Alternative pathway therapy for urea cycle disorders:
`
`Twenty years later
`
`
`
`Mark L. Batt1haw, MD, Robert B. MacArthur, PharmD, and Mendel Tuchman, MD
`
`
`
`
`
`led to the increased use of liver trans­
`
`
`
`plantation in treating patients with
`Alternative pathway therapy is currently an accepted treatment approach for
`
`
`
`
`
`
`
`
`neonatal-onset disease.5 It has also re­
`
`
`inborn errors of the urea cycle. This involves the long-term use of oral sodi­
`sulted in a search for the next genera­
`
`
`
`
`um phenylbutyrate, arginine supplements, or both, depending on the specific
`
`
`tion of treatment, potentially gene
`
`
`of acute hyp erammonemic crises with in­
`
`enzyme deficiency, and treatment
`
`6 However, until liver trans­
`therapy.
`
`
`
`
`travenous sodium benzoate/sodium phenylacetate plus arginine. A review of
`
`plantation becomes more widely ac­
`
`
`
`
`20 years of experience ,vith this approach illustrates the strengths and limita­
`
`cessible or gene therapy technically
`
`
`
`tions of this treatment. It has clearly decreased the mortality and morbidity
`
`
`&om these disorders, but they remain unacceptably high. The medications
`ASA Argtnlnosucclnlc acid
`
`
`
`
`
`are generally well tolerated, but severe accidental overdosage has been re­
`
`ASL Argtnlnosucclnlc acid lyase
`
`
`
`ported because of the in&equent use of the medication. There is also a differ­
`
`ASS Argtnlnosucclnlc acid synthetase
`
`
`CPS Cart>amyl phosphate synthetase I
`
`
`
`ence in their metabolism between newborns and older children that must be
`OTC Omlthlne transcarbamylase
`
`
`
`
`addressed in determining dosage. To avoid these complications it is recom­
`PA Phenytacetate
`
`
`mended that drug levels in blood be monitored routinely and that very spe­
`PAG Phenytacety1glutamlne
`PB Phenytbutyrate
`
`
`
`
`cific treatment protocols and oversight be followed to avoid overdoses. Final­
`PKU Phenytketonurta
`
`
`
`ly, it must be acknowledged that alternative pathway therapy has limited
`
`
`
`
`with ef­and must be combined effectiveness in preventing hyp erammonemia
`feasible, alternative pathway therapy
`
`
`
`
`
`
`
`fective dietary management. Therefore in children with neonatal-onset
`
`
`
`combined with protein restriction will
`
`
`
`disease or in those ,vith very poor metabolic control, liver transplantation
`
`remain the mainstay of therapy for in­
`
`
`
`should be considered. There should also be the continued search for innova­
`
`born errors of urea synthesis. Thus it
`
`
`
`tive therapies that may offer a more permanent and complete correction,
`is useful to review the first 2 decades
`
`
`such as gene therapy. (J Pediatr 2001; 138:S46-S55)
`
`
`of experience with this approach, dis­
`
`cuss the benefits and risks, and devel­
`
`
`op guidelines for best practice in the
`
`
`crises treatment of hyp erammonemic
`improvements in mortality
`
`The idea of using a detour around a significant
`
`
`
`
`
`
`congenital block within the urea cycle and morbidity, half of the children
`
`and in the chronic management of
`
`
`who survive neonatal-onset ornithine
`these disorders.
`
`was the result of a serendipitous ob­
`
`
`servation that occurred 20 years after
`
`transcarbamylase or carbamyl phos­
`1
`
`
`phate synthetase deficiency still die
`
`these disorders were first discovered
`
`
`
`before entering school, 5 and those
`
`and at a time when they carried a uni­
`
`
`versally fata.l prognosis.2 Although al­
`
`
`who survive have a high incidence of
`
`DISCOVERY
`4 This has
`
`developmental disabilities.
`
`ternative pathway therapy has led to
`
`THE SERENDIPITY OF
`
`The word serendipity was coined by
`
`From the Cbi/Jrr11'., NationaLMe'Jical Cmt-,; The Geol'ge W,uhinglon Univer;ity Schoo/of Af,Jici11e M'J Health
`
`
`
`
`
`Horace Walpo le after the characters
`in
`
`
`&ie,uu, Wa,bi119to11, DC, a11d Ruean,b Pharmacy, Col,m,bia U,uwr,ity, New York, Neu• Yo,-k.
`
`
`the fairy tale "The Three Princes of
`
`
`Drs. MacArthur and Tuchman are paid consultants of Ucy clyd Pharma.
`
`
`Serendip," who made fortunate and
`
`
`
`
`
`
`
`
`Reprint requests: Mark Batshaw, MD, Children's National Medical Center, 111 M;chigan Ave,
`7
`
`
`unexpected discoveries by accident.
`
`NV/, Washington, DC20010.
`
`In the case of alternative pathway
`
`Copyright© 200 I by Mosby, Inc.
`
`
`therapy, a serendipitous observation
`
`0022-3476/2001/$35.00 + 0 9/0/J I 1836
`was made by Dr Saul Brusilow. In
`
`doi: 10.1067/mpd.2001.111836
`
`S46
`
`Page 1 of 10
`
`Horizon Exhibit 2009
`
`Par v. Horizon
`IPR2017-01768
`
`

`

`THE JOURNAL OF PEDIATRICS
`VOLUME 138, NUMBER 1
`
`re-examining the data in an article
`on ketoacid therapy for citrullinemia
`(argininosuccinic acid synthetase de-
`ficiency) that we had previously
`published,8 he noted a previously
`unrecognized pattern; the more argi-
`nine that was given to the patient,
`the more nitrogen was excreted in
`the urine, and the lower the blood
`ammonia level was. He hypothesized
`that arginine was stimulating an al-
`ternative pathway for waste nitrogen
`excretion through increasing the
`synthesis of citrulline that contained
`a nitrogen from ammonia and could
`be rapidly excreted in the urine.9
`Shortly thereafter, we had the op-
`portunity of testing this theory on a
`newborn infant with argininosuc-
`cinic aciduria (argininosuccinic acid
`lyase deficiency), in whom the same
`principle would apply. In this case,
`with argininosuccinic acid as the
`waste nitrogen product, the provi-
`sion of 4 mmol/kg (800 mg/kg) of
`arginine hydrochloride intravenous-
`ly was associated with a remarkable
`and rapid fall in ammonia.10 The use
`of arginine to stimulate an alterna-
`tive pathway for waste nitrogen ex-
`cretion has remained a mainstay of
`therapy for these 2 disorders to the
`present day.
`A subsequent serendipitous obser-
`vation led to the use of sodium ben-
`zoate, sodium phenylacetate, and
`sodium phenylbutyrate. Dr Norman
`Radin, a biochemist from the Univer-
`sity of Michigan, was a National In-
`stitutes of Health site visitor for the
`Johns Hopkins Clinical Research
`Center in 1978, where we presented
`our research. We discussed the use of
`arginine to stimulate the urea cycle
`flux in argininosuccinic aciduria, and
`Dr Radin remarked that he remem-
`bered reading an article from around
`1914 in the Journal of Biological Chem-
`istry that spoke of using sodium ben-
`zoate to stimulate hippurate synthe-
`sis and decrease urea nitrogen
`excretion. After the site visit we went
`down to the bowels of the Welch
`
`medical library and found the arti-
`cle11 and a second article published a
`few years later that spoke of using
`sodium phenylacetate in a similar
`vein.12 Such was the birth of alterna-
`tive pathway therapy for urea cycle
`disorders.13
`
`PHARMACOLOGY OF
`ALTERNATIVE PATHWAY
`THERAPY
`
`In the subsequent 2 decades much
`has been learned about the pharmacol-
`ogy of these compounds and their effi-
`cacy in treating patients with acute and
`chronic hyperammonemia. Sodium
`benzoate, by its acylation of glycine to
`form hippurate and its renal clearance
`being fivefold the glomerular filtration
`rate, has the potential of removing 1
`mole of waste nitrogen for each mole of
`benzoate administered. This conver-
`sion occurs primarily in the liver and
`kidney by glycine N-acyltransferase.14
`Sodium phenylacetate is conjugated
`with glutamine by the enzyme phenyl-
`acetyl CoA:glutamine acyltransferase,
`to form phenylacetylglutamine, which
`is excreted by the kidney. This conver-
`sion also occurs in the liver and kid-
`ney.15 Glutamine contains 2 waste ni-
`trogen atoms, so 2 moles of nitrogen
`could be removed for each mole of
`phenylacetate administered. Sodium
`benzoate and sodium phenylacetate re-
`main useful in their intravenous for-
`mulation for the treatment of patients
`with acute hyperammonemia. Sodium
`phenylbutyrate (Buphenyl, Ucyclyd
`Pharma),16 which is first activated to
`its CoA ester, then metabolized by β-
`oxidation in the liver to phenylacetyl-
`CoA, and subsequently conjugated
`with glutamine, has replaced these
`compounds for chronic oral use. It has
`the advantage of lacking the disagree-
`able odor of the phenylacetate while
`maintaining its efficient nitrogen re-
`moval. For chronic oral use in patients
`who do not respond to or are intoler-
`ant of sodium phenylbutyrate, sodium
`
`BATSHAW, MACARTHUR, AND TUCHMAN
`
`phenylacetate with sodium benzoate
`oral liquid is also available from com-
`pounding pharmacies. This compound
`was previously marketed under the
`brand name of Ucephan.
`
`Intravenous Sodium Benzoate
`Green et al17 studied the pharmaco-
`kinetics of intravenous sodium ben-
`zoate in 4 neonates in hyperammone-
`mic coma, 3 caused by OTC deficiency.
`With a dose of 3.5 mmol/kg/d in 4 di-
`vided doses (460 mg/kg/d), a mean of
`84% of administered benzoate was
`found to be excreted as benzoate and
`hippurate in the urine. Serum benzoate
`concentrations ranged from 2.1 to 16
`mmol/L. Three of the 4 neonates
`cleared >50% of benzoate by the meta-
`bolic route (glycine conjugation). In
`these patients benzoate was not re-
`moved efficiently enough by peritoneal
`dialysis to prevent hippurate conver-
`sion, presumably because it is protein
`bound. Under these circumstances the
`addition of peritoneal dialysis was syn-
`ergistic with benzoate administration
`in the removal of ammonia. One
`neonate excreted only 12% of benzoate
`as hippurate. In this patient peritoneal
`dialysis served as the primary route of
`benzoate excretion. This may have
`been related to reduced renal clearance
`(caused by hypotension and impaired
`renal function) or limited benzoate
`conjugation capacity.
`Simell et al18 studied intravenous ad-
`ministration of sodium benzoate in 5
`older children with lysinuric protein
`intolerance who were clinically stable
`at the time of the study. Plasma ben-
`zoate levels peaked 2 hours after an in-
`fusion of 2 mmol/kg was started and
`had a half-life of 273 minutes. The
`mean plasma benzoate level at the end
`of the 90-minute infusion was 6
`mmol/L (range 5.2 to 7.0 mmol/L).
`Plasma hippurate peaked at 120 min-
`utes at 0.24 mmol/L (0.14 to 0.4
`mmol/L) and remained stable for the
`subsequent 3 hours. Less than 2% of
`administered sodium benzoate was ex-
`creted unchanged in urine over 24
`
`S47
`
`Page 2 of 10
`
`

`

`BATSHAW, MACARTHUR, AND TUCHMAN
`
`hours; hippurate excretion peaked in
`the first 6 hours, declining steadily
`thereafter. Total hippurate excretion
`accounted for 67% (35% to 126%) of
`benzoate administered.
`These studies suggest that sodium
`benzoate is useful for the acute treat-
`ment of patients with hyperammonemia.
`It is rapidly metabolized and removes
`significant amounts of nitrogen.19,20
`However, there are limitations; the re-
`sponse is stoichiometric, with a maxi-
`mum of 1 mole of nitrogen removed for
`each mole of benzoate administered,
`and the dosage is limited by potential
`toxicity. Thus when there is massive
`and incremental nitrogen accumula-
`tion, as during hyperammonemic coma
`(plasma ammonia >250 µmol/L), ben-
`zoate treatment will be insufficient,
`even when combined with sodium
`phenylacetate. However, it may well
`have a synergistic effect with dialysis.
`The wide variability of its metabolism
`in infants and neonates may result
`from immaturity of the acylation system
`in the liver and kidney. Between birth
`and 7 months of age, hepatic mitochon-
`drial glycine N-acyltransferase activity
`has been shown to vary from 5% to 40%
`of the peak activity, which occurs by 18
`months of age.21 This emphasizes the
`importance of monitoring plasma ben-
`zoate levels to avoid toxicity.
`
`Intravenous Sodium
`Phenylacetate
`Thibault et al22 studied the pharma-
`cokinetics of intravenous PA in adults
`with cancer. PA was administered as a
`bolus dose of 0.8 mmol/kg (150 mg/kg)
`followed by a continuous infusion.
`Serum PA levels displayed nonlinear
`pharmacokinetics. These authors indi-
`vidualized each patient’s continuous
`infusion dose based on parameters de-
`rived from the initial bolus infusion.
`No toxicity was associated with bolus
`administration of PA. During the con-
`tinuous infusion the authors were able
`to achieve stable phenylacetate con-
`centrations of 1 to 1.6 mmol/L, al-
`though induction of clearance and
`
`S48
`
`nonlinear kinetics made this difficult in
`some cases. Plasma PA concentrations
`of 5 to 7 mmol/L were associated with
`dose-limiting toxicity of reversible cen-
`tral nervous system depression, preced-
`ed by emesis. Simultaneous serum and
`cerebrospinal fluid sampling showed
`comparable phenylacetate levels and
`the presence of phenylacetylglutamine
`in the cerebrospinal fluid. This result
`suggests that PA can hasten the decline
`in glutamine and thereby ammonia
`from brain tissue, providing glutamine
`N-acylase is expressed in brain.
`In a subsequent study the authors
`used 2 daily bolus doses of 125 and 150
`mg/kg per dose, rather than a continu-
`ous infusion, to achieve the targeted
`plasma levels. Except for 1 occurrence,
`all cases of dose-limiting toxicity were
`seen at the 150 mg/kg per dose and
`were neurologic in nature. The peak
`PA levels associated with toxicity
`ranged from 2.7 to 5.5 mmol/L.
`Simell et al18 studied the use of intra-
`venous PA in lysinuric protein intoler-
`ance. Plasma PA levels 120 minutes
`after infusion of 2 mmol/kg (375 mg/kg)
`were less than plasma benzoate after
`an equimolar infusion, but the half-life
`of 254 minutes was similar. The plasma
`PA level was 4.8 mmol/L (range 3.7 to
`6.1 mmol/L) at the end of the infusion.
`Plasma PAG peaked at 270 minutes at
`0.48 mmol/L (range 0.22 to 1.06
`mmol/L), a level above that seen for
`hippurate after an equimolar infusion
`of sodium benzoate. Forty percent
`(range 15% to 110%) of infused PA
`was excreted as PAG in 24 hours,
`mostly within 12 hours. Peak plasma
`PA levels of 4 mmol/L were not asso-
`ciated with adverse clinical symptoms.
`
`Intravenous Sodium Benzoate
`and PA in Combination
`Brusilow et al23 reported on the effi-
`cacy of the combination of intravenous
`sodium benzoate and PA in the man-
`agement of intercurrent hyperam-
`monemic crises in children with urea
`cycle disorders. The dose used was 250
`mg/kg of each, given as a bolus over a
`
`THE JOURNAL OF PEDIATRICS
`JANUARY 2001
`
`2-hour period followed by a continu-
`ous infusion of 250 mg/kg per 24 hours
`until the ammonia level had stabilized.
`Eleven of 12 episodes were treated
`successfully, and measurement of the
`distribution of urinary nitrogen re-
`vealed that nitrogen in hippurate and
`PAG together accounted for an average
`of 60% of urinary waste nitrogen excre-
`tion. In 1 patient drug and metabolite
`plasma levels were described after a
`bolus and continuous infusion. For PA,
`maximal levels of 3 to 4 mmol/L were
`achieved after the bolus dose and re-
`mained within that range for approxi-
`mately 20 hours. PAG levels were not
`detected for 10 hours after the infusion,
`although the conjugation of phenylac-
`etate is known to occur rapidly.24 This
`was probably due to assay limitations.
`Plasma benzoate levels peaked at
`approximately 2.5 mmol/L after the
`bolus infusion and then descended
`to <1 mmol/L by approximately 15
`hours. Hippurate elevations were im-
`mediately detectable in plasma and re-
`mained so during the continuous infu-
`sion, in the range of 0.2 to 0.4 mmol/L.
`
`Sodium Phenylbutyrate
`In studying the intravenous pharma-
`cokinetics of sodium phenylbutyrate
`in adults, Piscitelli et al24 used doses
`of 600 to 2000 mg/m2 (equivalent to
`doses of approximately 25 to 90 mg/kg
`in a child) given intravenously over a
`30-minute period. Peak plasma PB
`concentrations ranged between 0.17
`and 0.30 mmol/L after the 600 mg/m2
`dose, between 0.3 and 0.6 mmol/L
`after the 1200 mg/m2 dose, and be-
`tween 0.6 and 1.0 mmol/L after the
`2000 mg/m2 dose. By 5 hours after the
`dose, levels were barely detectable in
`all cases, indicating first-order elimi-
`nation. Eighty percent of PB ap-
`peared in the urine as PAG. PA was
`detected in the urine within 10 min-
`utes of the start of the infusion, with a
`peak concentration 30 to 60 minutes
`after completion of the infusion of
`0.13 mmol/L, but levels were lower
`than when PA itself is infused. PA lev-
`
`Page 3 of 10
`
`

`

`THE JOURNAL OF PEDIATRICS
`VOLUME 138, NUMBER |
`
`BATSHAW, MACARTHUR, AND TUCHMAN
`
`Table I. Use of alternative pathway therapy during intermittent hyperammonemic crisis in patients with urea cycle disorders*
`
`Sustaining infusion
`
`Priming infusion
`
`0.250 g/kg/24 h
`or 5.5 g/m/24 h
`0.250 g/kg or 5.5 g/m?
`
`0.250 g/kg/24 h or
`6.5 g/m?/24 h
`0.250 g/kg or 5.5 g/m?
`
`——————————FeAllahReelSeelasbendeSERGEatsSondlee
`CPS or OTC deficiency
`Priming infusion
`0.250 g/kg or 5.5 g/m?
`0.250 g/kg or 5.5 g/m?
`0.20 g/kg (2 mL/kg) or
`4.0 g/m?
`0.20 g/kg (2 mL/kg)/24 h
`or 4.0 g/m/24 h
`0.60 g/kg (6 mL/kg) or
`Argininosuccinic acid
`12.0 g/m”
`synthetase deficiency
`Sustaining infusion—_0.250 g/kg/24 h or 0.250 g/kg/24 h or 0.60 g/kg (6 mL/kg)/24 h
`
`
`5.5 g/m?/24 h
`6.5 g/m?/24 h
`or 12.0 g/m?/24 h
`—
`0.60 g/kg (6 mL/kg)
`or 12.0 g/m?
`0.60 g/kg (6 ml/kg)/24 h or
`12.0 g/m?/24 h
`—
`0.250 g/kg or 5.5 g/m?
`0.250 g/kg or 5.5 g/m?
`Priming infusion
`Arginase deficiency
`
`Sustaining infusion—_0.250 g/kg/24 h or 0.250 g/kg/24 h or —
`
`5.5 g/m?/24 h
`6.5 g/m?/24 h
`
`Argininosuccinic acid
`lyase deficiency
`
`Priming infusion
`
`Sustaining infusion
`
`—
`
`*If the patientis already on therapy, primary infusion dose should be reduced or eliminated.
`
`els in plasma were also much lower
`(below Km) after intravenous PB
`than after PA infusion.
`Brusilow”® noted that children re-
`ceiving a diet containing 200 mg/kg/d
`nitrogen (1.25 g/kg/d protein) excrete
`94 mg urea nitrogen/kg/d, 47% ofdi-
`etary nitrogen. To excrete this amount
`of PAGnitrogenhecalculated that the
`child would require 525 mg/kg/d PA
`provided as phenylbutyrate. He subse-
`quently found that 80% to 90% ofthe
`predicted amount of PAG synthesized
`was excreted. In children with urea
`cycle disorders on a low-protein diet,
`PB administered orally at the recom-
`mended dose of 300 to 650 mg/kg/d
`yielded plasma levels of PA ranging
`from 0.026 to 1.87, of PB ranging from
`0 to 0.872, and of PAG ranging from
`0.093 to 3.15 mmol/L.
`
`Arginine
`Arginine free base is used as a long-
`term alternative pathway therapy to
`treat patients with ASS and ASLdefi-
`ciencies at a dose of 3 to 4 mmol/kg/d
`(500 to 700 mg/kg/d).26 This dose has
`been well tolerated and is associated
`with plasmaargininelevels 1.5-fold to
`twofold normal (mean 128 Umol/L). It
`also leads to further increases in plas-
`
`Page 4 of 10
`
`ma levels of citrulline (mean 3936
`mol/L) and ASA (907 Umol/L), re-
`spectively, which are already marked-
`ly elevated in these disorders. There is
`significant excretion ofcitrulline and
`ASA in urine, representing 33% to
`37% of waste nitrogen excretion in
`ASSdeficiency and 52% to 59% in
`ASLdeficiency. In ASL deficiency,
`arginine therapy combinedwith pro-
`tein restriction has proven very effec-
`tive for long-term control; in ASS defi-
`ciency the combination of arginine and
`phenylbutyrate has been effective.?”8
`
`ADVERSE EFFECTS OF
`ALTERNATIVE PATHWAY
`THERAPY
`
`Animal studies have suggested po-
`tential toxicities from benzoate admin-
`istration as a result of its completion
`for free CoA”? and impairmentof mi-
`tochondrial pathways and glycine pro-
`duction.*95! In addition, benzoate can
`theoretically displace bilirubin from
`high-affinity albumin bindingsites.
`However, sodium benzoate and other
`alternative pathway therapies have
`been found to be remarkably nontoxic
`in humansat the doses recommended
`
`to treat patients with urea cycle disor-
`ders. Laboratory studies of hematopoi-
`etic, renal, and hepatic function have
`been within normallimits, and patho-
`logic examination ofthe liver revealed
`normal hepatic tissue except for a
`small amount offibrosis,2° a finding
`similar to patients not receiving alter-
`native pathway therapy.*”
`During long-term PB therapy, 23%
`of menstruating females had irregular
`menses or became amenorrheic. De-
`creased appetite, taste disturbances,
`or disagreeable body odor occurred
`in approximately 4%. Abnormal elec-
`trolytes, increased serum hepatic en-
`zyme levels, hypoalbuminemia, and
`anemia may occurbutaredifficult to
`differentiate from the primary disease
`itself. A variety of gastrointestinal dis-
`orders, aplastic anemia, eccymoses,
`arrhythmias, renal tubular acidosis,
`depression, and rash have been re-
`ported rarely.5**4 Renal Fanconi syn-
`dromehas beenreportedin 2 patients,
`and a fewpatients were reported to
`have oral mucositis®®; there has been
`1 report of chronic pancreatitis.*>*
`Neurotoxicity as described in adults
`receiving PAis unlikely to occur, be-
`cause PAlevels remain lowafter PB
`administration.24
`
`$49
`
`

`

`BATSHAW, MACARTHUR, AND TUCHMAN
`
`In terms of side effects of the intra-
`venous bolus doses of sodium benzoate
`and PA used to treat patients with hy-
`perammonemic crises (Table I), there
`is an association with nausea and vom-
`iting during the infusion and with hy-
`pokalemia related to urinary loss that
`is enhanced by the excretion of the
`nonabsorbable hippurate and PAG. In
`children severe toxicity of intravenous
`benzoate and phenylacetate is rare, but
`overdoses (3 to 5 times normal doses)
`have led to symptoms that mimic hy-
`perammonemic crises including agita-
`tion, confusion, and hyperventilation.
`Plasma ammonia levels may not be
`elevated initially, but there is often sub-
`sequent rebound hyperammonemia.
`There was also a partially compensated
`metabolic acidosis and an increased
`anion gap (up to 52 meq/L). Two deaths
`caused by drug overdose have been re-
`ported; these were associated with cere-
`bral edema, hypotension, and cardiovas-
`cular collapse.26,36 The overdosage
`ranged from 750 mg/kg administered
`over a 10-hour period to 1750 mg/kg
`over an 18-hour period. These children
`had plasma benzoate and phenylacetate
`levels of approximately 10 mmol/L 4
`hours after infusion. This disproportion-
`ate increase in plasma levels is attribut-
`able to the nonlinear pharmacokinetics
`of phenylacetate. In adults plasma PA
`levels of >6 mmol/L were associated
`with confusion, lethargy, and emesis.37
`In children with ASS and ASL defi-
`ciency, at the recommended bolus dose
`of arginine hydrochloride (3 to 4
`mmol/kg/d), severe adverse events
`have not been reported, although hy-
`perchloremic acidosis has been noted,
`and extravasation of arginine hy-
`drochloride can cause tissue necrosis.
`A 21-month-old girl with short stature
`who was inadvertently given 18.5
`mmol/kg (3.9 g/kg; 300 mL of 10%
`arginine hydrochloride solution intra-
`venously) as a bolus to measure growth
`hormone stimulation had a fatal reac-
`tion.38 Within 30 minutes of the infu-
`sion she had gasping respiration and a
`cardiopulmonary arrest. She was re-
`
`S50
`
`suscitated but had suffered irreversible
`brain damage and died a few days later.
`She was found to be acutely hypona-
`tremic with a metabolic acidosis; the ar-
`rest was thought to have developed
`from an arrhythmia caused by the sud-
`den drop in pH.
`There is no proven toxicity from
`chronic therapy with arginine-free
`base, but there are some unresolved is-
`sues. One is that children with ASL
`deficiency have markedly enlarged liv-
`ers, and in some cases there is evidence
`of cirrhosis.32 Because arginine treat-
`ment leads to increases in both arginine
`and ASA in blood, there is the concern
`that it could potentially contribute to
`the hepatopathology. However, a re-
`view of patients treated before the use
`of arginine reveals hepatomegaly with
`fibrosis; thus it is unclear whether argi-
`nine is a contributing factor.39 The sec-
`ond concern is whether increasing cit-
`rulline or ASA levels in blood, and
`presumably brain, could have an ad-
`verse effect on intellectual function.
`Msall et al40 followed a few patients
`with ASS or ASL deficiency over time
`while they were receiving arginine
`therapy and found stable IQ scores
`over a number of years; however, these
`children all had pre-existing mental re-
`tardation resulting from neonatal hyper-
`ammonemic coma. An approach to de-
`creasing the levels of arginine, citrulline,
`and ASA could be to use a combination
`of phenylbutyrate and arginine therapy
`in both ASS and ASL deficiencies.
`
`OUTCOME MEASURES
`
`Although there have been no con-
`trolled studies, introduction of alterna-
`tive pathway therapy appears to have
`improved both biochemical control and
`neurologic outcome in patients with urea
`cycle disorders.4,26-28,41-43 There has
`been a reported decrease in the frequen-
`cy of episodes of hyperammmonemia,
`improved growth, and maintenance of
`intellectual function. Children with
`neonatal-onset ASS and ASL deficien-
`
`THE JOURNAL OF PEDIATRICS
`JANUARY 2001
`
`cies have better survival rates than those
`with OTC and CPS deficiency. The vast
`majority of both groups, however, have
`significant developmental disabilities.4,44
`The exception has been children who
`were treated prospectively rather than
`having been rescued from hyperam-
`monemic coma.42 These children, how-
`ever, remain at risk for potentially fatal
`hyperammonemic coma
`throughout
`their lives, and to our knowledge no pa-
`tient with neonatal onset of CPS or OTC
`deficiency has yet survived to adulthood.
`For this reason it is now advised that
`these children be considered for liver
`transplantation in their first year of life.
`For children with late-onset disease
`(including manifesting OTC-deficient
`heterozygotes) the outcome is better but
`variable.28 Most do well with alternative
`pathway therapy, and the neurodevelop-
`mental outcome is generally good. They
`too, however, remain at risk for hyper-
`ammonemic crises, which carry signifi-
`cant morbidity and mortality.45
`
`USE OF PB IN
`PREGNANCY
`
`Because more young women who are
`manifesting OTC-deficient heterozy-
`gotes or have other partial urea cycle
`defects are being treated with PB, the
`question of teratogenicity is raised.
`In certain ways PB treatment mimics
`maternal phenylketonuria. Pregnant
`women with PKU who are not follow-
`ing a low-phenylalanine diet have been
`found universally to produce children
`who have microcephaly and mental re-
`tardation; and congenital heart defects
`are also common.46 The cause is not
`clear, although animal models suggest
`that phenylalanine and its metabolites
`are teratogenic.47 Because PA is a
`metabolite of phenylalanine, it could
`conceivably lead to fetal damage in
`pregnant women receiving PB; the
`plasma levels of PA in these women are
`similar to those in women with mater-
`nal PKU.46 To this point only 1 preg-
`nancy has been reported in the litera-
`
`Page 5 of 10
`
`

`

`THE JOURNAL OF PEDIATRICS
`VOLUME 138, NUMBER |
`
`BATSHAW, MACARTHUR, AND TUCHMAN
`
`Table IT, Long-term treatment of urea cycle disorders
`
`abDISOPDerssannnntntnnnnsntennenCitNeacasnsnsstnstnsnennArgininefreebasemnsodiumphenylbutyratel
`CPS or OTC
`.170 g/kg/d or 3.8 g/m7/d
`~
`.450-.600 g/kg/d if <20 kg;
`deficiency
`9.9-13.0 g/m?/d in larger
`patients
`450-.600 g/kg/d if <20 kg;
`9.9-13.0 g/m?/d in larger
`patients
`May not be required
`
`~
`
`—
`
`.400-.700 g/kg/d or 8.8-15.4 g/m?/d
`
`-400-.700 g/kg/d or 8.8-15.4 g/m?/d
`
`Argininosuccinic acid
`synthetase deficiency
`
`Argininosuccinic acid
`lyase deficiency
`Arginase deficiency
`
`—
`
`—
`
`NAGSdeficiency*
`
`.170 g/kg/d or 3.8 g/m7/d
`
`-170 g/kg/d or 3.8 g/m7/d
`
`450-.600 g/kg/d if <20 kg;
`9.9-13.0 g/m?/din larger
`patients
`450-.600 g/kg/d if <20 kg;
`9.9-13.0 g/m?/d in larger
`
`NAGS, N-acetyl glutamate synthase.
`*Caloric requirement may be completed with a protein-free formula. In general, the minimum daily protein intake for growth was used: for 1 to 4
`months, 1.6 to 1.9 g/kg/d; for 4 to 12 months, 17 glkg/d.; for 1 to3 years, 1.4 g/ke/d. Daily protein intake may include an essential aminoacid formula.
`TIfintolerant ofphenylbutyrate, sodium benzoate + sodium phenylacetate can be given orally at a dose of .250 to .500 glkg/d each.
`*N-carbamylglutamate mayalso be given at a dose of .320to .650 g/kg/d.
`
`ture ofa patient receiving PB.*® This
`occurred in a woman with partial OTC
`deficiency who was placed on 280
`mg/kg/d phenylbutyrate beginning at
`11 weeks’ gestation. The plasma ammo-
`nia level was not well controlled on this
`
`dose, but at 33 weeks a child was well-
`born and was developing normally
`whenreportedat 2 years of age. How-
`ever, because the PB was notstarted
`until the end ofthefirst trimester, this
`case doesnotresolve the issue of poten-
`tial teratogenicity of PB. Recently, 2 ad-
`ditional pregnancies of OTC-deficient
`womenreceiving PB havebeen report-
`ed anecdotally to have resulted in well-
`born infants (James Leonard, personal
`communications). The actions of PB
`and PA ongenesilencing”? andcell dif-
`ferentiation”” support the need for more
`rigorous evaluationoffetal risk.
`
`OTHER Uses oF PA/PB
`
`Serendipity has also influenced the use
`of PA and PBin other disorders. These
`compoundshave beenstudied as poten-
`tial chemotherapeutic agents. PB and PA
`
`Page 6 of 10
`
`serve as inhibitors of histone deacetylase
`and inducers ofhistone acetylation that
`inhibit tumor growth; PA worksless well
`as a tumor suppressor than PB.%?
`PB and PAalso appear to enhance
`efficacy of other therapeutic agents
`includingretinoids, interferon-G, sura-
`min, 5-aza-2-deoxycytidine, and hydro-
`xyurea. The enhancement of hydroxy-
`urea makes them potentially valuable in
`increasing fetal hemoglobin production
`in hemoglobinopathies.*” They havealso
`been foundto increase cystic fibrosis
`transmembraneconductanceregulator
`trafficking in vitro by reducing Hsc70
`immunoreactivity and could therefore
`play a role in cystic fibrosis treatment.5!
`Howanyofthese otherbiologic effects
`of PA and PB mayaffect individuals
`with urea cycle disorders is uncertain.
`
`CURRENTLY RECOMMENDED
`ALTERNATIVE PATHWAY
`THERAPY
`
`Long-term Therapy
`
`The aim of long-term therapy has
`been to maintain metabolic control
`
`with plasma ammoniaconcentrations
`less than twice normal and plasma glu-
`taminelevels <1000 pmol/L.54? Table
`II summarizes the current recommen-
`dations for long-term alternative path-
`waytherapy in the different urea cycle
`disorders. Thebasic principles are to
`give adequate nitrogen intake for
`growth, to provide arginine (orcit-
`rulline) as a semiessential amino acid
`(except in arginase deficiency), and
`to stimulate alternative pathways
`of waste nitrogen excretion with PB
`(400 to 600 mg/kg/d) and arginine
`(400 to 700 mg/kg/d, in ASS and ASL
`deficiencies).
`
`Acute Management of
`Hyperammonemtc Crisis
`Early identification and treatment of
`intercurrent hyperammonemic episodes
`is essential, both because treatmentis
`moreeffective at lower ammonialevels
`and because neurologic outcome ap-
`pears to be a function ofdurationofse-
`vere hyperammonemia and coma./ If
`symptomshaveprogressed to vomiting
`and lethargy, plasma ammonialevels
`are usually >3 to 5 fold normal, and ag-
`
`SSI
`
`

`

`BATSHAW, MACARTHUR, AND TUCHMAN
`
`gressive treatment is warranted. The
`recommended approach to treating
`acute hyperammonemiais outlined in
`Table I. It focuses on ameliorating
`catabolism and removing nitrogen.
`Nonprotein calories are provided as
`glucose (8 to 10 mg/kg/min) with intra-
`venousfat emulsion added to maintain
`a caloric intake of >80 kcal/kg/d and to
`minimize endogenousproteolysis. For
`CPS and OTCdeficiencies,
`intra-
`venousarginine is given as 2 mL/kg of
`10% arginine hydrochloride (200
`mg/kg, Kabivitrum, Clayton, NC) over
`a 90-minute period followed by the
`same dose over a 24-hourperiod. For
`ASSand ASLdeficiencies, the priming
`dose of arginine hydrochloride is 6
`mL/kg (600 mg/kg) followed by the
`same dose over a 24-hourperiod. Sodi-
`um bicarbonate should be included in
`the intravenousfluids to neutralize the
`acidifying effects of arginine hy-
`drochloride. A priming dose of sodium
`benzoate/PA (250 mg/kg each; 5.5
`g/m? for older patients) is administered
`over a 90-minute period in 25 to 35
`mL/kg of 10% glucose (for older pa-
`tients 400 to 600 mL/m? of 10% glu-
`cose) followed by the same dose every
`24 hours until oral phenylbutyrate
`therapy can be restarted. Because
`vomiting often occurs during the prim-
`ing dose, Zofran (ondansetron hy-
`drochloride), a potent antiemetic, may
`be given intravenously (0.15 mg/kg) at
`the start of the infusion. Because | g of
`sodium benzoate contains 160 mgsodi-
`um and | g of PA contains 147 mg
`sodium, additional sodium does not
`need to be provided in the intravenous
`solution, because hypernatremia can
`develop. Conversely, hypokalemia can
`result from urinary potassium losses
`enhanced by hippurate and PAG ex-
`cretion, so potassium should be added
`to the intravenousfluid. If this treat-
`mentdoes notlead to a stabilizing or
`lowering of ammonia levels within 4
`hours, orif the patient’s clinical condi-
`tion worsens, hemodialysis should be
`started. Alternative pathway therapy
`
`$52
`
`Page 7 of 10
`
`should be continued during hemodial-
`ysis, because its effects appear to be
`additive. Notably, hemodialysis de-
`creases plasma concentrations of PAG
`and hippurate to approximately the
`sameextent as that of urea and creati-
`nine.!® Monitoring and treatment of
`increased intracranial pressure are also
`important.
`
`CONCLUDING REMARKS
`
`Alternative pathway therapy re-
`mains a mainstay of both acute and
`long-term tre