602

CLINICAL PRACTICE Neurological deterioration in young adults with phenylketonuria

7

patients with phenylketonuria who developed neurological disability in adolescence or early adult life are described. 4 had been diagnosed by routine neonatal screening and started a low phenylalanine diet in infancy. 3 were diagnosed in early childhood because of developmental delay, and then started dietary treatment. Dietary control deteriorated in later years and was withdrawn in mid to late childhood. The late neurological deterioration cannot be directly ascribed to poor compliance with or cessation of dietary treatment in this small, retrospective study—but other likely causes have been excluded and 2 patients showed a striking clinical improvement when a strict diet was resumed. Serial magnetic resonance images from one of these patients show abnormalities that appeared after cessation of dietary treatment and resolved after diet was resumed. If these findings are confirmed, strict dietary control into adult life would be indicated for at least some patients with

phenylketonuria. Introduction

phenylketonuria, caused by phenylalanine hydroxylase deficiency,1 may lead to severe mental retardation, often accompanied by microcephaly and epilepsy.2,3Tremor and pyramidal tract signs, particularly brisk reflexes with or without extensor plantar responses, are found in many patientsand 5% may develop spastic paraparesis.3 Extrapyramidal syndromes, including cogwheel rigidity and choreoathetosis, have also been Untreated

described. 4,5

Susceptibility to the damaging effects of the biochemical disturbance (hyperphenylalaninaemia, reduced tyrosine, and increased phenylketone production) is thought to be greatest in early childhood, during the phase of most rapid brain growth. The introduction of dietary treatment,6 combined with early diagnosis by routine testing in early infancy, strikingly altered the outcome of phenylketonuria in affected children.7 However, many paediatricians stopped dietary treatment in mid or late childhood, because the diet is difficult to follow, may increase the risk of nutritional problems if badly applied, and is expensive. In late treated or untreated patients progression of symptoms during late childhood and adult life was uncommon, but there have

been reports of late onset of spastic quadriparesis and epilepsy,8-lO and pathological studies have indicated that myelin abnormalities--quite distinct from those of early childhood9-may occur in older patients. Reports of falls in IQ after dietary treatment was withdrawn in late childhoodl1,12 have led many clinicians to encourage their patients to continue dietary restrictions into adult life. We describe 7 case histories which support such management.

Patients Patient 1. A woman, now aged 19 years, was diagnosed by routine neonatal screening and started a low phenylalanine diet at 13 days. Dietary control was good for 4 years (blood phenylalanine 60-400 pmol ’/1; normal range 40-80 jmol/1) but growth was poor (height below 3rd centile despite 3-7 kg birthweight). Control was less strict thereafter (400-700 jmol/1 to 10 years; 500-1300 lunolll thereafter). She attended normal school but at 12 years her intellectual ability seemed to deteriorate. At 13 she had 2 grand mal convulsions and was treated with sodium valproate; after the second fit she had mild ataxia and horizontal nystagmus. The next year she had another fit and pyramidal weakness of the right leg was noted; over the next year the ataxia worsened and tone increased in both lower limbs, with brisk reflexes and extensor plantar responses, but gait was not impaired. Routine investigations and cerebrospinal fluid (CSF) examination were normal, an electroencephalogram (EEG) showed generalised spike and slow-wave activity, and computed tomography (CT) showed mild white matter atrophy. She remained stable and stopped the diet at 18 years, after which blood phenylalanine concentration rose to 2000 lunol/1. Patient 2. A man, now aged 21, was diagnosed in infancy by routine urine testing and started dietary treatment at 7 weeks. Blood phenylalanine concentrations in early childhood are not known but he remained on a low phenylalanine diet to 16 years. Despite early treatment he required special schooling and at 16 years was considered unemployable because of low IQ and severe behavioural problems. 2 years later his gait was noted to be abnormal; on examination he had pyramidal weakness of both legs with hyperreflexia and extensor plantar responses. Routine investigations were normal apart from a blood phenylalanine concentration of 1334 umol/1.

ADDRESSES Institute of Neurology, National Hospital for Nervous Diseases, London (A J Thompson, MD, B D Youl, FRACP, B Kendall, FRCR, A J Lees, FRCP); Institute of Child Health, London (I Smith, FRCP); University College and Middlesex School of Medicine, London (D. Brenton, FRCP); Alder Hey Children’s Hospital, Liverpool (G Rylance, MRCP); and Children’s Hospital, Birmingham, UK (D C Davidson, FRCP) Correspondence to Dr A J. Thompson, Institute of Neurology, Queen Square, London WC1 N 3BG, UK.

603

TABLE I-SUMMARY OF PATIENTS WITH PHENYLKETONURIA AND LATE-ONSET NEUROLOGICAL DETERIORATION

Phe=phenylalanine, WISC=Wechsler Intelligence Scale for years), N D =not done *Still

on

Children

(at

age 5

diet

Patient 3. A 24-year-old man, diagnosed in infancy by routine urine testing, started dietary treatment at 6 weeks with good control (blood phenylalanine 60-400 Eunol/1) up to 8 years of age, but poor growth (height below 3rd centile despite 3-6 kg birthweight). The diet was continued until 18 years of age but was less strict after 10 years (600-800 J.l1llol/l, 10-14 years; 900-1100 umol/1 thereafter). Despite early treatment he required special education and at 16 years developed behavioural problems and generalised epilepsywhich was extremely difficult to control despite intensive anticonvulsant therapy-and was found to have intention tremor in both arms, brisk reflexes throughout, and flexor plantar responses. An EEG showed generalised slow-wave activity with occasional spikes and visual evoked responses showed bilateral symmetrical delays in latency, although concentration was poor. Patient 4. A man, now aged 22, was diagnosed by routine blood testing in the first week of life and started dietary treatment with good control for 6 years (blood phenylalanine 180-600 umol/1) but then deteriorated (300-900 lunol/1, 6 to 8 years; 600-1200 umol/1, 8-10 years; 900-1600 mol/1 thereafter). He attended a school for slow learners and diet was stopped at 19 years. Over the next 3 years he developed progressive difficulty walking; when recently referred he had brisk reflexes throughout, flexor plantar responses, and moderate ataxia. Patient 5. A 29-year-old male was diagnosed at 2 years of age when investigated for delayed walking and speech. He started a low phenylalanine diet a year later, but this was stopped after 4 years. Although less intelligent than his siblings, he attended a normal school and began work in a factory at the age of 16. 9 years later he noted that his left foot tended to turn inwards and stiffness in the foot gradually increased over the next 2 years. At 27, he had mild dysarthria with slow tongue movement, a broad-based gait, dystonia of both feet (greater on the left), dystonia of the left hand, generalised hypertonia, mild bilateral weakness of finger abduction, abnormally brisk reflexes with ankle clonus, and a left extensor plantar reflex. He had minimal intention tremor of both arms but no incoordination of the legs and no sensory disturbance. Routine investigations were all normal, including an autoantibody screen TABLE II--BIOCHEMICAL INVESTIGATIONS IN TWO PATIENTS WITH LATE NEUROLOGICAL SYMPTOMS

*0n normal diet, ton strict diet, #over 60% tetrahydrobtoptennn HVA= homo%anillic acid, 5HIAA = 5-hydroxytndoleacetic acid

serum copper and caeruloplasmin concentrations, except for a plasma phenylalanine concentration of 1386 J.ill1ol/l. Examination of the CSF, myelography, electroencephalography, and computed tomography showed no abnormalities. Visual evoked responses showed slightly delayed latencies bilaterally. Patient 6 A woman, now aged 24, was diagnosed at 15 months when investigated for developmental delay, hypersomnolence, and infantile spasms. Dietary treatment was started but control was poor (blood phenylalanine 330-1400 umol/1). She required special education, and diet was stopped at 7 years. She remained stable until aged 20 when she noted progressive difficulty walking and paraesthesia of her left hand. On examination she had spastic quadriparesis, with extensor plantar reflexes and dorsal column sensory loss in the legs. Over the next 6 months her walking deteriorated further, with increased spasticity of her legs such that

and

she became wheelchair-bound at the age of 21. Routine blood and CSF investigations, myelography, and computed tomography were normal except for a plasma phenylalanine concentration of 1533

umol/1. Patient 7. A 25-year-old man was diagnosed at 18 months when investigated for slow development and infantile spasms. Dietary treatment was started with good control to the age of 5 (blood

phenylalanine 180-600 jmol/1) which then deteriorated over the next 2 years (600-1000 umol/1). He attended a school for the severely handicapped and returned to a normal diet at the age of 7. He remained

stable until the age of 15 when he and had generalised hyperreflexia on developed examination. Routine investigations were normal. An EEG showed mild diffuse abnormality.

neurologically

severe tremor

Results Patient details are summarised in table I. 3 patients had first been diagnosed in childhood because of developmental delay. The other 4 had been detected by routine neonatal screening and started on a low phenylalanine diet from infancy; 3 were known to be well controlled on this diet up to 5 years of age by when, despite treatment, all had a low IQ. Dietary treatment continued to mid or late childhood. Late neurological deterioration was observed after cessation of diet in 5 patients and during dietary treatment (but with poorer control than before) in 2. All had signs of upper motor neuron damage; the presenting features were spastic paraparesis or quadriparesis in 5, epilepsy in 1, and a pronounced intention tremor in 1. No patient had any features to suggest defective tetrahydrobiopterin metabolism. In 5 (1, 3, 5-7), dihydropteridine reductase (DHPR) activities were normal (116-170 imol NADH/min per ml) and defective biopterin synthesis was excluded because total plasma biopterins were raised to 11-0-178 ng/ml-as would be expected in the presence of hyperphenylalaninaemia. Patients 1 and 5 underwent more detailed biochemical investigation (table II). CSF concentrations of 5-hydroxyindoleacetic acid (5HIAA, the major metabolite of serotonin) and, to a lesser extent, homovanillic acid (HVA, the major metabolite of dopamine) were reduced. CSF concentrations of total biopterins and total neopterins were raised; most biopterin was in the tetrahydro- form, a result typical of hyperphenylalaninaemia without defective biopterin metabolism. 13 6 patients have been followed for 3-10 years (mean 6) after the onset of late neurological deterioration. 4 were restarted on a low phenylalanine diet, associated with a slow but almost complete recovery in patient 6 (who was wheelchairbound but able to walk long distances unaided 2 years later), diminished behavioural problems in patient 2, and slightly reduced tremor in patient 7. However, patient 7 returned to a normal diet after 3 months. 5 years later (Dec, 1989) he re-attended with a striking clinical deterioration (impaired

604

c

B

A

D

T2-weighted M RIscans (SEZOOO/60) of patient 7 at two levels through the brain. Abnormal high-signal (light) areas, mainly periventricular and posterior, but with some discrete lesions Scan taken at 23 (A), 24 (B), and 25 (C) years of age; and at 2 months after resumption of strict diet (D). Axial

gait, right-sided dystonia, mild pyramidal weakness, increased tone in both legs, and extensor plantar responses) which largely resolved on resumption of a strict diet. Patient 5 showed no clinical improvement on diet and a trial of levodopa had little effect. Of the 2 patients who were still on a relaxed diet when late symptoms appeared, one (patient 1) stopped treatment at the age of 18 and showed no further deterioration 1 year later; the other (patient 3) remained on relaxed diet, with similar fits and behavioural problems 2 years later.

patients underwent Tz-weighted magnetic resonance imaging (MRI; Picker 0-5 tesla). All 6 had abnormal high-signal areas in the cerebral white matter, which varied in size and distribution but tended to be periventricular and more common in the posterior temporal and occipital white matter. Discrete high-signal areas were also seen in some subjects, and in patient 3 extended from the trigones above and laterally to the subcortical white matter in the gyral cores. Two patients had serial scans: in patient 5 there was little change in the scans over 2 years; there had been little change in symptoms and diet had not been changed. However, patient 7, who showed little change between 23 and 24 years of age (figure, A and B, respectively) when symptoms were stable, showed striking MRI changes 6

coincident with obvious clinical deterioration at 25 years (figure, C); after 2 months on a strict low phenylalanine diet symptoms and signs had improved, as had MRI appearances

(figure, D). Discussion

Despite varying periods of dietary treatment during childhood, these 7 patients with phenylketonuria developed neurological disease in either adolescence or early

Signs and symptoms of upper motor neuron dysfunction were a prominent feature, and in the 6 who underwent MRI all had abnormal high-signal areas (which adulthood.

indicates increased water content and/or an alteration in the macromolecular environment of water), restricted to the white matter. None had evidence of abnormal biopterin metabolism, a known cause of progressive neurological deterioration in phenylketonuria.13,14 MRI is a particularly helpful investigation in disorders of white matter, notably multiple sclerosis. In the patients we investigated, the lesions seen on MRI resembled appearances seen in patients with demyelinating disease. For patient 3, in whom the high signal extended to the gyral cores and into the occipital white the matter, pattern was similar to that of adrenoleukodystrophy. Could these similarities throw light upon the mechanisms underlying these disorders of

myelination? The underlying neuropathological changes in untreated phenylketonuria may vary; early reports found no abnormality apart from reduced brain size,16 but an abnormality of myelin is now recognised4,8,17,18 and similarities to leukodystrophies have been noted.18,19 Malamud9 described 4 patients with phenylketonuria who showed a striking neurological deterioration in early adult life, all of whom had extensive myelin loss with relative preservation of axis cylinders. By contrast, in 4 patients who died in childhood the changes were restricted to hypomyelination with numerous vacuoles interspersed among nerve fibres. Malamud suggested that while there was a failure in myelination in early childhood, some survivors developed active demyelination in late adolescence. A major difficulty in the

interpretation of pathological

605

studies is differentiation between primary demyelination or demyelination secondary to Wallerian axonal degeneration ’20 which may have similar appearances on light microscopy. Myelin turnover is increased in the brains of hyperphenylalaninaemic rats,21 and the rate of breakdown of myelin basic protein exceeds that of synthesis. However, the mechanisms by which the metabolic abnormality in untreated phenylketonuria causes intellectual and neurological deficits remain uncertain." In the serial MRI scans of patient 7 (see figure) new abnormalities which occurred in association with clinical deterioration resolved within 2 months of resumption of a strict diet. The most likely explanation for such transient areas of high signal is that they indicate increased extracellular water (oedema), as thought to occur in multiple sclerosis22-although intracellular water and inflammation cannot be excluded. What are the clinical implications of our findings? The answer largely depends on whether the late neurological signs and symptoms were caused by the phenylketonuria. We believe this is the most likely explanation, although a causal relation cannot be established by such a study. The late neurological findings were highly reminiscent of those seen in untreated subjects.3,4 Other likely causes of neurological damage were excluded and reintroduction of a strict diet led to clear improvement in two subjects and possible improvement in another-which implies that the neurological problems were related to metabolic control. Why should these patients with phenylketonuria have late neurological deterioration when most others do not? All 7 had either stopped, or relaxed, dietary treatment many years before their teens-but so have most subjects with phenylketonuria. It may be relevant that although 4 subjects had received a low phenylalanine diet from early infancy with good control up to school age, 3 of them had required special education-which indicates that neurological damage had already occurred in early childhood. In the other 3, the diagnosis of phenylketonuria was first made after neurological impairment had occurred. Children with phenylketonuria who are diagnosed in infancy and receive early dietary treatment still tend to have a lower IQ13,23 and more behavioural abnormalities24,25 than their peers and to have a lower IQ than their siblings.26 However, reduced intellect is particularly striking in subjects with high phenylalanine concentrations during the first few years of life,23 which indicates the importance of strict early dietary control. Is strict dietary control relaxed too soon, and who may be at special risk? The 4 patients detected at infancy were from a group of 912 similar subjects born in the UK between 1964 and 1977 and followed up prospectively by the Phenylketonuria Register,24 which indicates a 0-4% frequency of late neurological complications that arise in adolescence or early adulthood. But symptom-free adolescents and young adults who have returned to a normal diet often have unusually brisk tendon jerks, ankle clonus, and intention tremor13-so mild neurological impairment may be more common in early-treated subjects than is generally realised. Similar MRI abnormalities were recently reported in a group of children with phenylketonuria who did not have overt neurological disease but whose dietary control had been poor.27 The striking neurological features in the patients we describe may simply represent an extreme manifestation of a much more common defect. Identification of the cause may help to determine whether stricter early dietary control is needed, or whether

hyperphenylalaninaemia should be avoided

in adolescence

or

adulthood

by a continued low-phenylalanine diet.

We thank Dr Robert Leeming, General Hospital, Birmingham, for measurement of total biopterins and DHPR. Financial support was given by the Scarfe Trust (A. J. T.) and the Medical Research Council (I. S.).

REFERENCES A. Über Ausscheidung von Phenylbrenztraubensaure in den Ham Stoffwechselanomalie in Verbindung mit Imbezillitat. Z Physiol Chem 1934; 227: 169-87. 2. Jervis GA. Phenylpyruvic oligophrenia: introductory study of 50 cases of mental deficiency associated with excretion of phenylpyruvic acid. Arch Neurol Psychiatr (Chic) 1937; 38: 944-63. 3. Paine RS. The variability and manifestations of untreated patients with phenylketonuria (phenylpyruvic aciduria). Pediatrics 1957; 20: 290302. 4. Cowie VA. Phenylpyruvic oligophrenia. J Ment Sci 1951; 97: 505-31. 5. McCleod MD, Munro JF, Leddingham JG, Farquhar JW. Management of extrapyramidal manifestations of phenylketonuria with L-dopa. Arch Dis Child 1983; 58: 457-58. 6. Bickel H, Gerrard J, Hickmans EM. The influence of phenylalanine intake on the chemistry and behaviour of a phenylketonuric child. Acta Paediatr 1954; 43: 64-77. 7. Hudson FP, Mordaunt VL, Leahy I. Evaluation of treatment begun in the first three years of life in 184 cases of phenylketonuria. Arch Dis Child 1970; 45: 5-12. 8. Bechar M, Bornstein B, Elian M, Sandbank U. PKU presenting with an intermittent progressive course. J Neurol Neurosurg Psychiatry 1965; 28: 165-70. 9. Malamud N. Neuropathology of phenylketonuria. J Neuropathol Exp Neurol 1966; 25: 254-68. 10. Villasana D, Butler IJ, Wiliam JC, Roonngta SM. Neurological deterioration in an adult with phenylketonuria. J Inherited Metab Dis 1.

Folling

1989; 12: 451-59. 11. Cabalska B, Duczynska N, Borzymowska J, et al. Termination of dietary treatment in phenylketonuria. Eur J Pediatr 1977; 126: 253-62. 12. Smith I, Lobascher ME, Stevenson JE, et al. Effect of stopping low phenylalanine diet on intellectual progress of children with phenylketonuria. Br Med J 1978; ii: 723-26. 13. Smith I. The hyperphenylalaninaemias. In: Lloyd JK, Serten CR, eds. Genetic and metabolic disease in paediatrics. London: Butterworths, 1985: 166-210. 14. Smith I, Leeming RJ, Cavanagh NPC, Hyland K. Neurological aspects of biopterin metabolism. Arch Dis Child 1986; 61: 130-37. 15. Ormerod IEC, Miller DH, McDonald WI, et al. The role of NMR imaging in the assessment of multiple sclerosis and isolated neurological lesions: a quantitative study. Brain 1988; 110: 1579-616. 16. Corsellis JAN. The pathological report of a case of phenylpyruvic oligophrenia. J Neurol Neurosurg Psychiatry 1953; 16: 139-43. 17. Alvord EC, Stevenson LD, Vogel FS, Engle RL. Neuropathological fingings in phenylpyruvic oligophrenia (phenylketonuria). J Neuropathol Exp Neurol 1950; 9: 298-310. 18. Poser CM, van Bogaert L. Neuropathologic observations in phenylketonuria. Brain 1959; 82: 1-9. 19. Chrome L. The association of phenylketonuria with leukodystrophy. J Neurol Neurosurg Psychiatry 1962; 25: 149-54. 20. McDonald WI. The effects of experimental demyelination on conduction of peripheral nerve. A histological and electrophysiological study. Brain 1966; 86: 481-500. 21. Hommes FA, Ellen AG, Taylor EM. Turnover of fast components on myelin and myelin proteins in experimental hyperphenylalaninaemia. Relevance to termination of dietary treatment of human phenylalaninaemia. J Inherited Metab Dis 1982; 5: 21-27. 22. McDonald WI, Barnes D. Lessons from magnetic resonance imaging in multiple sclerosis. Trends Neurol Sci 1989; 12: 376-79. 23. Smith I, Beasley M, Ades EA. Intelligence and quality of dietary treatment in phenylketonuria. Arch Dis Child (in press). 24. Smith I, Beasley M. Intelligence and behaviour in children with early treated phenylketonuria. Eur J Clin Nutr 1989; 43: 1-5. 25. Smith I, Beasley M, Wolff OH, Ades EA. Behaviour disturbance in eight year old children with early treated phenylketonuria (PKU). J Pediatr

1988; 112: 403-08. 26. Williamson ML, Koch R, Azen C, Chang C. Correlates of intelligence test results in treated phenylketonuric children. Pediatrics 1981; 68: 161-67. 27. Bick U, Fahrendorf G, Ludolph A, Ullrich K. MR imaging of the brain in patients with hyperphenylalaninaemia. 27th Annual Symposium of Society for the Study of Inborn Errors of Metabolism, Munich, 1989

(abstr p023).