A phase III study of the safety and efficacy of viramidine versus ribavirin in treatment-naïve patients with chronic hepatitis C: ViSER1 results.

Abstract

Pegylated interferon (peg-IFN) and ribavirin (RBV) are effective in eradicating the hepatitis C virus in more than half of patients. However, anemia arising from RBV-induced hemolysis can prompt dose reductions and lower sustained virologic response (SVR) rates. In early clinical trials, Viramidine (VRD, renamed taribavirin), an RBV prodrug, was associated with less anemia and VRD given at 600 mg twice daily (BID) appeared to provide the best safety with comparable efficacy to RBV. The phase III Viramidine's Safety and Efficacy versus Ribavirin 1 (ViSER1) study randomized 972 treatment-naïve patients with chronic hepatitis C to fixed-dose VRD (600 mg BID) or weight-based RBV (1000 or 1200 mg/day), each given with peg-IFN alfa-2b at 1.5 microg/kg/week. The primary efficacy endpoint was SVR rate, and the primary safety endpoint was hemoglobin (Hb) event rate (percent of patients with Hb < 10 g/dL or at least a 2.5-g/dL decrease from baseline). SVR rates were 37.7% with VRD (244/647) and 52.3% with RBV (170/325). Thus, the ViSER1 study failed to demonstrate the primary noninferiority efficacy endpoint. Significantly fewer patients had Hb events with VRD (353/647; 54.6%) compared to those with RBV (272/325; 83.7%) (P < 0.001), and significantly fewer developed anemia (Hb < 10 g/dL) with VRD (34/647; 5.3%) compared to those with RBV (76/325; 23.5%) (P < 0.001). Conclusion: Fixed doses of VRD failed to demonstrate noninferiority to RBV in producing SVR rates. The incidence of anemia was approximately four-fold significantly lower with VRD than with RBV. These results suggest fixed-dose VRD given 600 mg BID is insufficient to treat patients with chronic hepatitis C; a weight-based dosing trial of viramidine is currently under way.

Full text

Page 1

A Phase III Study of the Safety and Efficacy of
Viramidine Versus Ribavirin in Treatment-Naı ¨ve
Patients with Chronic Hepatitis C: ViSER1 Results
Yves Benhamou,1Nezam H. Afdhal,2David R. Nelson,3Mitchell L. Shiffman,4Deanine G. Halliman,5Jamie Heise,5
Eric Chun,5and Paul J. Pockros6
Pegylated interferon (peg-IFN) and ribavirin (RBV) are effective in eradicating the hepatitis C
virus in more than half of patients. However, anemia arising from RBV-induced hemolysis can
prompt dose reductions and lower sustained virologic response (SVR) rates. In early clinical
trials,Viramidine(VRD,renamedtaribavirin),anRBVprodrug,wasassociatedwithlessanemia
andVRDgivenat600mgtwicedaily(BID)appearedtoprovidethebestsafetywithcomparable
efficacy to RBV. The phase III Viramidine’s Safety and Efficacy versus Ribavirin 1 (ViSER1)
studyrandomized972treatment-naı ¨vepatientswithchronichepatitisCtofixed-doseVRD(600
mg BID) or weight-based RBV (1000 or 1200 mg/day), each given with peg-IFN alfa-2b at 1.5
?g/kg/week. The primary efficacy endpoint was SVR rate, and the primary safety endpoint was
hemoglobin (Hb) event rate (percent of patients with Hb < 10 g/dL or at least a 2.5-g/dL
decrease from baseline). SVR rates were 37.7% with VRD (244/647) and 52.3% with RBV
(170/325). Thus, the ViSER1 study failed to demonstrate the primary noninferiority efficacy
endpoint. Significantly fewer patients had Hb events with VRD (353/647; 54.6%) compared to
thosewithRBV(272/325;83.7%)(P<0.001),andsignificantlyfewerdevelopedanemia(Hb<
10g/dL)withVRD(34/647;5.3%)comparedtothosewithRBV(76/325;23.5%)(P<0.001).
Conclusion:FixeddosesofVRDfailedtodemonstratenoninferioritytoRBVinproducingSVR
rates. The incidence of anemia was approximately four-fold significantly lower with VRD than
with RBV. These results suggest fixed-dose VRD given 600 mg BID is insufficient to treat
patients with chronic hepatitis C; a weight-based dosing trial of viramidine is currently under
way. (HEPATOLOGY 2009;50:717-726.)
C
chronic hepatitis C, resulting in overall sustained viro-
logic response (SVR) rates of 54%-56%.1,2However, ad-
herence to the prescribed regimen of peg-IFN and RBV,
especiallyduringthefirst12weeksoftreatment,iscritical
to viral clearance.3A recent study showed that when pa-
ombination therapy with pegylated interferon
alfa (peg-IFN) and ribavirin (RBV) is the stan-
dard of care for the treatment of patients with
tients infected with hepatitis C virus (HCV) genotype 1
maintained less than 80% of their cumulative RBV dose,
theSVRratewas52%comparedwith67%forthosewho
maintained more than 97% of their RBV dose.4
Patientcharacteristicssuchasethnicity(AfricanAmer-
ican) and coinfection with human immunodeficiency vi-
rus, as well as viral genotype (1, 4, 5, or 6) and adverse
effects, may impair treatment outcomes and lower SVR
Abbreviations: AE, adverse event; ALT, alanine aminotransferase; CI, confidence interval; FW, follow-up week; Hb, hemoglobin; HCV, hepatitis C virus; ITT,
intent-to-treat; peg-IFN, pegylated interferon; RBV, ribavirin; SVR, sustained virologic response; TW, treatment week; ViSER1, Viramidine’s Safety and Efficacy versus
Ribavirin 1 study; VRD, viramidine.
From1Service d’He ´pato-Gastroente ´rologie, Ho ˆpital La Pitie ´-Salpe ´trie `re, Paris, France;2Beth Israel Deaconess Medical Center, Harvard Medical School, Boston MA;
3Department of Medicine, Hepatology and Liver Transplantation, University of Florida, Gainesville, FL;4Hepatology Section, Virginia Commonwealth University
Medical Center, Richmond, VA;5Valeant Pharmaceuticals, Aliso Viejo, CA; and6Division of Gastroenterology/Hepatology, Scripps Clinic, La Jolla, CA.
Received December 1, 2008; accepted May 5, 2009.
The study was sponsored and supported by Valeant Pharmaceuticals, North America.
Address reprint requests to: Yves Benhamou, M.D., Ho ˆpital La Pitie ´-Salpe ´trie `re, Service d’He ´pato-Gastroente ´rologie, 47-84 Boulevard de l’Ho ˆpital, Paris, France
75651.
Copyright © 2009 by the American Association for the Study of Liver Diseases.
Published online in Wiley InterScience (www.interscience.wiley.com).
DOI 10.1002/hep.23073
717

Page 2

rates. Adverse effects, in particular dose-limiting hemolytic
anemia caused by RBV, are associated with lower SVR
rates.3,5-7Treatment-associatedanemiacancausefatigueand
affect quality of life, resulting in dose reductions.2,8,9Al-
thoughthemolecularmechanismofRBV-associatedantivi-
ralactivityremainsuncertain,itmaybedueinparttoitsability
to up-regulate double-stranded RNA-activated protein kinase
activity10;theanemiaappearstobeduetoanaccumulationof
the triphosphate form of RBV within erythrocytes, leading to
RBV-inducedoxidativestresswithsubsequentmembranedam-
agetoand,ultimately,destructionoferythrocytes.11,12Peg-IFN
therapy contributes to hemolytic anemia by suppressing the
productionoferythrocytesinthebonemarrow.13
Viramidine (VRD, renamed taribavirin) is a novel RBV
prodrug being developed for administration with peg-IFN
forthetreatmentofchronichepatitisC.VRDisaguanosine
analogue that preferentially targets the liver and is rapidly
converted to RBV by adenosine deaminase.14,15Pharmaco-
kinetic multiple-dose studies of VRD in nonhuman pri-
mates revealed that VRD-derived RBV is concentrated in
the liver in amounts three-fold higher than those seen after
administration of RBV.16A phase II dose-ranging compari-
son of VRD and RBV, both given in combination with
peg-IFN, evaluated VRD doses of 400, 600, and 800 mg
administered twice daily (BID) and RBV administered at a
weight-based dose of 1000 or 1200 mg/day.17Results
showedsimilarvirologicresponsewithVRD400,600,and800
mgBIDandRBV1000or1200mg/day,eachincombination
withpeginterferonalfa-2ainadoseof180?g/week.17
The current phase III, double-blind, multicenter Vira-
midine’s Safety and Efficacy versus Ribavirin 1 (ViSER1)
studywasperformedtocomparethesafetyandefficacyof
fixed-dose VRD (600 mg BID) and RBV (1000 or 1200
mg/day) when each is administered with peg-IFN alfa-2b
in treatment-naı ¨ve patients with compensated chronic
hepatitisC.TheVRDdoseof600mgBIDwaschosenfor
evaluation based on the earlier phase II study, which sug-
gested that this dose provided the best safety profile while
maintaining efficacy that appeared comparable to RBV.
Patients and Methods
Study Patients
The total planned enrollment for the study was 900
patients, with 600 randomized to receive VRD (Virami-
dine; Valeant Pharmaceuticals International, Aliso Viejo,
CA)and300toreceiveRBV(ribavirin,Rebetol;Schering
Corp., Kenilworth, NJ). Patients were eligible for inclu-
sion if they were treatment naı ¨ve, were at least 18 years of
age, had compensated chronic hepatitis C, had HCV
RNA levels ?2000 copies/mL based on NGI Super-
Quant assay (LabCorp, Burlington, NC), and showed
histologic changes consistent with chronic hepatitis C as
demonstrated on liver biopsy within the 2 years prior to
screening. Patients were also required to have elevated
alanine aminotransferase (ALT) levels at screening or
within the preceding 6 months or histologic evidence of
HCV infection and a detectable HCV RNA load. Pa-
tientswereexcludedfromthestudyiftheyhadpreviously
received IFN or peg-IFN with or without RBV. Other
reasonsforexclusionincludedlowhemoglobin(Hb)con-
centrations (?13 g/dL in men; ?12 g/dL in women),
neutropenia (absolute neutrophil count ?1200/mm3),
thrombocytopenia (?90,000 platelets/mm3), serum cre-
atinineconcentrations?2mg/dL,concomitantinfection
with hepatitis B virus or human immunodeficiency virus,
or chronic hepatic disease other than hepatitis C.
Study Objective
Theobjectiveofthisstudywastocomparethesafetyand
efficacy of VRD dose of 600 mg BID versus that of RBV
1000or1200mg/dayindivideddoseswhenadministeredin
combination with peg-IFN alfa-2b in treatment-naı ¨ve pa-
tients with compensated chronic hepatitis C.
Study Design
This study was a randomized, double-blind, multi-
center, phase III investigation conducted at 87 sites
worldwide. Patients were enrolled by investigators and
randomized to treatment groups using an interactive
voice response system (Fig. 1). Eligible patients were ran-
domlyassignedina2:1ratiotoreceiveVRD600mgBID
or RBV 1000 or 1200 mg/day in divided doses. Patients
weighing ?75 kg at screening were assigned to receive
RBV 1000 mg/day, and those weighing ?75 kg were
assigned to receive RBV 1200 mg/day. This RBV dose
regimen required weight-based dosing of RBV for all pa-
tients,regardlessofgenotype,andwasthestandardofcare
when this study was designed.
Thestudydrugwasinitiatedwithinthe24hoursafterran-
domization.Allpatientsalsoreceivedopen-labeltreatmentwith
subcutaneous peg-IFN alfa-2b (peginterferon alfa-2b, PegIn-
tron;ScheringCorp.,Kenilworth,NJ)1.5?g/kg/week.
Potential conflict of interest: Dr. Afdhal is a consultant for, advises, and received grants from Valeant, Echosens, GlaxoSmithKline and Vertex. He is a consultant for,
advises,receivedgrantsfrom,andisonthespeakers’bureauofSchering-Plough,Gilead,andNovartis.HereceivedgrantsfromValeantPharmaceuticals,Idenix,andQuest.
He is a consultant for and advises Biogen, Idera Pharmaceuticals, Boehringer Ingelheim, Human Genome Sciences, Biolex, Ono Pharmaceuticals, and Fibrogen. He is also
on the speakers’ bureau of Bristol Myers Squibb. Dr. Shiffman is a consultant for, advises, is on the speakers’ bureau of, and received grants from Roche. He advises Pfizer
and Vertex. He is a consultant for, is on the speakers’ bureau of and received grants from Schering-Plough. He is also a consultant for and received grants from Valeant. Dr.
Pockros is a consultant for and received grants from Valeant.
718BENHAMOU ET AL. HEPATOLOGY, September 2009

Page 3

Patients were stratified according to HCV RNA geno-
type (genotypes 2 or 3 versus genotypes 1, 4, 5, 6, mixed,
oruntypeable[genotypenon-2/3]),screeningHCVRNA
titers(?2millionor?2millioncopies/mL),andbaseline
weight (?75 or ?75 kg). Patients were assessed for dis-
ease stage using a nonstandard fibrosis scoring system.
The high number of patients entering the study with
long-term HCV infection required investigators to assess
fibrosis using either the Metavir or Knodell scoring sys-
tem, based on available biopsy records. Scores from both
five-point scales were then combined and expressed using
three descriptors: “cirrhosis” (Metavir and Knodell scores
of 4), “bridging fibrosis” (scores of 3 or 2), and “no ad-
vanced fibrosis” (scores of 1 or 0).
Study treatment was initiated on Day 1, and clinic visits
occurred at Treatment Weeks (TWs) 1, 2, 4, 8, 12, 18, and
24,aswellaspost-treatmentatFollow-upWeeks(FWs)4,8,
12, 18, and 24. Patients infected with genotype 2/3 were
treated for 24 weeks, and patients infected with genotypes
other than 2 or 3 (genotype non-2/3) were treated for 48
weeks. All patients were followed for 24 weeks after the end
of treatment to determine SVR rates. Only patients with
HCVgenotypenon-2/3attendedclinicvisitsatTWs30,36,
42,and48becauseoftheirlongerdurationoftreatment.All
patientsreceivedthreecapsulesofstudydrugtwicedailywith
their morning and evening meals.
Thesamplesizedeterminationwascalculatedbasedon
theprimaryefficacyvariable,whichwastheproportionof
patients with SVR at the end of the 24-week follow-up.
The expected SVR rate in the test group (VRD) was esti-
mated to be equal to or within 2% of the rate in the
control group (RBV). Therefore, the power and sample
size calculation assumed that the SVR rate in the VRD
group was equal to or slightly lower than the rate in the
RBVgroup.The900-patientsamplesize(n?600,VRD
group;n?300,RBVgroup)allowedrejectionofthenull
hypothesis with a power of 92% when the true SVR rates
were the same between treatment groups and a power of
80% when the true SVR rate for the VRD group was 2%
lower than the RBV group, based on the primary modi-
fied intent-to-treat (ITT) population.
The protocol was approved by the Investigational
Review Boards of the participating institutions and
conforms to the ethical guidelines of the Declaration of
Helsinki (modified in 2004). All patients provided
written informed consent. The study was conducted in
accordance with the provisions of Good Clinical Prac-
tices.
Efficacy Assessments
The primary efficacy endpoint was SVR rate, defined
as the proportion of patients with undetectable plasma
HCV RNA levels (?100 copies/mL) at the end of the
24-week follow-up. Additional predetermined efficacy
variables included the number and proportion of patients
with undetectable HCV RNA load or at least a 2-log10
decrease from baseline at TWs 4, 12, and 24 (all geno-
types) and TW 48 (genotypes non-2/3), as well as FWs 4,
12, 18, and 24 by treatment; normalization of ALT levels
atTW24andFW24bytreatment;andtheproportionof
patients achieving SVR by genotype, baseline viral load,
and other clinical and demographic characteristics.
Lackofefficacywasdefinedaslessthana2-log10dropor
detectable levels of HCV RNA (copies/mL) at TWs 12 or
24.Relapserateswerecalculatedbymeasuringproportionof
respondingpatientswhoseplasmaHCVlevelschangedfrom
undetectable at end of treatment to detectable at FW 24.
Fig. 1. ViSER1 study schematic. *Took
place within 30 days of screening visit.
HEPATOLOGY, Vol. 50, No. 3, 2009 BENHAMOU ET AL. 719

Page 4

Safety Assessments
Theprimarysafetyvariablewasthepercentageofpatients
experiencing an Hb event at any time during treatment. An
HbeventwasdefinedasHbconcentrations?10g/dLorat
leasta2.5-g/dLdecreasefrombaselineoratanypointduring
treatmentifnobaselinemeasurementwasavailable.Patients
whodiscontinuedtreatmentearlywerecategorizedashaving
an event if their Hb concentrations met this criterion, and
patients without Hb measurements during treatment were
considered not to have had an event.
Additionalanalysesfortheprimarysafetyvariablewere
performed for subgroups that were based on HCV geno-
type, baseline plasma HCV RNA titers, and baseline
weight. The proportion of patients with Hb concentra-
tions ?10 g/dL during treatment and with concentra-
tions ?10 g/dL or at least a 2.5 g/dL decrease from
baseline at each scheduled visit were determined by treat-
ment group for each of these subgroups.
If a dose reduction of study drug (VRD or RBV) was
required because of an adverse event (AE), the dose was re-
duced to 600 mg once daily, to be taken in the morning.
Once the dose was reduced, no increase was allowed for the
durationofthestudy,butpatientscouldremaininthestudy.
IfanAEnecessitatedinterruptionofanystudydrugformore
than 14 consecutive days, or if a patient experienced a life-
threatening AE, all study drugs were discontinued and the
patient was removed from the study.
Statistical Analyses
This study had two coprimary measures of safety and
efficacy. Safety analyses were conducted with data from
the ITT population. The primary efficacy analysis was
performed using data from the ITT and per-protocol
populations; the secondary efficacy analysis considered
data from the per-protocol population only. Hochberg’s
method was used to simultaneously control the overall
experiment-wise Type I error rate at 0.05 for the two
coprimary measures of safety and efficacy, with equal pri-
ority. All other statistical analyses were performed using a
two-sided hypothesis test at the overall 5% level of signif-
icance. The stratification factors were those used for ran-
domization (i.e., genotype, screening HCV RNA, and
baseline weight) with a binary cutoff. If the upper limit of
the confidence interval (CI) for the difference between
RBV and VRD was less than 12%, the VRD group was
considered noninferior to the RBV group. Continuous
data were summarized using descriptive statistics unless
otherwise noted: n, mean, standard deviation, median,
minimum (min), and maximum (max). Categoric vari-
ables were summarized using frequency counts and per-
centages. All analyses were conducted using SAS (Cary,
NC) version 8.2 or later.
The data for this study were managed by the investiga-
torsandsponsor.Thestatisticalanalysiswascompletedby
the sponsor. The authors had access to the clinical study
report and have either written or provided intellectual
input to the manuscript.
Results
Study Patients
Enrollment began in December 2003, and the last pa-
tient completed the study on December 28, 2005. A total
of 972 patients with HCV were randomized in a 2:1 ratio
to receive peg-IFN alfa-2b in combination with either
VRD 600 mg BID (n ? 647) or RBV 1000 mg/day or
1200mg/day(n?325).Allrandomizedpatientsreceived
at least one dose of study drug and were included in the
ITT population and analyzed for safety and efficacy. A
total of 73% (470/647) of the patients in the VRD group
and 76% (246/325) of the patients in the RBV group
completed the study (Fig. 2); 177 (27.4%) patients in the
VRD group and 79 (24.3%) in the RBV group discon-
tinuedstudydrugduringtreatment.Themostcommonly
cited reasons for discontinuation in the VRD and RBV
groups were AEs (9.6% and 9.8%, respectively) and lack
ofefficacy(11.7%and7.4%,respectively).Baselinechar-
acteristics were similarly distributed between the two
treatment groups (Table 1).
Efficacy
Sustained Virologic Response. The overall SVR rate
at FW 24 in the ITT population was lower among VRD-
treated patients (37.7%) than RBV-treated patients
(52.3%);differenceinproportionsbetweenRBVandVRD
0.15 (95% CI 0.09, 0.21) (Table 2). The upper limit of the
95% CI for the difference between the VRD and RBV
groupsexceeded12%;thus,thenoninferiorityofVRDcom-
pared with RBV was not demonstrated. Mean changes in
viral load from baseline over the course of the study were
lowerintheVRDgroupthantheRBVgroupateverystudy
visit. Patients in the VRD group had lower rates than the
RBV group of undetectable HCV RNA at TWs 12 and 24
(Fig. 3). Relapse rates were 30.8% in the VRD group and
21.3% in the RBV group (P ? 0.0003).
SubgroupAnalysis. LowerSVRrateswereseeninthe
VRD group compared to the RBV group for the follow-
ing demographic and disease characteristics, which have
been known to affect SVR: genotype, baseline viral load,
sex, age, body weight, African-American ethnicity, and
disease stage (Table 2). Interestingly, age was an impor-
tant variable for response in comparing VRD to RBV. In
patients younger than 45 years, efficacy of VRD was sim-
ilar to that of RBV (51% versus 56%, respectively), but
VRD clearly was less efficacious in patients older than 45
720BENHAMOU ET AL. HEPATOLOGY, September 2009

Page 5

years.Inthesepatients,responsetoVRDwasonly26.9%
compared with 50.9% seen with RBV (P ? 0.004).
Apost-hocsubgroupanalysesindicatedthatgreaterexpo-
sure to VRD (per kilogram of body weight) was associated
withhigherSVRrates(Fig.4).TheSVRrateamongpatients
receiving VRD at doses greater than 18 mg/kg was 48.8%
(78/160);incontrast,theSVRrateamongthosereceiving13
mg/kg or less was 28.8% (38/132).
ALT Normalization. ComparedwiththeRBVgroup,
theVRDgrouphadfewerpatientswithabnormalALTlev-
els at baseline. A greater proportion of patients receiving
RBV achieved normal ALT values at every study visit.
Safety
Anemia. The primary safety endpoint was the pro-
portion of patients with an Hb event, defined as Hb
concentrations ?10 g/dL or at least a 2.5-g/dL de-
crease from baseline, at any time during the treatment
period. Significantly fewer patients in the VRD group
(5.3%, 34/647) had severe anemia (Hb ? 10 g/dL)
than did those in the RBV group (23.4%, 76/325);
odds ratio ? 0.16; P ? 0.001 (Fig. 5A). Fewer patients
in the VRD group had an Hb event than did those in
the RBV group (54.6%, 353/647 versus 83.7%, 272/
325, respectively); odds ratio ? 0.21; P ? 0.001. At
eachstudyvisit,fewerHbreductionsofatleast25%from
baseline occurred in VRD-treated patients (range, 0.3%-
8.7%) than in RBV-treated patients (range, 3.4%-
28.9%) (Fig. 5B). Mean Hb concentrations decreased
from baseline in both treatment groups throughout the
study,althoughthemeandecreasewasgreaterintheRBV
group (Fig. 5C). A post-hoc analysis of patients with se-
vere anemia (Hb? 10/dL) assigned to weight-based dose
subgroups was performed to correspond with the efficacy
subgroup analysis (Fig. 5D). Higher rates of anemia
were seen with increased exposure to VRD (range, 3.8%-
6.9%) but were smaller compared with increased expo-
sure to RBV (range, 11.1%-36.8%).
Fig. 2. Disposition of patients during treatment. *Completed per case study form but did not meet the criteria for minimum treatment weeks.
HEPATOLOGY, Vol. 50, No. 3, 2009 BENHAMOU ET AL.721