disease. Still, the small sample size with at times only two patients per metabolic center and the retrospective character of this survey are clear limitations of our study. In the future, prospective studies in a larger group of hepatorenal tyrosinaemia patients should be performed including quality of life issues. The present survey was neither controlled nor randomized. In future studies, a waiting-control group could be implemented following an unaltered diet before being randomized to strict and relaxed diet groups. This would also avoid significant age differences between relaxed and strict diet groups. Age can be a confounder in terms of protein tolerance [14] and compliance.
In conclusion, this study shows non-inferiority of a simplified diet with 3 categories compared to a strict diet in terms of tyrosine levels. We advocate that neither weighing of foods nor calculation of protein or tyrosine intake is necessary. Phenylalanine levels were even higher which may have a positive impact on neurocognitive development. Simplification of dietary recommendations may improve the compliance and quality of life of patients and their families. Nonetheless, relaxation of the diet should be regularly monitored in order to assess metabolic control and not to risk organ dysfunction as well as malignancy.
are inappropriate, whereas butter, oil, cream, crème fraiche, sour cream, vegetables, fruits, low-protein bread, pasta, pastries as well as low-protein milk are regarded as appropriate. There was no general consensus on normal milk, yoghurt, cheese, normal bread, and pastries, legumes, nuts and seeds, so it was agreed that these may be taken occasionally. In 10/39 cases, families did not calculate tyrosine, whereas no patient calculated phenylalanine intake. In 17/39 families, food was weighed and calculated and 9 families stated to calculate protein intake ( Figure A1a,b). 19/39 cases divided amino acid mixture in 3 portions. Two patients took the amino acid mixture in a single portion, whereas 14 patients divided the daily intake in two portions. Special low-protein products were taken by 15 patients regularly. 15 patients stated to take them irregularly, whereas 6 patients rarely or never used low-protein products ( Figure A2). Normal bread was eaten by 21 patients, 11 patients consumed normal bread irregularly and 7 rarely or never ate normal bread. Normal pastries were taken rarely or never by 18 patients, 11 patients took normal pastries regularly and 7 cases irregularly. 25 patients consumed protein-rich food rarely or never while 3 patients ate it regularly and 11 families admitted an irregular intake ( Figure A3).  Adherence to diet was found in 23 patients below the age of 7 years, 8 families did not consistently stick to diet and one case was not compliant. 6 patients beyond the age of 7 years showed no compliance, whereas in 13 cases an inconsistent adherence to diet was observed. 7 children older than 7 years of age regularly adhered to the diet. Regular intake of nitisinone was reported in 35/42 patients, 5 patients showed an irregular intake of nitisinone and one patient had no compliance towards medication ( Figure A4).