Neuropsychological outcome of experimental manipulation of phenylalanine intake in treated phenylketonuria

The impact of metabolic control on cognition, neurophysiology, and well-being in PKU: A systematic review and meta-analysis of the within-participant literature

Phenylketonuria (PKU) is a metabolic disease where Phenylalanine (Phe) rises much above normal levels. Cross-sectional and correlational studies provide valuable information on the importance of maintaining low blood-Phe to achieve good outcomes, but they may be confounded, at least partially, by differences in participant demographics. Moreover, the effect of Phe at older ages is difficult to ascertain because of strong associations between Phe levels across ages. Within-participant studies avoid confounding issues. We have reviewed these studies. We followed PRISMA guidelines to search the literature for studies reporting the impact of Phe changes within participants. Phe was either increased or decreased through diet relaxation/resumption or through pharmacological interventions. Forty-six separate articles reported, singly or in combination, results on cognition (N = 37), well-being (N = 22) and neurophysiological health (N = 14). For all studies, we established, in a binary way, whether a benefit of lower Phe was or was not demonstrated and compared numbers showing benefit versus a null or negative outcome. We then analyzed whether critical parameters (e.g., length of the study/condition for the change, size of Phe change achieved) influenced presence or absence of benefit. For a subset of studies that reported quantitative cognitive outcomes, we carried out a meta-analysis to estimate the size of change in cognitive performance associated with a change in Phe and its significance. There were significantly more studies with benefits than no benefits, both for cognitive and well-being outcomes, and a trend in this direction for neurophysiological outcomes. The meta-analysis showed a highly significant effect size both overall (0.55) and when studies with adults/adolescents were considered separately (0.57). There was some indication that benefits were easier to demonstrate when differences in Phe were larger and achieved across a longer period, but these effects were not always consistent. These results reinforce results from the literature by demonstrating the importance of lower Phe in children as well as in adolescents and adults, even when confounding factors in group composition are eliminated. The field would benefit from further studies where Phe levels are contrasted within-participants to ascertain how much Phe needs to be changed and for how long to see a difference and which measures demonstrate a difference (e.g., which cognitive tasks).

Dietary interventions for phenylketonuria

BACKGROUND

Phenylketonuria is an inherited disease treated with dietary restriction of the amino acid phenylalanine. The diet is initiated in the neonatal period to prevent learning disability; however, it is restrictive and can be difficult to follow. Whether the diet can be relaxed or discontinued during adolescence or should be continued for life remains a controversial issue, which we aim to address in this review. This is an updated version of a previously published review.

OBJECTIVES

To assess the effects of a low-phenylalanine diet commenced early in life for people with phenylketonuria. To assess the possible effects of relaxation or termination of the diet on intelligence, neuropsychological outcomes and mortality, growth, nutritional status, eating behaviour and quality of life.

SEARCH METHODS

We searched the Cochrane Cystic Fibrosis and Genetic Disorders Group Trials Register comprising references identified from comprehensive electronic database searches, handsearches of relevant journals and abstract books of conference proceedings. Most recent search of the Inborn Errors of Metabolism Trials Register: 30 April 2020.

SELECTION CRITERIA

All randomised or quasi-randomised controlled trials comparing a low-phenylalanine diet to relaxation or termination of dietary restrictions in people with phenylketonuria.

DATA COLLECTION AND ANALYSIS

Two authors independently assessed study eligibility and methodological quality, and subsequently extracted the data.

MAIN RESULTS

We included four studies in this review (251 participants), and found few significant differences between treatment and comparison groups for the outcomes of interest. Blood phenylalanine levels were significantly lower in participants with phenylketonuria following a low-phenylalanine diet compared to those on a less restricted diet, mean difference (MD) at three months -698.67 (95% confidence interval (CI) -869.44 to -527.89). Intelligence quotient was significantly higher in participants who continued the diet than in those who stopped the diet, MD after 12 months 5.00 (95% CI 0.40 to 9.60). However, these results came from a single study.

AUTHORS' CONCLUSIONS

The results of non-randomised studies have concluded that a low-phenylalanine diet is effective in reducing blood phenylalanine levels and improving intelligence quotient and neuropsychological outcomes. We were unable to find any randomised controlled studies that have assessed the effect of a low-phenylalanine diet versus no diet from diagnosis. In view of evidence from non-randomised studies, such a study would be unethical and it is recommended that low-phenylalanine diet should be commenced at the time of diagnosis. There is uncertainty about the precise level of phenylalanine restriction and when, if ever, the diet should be relaxed. This should be addressed by randomised controlled studies; however, no new studies are expected in this area so we do not plan to update this review.

Neuropsychological Profile of Children with Early and Continuously Treated Phenylketonuria: Systematic Review and Future Approaches

OBJECTIVE

To provide a comprehensive systematic review of the literature by examining studies published on all cognitive aspects of children with early and continuously treated phenylketonuria (ECT-PKU) included in the databases Medline, PsycINFO, and PsycARTICLE.

METHOD

In addition to a classical approach, we summarized methodology and results of each study in order to discuss current theoretical and methodological issues. We also examined recent advances in biochemical markers and treatments of PKU, with implications for future research on metabolic control and its role as a determinant of neuropsychological outcome.

RESULTS

Consistent with previous reviews, the hypothesis of a specific and central executive impairment in children with ECT-PKU was suggested. However, findings are inconclusive regarding the nature of executive impairments as well as their specificity, impact on everyday life, persistence over time, and etiology.

CONCLUSION

Given the current state of the science, we suggest future directions for research that utilizes a developmental and integrative approach to examine the effects of recent advances in biochemical markers and treatment of PKU. (JINS, 2019, 25, 624-643).

Efficacy and safety of sapropterin dihydrochloride in patients with phenylketonuria: A meta-analysis of randomized controlled trials

AIMS

The aim of the present meta-analysis was to evaluate the efficacy and safety of sapropterin dihydrochloride in phenylketonuria (PKU) patients.

METHODS

The following databases were searched for randomized controlled trials (RCT) regarding PKU patients treated with sapropterin dihydrochloride: PubMed, Embase, Cochrane Library and clinicaltrials. Two authors independently selected studies, assessed the risk of bias and extracted data. The meta-analysis was performed in RevMan 5.3 provided by the Cochrane Collaboration.

RESULTS

Four studies met the inclusion criteria. In PKU patients with low blood phenylalanine (Phe) concentration, no significant difference was indicated for the decrease of Phe level (weighted mean difference (WMD) = -7.75 μmol L-1 ; 95% confidence intervals (CI): -82.63 to 67.13, P = 0.84, I2  = 0%), however, the dietary Phe tolerance was significantly improved in the sapropterin group (WMD = 19.89 mg kg-1  d-1 ; 95% CI: 10.26 to 29.52, P < 0.0001, I2  = 0%). In PKU patients with high blood Phe level, sapropterin showed a significant lowering in blood Phe concentration (WMD = -225.31 μmol L-1 ; 95% CI: -312.28 to -138.34, P < 0.00001, I2  = 0%). There was no significant difference for adverse events.

CONCLUSIONS

Sapropterin could bring benefit for PKU patients with high or low Phe level, due to Phe reduction in a short time or dietary Phe tolerance improvement respectively. Sapropterin has an acceptable safety profile.

A comparison of phenylketonuria with attention deficit hyperactivity disorder: do markedly different aetiologies deliver common phenotypes?

Phenylketonuria (PKU) is a well-defined metabolic disorder arising from a mutation that disrupts phenylalanine metabolism and so produces a variety of neural changes indirectly. Severe cognitive impairment can be prevented by dietary treatment; however, residual symptoms may be reported. These residual symptoms appear to overlap a more prevalent childhood disorder: Attention Deficit/Hyperactivity Disorder (ADHD). However, the aetiology of ADHD is a vast contrast to PKU: it seems to arise from a complex combination of genes; and it has a substantial environmental component. We ask whether these two disorders result from two vastly different genotypes that converge on a specific core phenotype that includes similar dysfunctions of Gray's (Gray, 1982) Behavioural Inhibition System (BIS), coupled with other disorder-specific dysfunctions. If so, we believe comparison of the commonalities will allow greater understanding of the neuropsychology of both disorders. We review in detail the aetiology, treatment, neural pathology, cognitive deficits and electrophysiological abnormalities of PKU; and compare this with selected directly matching aspects of ADHD. The biochemical and neural pathologies of PKU and ADHD are quite distinct in their causes and detail; but they result in the disorder in the brain of large amino acid levels, dopamine and white matter that are very similar and could explain the overlap of symptoms within and between the PKU and ADHD spectra. The common deficits affect visual function, motor function, attention, working memory, planning, and inhibition. For each of PKU and ADHD separately, a subset of deficits has been attributed to a primary dysfunction of behavioural inhibition. In the case of ADHD (excluding the inattentive subtype) this has been proposed to involve a specific failure of the BIS; and we suggest that this is also true of PKU. This accounts for a substantial proportion of the parallels in the superficial symptoms of both disorders and we see this as linked to prefrontal, rather than more general, dysfunction of the BIS.

Dietary interventions for phenylketonuria

BACKGROUND

Phenylketonuria is an inherited disease treated with dietary restriction of the amino acid phenylalanine. The diet is initiated in the neonatal period to prevent mental handicap; however, it is restrictive and can be difficult to follow. Whether the diet can be relaxed or discontinued during adolescence or should be continued for life remains a controversial issue, which we aim to address in this review.

OBJECTIVES

To assess the effects of a low-phenylalanine diet commenced early in life for people with phenylketonuria. To assess the possible effects of relaxation or termination of the diet on intelligence, neuropsychological outcomes and mortality, growth, nutritional status, eating behaviour and quality of life.

SEARCH STRATEGY

We searched the Cochrane Cystic Fibrosis and Genetic Disorders Group Trials Register comprising references identified from comprehensive electronic database searches, handsearches of relevant journals and abstract books of conference proceedings.Most recent search of the Inborn Errors of Metabolism Trials Register: 05 March 2009.

SELECTION CRITERIA

All randomised or quasi-randomised controlled trials comparing a low-phenylalanine diet to relaxation or termination of dietary restrictions in people with phenylketonuria.

DATA COLLECTION AND ANALYSIS

Two authors independently assessed study eligibility and methodological quality, and subsequently extracted the data.

MAIN RESULTS

We included four studies in this review (251 participants), and found few significant differences between treatment and comparison groups for the outcomes of interest. Blood phenylalanine levels were significantly lower in participants with phenylketonuria following a low-phenylalanine diet compared to those on a less restricted diet, mean difference (MD) at three months -698.67 (95% confidence interval (CI) -869.44 to -527.89). Intelligence quotient was significantly higher in participants who continued the diet than in those who stopped the diet, MD after 12 months 5.00 (95% CI 0.40 to 9.60). However, these results came from a single study.

AUTHORS' CONCLUSIONS

The results of non-randomised studies have concluded that a low-phenylalanine diet is effective in reducing blood phenylalanine levels and improving intelligence quotient and neuropsychological outcomes. We were unable to find any randomised controlled studies that have assessed the effect of a low-phenylalanine diet versus no diet from diagnosis. In view of evidence from non-randomised studies, such a study would be unethical and it is recommended that low-phenylalanine diet should be commenced at the time of diagnosis. There is uncertainty about the precise level of phenylalanine restriction and when, if ever, the diet should be relaxed. This should be addressed by randomised controlled studies.

Efficacy of sapropterin dihydrochloride (tetrahydrobiopterin, 6R-BH4) for reduction of phenylalanine concentration in patients with phenylketonuria: a phase III randomised placebo-controlled study

BACKGROUND

Early and strict dietary management of phenylketonuria is the only option to prevent mental retardation. We aimed to test the efficacy of sapropterin, a synthetic form of tetrahydrobiopterin (BH4), for reduction of blood phenylalanine concentration.

METHODS

We enrolled 89 patients with phenylketonuria in a Phase III, multicentre, randomised, double-blind, placebo-controlled trial. We randomly assigned 42 patients to receive oral doses of sapropterin (10 mg/kg) and 47 patients to receive placebo, once daily for 6 weeks. The primary endpoint was mean change from baseline in concentration of phenylalanine in blood after 6 weeks. Analysis was on an intention-to-treat basis. The study is registered with ClinicalTrials.gov, number NCT00104247.

FINDINGS

88 of 89 enrolled patients received at least one dose of study drug, and 87 attended the week 6 visit. Mean age was 20 (SD 9.7) years. At baseline, mean concentration of phenylalanine in blood was 843 (300) micromol/L in patients assigned to receive sapropterin, and 888 (323) micromol/L in controls. After 6 weeks of treatment, patients given sapropterin had a decrease in mean blood phenylalanine of 236 (257) micromol/L, compared with a 3 (240) micromol/L increase in the placebo group (p<0.0001). After 6 weeks, 18/41 (44%) patients (95% CI 28-60) in the sapropterin group and 4/47 (9%) controls (95% CI 2-20) had a reduction in blood phenylalanine concentration of 30% or greater from baseline. Blood phenylalanine concentrations fell by about 200 micromol/L after 1 week in the sapropterin group and this reduction persisted for the remaining 5 weeks of the study (p<0.0001). 11/47 (23%) patients in the sapropterin group and 8/41 (20%) in the placebo group experienced adverse events that might have been drug-related (p=0.80). Upper respiratory tract infections were the most common disorder.

INTERPRETATION

In some patients with phenylketonuria who are responsive to BH4, sapropterin treatment to reduce blood phenylalanine could be used as an adjunct to a restrictive low-phenylalanine diet, and might even replace the diet in some instances.

Infusionstherapie und Ernährung von Risikogruppen

Besondere Situationen erfordern ein besonderes Vorgehen. Während bisher das »Standardvorgehen« bezüglich der Ernährung von pädiatrischen Patienten dargestellt wurde, beschäftigt sich das vorliegende Kapitel mit »Sondersituationen« der pädiatrischen Infusionstherapie und Ernährung. Behandlungssituationen, die ein besonderes Vorgehen bei der Therapie oder spezielle Aufmerksamkeit bei der Anpassung der Ernährung erfordern, entstehen in der Regel durch 4 mögliche Situationen:

spezifische Physiologie von Patientengruppen (z. B. Früh- oder Neugeborene),

Auswirkungen von therapeutischen Maßnahmen (z. B. Operationen),

Pathophysiologie von Erkrankungen (z. B. angeborene Stoffwechselerkrankungen, Erkrankungen des onkologischen, rheumatischen oder atopischen Formenkreises, Anorexia nervosa, Bulimie oder Adipositas) oder

besondere körperliche Belastungen [z. B. (Leistungs-)Sport].

Bekannte Strategien werden systematisch und prägnant dargestellt und diskutiert.

Die Beschäftigung mit der Ernährung von »Risikogruppen« übt das Erkennen und den Umgang von potenziellen Gefahrensituationen bei der Verordnung von bilanzierter Ernährung. So sollte auch derjenige von dem Kapitel profitieren, der sich mit den behandelten Patientengruppen, Situationen, Erkrankungen üblicherweise nicht beschäftigen muss.

Inborn Errors in Metabolism and 4-Boronophenylalanine–Fructose-Based Boron Neutron Capture Therapy

Infusions of boronophenylalanine-fructose complex (BPA-F), at doses up to 900 mg/kg of BPA and 860 mg/kg of fructose, have been used to deliver boron to cancer tissue for boron neutron capture therapy (BNCT). In patients with phenylketonuria (PKU), phenylalanine accumulates, which is harmful in the long run. PKU has been an exclusion criterion for BPA-F-mediated BNCT. Fructose is harmful to individuals with hereditary fructose intolerance (HFI) in amounts currently used in BNCT. The harmful effects are mediated through induction of hypoglycemia and acidosis, which may lead to irreversible organ damage or even death. Consequently, HFI should be added as an exclusion criterion for BNCT if fructose-containing solutions are used in boron carriers. Non-HFI subjects may also develop symptoms, such as gastrointestinal pain, if the fructose infusion rate is high. We therefore recommend monitoring of glucose levels and correcting possible hypoglycemia promptly. Except for some populations with extremely low PKU prevalence, HFI and PKU prevalences are similar, approximately 1 or 2 per 20,000.

Motor function under lower and higher controlled processing demands in early and continuously treated phenylketonuria

This study examined motor control in 61 early and continuously treated patients with phenylketonuria (PKU) and 69 control participants, aged 7 to 14 years. The pursuit task demanded concurrent planning and execution of unpredictable movements, whereas the tracking task required a highly automated circular movement that could be planned in advance. PKU patients showed significantly poorer motor control in both tasks compared with control participants. Deficits were particularly observed for younger patients (age < 11 years). Differences between control participants and PKU patients were significantly greater in the pursuit task compared with the tracking task, indicating more serious deficits when a higher level of controlled processing is required. Correlations with historical phenylalanine levels indicated a later maturation of the level of control required by the pursuit task compared with the tracking task.

Inhibition of prepotent responding and attentional flexibility in treated phenylketonuria

Inhibition of prepotent responding and attentional flexibility were assessed in 58 early and continuously treated phenylketonuria (PKU) patients and 69 controls, age 7 to 14 years. A computerized task was used requiring participants to process consecutive stimuli according to various attentional sets. Analysis of error rate suggested poorer inhibition of prepotent responding in PKU patients compared with controls. No influence of concurrent plasma phenylalanine (phe) was shown, neither in the younger (age < 11 years) nor in the older participants (age > or = 11 years). Analysis of error rate provided strong evidence for poorer attentional flexibility in PKU patients compared with controls. The difference between attentional flexibility in controls and PKU patients could mainly be attributed to younger PKU patients, with concurrent phe levels higher than 360 micromol/L. Younger PKU patients with phe levels below 360 micromol/L performed at the same level as age-matched controls. Performance of PKU patients was strongly associated with phe levels in age periods during the first 10 years of life, which are characterized by a strong development of executive functioning (ages 2-7 and age 9). High phe levels during these age periods could delay development of inhibitory control and attentional flexibility. With regard to treatment, analyses with lifetime and concurrent phe levels support strict dietary control throughout the first decade of life, after which the phe-restricted diet can be relaxed. However, based on the evidence that development of specific executive functions continues until approximately age 12, it is recommended to maintain phe levels below 360 micromol/L throughout early adolescence.

Short-term dietary interventions in children and adolescents with treated phenylketonuria: effects on neuropsychological outcome of a well-controlled population

This study addressed two questions: is there an effect of dietary interventions that induce relatively small changes in phenylalanine (Phe) concentration on neuropsychological outcome of early- and continuously treated phenylketonuria (PKU) patients, and are there differences in effects for PKU children and adolescents? To answer the first question, the effect of a short-term dietary intervention (1-2 weeks) was compared for patients whose Phe concentrations increased vs those whose Phe concentrations decreased. Controls were tested twice to control for learning effects. To answer the second question, the effect of dietary interventions was examined in younger patients (aged 7-10 years) and older patients (aged 11-14 years). The effect of dietary interventions was determined with three neuropsychological tasks: one requiring sustained attention; a second demanding maintenance in working memory; and a third in which complex operations were performed in working memory. Relatively small fluctuations in Phe concentration were found to influence neuropsychological task performance of PKU patients. Patients whose Phe concentrations decreased at the second assessment showed generally more improvement than controls. Patients whose Phe concentrations had increased showed minimal improvement or deterioration of task performance. The strongest effects were observed when sustained attention and manipulation of working memory content were required. There were some indications of a greater sensitivity of younger patients to fluctuations in Phe concentration.

The neuropsychological profile of early and continuously treated phenylketonuria: orienting, vigilance, and maintenance versus manipulation-functions of working memory

In this paper, we review neuropsychological test results of early and continuously treated Phenylketonuria (PKU) patients. To increase insight into the neuropsychological profile of this population, we have attempted to place the results within an attentional network model [Images of the mind, 1994], which proposes interacting but dissociable attentional networks for orienting, vigilance, and executive control of attention. Executive control of attention is discussed against the background of the process-specific theory of working memory (WM) [Handbook of neuropsychology, 1994], which postulates a distinction between the 'maintenance'-function of WM and the 'manipulation and monitoring'-function. Neuropsychological results are presented for 67 early and continuously treated PKU patients and 73 controls aged 7-14 years. Four neuropsychological tasks were employed to measure orienting, mnemonic processing, interference suppression, and top-down control in visual search. No differences were found in orienting and the maintenance-function of WM. In addition to previously reported impairments in sustained attention/vigilance and inhibition of prepotent responding, PKU patients exhibited deficits when top-down control was required in a visual search task, but showed no impairment when interference suppression was required. It is discussed how the specific neuropsychological impairments in PKU may be a consequence of mid-dorsolateral prefrontal cortex (DLPFC) dysfunctioning due to deficiencies in catecholamine modulation.

Neonatal screening and long-term follow-up of phenylketonuria: the French database

BACKGROUND

In France, neonatal screening of phenylketonuria (PKU) started in 1966. A national association was created in 1978 in order to organise the neonatal screening program and to control the efficacy of the screening and patients' follow-up.

AIMS

To evaluate the results of the French PKU screening program in terms of hyperphenylalaninaemia epidemiology, efficacy of the screening procedure, management and outcome of the patients.

STUDY DESIGN

The national database has been filled-up first with the answers to questionnaires that were sent each year by the PKU patients' physicians, and second with the results of an additional inquiry, which was set up in 1994 in order to investigate diagnosis, treatment, and school outcome of all French PKU patients.

RESULTS

PKU was diagnosed in 81.6% of patients with hyperphenylalaninaemia (HPA), non-PKU HPA in 17.2% and cofactor deficiency in 1.1%. From 1980, incidence of PKU has been stable: 1 per 17,124 live births. Sensitivity of the screening procedure was 99.3%. Age at diet initiation regularly decreased to reach 14 days as a median in 1996. Until 1990, median age at diet discontinuation was 6 years of age. Later, strict diet was continued longer (at least, up to 8-10 years). PKU patients who entered to secondary school at normal age were characterised by an earlier age at diagnosis and at diet initiation and a later age at diet discontinuation, compared to those who entered 1 year or more behind normal age.

CONCLUSION

These data confirm the benefit of a nationwide organised screening program. They emphasise the importance of an early neonatal diagnosis and diet initiation in PKU patients and are consistent with the benefit of a longer period of strict diet in childhood.

Dietary interventions for phenylketonuria

BACKGROUND

Phenylketonuria is an inherited disease for which the main treatment is the dietary restriction of the amino acid phenylalanine. The diet has to be initiated in the neonatal period to prevent or reduce mental handicap however the diet is very restrictive and unpalatable and can be difficult to follow. Whether the diet can be relaxed or discontinued during adolescence or should be continued for life remains a contraversial issue which we aim to address in this review.

OBJECTIVES

To assess the effects of a phenylalanine restricted diet commenced early in life for patients with phenylketonuria. To assess the possible adverse effects of relaxation or termination of the diet on intelligence, neuropsychological outcomes and mortality, and to assess the effect on growth, nutritional status and eating behaviour and quality of life.

SEARCH STRATEGY

We searched the Cochrane Cystic Fibrosis and Genetic Disorders Trials Register which is a specialist trials register which comprises references identified from comprehensive electronic database searches, handsearching relevant journals and handsearching abstract books of conference proceedings. Additional studies were identified from handsearching the Journal of Inherited Metabolic Disease (from inception, 1978 to 1998). The manufacturers of dietary products for phenylketonuria were also contacted. Date of the most recent search of the Group's specialised register: November 1999.

SELECTION CRITERIA

All randomised or pseudorandomised controlled trials comparing a phenylalanine restricted diet to either relaxation or termination of dietary restrictions in patients with phenylketonuria.

DATA COLLECTION AND ANALYSIS

Two reviewers independently assessed the trial eligibility, methodological quality and extracted the data.

MAIN RESULTS

Four studies were included in this review including a total of 251 patients. Few statistically significant differences were found between treatment and comparison groups for any of the outcomes apart from for blood phenylalanine level and intelligence quotient. Blood phenylalanine levels were significantly lower in those subjects with phenylketonuria following a phenylalanine restricted diet compared to those on a less restricted or relaxed diet (weighted mean difference at three months -672.203, 95% Confidence interval (CI) -813.799 to - 530.608). Intelligence quotient was significantly higher in subjects who continued on the phenylalanine restricted diet compared to those who terminated the diet (weighted mean difference after 12 months -5.00, 95% CI -9.595 to -0.405). However this is based on the results of only one study.

REVIEWER'S CONCLUSIONS

The results of non-randomised studies have concluded that a phenylalanine restricted diet is effective in reducing blood phenylalanine levels and improving intelligence quotient and neuropsychological outcome. No randomised controlled trials have assessed the effect of a phenylalanine restricted diet versus no dietary restrictions from diagnosis. In view of evidence from non-randomised studies, such a trial would now be unethical and it is recommended that phenylalanine restricted diet should be commenced at the time of diagnosis. There is uncertainty about the precise level of phenylalanine restriction and when, if ever, the diet should be relaxed. These questions should be addressed by randomised controlled trials with careful consideration given to which patients to include.