`Phenylbutyrate in Healthy Adults and
`Adults with Cirrhosis
`
`Brendan M. MCGuire,1 Igor A. Zupanets,2 Mark E. Lowe,3 Xunjun Xiao,3 Vasyliy A. Syplyviy,2
`Jon Monteleone,4 Sharron Gargosky,5 Klara Dickinson,5 Antonia Martinez,5
`Masoud Mokhtarani,5 and Bruce F. Scharschmidt5
`
`Phenylbutyric acid (PBA), which is approved for treatment of urea cycle disorders (UCDs)
`as sodium phenylbutyrate (NaPBA), mediates waste nitrogen excretion via combination of
`PBA-derived phenylacetic acid with glutamine to form phenylactylglutamine (PAGN) that
`is excreted in urine. Glycerol phenylbutyrate (GPB), a liquid triglyceride pro-drug of PBA,
`containing no sodium and having favorable palatability, is being studied for treatment of
`hepatic encephalopathy (HE). In vitro and clinical studies have been performed to assess
`GPB digestion, safety, and pharmacology in healthy adults and individuals with cirrhosis.
`GPB hydrolysis was measured in vitro by way of pH titration. Twenty-four healthy adults
`underwent single-dose administration of GPB and NaPBA and eight healthy adults and
`24 cirrhotic subjects underwent single-day and multiple-day dosing of GPB, with metabo-
`lites measured in blood and urine. Simulations were performed to assess GPB dosing at
`higher levels. GPB was hydrolyzed by human pancreatic triglyceride lipase, pancreatic
`lipase-related protein 2, and carboxyl-ester lipase. Clinical safety was satisfactory. Com-
`pared with NaPBA, peak metabolite blood levels with GPB occurred later and were lower;
`urinary PAGN excretion was similar but took longer. Steady state was achieved within 4
`days for both NaPBA and GPB; intact GPB was not detected in blood or urine. Cirrhotic
`subjects converted GPB to PAGN similarly to healthy adults. Simulations suggest that
`GPB can be administered safely to cirrhotic subjects at levels equivalent to the highest
`approved NaPBA dose for UCDs. Conclusion: GPB exhibits delayed release characteristics,
`presumably reflecting gradual PBA release by pancreatic lipases, and is well tolerated
`in adults with cirrhosis, suggesting that further clinical testing for HE is warranted.
`(HEPATOLOGY 2010;51:2077-2085)
`
`Abbreviations: AE, adverse event; AUC, area under the concentration versus time curve; CEL, carboxyl ester lipase; Cmax, maximum plasma concentration;
`GPB, glycerol phenylbutyrate; HE, hepatic encephalopathy; MELD, model for end-stage liver disease; NaPBA, sodium phenylbutyrate; PAA, phenylacetic acid;
`PAG, phenylacetylglycine; PAGN, phenylacetylglutamine; PBA, phenylbutyric acid; PBG, phenylbutyrylglycine; PBGN, phenylbutyrylglutamine; PK,
`pharmacokinetic; PLRP2, pancreatic lipase-related protein 2; PTL, pancreatic triglyceride lipase; UCD, urea cycle disorder.
`From the 1University of Alabama, Birmingham, AL; the 2National University of Pharmacy and Kharkiv National Medical University, Kharkiv, Ukraine; the
`Division of 3Pediatric Gastroenterology, Hepatology, and Nutrition, University of Pittsburgh, Pittsburgh, PA; 4Pharsight Corp., Cary, NC; and 5Hyperion
`Therapeutics, Inc., South San Francisco, CA.
`Received October 13, 2009; accepted January 15, 2010.
`Clinicaltrials.gov registration numbers: protocol UP 1204-001, ID# NCT00977600; protocol UP 1204-002, ID# NCT00986895. Studies for UP 1204-001
`and UP 1204-002 provide assurance that informed consent in writing was obtained from each patient and the study protocols conformed to the ethical guidelines
`of the 1975 Declaration of Helsinki as reflected in a prior approval by the appropriate institutional review committee.
`Address reprint requests to: Bruce F. Scharschmidt, M.D., Senior Vice President and Chief Medical Officer, Hyperion Therapeutics, 601 Gateway Boulevard,
`Suite 200, South San Francisco, CA 94080. E-mail: bruce.scharschmidt@gmail.com; fax: 650-745-3581.
`Copyright VC 2010 by the American Association for the Study of Liver Diseases.
`Published online in Wiley InterScience (www.interscience.wiley.com).
`DOI 10.1002/hep.23589
`Potential conflict of interest: K. Dickinson, A. Martinez, M. Mokhtarani, S. Gargosky, and B. F. Scharschmidt are employees of Hyperion or were at the time of the
`study. None of the other authors have a financial interest in Hyperion, although payments were made by Hyperion to Pharsight Corporation (Jon Monteleone), which
`performed the pharmacokinetic modeling, to University of Pittsburgh (Mark Lowe) for the studies of in vitro digestion of glycerol phenylbutyrate, and to the National
`University of Pharmacy and Kharkiv National Medical University, Kharkiv, Ukraine (Igor Zupanets, V. Syplyviy), where the clinical studies were conducted. Drs. McGuire,
`Monteleone, and Lowe are consultants for and received grants from Hypersion. Dr. Syplyviy received grants from Hypersion. Drs. Gargosky, Scharschmidt, Martinez, and
`Dickinson own stocks in Hypersion. Dr. Mokhtarani owns stock in Pfizer.
`Additional Supporting Information may be found in the online version of this article.
`
`2077
`
`
`
`2078 MCGUIRE ET AL.
`
`HEPATOLOGY, June 2010
`
`Glycerol phenylbutyrate (GPB) [or glyceryl tri-
`
`(4-phenylbutyrate), also referred to as HPN-
`100] is an oral
`investigational agent under
`development
`for hepatic encephalopathy (HE) and
`urea cycle disorders (UCDs). It is a pro-drug of phe-
`nylbutyric acid (PBA), currently marketed as sodium
`phenylbutyrate (NaPBA), for the treatment of UCDs.
`It consists of glycerol with three molecules of PBA
`linked as esters. GPB is a pale yellow, nearly odorless
`and tasteless oil, whereas NaPBA has palatability
`issues, high sodium content, and high pill burden.
`The maximum approved daily dose of NaPBA (20 g)
`corresponds to 40 tablets containing 2,400 mg of
`sodium, which exceeds
`the daily
`allowance
`of
`2,300 mg/day recommended in the US Department of
`Health and Human Services Dietary Guidelines
`for
`Americans, 2005 for the general population and 1,500
`mg/day for individuals with hypertension or sodium
`retaining states.1 The corresponding dose of GPB is
`17.4 mL, which contains no sodium.
`NaPBA mediates excretion of waste nitrogen as
`shown in Fig. 1. PBA is absorbed from the intestine
`and converted by way of b-oxidation to the active
`moiety, phenylacetic acid (PAA). PAA is conjugated
`with glutamine in the liver and kidney by way of N-
`acyl coenzyme A-L-glutamine N-acyltransferase to form
`phenylacetylglutamine (PAGN).2 Like urea, PAGN
`
`incorporates two waste nitrogens and is excreted in the
`urine.3
`Because GPB contains no sodium and may be better
`tolerated than NaPBA,
`its safety and pharmacology
`were studied in healthy adults and adults with cirrho-
`sis, as was the handling of GPB by human pancreatic
`lipases. Monte Carlo simulations were performed to
`assess metabolites blood levels and therefore clinical
`safety at doses approximating the highest approved
`dose of NaPBA for treatment of UCDs.
`
`Materials and Methods
`In Vitro Hydrolysis of GPB by Pancreatic
`Enzymes
`Recombinant human pancreatic triglyceride lipase
`(PTL), pancreatic lipase-related protein 2 (PLRP2),
`colipase,
`and carboxyl
`ester
`lipase
`(CEL) were
`expressed in yeast and purified as described.4-7 Lipase
`activity against GPB was measured by titration of the
`
`released fatty acid (PBA) at 23
`C using a Radiometer
`TIM 854 pH-stat.8 The assay buffer contained 0.5 mL
`(550 mg) of emulsified GPB and 1 mM Tris-HCl (pH
`8.0), 2 mM CaCl2, 150 mM NaCl, and 0.5 or 4 mM
`sodium taurodeoxycholate for PTL and PLRP2 or
`10 mM sodium cholate for CEL assays. PTL activity
`was determined with 3 lg of PTL 6 3 lg of colipase
`
`Fig. 1. Urea cycle and removal of
`waste nitrogen as hippuric acid follow-
`ing administration of sodium benzoate
`and as phenylacetylglutamine following
`administration of sodium phenylbuty-
`rate. [Adapted from Summar and Tuch-
`man, J Pediatr 2001;138(Suppl.):S6-
`S10.]
`
`
`
`HEPATOLOGY, Vol. 51, No. 6, 2010
`
`MCGUIRE ET AL.
`
`2079
`
`added at time zero. PLRP2 activity was determined
`with 10 lg of PLRP2 6 10 lg of colipase added at
`time zero. CEL activity was determined with 10 lg of
`CEL in the absence of colipase. Each reaction was
`monitored for 5 minutes. The reaction rate was deter-
`mined from the slope of the linear curve. The rate of
`100 mM NaOH titration during the assay was set to
`maintain a constant pH of 8.0 for PTL and PLRP2
`and 50 mM NaOH for CEL. The activity of PTL and
`PLRP2 against tributyrin and triolein in 1 mM Tris-
`HCl (pH 8.0), 2 mM CaCl2, 150 mM NaCl, and 4
`mM sodium taurodeoxycholate and of CEL against
`tributyrin and triolein in the same buffer with 10 mM
`sodium cholate and no taurodeoxycholate was deter-
`mined using the same methodology.
`
`Study Design and Treatments
`UP 1204-001. This was a phase 1, randomized,
`crossover, open-label study designed to assess safety,
`tolerability, pharmacokinetic (PK) equivalence, and
`bioequivalence in healthy adult subjects. Intravenous
`AMMONUL (a 10%/10% solution of sodium phenyl-
`acetate and sodium benzoate) and a formulated oral
`preparation of GPB were administered in addition to
`GPB (unformulated) and NaPBA, but only the results
`for NaPBA and unformulated GPB are reported in
`this study. Subjects received a single dose of either
`NaPBA or GPB on separate dosing days, at least 7
`days apart. NaPBA and GPB were administered at a
`dose equivalent to 3 g/m2 of PBA. PK samples were
`taken predose and 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, 8, 12,
`24, and 48 hours postdose. Urine was collected from
`0-4, 4-8, 8-12, and 12-24 hours postdose. PK variables
`were calculated for PBA, PAA, phenylacetylglycine
`(PAG), PAGN, phenylbutyrylglycine (PBG), and phe-
`nylbutyrylglutamine (PBGN). A test for intact GPB
`was also conducted in subjects receiving GPB.
`Bioequivalence was assessed by calculating 90% con-
`fidence intervals
`for
`the ratio of geometric means
`between test and reference treatments. The ratios and
`confidence intervals were calculated in an analysis of
`variance model
`for log-transformed pharmacokinetic
`variables including treatment, period, and the treat-
`ment by period interaction as fixed effects and subject
`as a random effect.
`UP 1204-002. This was an open-label study of the
`safety and PK equivalence of GPB in subjects with cir-
`rhosis (Child-Pugh score A, B, or C [n ¼ 8 in each
`group]) compared with age- and sex-matched healthy
`subjects with normal hepatic function (n ¼ 8). Sub-
`jects received a single oral GPB dose (100 mg/kg/day)
`
`on day 1, two doses per day (12 hours apart) on days
`8-14 (200 mg/kg/d), and a single dose on day 15 (100
`mg/kg/d). The single oral dose on day 1 was a fasting
`dose, whereas the first dose on day 8 was given with a
`meal. The last GPB dose was administered on the
`morning of day 15 and was followed by 48 hours of
`plasma PK sampling and urine collection. PK blood
`samples were drawn at 0.25, 0.5, 1, 1.5, 2, 3, 4, 6, 8,
`12, and 24 hours postdose on days 1, 8, and 15, and
`at 48 hours after dosing on days 1 and 15. Urine was
`collected from 0-4, 4-8, 8-12, and 12-24 hours post-
`dose on days 1, 8, and 15 and at 24-48 hours post-
`dose on days 1 and 15. PK samples were drawn fasting
`prior to the morning dose (trough) and 2 hours post-
`dose on days 9-14. A 12-lead electrocardiogram was
`performed at screening on days 0 and 7, 2 hours post-
`dose on days 1 and 15 (between 9:00 AM and 10:00
`AM), and at follow-up (7 days after day 15).
`
`Pharmacokinetic Analyses
`Plasma and urine PK parameters were calculated for
`all subjects and summarized with descriptive statistics
`(number of patients, mean, standard deviation, me-
`dian, minimum, and maximum). PK parameters were
`calculated using time concentration profiles for each
`subject, including area under the concentration versus
`time curve from time 0 (predose) to 24 hours (AUC0-
`24), calculated using the linear trapezoidal rule; maxi-
`mum plasma concentration at
`steady state (Cmax);
`and the time of maximum plasma concentration at
`steady state. The amount of PAGN excreted in urine
`over 24 hours was calculated from urinary concentra-
`tion (by multiplying the urinary volume with urinary
`concentrations).
`
`Pharmacokinetic Modeling/Dosing Simulations
`Monte Carlo simulations were performed to predict
`the average and uncertainty (5% and 95% prediction
`intervals) for simulated plasma PBA, PAA, and PAGN
`concentrations
`in a hypothetical clinical
`trial with
`5,000 cirrhotic subjects dosed with GPB at 9 mL
`(9.9 g) twice daily. A concentration time profile was
`developed for each analyte corresponding to the mean
`as well as the 5% of patients with the highest and low-
`est levels.
`The population PK model and corresponding PK
`parameter estimates used for the Monte Carlo simula-
`tions were developed using Nonmem VI (NONMEM;
`ICON Development Solutions, Ellicott City, MD)
`and PK data from protocols UP 1204-001 and UP
`1204-002 and a phase 2 study in UCD patients (pro-
`tocol UP 1204-003).9 Simulations were preformed
`
`
`
`2080 MCGUIRE ET AL.
`
`HEPATOLOGY, June 2010
`
`Table 1. Activity of Pancreatic Lipases Against GPB
`
`Lipase (U/mg)
`
`With Colipase
`
`Without Colipase
`
`0.5 mM sodium taurodeoxycholate
`PTL (3 lg)
`PLRP2 (20 lg)
`4 mM sodium taurodeoxycholate
`PTL (3 lg)
`PLRP2 (20 lg)
`10 mM sodium cholate
`CEL (10 lg/mL)
`
`618
`35
`
`592
`22
`
`342
`32
`
`42
`11
`
`249
`
`Values are the average of 2 to 3 determinations and are expressed as lmole
`FA released/minute/mg protein or U/mg.
`
`using Trial Simulator software (TS2; Pharsight Corpo-
`ration Inc., Mountain View, CA), assuming dosing at
`8:00 AM and 6:00 PM (to coincide with breakfast and
`dinner), 7 days of dosing to ensure steady state con-
`centrations were achieved, and frequent
`sampling
`(daily samples at 0, 0.25, 0.5, 0.75, 1, 2, 3, 4, 5, 6, 7,
`8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21,
`22, 23, and 24 hours). Because body surface area was
`a significant demographic covariate for clearance and
`volume of distribution parameters in the PK model,
`simulation was used to generate this demographic vari-
`able for each of the 5,000 hypothetical patients.
`
`Results
`In Vitro Hydrolysis of GPB by Pancreatic
`Enzymes
`PTL, PLRP2, and CEL all hydrolyzed GPB (Table
`1). The specific activity (lmole fatty acid released/
`min/mg protein or U/mg) of PTL (600 U/mg) was
`27-fold higher than that of PLRP2 (22 U/mg)
`when both were assayed in the presence of colipase
`and 4 mM sodium taurodeoxycholate and 2.4-fold
`higher than that of CEL (250 U/mg). For compari-
`son, under the same assay conditions,
`the activity
`against tributyrin was 4,600 6 30 U/mg for PTL, 200
`6 9.0 U/mg for PLRP2, and 260 6 12.0 U/mg for
`CEL and against triolein was 1,600 6 153 U/mg for
`PTL, 120 6 40 for PLRP2, and 30.3 6 6.0 for CEL.
`
`Patient Demographics and Disposition
`Twenty-four healthy adults were enrolled in protocol
`UP 1204-001, of whom 22 received each of the study
`drugs and completed the study according to the
`protocol.
`Thirty-two subjects met the entry criteria and en-
`rolled in protocol UP 1204-002 (Supporting Table 1).
`All subjects completed the study and were included in
`the
`analyses. Subject groups were generally well
`
`matched. There were more men than women in
`Child-Pugh groups A and B and equal numbers of
`men and women in Child-Pugh C and the healthy
`volunteer groups. None of the subjects in Child-Pugh
`A or the healthy volunteer group had HE or ascites.
`All subjects in Child-Pugh group B had mild ascites
`and stage I HE, and all subjects in Child-Pugh group
`C had mild or moderate ascites and stage I or II HE.
`Serum albumin, creatinine, and international normal-
`ized ratio were similar in all subject groups. Serum bil-
`irubin increased with the degree of hepatic impair-
`ment,
`ranging from 0.74 mg/dL in the healthy
`volunteers to 3.46 mg/dL in Child-Pugh group C.
`Mean (standard deviation) Child-Pugh and model for
`end-stage liver disease (MELD)
`scores,
`respectively,
`increased commensurate with Child-Pugh grade (A ¼
`5.8 [0.5] and 7.3 [1.3]; B ¼ 8.3 [0.5] and 8.6 [2.1];
`C ¼ 10.6 [0.5] and 12.6 [2.8]) among the cirrhotic
`subjects, and all 32 subjects had negative drug screens
`and alcohol breath test results at all assessments.
`
`Safety and Tolerability (Supporting Table 2)
`Protocol UP 1204-001. Twenty-one adverse events
`(AEs) were reported by 10 subjects while receiving
`NaPBA compared with six AEs by two subjects while
`receiving GPB. The most
`frequently reported AEs
`with NaPBA were dizziness (n ¼ 5), headache (n ¼
`4), and nausea (n ¼ 3). One patient reported epigas-
`tric discomfort and one patient
`reported vomiting
`(n ¼ 2) while taking GPB.
`Protocol UP 1204-002. There were no SAEs or
`AEs leading to withdrawal during the study. Overall,
`AEs were reported in 26 of 32 subjects. Among
`healthy volunteers, five of eight reported AEs, whereas
`seven of eight subjects in each of
`the Child-Pugh
`groups reported AEs. The most common system organ
`class was investigations (18 subjects);
`increased body
`temperature was reported by 10 subjects with cirrhosis
`and decreased platelet count was recorded for four
`subjects in Child-Pugh group A and one subject in
`healthy volunteer group D. Other common classes of
`AEs included gastrointestinal complaints (n ¼ 11) and
`nervous system disorders (n ¼ 8), particularly head-
`ache (n ¼ 7). Most AEs were considered not related
`(n ¼ 9) or possibly related (n ¼ 20) to the study
`medication, and no AEs were considered definitely
`related. Analysis of vital signs including oral tempera-
`ture did not reveal clinically or statistically significant
`changes from baseline. The highest mean temperature
`recorded in any treatment group at any time was
`
`37.2
`C, and the highest temperature recorded in any
`
`
`
`HEPATOLOGY, Vol. 51, No. 6, 2010
`
`MCGUIRE ET AL.
`
`2081
`
`Analyte
`
`Treatment
`
`Cmax (lg/mL)
`Study UP 1204-001: healthy volunteers (single dose ¼ 3 g/m2/day PBA mole equivalent)
`PBA
`NaPBA
`221.0 (44.0)
`GPB
`37.0 (21.74)
`NaPBA
`58.8 (10.37)
`GPB
`14.9 (6.86)
`NaPBA
`63.1 (7.14)
`GPB
`30.1 (8.95)
`Study UP 1204-002: healthy volunteers and cirrhotic patients (dose ¼ 100 mg/kg twice daily)*
`PBA
`Child-Pugh A
`42.81 (25.53)
`Child-Pugh B
`41.83 (26.22)
`Child-Pugh C
`44.33 (21.50)
`Volunteers
`29.80 (14.15)
`Child-Pugh A
`33.15 (14.66)
`Child-Pugh B
`30.85 (19.82)
`Child-Pugh C†
`53.08 (64.49)
`Volunteers
`25.52 (16.05)
`Child-Pugh A
`37.67 (9.33)
`Child-Pugh B
`38.10 (15.20)
`Child-Pugh C
`43.09 (15.27)
`Volunteers
`46.27 (15.07)
`
`Tmax (hours)
`
`0.9 (0.61)
`2.4 (0.8)
`3.9 (0.3)
`4.0 (1.0)
`3.2 (0.4)
`4.0 (0.8)
`
`2.25 (0.65)
`2.88 (0.95)
`3.13 (1.55)
`3.00 (0.76)
`3.75 (1.16)
`4.50 (1.69)
`4.75 (2.38)
`3.63 (0.52)
`3.88 (0.99)
`4.00 (1.69)
`5.25 (2.05)
`4.25 (0.71)
`
`538.2 (111.6)
`133.5 (74.7)
`266.6 (57.4)
`70.90 (36.04)
`379.9 (59.4)
`278.1 (99.1)
`
`131.65 (74.06)
`189.51 (141.55)
`192.08 (82.90)
`132.68 (45.24)
`168.80 (155.76)
`252.35 (299.13)
`579.85 (1150.54)
`130.48 (121.45)
`335.10 (145.63)
`466.89 (359.11)
`578.41 (540.32)
`550.89 (171.53)
`
`Table 2. Plasma Pharmacokinetics of PBA, PAA, and PAGN in Clinical Studies UP 1204-001 and UP 1204-002
`AUCt (lg h/mL)
`
`PAA
`
`PAGN
`
`PAA
`
`PAGN
`
`Values are expressed as the mean (standard deviation).
`Abbreviations: AUCt, area under the plasma concentration curve from time 0 to the last measurable plasma concentration Cmax, maximum plasma concentration;
`Tmax, time of maximum plasma concentration.
`*Protocol UP 1204-002 involved administration of glycerol phenylbutyrate (GPB) only and did not include a sodium phenylbutyrate (NaPBA) comparator arm.
`AUC values represent the AUC from time 0 to the last measurable plasma concentration. (AUCt [(lg/mL)/hour]).
`†The mean PAA Cmax and AUCt are considerably higher than other groups in group C due to one subject who exhibited unusually high level of PAA.
`
`
`C.
`subject at any time point was 38.2
`individual
`Abnormal laboratory safety findings were common in
`subjects with hepatic impairment. There was no con-
`sistently observed pattern among hematology, coagula-
`tion, or chemistry (including liver enzymes), and
`changes after 7 days of dosing with GPB were clini-
`cally insignificant. Clinically significant changes
`in
`electrocardiogram were not observed with GPB dosing,
`nor were changes observed in the QTc intervals.
`
`Pharmacokinetic Analyses
`Protocol UP 1204-001. NaPBA resulted in higher
`plasma levels (both Cmax and AUC) of PBA, PAA, and
`PAGN than GPB; the 90% confidence intervals for the
`ratio of geometric means of each metabolite following
`GPB compared with NaPBA extended below the com-
`monly used lower bioequivalence level of 0.8 (Fig. 2
`and Table 2). The mean plasma half-lives of PBA,
`PAA, and PAGN were 0.7 (6 0.1) hours, 1.2 (6 0.2)
`hours, and 1.7 (6 0.5) hours,
`respectively, after
`NaPBA administration. The mean plasma half-life of
`PBA after GPB administration was 1.9 (6 1.7) hours
`and ranged from 0.8 to 7.4 hours. The plasma half-
`lives for PAA and PAGN after GPB administration
`ranged from 1.0 to 1.8 hours and 1.9 to 16.9 hours,
`respectively.
`
`Urinary excretion of PAGN was higher following
`NaPBA than following GPB (Table 3). However, uri-
`nary collection of PAGN was incomplete at 24 hours
`following GPB dosing, as PAGN was still detectable in
`plasma at 24 hours. In contrast, PAGN following
`NaPBA dosing was undetectable in the plasma by 24
`hours and PAGN elimination considered complete by
`24 hours. Taking into account the pattern of urinary
`excretion and plasma levels following the single dose
`arm of study UP 1204-002, the 0-48 hours urine col-
`lection was split into 0-24 and 24-48 hours to calcu-
`late the percentage of urinary PAGN that occurred af-
`ter 24 hours (Table 4). It is estimated that 15% of
`urinary PAGN excretion in patients on GPB occurred
`beyond 24 hours, the time the collection was termi-
`nated for study UP 1204-001. When corrected for the
`
`Table 3. Urinary Output of PAGN in Study UP 1204-001
`(Healthy Volunteers)
`
`NaPBA
`
`HPN-100
`
`HPN-100c*
`
`PAGN (0-24 hours)
`amount excreted
`
`4,905.0 (1414)
`
`4,130.3 (925)
`
`4,749.9
`
`All values are expressed as the mean (standard deviation). Single dose:
`3 g/m2/day PBA mole equivalent.
`*HPN-100c value is corrected for approximately 15% under collection of uri-
`nary PAGN. PAGN was detectable in plasma samples of subjects receiving GPB
`but not NaPBA after the 24-hour time point, indicating that urinary collection of
`PAGN was incomplete at 24 hours following GPB dosing.
`
`
`
`2082 MCGUIRE ET AL.
`
`HEPATOLOGY, June 2010
`
`Table 4. Urinary Output of PAGN in Study UP 1204-002 (Healthy Volunteers and Cirrhotic Patients)
`
`Child-Pugh A (n 5 8)
`
`Child-Pugh B (n 5 8)
`
`Child-Pugh C (n 5 8)
`
`Healthy Volunteers (n 5 8)
`
`PAGN after dosing on day 1 (0-24 hours)
`Amount excreted (lmol)
`Proportion excreted post-24 hours (%)*
`
`PAGN after dosing on day 1 (0-48 hours)
`Amount excreted (lmol)
`Mole % of dose excreted
`
`PAGN after dosing on day 8 (0-12 hours)†
`Amount excreted (lmol)
`Mole % of dose excreted
`
`15,553 (4,201)
`15.5 (7.5)
`
`16,847 (4,326)
`14.5 (13.3)
`
`19,291 (12,054)
`8.8 (6.0)
`
`14,903 (4,292)
`16.9 (10.5)
`
`18,386 (4,961)
`47.1 (10.4)
`
`19,902 (4,505)
`44.9 (9.7)
`
`20,854 (15,281)
`48.5 (29.4)
`
`17,869 (4,312)
`42.2 (11.4)
`
`16,068 (7,397)
`40.6 (16.4)
`
`13,179 (5,786)
`29.9 (13.0)
`
`15,428 (6,519)
`44.6 (24.2)
`
`10,195 (4,189)
`24.5 (11.4)
`
`PAGN after dosing on day 15 (0-48 hours)
`Amount excreted (lmol)
`31,431 (15,291)
`25,152 (11,426)
`Mole % of dose excreted
`79.6 (30.5)
`58.2 (29.2)
`All values are expressed as the mean (standard deviation). Dose ¼ 100 mg/kg HPN-100 twice daily.
`*Values do not represent % dose; instead, they represent the amount of PAGN yet to be eliminated in urine expressed as a percentage of total PAGN eliminated
`over 48 hours. For example, 15,553/18,386*100 ¼ 15.5% for Child-Pugh A patients.
`†Urine collection was performed after the first HPN-100 dose (100 mg/kg) in the morning.
`
`30,752 (20,860)
`85.0 (65.1)
`
`28,716 (8223)
`68.6 (21.9)
`
`estimated 15% excreted after 24 hours, PAGN excre-
`tion was essentially the same following NaPBA and
`GPB administration (Table 3).
`Protocol UP 1204-002. Intact GPB was not
`detected in the systemic circulation; nor were the
`minor metabolites PAG, PBG, and PBGN. AUC0-t
`and Cmax for PBA and PAA tended to be higher in
`Child-Pugh groups B and C than in Child-Pugh group
`A or in the healthy subjects group, but these changes
`were not
`statistically significant
`(Table 2). Plasma
`PAGN levels did not differ among the study groups.
`No consistent differences between cirrhotic subjects
`and healthy subjects were observed in the plasma PK
`variables examined on days 1 (single dose and fasting)
`or 15 (after multiple doses and at steady state). There
`were also no statistically significant differences in the
`PK characteristics of GPB when given after fasting
`(day 1) or with a meal (day 8).
`Plasma PBA concentrations returned to near predose
`levels between doses during multiple dosing days 8-15
`and did not reach steady state. By contrast, PAA and
`PAGN predose concentrations
`increased during the
`first 2 to 4 days of multiple dosing but did not increase
`consistently thereafter, indicating that steady state had
`been reached (Fig. 3). After dosing on day 15, the
`extent of exposure to PAA, but not PBA, significantly
`correlated with hepatic impairment,
`increasing with
`worsening MELD score. During multiple dosing, PAA
`accumulation in Child-Pugh C cirrhotic
`subjects
`exceeded that in other groups. However, this trend was
`attributable to a single Child-Pugh C subject
`that
`showed unusually high levels of PAA assessed as Cmax
`and AUC (208.8 lg/mL and 2,245.51 [(lg/mL)/
`hour], respectively) after GPB administration, com-
`
`subject, who
`subjects. This
`pared with all other
`received GPB at a dose of 7 mL twice daily, exhibited a
`clinical profile similar to the other Child-Pugh C sub-
`jects. Reanalysis omitting this subject’s data resulted in
`mean PAA Cmax and AUC levels for the Child-Pugh C
`group similar to other subject groups. During repeated
`dosing, similar but less profound patterns of increased
`PAA levels compared with their group mean were
`noted in one healthy subject (AUC 420.32 [(lg/mL)/
`hour], Cmax 61.31 lg/mL) and one Child-Pugh B sub-
`ject (AUC 938.85 [(lg/mL)/hour], Cmax 65.40 lg/
`mL).
`Urinary PAGN. PAGN was the major metabolite
`excreted: 42%-49% of the GPB dose administered was
`excreted as PAGN on day 1, 25%-45% on day 8, and
`58%-85% on day 15 (Tables 3 and 4). Very low
`amounts of PBA and PAA were excreted in the urine
`(0.05% of the total GPB dose). There were no stat-
`istically significant differences in the amount of PAGN
`excreted between any of the Child-Pugh groups and
`the healthy subjects. Urinary PAGN excretion was sig-
`nificantly greater in all groups after multiple dosing
`compared with single dosing, a result consistent with
`the larger daily GPB doses and higher plasma PAA
`and plasma PAGN observed during the first 2 to 4
`days of multiple dosing, after which steady state
`appeared to have been reached.
`
`Dosing Simulation
`Simulations of 9 mL bid dosing were consistent with
`the pharmacokinetic findings observed in protocol UP
`1204-002. PBA levels did not accumulate with repeated
`dosing, and PBA trough concentrations were predicted
`to be at or near baseline levels. PAA and PAGN, by
`
`
`
`HEPATOLOGY, Vol. 51, No. 6, 2010
`
`MCGUIRE ET AL.
`
`2083
`
`GPB is likely similar to dietary triglycerides. Because
`PLRP2 and CEL, unlike PTL, are expressed at birth,
`the findings further suggest that GPB may be digested
`by newborns, which is of particular
`relevance to
`UCDs that can present shortly after birth.12-20 This
`requires confirmation in clinical studies.
`The safety and tolerability of GPB with short-term
`dosing was generally satisfactory and comparable to
`NaPBA. There were no SAEs or deaths and no clini-
`cally significant changes in laboratory parameters
`Metabolite peak blood levels were lower after single-
`dose
`administration to healthy
`adults of GPB
`
`Fig. 2. Twenty-four-hour mean plasma concentration time profile of
`metabolites PAA (~), PAGN (*), and PBA (h) following administra-
`tion of either NaPBA at a dose of 3 g/m2 or the equivalent dose of
`PBA as GPB in clinical study UP 1204-001. The concentrations have
`been normalized by body surface area and expressed in molar units
`to allow direct comparison of the PK profiles between the analytes.
`
`contrast, did exhibit accumulation with repeated dosing
`and achieved steady state within 4 days. Simulations
`projected the median PAA concentration to be several-
`fold lower than the no observed adverse event level in
`primates (331 lg/mL [results not shown]) or the levels
`reported to be associated with neurological symptoms
`in human studies10,11 (Fig. 4).
`
`Discussion
`
`GPB was hydrolyzed by all pancreatic enzymes
`tested, including PTL, CEL, and PLRP2 in order of
`specific activity. Compared with GPB, the specific ac-
`tivity of PTL is 8-fold higher against
`tributyrin,
`which has the same fatty acid chain length as GPB but
`lacks the phenyl group, and is 2.5-fold higher against
`triolein, a major dietary fat and physiological substrate.
`These findings suggest that the intestinal handling of
`
`Fig. 3. Concentrations of plasma metabolites PBA, PAA, and PAGN
`following multiple-day (100 mg twice daily for 7 days) administration
`of GPB in clinical study UP 1204-002. Group A (l) ¼ Child-Pugh A.
`Group B (^) ¼ Child-Pugh B. Group C (n) ¼ Child-Pugh C. Note
`that the higher PAA levels in the Child-Pugh C group is due primarily
`to one subject with unusually high PAA levels.
`
`
`
`2084 MCGUIRE ET AL.
`
`HEPATOLOGY, June 2010
`
`PAA is of particular interest in that neurological tox-
`icity has been reported in cancer subjects administered
`high doses of PBA or PAA intravenously and were asso-
`ciated with PAA blood levels ranging from 499-
`1,285 lg/mL.10,11 Although the blood levels observed
`in cirrhotic subjects were well below the levels reported
`by Thibault and colleagues to be associated with neuro-
`logical symptoms, Monte Carlo simulation analyses
`were conducted to further assess the anticipated PAA
`plasma concentrations in a hypothetical clinical trial
`involving 5,000 cirrhotic subjects administered GPB in
`a daily dose of 9 mL bid, which is slightly above the
`equivalent maximum approved dose of NaPBA for
`UCD patients (20 g/day of NaPBA is equivalent to
`17.4 mL of GPB). As depicted in Fig. 4, Monte Carlo
`simulation indicated that even the highest 5% of pre-
`dicted PAA concentrations were well below those asso-
`ciated with toxicity in phase 1 oncology studies.10,11
`Urinary PAGN is also of particular interest because it is
`stoichiometrically related to nitrogen scavenging. Cirrhotic
`subjects excreted essentially the same amount of PAGN as
`healthy adults, indicating that GPB should enhance excre-
`tion of waste nitrogen and lower ammonia in patients with
`cirrhosis. Assuming that dietary protein is 16% nitrogen
`by weight, that approximately 47% of dietary nitrogen is
`excreted as waste nitrogen, and that 60% conversion of
`PBA delivered as GPB is converted to urinary PAGN, then
`a GPB dose of 9 mL twice daily—which is essentially the
`same as the maximum approved dose in UCD patients—
`would be expected to mediate excretion of waste nitrogen
`associated with 26 g dietary protein, or 0.4 g/kg body
`weight for a 70-kg adult male.16 This would represent a
`substantial contribution to waste nitrogen excretion in cir-
`rhotic subjects, in whom urea synthetic capacity is known
`to be impaired, and would likely compare favorably with
`the presumed reduction in ammonia production with
`agents commonly used for HE such as nonabsorbable di-
`saccharides or antibiotics.21,22 The effect of nonabsorbable
`disaccharides or antibiotics on nitrogen metabolism is diffi-
`cult to measure, and there is not a clinically useful bio-
`marker for these agents such as urinary PAGN for GPB.22
`These considerations, in conjunction with the results of a
`phase 2 study that suggest that GPB is at least as effective
`as NaPBA in lowering blood ammonia in UCD patients,9
`suggest that GPB has the potential to lower blood ammo-
`nia in patients with cirrhosis and warrants further explora-
`tion for the treatment of HE.
`
`References
`1. US Department of Health and Human Services. Dietary Guidelines for
`Americans, 2005. http://www.health.gov/DietaryGuidelines/dga2005/
`document/default.htm.
`
`Fig. 4. Simulated median plasma PAA concentrations projected to
`occur with continuous dosing of GPB at a dose of 9 mL twice daily for
`7 days in 5,000 patients with cirrhosis. The straight, solid horizontal
`line corresponds to a concentration of 331 lg/mL, a conservative
`estimate of PAA levels expected not to be associated with AEs based
`testing. The median PAA Cmax at steady state levels
`on preclinical
`(solid line) and the highest 95% prediction interval (upper dashed
`line) were both below this level, as was the lowest 95% interval (lower
`dashed line).
`
`compared with NaPBA, and urinary excretion of
`PAGN was prolonged. These findings are consistent
`with the gradual release of PBA by pancreatic lipases
`when delivered in the form of GPB compared with
`NaPBA, and the absence of intact GPB in blood or
`urine suggests that intestinal breakdown of GPB is
`complete. Although the