Multicenter, Randomized Phase 2B Study to Evaluate the Efficacy, Safety and Tolerability of OCR-002 (ornithine phenylacetate) in Hospitalized Patients with Cirrhosis and Associated Hyperammonemia with an Episode of Hepatic Encephalopathy (STOP-HE Study)

Hyperammonemia and Associated Hepatic encephalopathy
Hepatic encephalopathy (HE) is the most common complication of cirrhosis; it
has a significant negative effect on survival even after liver transplantation
(Bustamante 1999) and is associated with irreversible impairment in cognitive
function (Garcia-Martinez 2011). An estimated 60*70% of cirrhotic patients have
at least subtle signs of neurocognitive impairment, and HE is the principal
diagnosis in hospitalized patients. Overt HE has a prevalence of approximately
30% in the cirrhotic population, and accounts for about 150,000
hospitalizations annually in the United States (Al Sibae 2009).
HE is a neuropsychiatric disorder that occurs when gut-derived toxins,
primarily ammonia, bypass a failing liver that would normally detoxify such
agents, enter the circulation, and then cross the blood-brain barrier,
resulting in impairment of neurotransmission and central nervous system (CNS)
function. HE can arise in the setting of acute liver failure, chronic
progressive liver disease resulting in cirrhosis, or as a result of portocaval
shunting in the absence of any true liver disease. The severity of HE
correlates with elevated ammonia levels (Ong 2003, Bernal 2007), supporting the
need for novel, safe, and effective ammonia-lowering therapies to treat as well
as to prevent episodes of HE.
Ammonia is generated in the intestines from nitrogenous compounds from the
diet, deamination of glutamine by glutaminase, and metabolism of nitrogenous
substances by colonic flora. In normal circumstances, most ammonia is
metabolized to urea in the liver. Portal-systemic shunts and liver failure
cause a rise in blood ammonia that may affect brain function by inducing
several disturbances in astrocytes; these may impair mitochondria and the
glutamate-glutamine trafficking between neurons and astrocytes. Skeletal muscle
is capable of decreasing blood ammonia by metabolizing ammonia to glutamine.
The kidney has also an important role in determining blood ammonia by excreting
urea in the urine (Cordoba and Minguez, 2008).
In healthy people, ammonia is produced mostly from ingested protein, as well as
endogenous protein catabolism in skeletal muscle. Protein digestion is not 100%
efficient, and even though most protein is absorbed as amino acids, some amino
acids remain unabsorbed within the lumen of the gut and reach the colon.
Certain colonic bacteria, mainly coliforms and anaerobes, convert ingested
amino acids into ammonia. The ammonia is then absorbed, and approximately 90%
of it enters the portal circulation and is brought to the liver where it gets
converted to urea via the urea cycle and excreted in urine. Only about 10% of
absorbed ammonia normally bypasses the liver and is metabolized by peripheral
tissues, including kidney, heart, and brain (Meijer 1990). Liver failure is the
most common cause of elevated ammonia, and can result from numerous conditions
from cirrhosis to inborn genetic disorders such as urea cycle disorders.
In patients with inherited deficiencies of the urea cycle, the liver partially
compensates by converting ammonia and glutamate into glutamine, an action
catalyzed by the enzyme glutamine synthetase. This is an incomplete response to
the problem, as glutamine is converted back to glutamate and ammonia in other
organs via glutaminase.
In patients with cirrhosis, arterial ammonia concentrations can be increased 2*
3 fold. Patients are at risk of developing multifactorial disturbances in
neuropsychiatric and neurocognitive function: the chronic elevation in
circulating ammonia induces brain swelling as well as multiple alterations in
neurotransmitter function, including increases in gamma-aminobutyric acid,
endogenous benzodiazepine-like compounds, endogenous opioids, and
pro-inflammatory cytokines; and decreases in serotonergic neurotransmitters and
endogenous antioxidants (Al Sibae 2009). Additionally, the peripheral immune
system communicates with the brain in response to infection and inflammation.
Astrocytes and microglial cells release cytokines in response to injury or
inflammation (Prakesh and Mullen, 2010). Findings from studies in rats have
indicated that the rise in blood levels of tumor necrosis factor (TNF) that
occurs during inflammation stimulates glial cells to secrete the cytokines
interleukin(IL)-1 and IL-6. TNF also compromises the endothelial blood*brain
barrier and IL-1* affects the integrity of the glial side of the blood-brain
barrier. Both TNF and IL-6 enhance fluid-phase permeability of isolated brain
endothelial cells in vitro, and TNF also increases the diffusion of ammonia
into astrocytes (Prakesh and Mullen, 2010). These effects accentuate and
contribute to manifestations of hepatic encephalopathy.
Cirrhotic patients with HE often suffer sleep-wake abnormalities, cognitive
changes, and impaired motor function that impact their day-to-day lives and can
become severe enough to lead to coma and require hospitalization. Acute or
overt HE (HE episode resulting in hospitalization) affects about 150,000
patients per year in the United States, and there are no FDA-approved effective
parenteral treatments. The current standard of care in HE treatment focuses on
reducing ammonia uptake from the gastrointestinal (GI) tract and treating the
precipitating factor. In patients hospitalized with HE, a reversible
precipitating factor is identified in approximately 50% of cases. These factors
include GI bleeding, infection, dehydration, electrolyte disturbances, an
increase in dietary protein load, and use of sedative-hypnotic drugs.

Dietary protein restriction had long been advocated as a strategy to reduce
circulating ammonia in these patients. However, recent data have shown that
this strategy is not effective in preventing HE and may harm these patients by
making them more prone to muscle wasting (Cordoba 2004). Pharmacological
approaches include non-absorbed disaccharides (e.g. lactulose) and antibiotics,
(e.g. neomycin, rifaximin), which reduce ammonia production in the GI tract.
These strategies have been shown to be effective in preventing recurrent
episodes of HE, but not in the acute care of hospitalized cirrhotic patients
with HE (Bass 2010, Als-Nielsen 2004). L-ornithine L-aspartate (LOLA) is
available as an IV product in Europe and Asia and has shown to benefit patients
by trapping circulating ammonia in the form of glutamine (Kircheis 1997),
although benefit in acute liver failure was not demonstrable (Acharya et al.,
2009). Phenylacetate (PAA) and its prodrug phenylbutyrate have been used
successfully to reduce ammonia in patients with genetic urea cycle disorders as
these patients have very high circulating glutamine levels. The approach of
using only PAA or phenylbutyrate to reduce high ammonia loads is not expected
to work as effectively in patients with chronic liver disease, because these
patients typically have lower circulating glutamine levels (reduced expression
of glutamine synthetase), although recent data of oral prophylaxis to prevent
recurrent hepatic encephalopathy with glycerol phenylbutyrate have shown
promising results (Rockey et al., 2012). Additionally, the risk of chronic
treatment and sustained glutamine depletion in cirrhotic patients with already
poor lean muscle mass remains a concern.

OCR-002 (ornithine phenylacetate) in the form of the phenylacetate (PAA) salt
of L-ornithine (ORN) is a single new chemical entity that allows for
alternative pathways for the excretion of ammonia in the setting of cirrhosis
potentially producing an enhanced effect on elimination of ammonia (Jalan
2007). ORN stimulates the activity of glutamine synthetase, inducing body
muscle to trap circulating ammonia in the form of glutamine, which is a
non-toxic carrier of ammonia. Glutamine is then conjugated with PAA to form
phenylacetylglutamine (PAGN), which is excreted in urine. This strategy
prevents the eventual recirculation and degradation of glutamine by glutaminase
and avoids re-formation of ammonia.
Because OCR-002 works simultaneously on stimulation of glutamine production
(via ORN) and glutamine conjugation and excretion (via PAA), this approach is
expected to be effective in removing circulating ammonia.
L-ornithine is a precursor for glutamate production and reduces ammonia via
inducing glutamine production in muscle. In chronic liver diseases, increasing
glutamine production has not proven to be a durable ammonia-reducing strategy
due to the glutaminase enzyme back converting glutamine into the component
parts of glutamate and ammonia.

OCR-002 is the phenylacetate salt of L-ornithine, providing both L ornithine
(ORN) and phenylacetate (PAA). PAA covalently binds glutamine via enzymatic
conjugation and prevents it from being back converted into glutamate and
ammonia. OCR-002 thus allows for alternative pathways for the efficient
excretion of ammonia.
OCR-002 has been granted orphan drug status by the United States (US) Food and
Drug Administration (FDA) and was granted fast track designation for the
treatment of hyperammonaemia and resultant hepatic encephalopathy.
1.2 Rationale for the Investigational Use of OCR-002 in HE

Rationale for Dose Selection
Ocera Therapeutics Inc. (Palo Alto, CA) has conducted two clinical trials
evaluating safety, tolerability and pharmacokinetics in healthy volunteers
(n=48) and stable cirrhotic patients (n=43) (OCR-002 HV201 and OCR-002 HE201)
[Total patients dosed with OCR-002 was 77].
Clinical study OCR-002 HV201 was designed as a randomized, placebo-controlled,
dose-ranging study, evaluating doses (IV infusion) of 1, 3, 10, 20 and 30g
administered over 4h, and 30, 40 and 60g administered over 24h in healthy
subjects. Following a single ascending dose phase, a multiple ascending dose
phase (5 days of consecutive administration) for doses of 1, 3, 10 and 20g
administered over 4h was assessed. Clinical study OCR-002 HE201 was designed as
a randomized, placebo-controlled dose ranging study, evaluating doses (IV
infusion) of 1, 3, 10, 20 and 40g administered over 4h and 10, 20 and 40g
administered over 24h in stable cirrhotic patients.

Key findings from both studies are summarized as follows:
* There were no deaths, SAE or AE discontinuations.
* Dose-related increases in the incidence and severity of adverse events was
observed.
* Commonly occurring adverse events that appeared to be related to dose and
phenylacetate Cmax included dizziness, somnolence, headache, nausea/vomiting
and tinnitus.
* There was no evidence to suggest that AEs increased or new AEs occur with
multiple dosing.
* With respect to blood pressure, there were no apparent trends from placebo
with respect to mean changes from baseline; however at dose levels *20g/4h,
several subjects experienced clinically significant orthostatic hypotension
(this is probably not relevant to current dosing paradigm [24h continuous
infusion] as the effect was noted at 6 times planned infusion rate).
* 24h Holter monitor recording identified dose and exposure-related increases
in mean heart rate at infusion rates of 2.5g/h and higher (* 3-fold higher rate
than the proposed highest dose in current study). In healthy volunteers, these
effects were accompanied by modest effects on ventricular repolarization as
measured by QTcF, however were confounded by the inability of the Fridericia
formula to appropriately correct for increases in heart rate. The increase in
heart rate was also demonstrated in cirrhotic patients at infusion rates >
2.5g/h (more than 3-fold higher rate than proposed highest dose in current
study). Importantly, no associated effects on ventricular repolarization as
measured by QTcF were recorded in this population.
* 20g administered over 4 hours was the maximum tolerated dose (MTD). Doses of
10g/4h (2.5g/h) and below were well tolerated in both study populations. For
24-hour infusions, 40g administered over 24 hours was the MTD.
* There were no clinically significant abnormal laboratory parameters and no
patient had any clinically significant changes in physical examination findings
or mean change in body weight between baseline and follow up.
* Doses higher than 10g/4h were associated with adverse effects that appear
related to phenylacetate plasma concentrations (Cmax). Thus slower 24-hour
infusions were considered optimal for the present study.
* A slower infusion rate (24hr-infusion) avoids Cmax related adverse effects
while providing similar waste nitrogen excretion as measured by urinary output
of PAGN. 40g administered over 24 hours was the MTD for continuous infusion,
and in the current study the highest planned dose level (20g/24h) is half of
this MTD.

Both 10g/24h and 20 g/24h as a continuous infusion of 0.417 g/h (or 0.833 g/h),
were well tolerated in the HE-201 (compensated cirrhosis) study. The OCR-002
dose of 10g/24h has also been well tolerated in open-label Part A of the study
reported by Cordoba and colleagues (Ventura-Cots, ISHEN 2012) in 10 patients
without hepatic encephalopathy who had cirrhosis and acute gastrointestinal
bleeding (7 Child-Pugh A; 3 Child-Pugh B) with a salutary effect on ammonia,
and no intolerance or safety issues. Preliminary blinded safety data from the
double-blind, placebo-controlled portion of this study (Part B of OP-GIB) in 17
patients including Child-Pugh A, B, C cirrhotics with upper gastrointestinal
bleeding receiving an OCR-002 dose of 10g/24h for 5 days has shown a benign
safety profile to date.

However, in the setting of Study OCR002-HE209 wherein patients are hospitalized
with acute overt hepatic encephalopathy utilization of higher dose levels than
10 g/24h OCR-002 seems necessary to provide potential benefit. Newly available
pharmacokinetic data from OP-GIB Part A and Part B (currently in aggregate n=21
patients) indicated that average plasma phenylacetate exposure (Cave) was
substantially below that achieved with glycerol phenylbutyrate (RavictiTM) at
the dosage (6.6 g twice daily) employed in their seminal HALT-HE study which
showed statistically significant reduction in HE episodes via the same
mechanism of action, ammonia reduction via urinary PAGN elimination (Ghabril
2013; Rockey 2014). Specifically, in the OP-GIB study plasma PAA Cave was 30 µg/
mL (mean), which is more than 3-fold lower.
OCR-002 has a molar ratio of ornithine to phenylacetate of 1:1, which given the
nearly identical molecular weight for ornithine and phenylacetate means that 10
g of OCR-002 contains 5 g of phenylacetate, and 20 g of OCR-002 contains 10 g
of phenylacetate. Essentially the above OP-GIB PK results confer a need to
provide OCR-002 dosages in the range of 20 g/24h to achieve the anticipated
optimal effect on hepatic encephalopathy. This dose is well within the MTD
delineated above for OCR-002, and would provide a phenylacetate dosage (10
g/24h) which is within the approved labeling in terms of phenylacetate content
for intravenous Ammonul®(sodium phenylacetate, sodium benzoate for urea cycle
disorders), in the range of the RavictiTM dosage employed for the HALT-HE study
which demonstrated efficacy in prevention of HE episodes, and less than that
employed in oncology studies wherein phenylacetate infusions were maintained
for 14 days (Thibault 1995; Chang 1999). Further details are provided below.
The recommended adult dose of Ammonul® (sodium phenylacetate and sodium
benzoate intravenous injection for urea cycle disorders) is 5.5 g/m2 infused
over 1.5 to 2 hours as a loading dose followed by 5.5 g/m2 infused over 24
hours as the maintenance dose. With an estimated average body surface area
(BSA) of 1.73 m2 for adults, the dose for sodium phenylacetate in Ammonul®
given in the first day is approximately 19 g (16.4 g phenylacetate), nearly
double the phenylacetate content at the highest dose level in the current
Protocol OCR002-HE209. Similarly, the currently prescribed maximum total daily
dose for RavictiTM (glycerol phenylbutyrate, GPB) for urea cycle disorders is
17.5 mL, which is equivalent to approximately 17.9 g of phenylbutyrate (PBA),
and equivalent to approximately 14.8 g of phenylacetate (PAA) per day based on
direct molecular weight conversion.
In the HALT-HE clinical trial of RavictiTM as HE prophylaxis in patients with
cirrhosis and a prior history of HE, a 6 mL twice daily dose of GPB was found
to significantly reduce the proportion of patients with an HE event,
correlating with significant reduction of plasma ammonia levels from baseline
(Rockey 2014). This dose of 6 mL of RavictiTM twice daily translates to
approximately 12.2 g PBA dose per day, and 10.2 g PAA dose per day assuming
complete bioavailability. This is essentially the same PAA content as the
highest dose level of OCR-002 (20 g/24h) proposed in Study OCR002-HE209.
As mentioned above, pharmacokinetic data from Study OP-GIB obtained to date
show that plasma exposure to phenylacetate following OCR-002 administration at
10 g/24h is substantially lower than that observed following administration of
RavictiTM or Ammonul® at doses prescribed in the product labels or in relevant
literature as summarized in the table in section 1.2 of the protocol. We
interpret this finding to be the result of a lower phenylacetate dosage (5 g)
with OCR-002 (infused at 10 g/24h) than provided with either RavictiTM or
Ammonul® and the fact that OCR-002 is administered as a slow continuous
infusion without a loading dose.

The summary of the data (Table 1 protocol section 1.2) shows that plasma PAA
exposure following administration of OCR-002 at a rate of 10g/24h was
substantially lower than after Ammonul® or RavictiTM administration, even after
PAA dose normalization. It is well documented that PAA disposition follows
non-linear kinetics, with saturable conjugation of PAA with glutamine to form
phenylacetylglutamine (PAGN). The IV loading dose of Ammonul® and an oral dose
of RavictiTM would be expected to yield initial higher plasma PAA
concentrations due to potential saturation of the conjugation pathway than an
equivalent phenylacetate dose administered as a constant slow infusion. The
total body clearance (CL) of PAA after OCR-002 administration in the OP-GIB
study was apparently much higher than that after Ammonul® administration,
suggesting that the OCR-002 (PAA) dosing rate may not be high enough at 10
g/24h (5g/24h PAA) to maintain efficient conjugation of PAA with glutamine to
form PAGN. This is shown by the lower plasma PAGN concentrations observed
after OCR-002 administration in Study OP-GIB (mean Cave: 19.2 *g/mL) as
compared to RavictiTM (mean Cave: 46.3 *g/mL) (Ghabril 2013). Thus it was
considered necessary in Study OCR002-HE209 to consider a dose of OCR-002 of 20
g/24h (10 g PAA) as the highest dose level.

Furthermore, the OCR-002 dosage for each patient will be predicated upon dose
level adjustment per severity of liver impairment. By analogy with
pharmacokinetic data in the RavictiTM 2013 product label summarizing hepatic
impairment studies in patients without overt encephalopathy, participants in
the current study with greater degrees of hepatic synthetic and portal
impairment will receive lower dosages of OCR-002, specifically 15 g/24h and 10
g/24h, respectively, because (per the RavictiTM label) Child-Pugh class B and C
cirrhotics showed a progressive increase (about 30%) in PAA plasma exposure
when compared to Child-Pugh class A. Given the decades-long experience with
intravenous phenylacetate in various clinical settings (urea cycle disorders,
oncology trials) and available OCR-002 data to date, this dose level adjustment
by extrapolation is not undertaken because of any observed medical or clinical
issue but rather to assure a considered, conservative, and temperate approach
and to provide objective reassurance regarding dosing of medically delicate
hospitalized patients with greater degrees of hepatic impairment in this
specific context.

Notably, Thibault et al. (1994) previously described PAA kinetics in cancer
patients and suggested a threshold for CNS effects of PAA of >500 µg/mL. This
level of PAA concentration is approximately 10-fold higher than the mean PAA
concentrations observed when administering 10g OCR-002 over 24h, and well above
that observed at 20 g/24h. The liver is not the primary organ responsible for
metabolism of phenylacetic acid to PAGN since renal tissue metabolic activity
accounts for the highest percentage of phenyl CoA-acyltransferase activity with
the ratio of renal to liver metabolism of approximately 3:1. Additionally, in
preclinical models of acute liver failure (ALF) (anhepatic porcine model)
OCR-002 was effective in lowering ammonia levels and was metabolized even in
the absence of a liver (Ytrebo et al., 2009).

These data taken together combined with the a priori OCR-002 dose level
adjustment predicated on hepatic synthetic and portal impairment point score at
entry provide a strong level of assurance that an appropriate benefit/risk will
be maintained for participating patients hospitalized with hepatic
encephalopathy in acute need of intervention. The Independent Data Monitoring
Committee will review pharmacokinetic data as delineated in Section 3 and are
empowered to provide dose level adjustment while the study is ongoing as
outlined in Section 5.6 in the case of over-exposure or suboptimal dosage.

The primary objective of this study is the following:
**To evaluate the efficacy of OCR-002 for treatment of an acute hepatic
encephalopathy
episode in cirrhotic patients requiring hospitalization
**To evaluate the safety and tolerability of OCR-002 in hospitalized cirrhotic
patients with
an acute episode of hepatic encephalopathy

The secondary objectives include the following:
**To confirm the pharmacokinetic (PK) profile of OCR-002 in this patient
population
**To assess the kinetics of reduction of plasma ammonia with OCR-002 and
excretion of
phenylacetylglutamine (PAGN) in urine

This is a Phase 2B, multicenter, randomized, double-blind, placebo-controlled
study evaluating OCR-002 administered via continuous 24-h intravenous (IV)
infusion to hospitalized patients with cirrhosis, hyperammonemia, and an acute
episode of hepatic encephalopathy Stage 2-4,
inclusive (Appendix A, Hepatic Encephalopathy Staging Tool). A total of
approximately 140 unique patients (± 10%) will be stratified at time of
randomization by 1) model for end-stage liver disease (MELD) score (less than
or equal to 30 versus > 30), and 2) Hepatic Encephalopathy
Staging Tool (Stage 2 versus Stage 3/4) using an interactive voice/web-response
randomization system. A third stratification (in North America) will group
investigational centers into two categories as follows: liver transplantation
centers performing * 70 transplants per year, and liver
transplantation centers performing < 70 transplants per year or centers not
performing liver transplantation. Outside of North America stratification will
be employed with the same MELD and Hepatic Encephalopathy Staging Tool
thresholds delineated above. Patients will be
randomized (1:1) to OCR-002 or placebo given as a continuous infusion for 5
days which will be administered on top of standard of care inclusive of
lactulose/lactitol per usual institutional practice. Breakthrough HE occurring
during prophylaxis with rifaximin for * 1 week duration is
allowed (de novo treatment of HE episode with rifaximin is not allowed). There
is no protocol stipulation for in-hospital location of participating patients
who may be hospitalized for instance in the intensive care unit, step-down
unit, hospital floor etc, according to medical condition and
usual clinical triage. Patients can be discharged from hospital if medically
appropriate before 5 days (120 hours) of continuous infusion is completed. For
patients remaining in hospital for standard of care safety/efficacy is assessed
24-hours post end-of-last-infusion, for patients
subsequently discharged during the follow-up period an additional evaluation is
performed immediately prior to discharge, and all patients have a follow-up
visit scheduled 2 weeks after cessation of IV infusion of study drug (Study Day
19). Assessments of hepatic encephalopathy, specifically the Hepatic
Encephalopathy Staging Tool, Glasgow Coma Scale, and modified orientation log
will be evaluated by specifically trained investigators, subinvestigators,
nurse practitioners, or physician*s assistants. The Physician Overall Treatment
Evaluation of Hepatic Encephalopathy and Physician Ranked Assessments (specific
items) performed in the time interval between end-of-final-infusion and 3 hours
post end-of-final-infusion must be performed by a physician, and must be the
same assessor as the Screening or Baseline evaluator so that the ranking
relative to pretreatment status (same, better, worse) will be accurate. OCR-002
or matching placebo will be infused IV for 5 days (500 mL/24h at a rate of 20.8
mL/h); study drug dosage will be predicated upon Baseline calculation of
hepatic synthetic and portal
elements (depicted below). The interactive voice/web-response randomization
system will also autocalculate the Child-Pugh score for patients based on the
entries for hepatic encephalopathy stage (Appendix A) plus the hepatic
synthetic and portal elements of Child-Pugh tabulated below (for consistency of
measure).

Parameter Points (circle patient*s status for each item then total all 4 items
below)

1
2
3
Ascites
None Mild/Moderate (diuretic-responsive) Tense
(diuretic-refractory)
Total bilirubin (mg/dL) < 2 (< 34 µmol/L) 2-3 (34-51 µmol/
L) > 3 (> 51 µmol/L)
Albumin (g/dL) > 3.5 (> 35 g/L) 2.8-3.5 (28-35
g/L) < 2.8 (< 28 g/L)
PT (sec prolonged) or INR < 4
4-6 > 6
<
1.7

1.7-2.3
> 2.3
Patient Total Score (all 4 items added together)

For 4-6 points using the above scoring system the OCR-002 dose will be 20 g/24h
(0.833 g/h); for 7-9 points 15 g/24h (0.625 g/h); and for 10-12 points 10 g/24h
(0.417 g/h). The unblinded pharmacist will adjust study drug concentration in
the infusate (40 mg/mL, 30 mg/mL, or 20 mg/mL for 4-6 points, 7-9 points, and
10-12 points, respectively) per instructions from the interactive voice/web
response randomization system so that 500 mL infused at a continuous rate over
24 hours (20.8 mL/h) will provide the appropriate dosage by level of hepatic
decompensation. This approach is intended for best dose level assignment which
is related to the above tabulated elements of liver function and is congruent
with the level of cognition in the PK population used in common practice for
dosage considerations. Dosing may be stopped at any time for safety or
tolerability reasons.
Blood and urine will be collected frequently for PK assessments beginning at
baseline and continuing throughout the dosing period, up to 3 hours following
cessation of dosing (end-of-final infusion).
For all patients HE will be assessed by the Hepatic Encephalopathy Staging Tool
(Appendix A), Glasgow Coma Scale, and modified orientation log prior to the
start of infusion, twice daily (7 a.m. and 5 p.m. with a ± 1 hour window)
during infusion and 3 hours post end-of-infusion, and Day 19.
An Independent Data Monitoring Committee will conduct a review of safety data
monthly for each of the first 3 months of the study, every 2 months for the
next 6 months and every 3 months thereafter until study completion (unless no
new data); the review will also include assessment of survival, and an analysis
of survival between arms may be performed if 10 or more deaths have occurred.
Additionally following recruitment of at least 6-10 patients, PK data will be
reviewed monthly on an ongoing basis by a third party expert
pharmacokineticist; individual patient PK data will be blinded and not be
shared with personnel involved in study conduct in order to assure that the
study blind is maintained. As PK data are generated on a monthly basis during
the trial, the PK findings of the third party expert pharmacokineticist will be
reviewed in conjunction with safety by the Data Monitoring Committee (see
above).
An interim analysis of the primary efficacy endpoint will be conducted after
the first 37 endpoints are observed to assess the sample size and power of the
study, including the impact of censoring due to liver transplant or death. The
Haybittle procedure will be utilized for purpose of declaring a statistically
significant difference between treatment groups using alpha=0.001, thereby
preserving the remaining alpha (0.049) for the end of treatment analysis
(Haybittle 1971). The interim analysis will be prepared by a third party
statistical group, shared with the Independent Data Monitoring Committee, and
may also include PK/pharmacodynamic (PD) analyses. Summary outputs will not
unblind treatment assignment at the individual patient level.

Patients will be randomized 1:1 in one of the following treatment arms (70
patients per arm):

-treatment with OCR-002+standard of care
-treatment with placebo+ standard of care

Administration of OCR-002 or placebo will last for up to 5 days (120 hours)
with brief assessments through 3 hours after the end of the final intravenous
infusion (nominal day 6/7), and for those patients still in the hospital, an
assessment 24 hours after the end of the final intravenous infusion.
Additionally there is a follow-up visit 14 days after the last study drug
infusion (nominal Day 19). Thus total study participation lasts for about 19
days, though most assessments are within 6/7 days.

Patients will be randomized 1:1 in one of the following treatment arms (70
patients per arm):

-treatment with OCR-002+standard of care
-treatment with placebo+ standard of care

Administration of OCR-002 or placebo will last for up to 5 days (120 hours)
with brief assessments through 3 hours after the end of the final intravenous
infusion (nominal day 6/7), and for those patients still in the hospital, an
assessment 24 hours after the end of the final intravenous infusion.
Additionally there is a follow-up visit 14 days after the last study drug
infusion (nominal Day 19). Thus total study participation lasts for about 19
days, though most assessments are within 6/7 days.

Screening/baseline:
vital signs (blood pressure, pulse rate, respiratory rate, and body
temperature); height and weight; brief examinations of mental functioning to
assess orientation (Modified Orientation Log), investigator (physician)
evaluation of awareness and responsiveness (Hepatic Encephalopathy Staging
Tool), and alertness (Glasgow Coma Scale); physical and neurological (nervous
system) examination; a review of medical history and medications; and
laboratory evaluations performed for kidney function, liver function,
hematology, chemistry including blood ammonia measurement, blood alcohol at
screen, pregnancy test in women, analysis of urine and urine drug screen.
12-lead electrocardiograms (ECGs) to measure the electrical activity of your
heart will be performed. A blood and urine sample will be drawn as a reference
for study drug concentrations during infusion right before starting the study
drug (active or placebo).

Treatment Period:
Study drug intravenous infusion lasts up to 5 days (120 hours) for full dosing
but may be shorter if hospital discharge is planned earlier. Two times on each
day of study drug infusion vital signs and brief examinations of mental
functioning will be performed as described above, and a sample from the vein
for blood ammonia measurement will be drawn. Each day of infusion 12-lead ECGs
will be performed and there will be complete collection of all urine for
metabolites of study drug (which measures how your body disposes of study
drug). Each day of study drug infusion laboratory evaluations will be performed
of kidney function, liver function, hematology, chemistry, analysis of urine,
and blood collected for plasma concentration of study drug metabolites.
Physical and neurological (nervous system) exam and review of medications will
also be done daily. These assessments will all be performed at the time of
hospital discharge if this occurs before Day 6/7. For patients who are still in
the hospital 24 hours after the end of the last study drug infusion, laboratory
evaluations will be performed of kidney function, liver function, hematology,
chemistry, analysis of urine, physical and neurological (nervous system) exam,
vital signs, brief examinations of mental functioning will be performed as
described above, and review of medications will be done. Monitoring for any
possible adverse effects will continuously be done during the Treatment Period.

Follow-up Visit:
If a patient is discharged from hospital during the follow-up period an
additional evaluation will be performed immediately prior to discharge (brief
examinations of mental functioning as described above and review of
medications). A follow-up visit will be conducted for all patients who complete
the full course of treatment or who do not complete the full course of
treatment but remain in the study and will be done approximately 14 days after
the end of the final (last) infusion (19 days after admission into the study).
At this visit review of current medications and brief examinations of mental
functioning will be performed as described above.

Taking the study drug may involve unknown risks to a pregnant woman, an unborn
baby or a nursing infant. If the patient is pregnant or is breastfeeding a
child, this patient cannot take part in this study. If the patient is a woman
who could become pregnant, the serum (blood) pregnancy test must be negative at
the time of Screening for the study. The patient should not become pregnant nor
breastfeed a baby while taking part in this study because the active drug in
this study might affect and involve risks to a pregnant woman, unborn child, or
a baby in ways that are currently not foreseeable.

If the patient agrees to take part in this study, there may not be direct
benefits. The researchers cannot guarantee that the patient will benefit from
participation in this research.

There may possibly be other side effects that are unknown at this time 9for
adverse events, please see section E9).