Eur J Pediatr (2008) 167:305–309
`DOI 10.1007/s00431-007-0486-z
`
`ORIGINAL PAPER
`
`Hypothesis: proposals for the management of a neonate
`at risk of hyperammonaemia due to a urea cycle disorder
`
`James Vivian Leonard & Martin Peter Ward Platt &
`Andrew Alan Myles Morris
`
`Received: 15 January 2007 / Revised: 19 March 2007 / Accepted: 20 March 2007 / Published online: 14 April 2007
`# Springer-Verlag 2007
`
`Abstract It is difficult to prevent hyperammonaemia in
`patients with urea cycle disorders that present
`in the
`newborn period. This is true, even if treatment is started
`prospectively because of an affected relative. We propose
`several additional measures that could be used in conjunc-
`tion with conventional therapy to improve the metabolic
`control. Catabolism could be reduced by delivering the
`babies by elective caesarean section, by starting intravenous
`glucose immediately after delivery and, possibly, by using
`"-blockers or octreotide and insulin. The effectiveness of
`sodium benzoate and sodium phenylbutyrate might be
`increased by giving phenobarbital
`to the mother before
`delivery and subsequently to the baby to induce the
`enzymes for conjugation. We would expect the proposed
`measures to reduce the risk of hyperammonaemia and to
`improve the outcome for these patients. They have not,
`however, previously been used in this context, so families
`would need to be counselled carefully and controlled
`studies should be undertaken.
`
`J. V. Leonard
`Clinical and Molecular Genetics Unit, Institute of Child Health,
`University College London,
`London, UK
`
`M. P. Ward Platt
`Neonatal Unit, Royal Victoria Infirmary,
`Newcastle-upon-Tyne, UK
`
`A. A. M. Morris (*)
`The Willink Biochemical Genetics Unit,
`Royal Manchester Children’s Hospital,
`Pendlebury, Manchester M27 4HA, UK
`e-mail: Andrew.Morris@cmmc.nhs.uk
`
`Keywords Hyperammonaemia . Catabolism .
`Sodium benzoate . Octreotide . Phenobarbital
`
`Abbreviations
`ACTH Adrenocorticotrophic hormone
`CPS
`Carbamyl phosphate synthetase
`ECMO Extracorporeal membrane oxygenation
`NAG
`N-aceylglutamate
`OTC
`Ornithine transcarbamylase
`
`The prognosis for urea cycle disorders presenting in the
`neonatal period is generally poor [16, 24]. In one small
`study, results were less bad for a heterogeneous group of
`patients that
`included organic acidaemias [18]. Other
`studies have suggested that even moderate hyperammonae-
`mia is very commonly associated with a poor outcome [3,
`14, 17]. The outcome of urea cycle disorders has recently
`been reviewed [6]. Until more detailed information is
`available about ways of
`reducing the morbidity and
`mortality, every effort should be made to keep the plasma
`ammonia concentration as low as possible. It is, however,
`unusual to make the diagnosis clinically before the patient
`already has marked hyperammonaemia (> 400 μmol/l).
`Prospective treatment is possible if a baby is known to
`be at risk of neonatal hyperammonaemia because of the
`family history. The history will generally be of hyper-
`ammonaemia in a sibling but, for ornithine transcarbamy-
`lase deficiency, more distant
`relatives are relevant.
`Outcomes have always been better for prospectively treated
`patients [15]. Nevertheless, it can still be difficult to keep
`the plasma ammonia concentration below 400 μmol/l with
`conventional therapy, even if started prospectively. This is
`particularly true for carbamyl phosphate synthetase (CPS)
`
`

`

`306
`
`Eur J Pediatr (2008) 167:305–309
`
`deficiency (OMIM #237300) and ornithine transcarbamy-
`lase (OTC) deficiency (OMIM #311250).
`In this paper we suggest some new ideas about
`management of babies at risk of having a urea cycle disorder
`that may help to control hyperammonaemia.
`
`Metabolic events around the time of birth
`
`In utero, the fetus receives a constant supply of glucose,
`which crosses the placenta down a concentration gradient,
`facilitated by the GLUT1 glucose transporter. Fetal insulin
`concentrations are high and glucagon concentrations are
`low [12]. Around the time of birth, babies show a number
`of hormonal changes. First, catecholamines and cortisol are
`released in response to the stress of the birth process.
`Subsequently,
`there is the hormonal
`response to the
`interruption of the supply of glucose from the placenta.
`This response has two phases: within the first postnatal
`half-hour,
`the physiological
`fall
`in the blood glucose
`concentration triggers a surge in the glucagon concentration
`and a fall in the insulin concentration [12, 22]. There is then
`a more prolonged phase of catabolism,
`the severity of
`which depends on the length of time to reach adequate
`volumes of feed.
`These changes increase the rate of proteolysis, gluco-
`neogenesis and the urea cycle so that
`inborn errors of
`catabolic pathways commonly present
`in the immediate
`post-natal period, even if milk feeds are withheld.
`
`Investigations
`
`Investigation of the new baby is simplified if the diagnosis
`is known in the index case. If the index case died, the
`diagnosis may be uncertain: further studies should be
`undertaken urgently if appropriate samples have been
`saved.
`If there is a risk of neonatal hyperammonaemia, it is
`extremely helpful to know whether the baby is affected
`before birth so that appropriate plans can be made. Affected
`babies should be transferred in utero or soon after delivery
`to a centre with facilities for full intensive care, whereas
`babies known to be unaffected can stay locally and be
`treated normally.
`If prenatal diagnosis has not been
`undertaken earlier, the mother may agree to amniocentesis
`late in pregnancy. In some cases, however, it is not possible
`to determine whether the fetus is affected (for example in
`CPS or OTC deficiencies if molecular genetic studies have
`not been undertaken on the index case). In these cases,
`serial ammonia measurements and specific tests will be
`needed following delivery.
`
`Management
`
`Intensive care
`
`The standard management of a hyperammonaemic neonate
`is discussed elsewhere [23] and will only be summarised
`here. As any patient at risk of severe hyperammonaemia
`may deteriorate rapidly, careful thought should be given to
`their transfer to ICU. If there is any doubt, patients should
`be ventilated electively with good vascular access, usually
`via a multi-lumen central
`line. Initially babies may be
`dehydrated but rehydration should be cautious because of
`the sodium load from the medication and the risk of
`cerebral oedema.
`
`Feeding and intravenous therapy
`
`Milk feeds are probably less significant than endogenous
`catabolism in precipitating hyperammonaemia. Neverthe-
`less, protein should be withheld initially in a baby at risk of
`a severe urea cycle disorder (i.e.,
`if the index patient
`presented within the first few days). Unless there is severe
`hyperammonaemia (>200 μmol/l), however, some milk
`should gradually be introduced, starting within about 48
`hours of birth. Withholding protein for longer will only
`exacerbate the catabolic state.
`If the index patient presented after the first week, the
`new baby should be given the minimum safe level of
`protein intake from birth (approximately 1.5 g/kg/day).
`Breast feeding should be allowed under these circumstances
`with top-up feeds of a low protein formula over the first 48
`hours to minimise catabolism. It is particularly difficult to
`decide on the intake for females with OTC deficiency as the
`severity of their disorder cannot be predicted: a few females
`with unfavourable lyonisation will develop neonatal hyper-
`ammonaemia. Careful monitoring of the plasma ammonia
`concentration is essential, and if
`there is any doubt
`treatment with drugs should be started (see below).
`
`Prevention of catabolism: Mode of delivery
`
`The stress of birth is the first perinatal stimulus to
`catabolism, associated with the release of cortisol and
`catecholamines. Hyperammonaemia does not occur till a
`number of hours after delivery but once catabolism has
`started it may be difficult to reverse in babies with severe
`urea cycle disorders. Delivery by an elective caesarean
`section is associated with lower catecholamine and cortisol
`levels compared to vaginal delivery [4] and this is the only
`process known to reduce the stress associated with being
`born. If a fetus is known to have a severe urea cycle
`disorder, the option of elective caesarian section should be
`discussed with the parents, along with the plans for
`
`

`

`Eur J Pediatr (2008) 167:305–309
`
`307
`
`subsequent neonatal management, so that they can make an
`informed choice.
`
`Prevention of catabolism: Intravenous glucose
`
`The next stimulus to catabolism is the fall in blood glucose
`once the placental supply is interrupted. This stimulus to
`catabolism could largely be prevented by an intravenous
`infusion of glucose at 6 to 8 mg/kg/min. The infusion
`would, however, have to start immediately after delivery,
`for example, through an umbilical catheter, because the
`hormonal mediators of catabolism are released with 30
`minutes of delivery (see above). Though glucose can be
`given orally (for example, as a soluble glucose polymer), it
`is not possible to rely on the intake being sufficient to
`prevent catabolism completely. Moreover, small babies may
`only tolerate relatively small quantities without developing
`diarrhoea. If hyperglycaemia develops, it is important to
`start an insulin infusion rather than cutting back on the
`glucose infusion. The energy intake could be increased by
`adding a fat emulsion, either orally or intravenously. Long
`chain fatty acids are important
`in activating carnitine
`the rate limiting step in β-
`palmitoyltransferase type I,
`oxidation. Fat emulsions have not been used extensively in
`neonates with inborn errors but they should be safe in urea
`cycle disorders.
`
`Pharmacological prevention of catabolism
`
`There have been no published reports of babies with urea
`cycle disorders managed prospectively with all the aggres-
`sive strategies discussed above. It is, therefore, uncertain
`whether these measures will completely prevent catabolism.
`Moreover,
`there will be some cases in which these
`measures cannot be implemented fully. It would, therefore,
`be helpful to have drugs that could prevent catabolism.
`Adrenergic blockade has been shown to reduce catabolism
`in patients with severe burns [8] and it may have a role in
`neonates with inborn errors of metabolism.
`We have already mentioned the need for insulin if
`glucose infusions lead to hyperglycaemia. There may also
`be a role for octreotide combined with insulin. Octreotide is
`a long acting analogue of somatostatin: both agents
`suppress insulin secretion and octreotide is now widely
`used for
`the neonatal and long term management of
`congenital hyperinsulinism [10]. It is given as a continuous
`subcutaneous (or intravenous) infusion in doses of 5–
`25 mcg/kg/day and it appears to be safe in these babies.
`Octreotide suppresses the secretion of glucagon, growth
`hormone and ACTH in addition to insulin. If given alone,
`octreotide would probably increase protein catabolism by
`lowering insulin levels. If it is combined with infusions of
`glucose and insulin, however, octreotide is likely to
`
`its effects on
`suppress protein catabolism because of
`glucagon, growth hormone and cortisol. Evidence support-
`ing this has been obtained in adults with multiple organ
`dysfunction syndrome [2].
`
`N-carbamylglutamate
`
`Flux through the urea cycle is regulated by the synthesis of
`N-acetylglutamate (NAG), an allosteric activator of CPS.
`NAG synthase deficiency is a rare cause of neonatal
`hyperammonaemia but an important one to consider, because
`patients respond to treatment with N-carbamylglutamate, an
`analogue of NAG [21]. N-carbamylglutamate can also lower
`ammonia concentrations in some patients with partial CPS
`deficiency and organic acidaemias [5, 13]. If a neonate has
`died of hyperammonaemia without these conditions being
`excluded, N-carbamylglutamate should be given to future
`siblings, either prophyllactically from birth or immediately if
`ammonia concentrations start to rise.
`
`Alternative pathways for nitrogen excretion: inducing
`the enzymes of conjugation
`
`Large doses of arginine are used in patients with citrulli-
`naemia and argininosuccinic aciduria to promote nitrogen
`excretion as citrulline or argininosuccinate. Sodium benzo-
`ate and sodium phenylbutyrate (or sodium phenylacetate)
`are used in all severe urea cycle disorders to exploit
`alternative pathways for nitrogen excretion [23]. At the
`recommended doses, these drugs appear to be safe and
`reasonably effective in older subjects. There are, however,
`few published data about plasma concentrations, the rates
`of conjugation and nitrogen elimination in neonates. Neo-
`nates are usually given loading doses followed by contin-
`uous infusions. Despite these, the available data suggest
`that maximal rates of conjugation are not achieved for
`several days. Figure 1 shows the concentrations of
`ammonia, benzoate and hippurate over the first two weeks
`in a neonate with argininosuccinic aciduria: maximal
`hippurate concentrations are only reached on day 9. The
`conversion of benzoate to hippurate depends first on
`benzoate:CoA ligase and then on benzoyl-CoA:glycine N-
`acyltransferase. In neonatal rodents, these enzyme cannot
`be induced by benzoate itself but they can be induced by
`phenobarbital [1, 19]. The rate of conjugation in neonates
`with urea cycle disorders might, therefore, be increased by
`giving phenobarbital as well as sodium benzoate. Pheno-
`barbital has been used extensively in neonates but its use in
`this context should, of course, be closely monitored to
`establish whether the observations in rodents are also true
`in humans.
`Ideally, one would like the conjugating enzymes to have
`maximal activity at the time of birth. It may be possible to
`
`

`

`Eur J Pediatr (2008) 167:305–309
`
`Hippurate
`
`Benzoate
`
`Ammonia
`
`10000
`
`1000
`
`100
`
`10
`
`Concentrations µmol/l
`
`308
`
`Fig. 1 Serial concentrations of
`ammonia, benzoate and hippu-
`rate over the first two weeks in a
`neonate with argininosuccinic
`aciduria. Concentrations are on
`a logarithmic scale
`
`1
`
`4
`
`6
`
`8
`
`10
`12
`Age (days)
`
`14
`
`16
`
`18
`
`If
`achieve this by inducing the enzymes in utero.
`phenobarbital induces increased activity of benzoate:CoA
`ligase and benzoyl-CoA:glycine N-acyltransferase in neo-
`nates, it might also induce increased activity before birth. It
`might, therefore, be helpful to give the mother phenobar-
`bital for a few days before delivery if the fetus is known to
`have a severe urea cycle disorder. Phenobarbital has been
`given to mothers in the past to induce the conjugation of
`bilirubin in neonates, before the introduction of photo-
`therapy simplified the management of neonatal jaundice
`[20]. Obviously,
`the use of phenobarbital would need
`careful monitoring and the experimental nature of this
`treatment should be discussed fully with the family.
`Sodium benzoate appears not to induce its enzymes of
`conjugation but sodium phenylbutyrate may do so, as it
`induces the expression of many genes [11]. If phenyl-
`butyrate can be shown to induce its own conjugation, it
`may be helpful to give the mother this drug as well as
`phenobarbital.
`
`Dialysis
`
`Early diagnosis and intensive therapy are important and, if
`the metabolic disturbance cannot be controlled quickly,
`dialysis should be started early. The various alternatives for
`dialysis have been reviewed elsewhere [23]. ECMO/
`haemodialysis and haemodiafiltration are the most effective
`[7, 23] but they are technically more demanding. Although
`somewhat less effective, continuous veno-venous haemofil-
`tration is widely available and less likely to cause serious
`complications [9]. Peritoneal dialysis lowers ammonia
`concentrations more slowly. Exchange transfusion has no
`place in the management of hyperammonaemia.
`
`Neuroprotection
`
`Various forms of treatment have been proposed to reduce
`the adverse effects of hyperammonaemia on the brain.
`Potential neuroprotective measures include hypothermia
`[26] and NMDA receptor antagonists, such as memantine
`[25]; their effectiveness is not yet clear. Obviously, trials
`should be undertaken in patients with established hyper-
`ammonaemia before they can be justified in those at risk of
`hyperammonaemia.
`
`Conclusions
`
`to control
`is often difficult
`it
`With standard treatment,
`hyperammonaemia in patients with severe urea cycle
`defects presenting in the newborn period, particularly CPS
`and OTC deficiencies. We suggest some additional mea-
`sures that may help to control plasma ammonia concen-
`trations. The measures we propose include giving the
`mother phenobarbital
`for a few days before delivery,
`delivering the baby by an elective caesarean section and
`managing the baby aggressively, with intravenous glucose
`from birth and drugs started soon afterwards. Several of
`these ideas have not yet been tried so that controlled studies
`need to be done to confirm that they are both safe and
`effective. Obviously, the situation needs to be discussed
`fully with the family in advance and a detailed management
`plan needs to be agreed by the metabolic specialist,
`obstetrician and neonatologist. It is likely that the some of
`the proposals are also applicable to other disorders that
`present with early decompensation in the neonatal period,
`such as organic acidaemias.
`
`

`

`Eur J Pediatr (2008) 167:305–309
`
`309
`
`Acknowledgements We would like to thank Dr Claude Bachmann
`for valuable discussion and particularly for permission to publish
`Fig. 1. We are also grateful to Dr Khalid Hussain for his help with
`this work.
`
`References
`
`1. Ali A, Qureshi IA (1992) Benzoyl-CoA ligase activity in the liver
`and kidney cortex of weanling guinea pigs treated with various
`inducers: relationship with hippurate synthesis and carnitine
`levels. Dev Pharmacol Ther 18:55–64
`2. Arnold J, Campbell IT, Hipkin LJ et al (1995) Manipulation of
`substrate utilization with somatostatin in patients with secondary
`multiple organ dysfunction syndrome. Crit Care Med 23:71–77
`3. Bachmann C (2003) Long-term outcome of patients with urea
`cycle disorders and the question of neonatal screening. Eur J
`Pediatr 162(Suppl 1):S29–S33
`4. Falconer AD, Poyser LM (1986) Fetal sympatho-adrenal mediated
`metabolic responses to parturition. Br J Obstet Gynaecol 93:747–753
`5. Gebhardt B, Dittrich S, Parbel S, Vlaho S, Matsika O, Bohles H
`(2005) N-carbamylglutamate protects patients with decompen-
`sated propionic aciduria from hyperammonaemia. J Inherit Metab
`Dis 28:241–244
`6. Gropman AL, Batshaw ML (2004) Cognitive outcome in urea
`cycle disorders. Mol Genet Metab 81(Suppl 1):S58–S62
`7. Haller M, Henzler-Le Boulanger A, Sass JO, Brandis M,
`Zimmerhackl LB (2005) Successful extracorporeal treatment of
`a male with hyperammonaemic coma. Nephrol Dial Transplant
`20:453–455
`8. Herndon DN, Hart DW, Wolf SE, Chinkes DL, Wolfe RR (2001)
`Reversal of catabolism by beta-blockade after severe burns. N
`Engl J Med 345:1223–1229
`9. Hiroma T, Nakamura T, Tamura M, Kaneko T, Komiyama A
`(2002) Continuous venovenous hemodiafiltration in neonatal
`onset hyperammonemia. Am J Perinatol 19:221–224
`10. Hussain K (2005) Congenital hyperinsulinism. Semin Fetal
`Neonatal Med 10:369–376
`11. Kern RM, Yang Z, Kim PS, Grody WW, Iyer RK, Cederbaum SD
`(2007) Arginase induction by sodium phenylbutyrate in mouse
`tissues and human cell lines. Mol Genet Metab 90:37–41
`12. Ktorza A, Bihoreau MT, Nurjhan N, Picon L, Girard J (1985)
`Insulin and glucagon during the perinatal period: secretion and
`metabolic effects on the liver. Biol Neonate 48:204–220
`
`13. Kuchler G, Rabier D, Poggi-Travert F et al (1996) Therapeutic use
`of carbamylglutamate in the case of carbamoyl-phosphate synthe-
`tase deficiency. J Inherit Metab Dis 19:220–222
`14. Maestri NE, Clissold D, Brusilow SW (1999) Neonatal onset
`ornithine transcarbamylase deficiency: a retrospective analysis.
`J Pediatr 134:268–272
`15. Maestri NE, Hauser ER, Bartholomew D, Brusilow SW (1991)
`Prospective treatment of urea cycle disorders. J Pediatr 119:923–928
`16. Msall M, Batshaw ML, Suss R, Brusilow SW, Mellits ED (1984)
`Neurologic outcome in children with inborn errors of urea
`synthesis. Outcome of urea-cycle enzymopathies. N Engl J Med
`310:1500–1505
`17. Nicolaides P, Liebsch D, Dale N, Leonard J, Surtees R (2002)
`Neurological outcome of patients with ornithine carbamoyltrans-
`ferase deficiency. Arch Dis Child 86:54–56
`18. Picca S, Dionisi-Vici C, Abeni D et al (2001) Extracorporeal
`dialysis in neonatal hyperammonemia: modalities and prognostic
`indicators. Pediatr Nephrol 16:862–867
`19. Qureshi
`IA, Lebel S, Letarte J (1989) Development and
`inducibility of the hepatic and renal hippurate-synthesizing system
`in sparse-fur (spf) mutant mice with ornithine transcarbamylase
`deficiency. Biochem Int 19:657–666
`20. Ramboer C, Thompson RP, Williams R (1969) Controlled trials of
`phenobarbitone therapy of neonatal jaundice. Lancet 1:966–968
`21. Schubiger G, Bachmann C, Barben P, Colombo JP, Tonz O,
`Schupbach D (1991) N-acetylglutamate synthetase deficiency:
`diagnosis, management and follow-up of a rare disorder of
`ammonia detoxication. Eur J Pediatr 150:353–356
`22. Sperling MA, DeLamater PV, Phelps D, Fiser RH, Oh W, Fisher
`DA (1974) Spontaneous and amino acid-stimulated glucagon
`secretion in the immediate postnatal period. Relation to glucose
`and insulin. J Clin Invest 53:1159–1166
`23. Summar M (2001) Current strategies for the management of
`neonatal urea cycle disorders. J Pediatr 138:S30–S39
`24. Uchino T, Endo F, Matsuda I
`(1998) Neurodevelopmental
`outcome of long-term therapy of urea cycle disorders in Japan.
`J Inherit Metab Dis 21(Suppl 1):151–159
`25. Vogels BA, Maas MA, Daalhuisen J, Quack G, Chamuleau RA
`(1997) Memantine, a noncompetitive NMDA receptor antagonist
`improves hyperammonemia-induced encephalopathy and acute
`hepatic encephalopathy in rats. Hepatology 25:820–827
`26. Whitelaw A, Bridges S, Leaf A, Evans D (2001) Emergency
`treatment of neonatal hyperammonaemic coma with mild systemic
`hypothermia. Lancet 358:36–38
`
`