Randomized Trial of Peginterferon alfa-2b and
Ribavirin for 48 or 72 Weeks in Patients
with Hepatitis C Virus Genotype 1 and
Slow Virologic Response
Maria Buti,1Yoav Lurie,2Natalia G. Zakharova,3Natalia P. Blokhina,4Andrzej Horban,5
Gerlinde Teuber,6Christoph Sarrazin,6Ligita Balciuniene,7Saya V. Feinman,8Rab Faruqi,9
Lisa D. Pedicone,9and Rafael Esteban1; for the SUCCESS Study Investigators
The benefit of extending treatment duration with peginterferon (PEG-IFN) and ribavirin
(RBV) from 48 weeks to 72 weeks for patients with chronic hepatitis C genotype 1 infec-
tion has not been well established. In this prospective, international, open-label, random-
ized, multicenter study, 1,428 treatment-naı ¨ve patients from 133 centers were treated with
PEG-IFN alfa-2b (1.5 lg/kg/week) plus RBV (800-1,400 mg/day). Patients with detectable
hepatitis C virus (HCV) RNA and a ?2-log10drop in HCV RNA levels at week 12 (slow
responders) were randomized 1:1 to receive 48 weeks (n 5 86) or 72 weeks (n 5 73) of
treatment. Sustained virologic response (SVR) rates were 43% in slow responders treated
for 48 weeks and 48% in slow responders treated for 72 weeks (P 5 0.644). Relapse rates
were similar in slow responders treated for 48 or 72 weeks (47% versus 33%, P 5 0.169).
The safety profile was similar in both treatment arms; serious adverse events leading to dis-
continuation of treatment were observed in 3.5% of slow responders treated for 48 weeks
and 8.2% of those treated for 72 weeks. Among slow responders with a <2-log drop in
HCV RNA at week 8, SVR was 39% in the 72-week arm and 19% in the 48-week arm.
Conclusion: These data suggest that 48 weeks of therapy with PEG-IFN alfa-2b plus RBV
(800-1,400 mg/day) should remain a standard-of-care treatment for treatment-naı ¨ve G1
slow responders. (HEPATOLOGY 2010;52:1201-1207)
(CHC) genotype 1 (G1) infection and achieves sus-
tained virologic response (SVR) in 40%-52% of
treated patients.1-4Because these response rates are
largely unsatisfactory, an individualized approach to
treatment of hepatitis C is increasingly being adopted.
In this approach, total treatment duration is deter-
mined according to the first time during therapy that
hepatitis C virus (HCV) RNA becomes undetectable.
eginterferon (PEG-IFN) alfa-2a or alfa-2b plus
ribavirin (RBV) for 48 weeks is the standard of
care for patients with chronic hepatitis C
The length of time that each patient maintains unde-
tectable HCV RNA while on treatment is directly cor-
related to the likelihood of SVR.4,5
Several studies adopting this approach have shown
that extending therapy to 72 weeks may increase SVR
rates in selected G1 patients.6-11However, the on-treat-
ment virologic criteria used to select patients for
extended therapy vary across studies. Accurately defining
which G1 patients will benefit from extended treatment
is important, because prolonged treatment is associated
with increased adverse events and higher costs.
Abbreviations: cEVR, complete early virologic response; CHC, chronic hepatitis C; CI, confidence interval; G1, genotype 1; HCV, hepatitis C virus; PEG-IFN,
peginterferon; RBV, ribavirin; SVR, sustained virologic response.
From the1Liver Unit, Valle d’Hebron (Ciberehd) University Hospital, and Biomedical Research Center Network in the Area of Hepatic and Digestive Disorders of
the Carlos III Health Institute, Barcelona, Spain; the2Sourasky Medical Center, Tel Aviv, Israel; the3St. Petersburg Municipal Center of Prophylactic AIDS and
Obvodny Channel, St. Petersburg, Russia;4Clinical Infection Hospital #1, Volokolamskoye Shosse, Moscow, Russia; the5Warsaw Medical University & Hospital of
Infectious Diseases, Warsaw, Poland;6J.W. Goethe University Hospital, Frankfurt, Germany; the7Vilnius University Hospital of Tuberculosis and Infection Diseases,
Santarisˇke ˙s, Lithuania;8Mount Sinai Hospital, Toronto, Ontario, Canada; and9Schering-Plough Corporation, Whitehouse Station, NJ.
Received April 30, 2010; accepted June 10, 2010.
Supported by Schering-Plough Corporation (currently Merck & Co., Inc.).
Address reprint requests to: Dr. Maria Buti, Paseo Valle de Hebron 119-129, Servicio de Hepatologi, Hospital Vall d’Hebro ´n and CIBERehd del Instituto
Carlos III, Barcelona, Spain 08036. E-mail: firstname.lastname@example.org; fax: (34)-934274495.
The aim of this large, randomized, international,
multicenter study—known as the SUCCESS study—
was to determine whether an increase of treatment du-
ration to 72 weeks instead of the standard 48 weeks
was associated with an increase in SVR in patients
with HCV G1 who showed a slow virologic response.
Patients and Methods
Patient Selection. Patients were eligible for enroll-
ment if they were treatment-naı ¨ve, aged 18-70 years,
and had compensated CHC (anti-HCV–positive with
detectable HCV RNA). All patients had alanine ami-
notransferase levels above the upper limit of normal
and a liver biopsy performed within 18 months prior
to screening that confirmed a histologic diagnosis of
chronic hepatitis. Key exclusion criteria were body
weight >125 kg; coinfection with hepatitis B virus,
human immunodeficiency virus, or both; or any cause
of liver disease other than CHC.
Study Design. This was a prospective, open-label,
randomized, multicenter, international study conducted
in 133 centers between December 2004 and May 2008.
Patients received PEG-IFN alfa-2b 1.5 lg/kg/week
(PegIntron; Schering-Plough, Kenilworth, NJ) plus
RBV 800-1,400 mg/day (Rebetol; Schering-Plough)
according to body weight (800 mg for patients weigh-
ing <65 kg; 1,000 mg for patients weighing 65-85 kg;
1,200 mg for patients weighing 85-105 kg; and 1,400
mg for patients weighing >105 kg but <125 kg). All
patients were treated for an initial 12-week period,
and further treatment duration was set in accordance
with week 12 HCV RNA levels. According to the cur-
rent clinical guidelines and standard of care,12patients
with a <2-log decline from baseline at week 12 were
withdrawn from treatment, whereas those with unde-
tectable HCV RNA (complete early virologic response
[cEVR]) were treated for an additional 36 weeks
(group C; total of 48 weeks of treatment).
Patients with detectable HCV RNA and a ?2-log
drop at week 12 (partial early virologic response) con-
tinued to receive the same treatment regimen until
week 24. At week 24, patients with detectable HCV
patients with undetectable HCV RNA at week 24
were considered slow responders and randomized 1:1
to treatment for an additional 24 weeks (group A;
total of 48 weeks of therapy) or 48 weeks (group B;
total of 72 weeks of therapy). Randomization was per-
formed independent of sponsor and investigators
through a data fax response system using a computer-
generated randomization scheme in blocks of four
(Everest Clinical Research Services, Markham, Ontario,
Canada). Study groups were stratified by center. Stand-
ard criteria were employed for dose reduction and
treatment discontinuation in patients experiencing he-
matologic toxicity. Compliance was monitored by
comparing the amounts of dispensed and returned
medication to determine whether treatment had been
taken per protocol in the preceding period.
The study was conducted in accordance with princi-
ples of Good Clinical Practice and was approved by
the appropriate institutional review boards and regula-
tory agencies. All patients provided voluntary written
informed consent prior to trial entry. The study spon-
sor and the academic principal investigators (MB and
RE) were responsible for the study design, protocol,
statistical analysis plan, and data analysis. The princi-
pal investigators had unrestricted access to the data
and wrote the manuscript, and the sponsor performed
the statistical analysis. All authors approved the final
draft of the manuscript. This study is registered with
clinicaltrials.gov as NCT00265395.
Patient Evaluation. HCV RNA analyses were per-
formed at a central laboratory using quantitative
(COBAS Taqman, Roche) assay with a lower limit of
quantitation of 30 IU/mL. HCV RNA levels were eval-
uated at screening, baseline, and treatment weeks 4, 8,
12, 24, 48, and 72 (group B) and at week 24 follow-
up. Trugene HCV Genotyping (Bayer HealthCare LLC,
Tarrytown, NY) was used to determine HCV genotype.
Study Endpoints. The study was powered to assess
the superiority of 72 weeks compared with 48 weeks
of treatment in slow responders. The intent-to-treat
population included all patients who were randomized
and received at least 1 dose of either PEG-IFN alfa-2b
or RBV. The primary efficacy analysis was to compare
the percentage of slow responders attaining SVR
(undetectable HCV RNA 24 weeks after receiving the
last dose of therapy) when treated for the longer
View this article online at wileyonlinelibrary.com.
Potential conflict of interest: Dr. Buti and Dr. Esteban have received grants for development of educational presentations (including speakers bureau) from
Schering-Plough Corporation. Dr. Sarrazin is a consultant for, advises, is on the speakers’ bureau of, and received grants from Essex, Schering-Plough, and Roche.
Dr. Pedicone owns stock in Schering-Plough. Dr. Faruqi is a consultant for Schering-Plough. Dr. Teuber is a consultant for, is on the speakers’ bureau of, and
received grants from Essex. She also received grants from Roche.
C 2010 by the American Association for the Study of Liver Diseases.
1202BUTI ET AL.HEPATOLOGY, October 2010
duration of 72 weeks with the standard treatment du-
ration of 48 weeks in patients with slow virologic
response (group A versus group B). Secondary end-
points were end-of-treatment virologic response (unde-
tectable HCV RNA at the end of therapy), relapse
rates (end-of-treatment response, but with detectable
HCV RNA at the end of the 24-week follow-up pe-
riod), and safety and tolerability. Positive and negative
predictive values for a ?2-log decline in HCV RNA
at week 8 were calculated. All patients who received at
least 1 dose of either PEG-IFN alfa-2b or RBV were
included in the safety analysis. The modified World
Health Organization grading system was used to grade
the severity of adverse events. Investigators were re-
sponsible for assigning the relatedness to treatment for
each adverse event.
Sample Size. It was estimated that 120 slow res-
ponders would be required to detect a difference in
SVR rates of 25% between groups A and B (i.e., 45%
in group A and 70% in group B, with a 2-sided alpha
of 0.05) with at least 80% power. Based on the expec-
tation that approximately 10% of patients would meet
the slow-responder criterion, total enrollment was esti-
mated at 1200 patients.
Statistical Analysis. The primary efficacy analysis
was an asymptotic z test with a null hypothesis of no
difference in the rate of SVR in slow responders
between groups A and B. In addition, the two-sided
95% confidence interval (CI) for the difference in
SVR rates was used to estimate the degree of variability
between the two groups. Similar methods were applied
to secondary efficacy analyses. Continuous variables
were summarized using descriptive statistics, and cate-
gorical variables were summarized using frequency
counts and percentages. Whenever appropriate, P values
and 95% CIs were calculated for the relevant statistics.
A predefined ‘‘per protocol’’
patients who received study medication, did not devi-
ate significantly from the entry criteria, and did not
take any prohibited medication. Additionally, an ad
hoc analysis included all treated patients who met the
criteria for fast or slow response and who completed
the assigned duration of therapy.
Patient Population. The study enrolled 1,428 treat-
ment-naı ¨ve patients with CHC G1 infection at 133
study sites. Of the 1,428 patients enrolled, 1 did not
receive the study drug; thus, 1,427 patients received
treatment per protocol. In total, 159 patients (11.1%)
met the slow responder criteria and were randomized
to 48 (n ¼ 86) or 72 (n ¼ 73) weeks of treatment.
Of the remaining patients, 816 (57.2%) attained a
cEVR and were treated for 48 weeks (group C) and
452 (31.7%) either withdrew from the study at week
12 or were nonresponders, and treatment was stopped
as defined in the protocol (Fig. 1). Patient demo-
graphics were generally similar across the treatment
groups (Table 1). A baseline viral load >800,000 IU/
mL was more common in slow responders than
patients with cEVR.
In total, eight patients in group A and 17 patients
in group B failed to complete the study (Fig. 1). All
eight patients in group A discontinued treatment
between weeks 24 and 48. Of the 17 slow responders
in group B who failed to complete treatment, 11 dis-
continued treatment between weeks 24 and 48, and
six discontinued treatment between weeks 49 and 72.
Reasons for discontinuation in group B were adverse
events (n ¼ 6 [4 prior to week 48 and 2 after week
48]), lost to follow-up (n ¼ 1 [prior to week 48]), did
not wish to continue (n ¼ 6 [4 prior to week 48 and
2 after week 48]), noncompliance (n ¼ 2 [1 prior to
week 48 and 1 after week 48]), and other reasons
(n ¼ 2 [1 prior to week 48 and 1 after week 48]).
Virologic Response. In total, 721 of 1,427 (50.5%)
patients who received treatment per protocol attained
an SVR. Among patients with cEVR, 27.5% had
undetectable HCV RNA at week 4 of treatment (rapid
virologic response), and 71.4% had undetectable HCV
RNA at week 8.
In the intent-to-treat analysis, SVR rates were simi-
lar in groups A and B (43% versus 48%; P ¼ 0.6445)
and higher in group C (80%; P < 0.0001 versus
group A) (Fig. 2). End-of-treatment response was
83%, 70%, and 89% in groups A, B, and C, respec-
tively. Relapse rates were 47% for group A and 33%
for group B; however, the difference was not statisti-
cally significant (P ¼ 0.1699). Relapse rates were sig-
nificantly lower for group C compared with group A
(10% versus 47%; P < 0.0001).
Among adherent patients (those who received
?80% of the planned dose of each drug for ?80% of
the assigned treatment duration), SVR rates were 44%
for group A and 57% for group B (P ¼ 0.20); SVR
rates in the per-protocol population were 44% and
49%, respectively (P ¼ 0.63). Similarly, SVR rates in
the completers population were 46% and 57% (P ¼
0.28), and relapse rates were 47.1% and 28.9% in the
48- and 72-week treatment arms, respectively.
Variables Associated with SVR. Slow responders
<40 years old were significantly more likely to attain
an SVR compared with those >60 years old (odds
HEPATOLOGY, Vol. 52, No. 4, 2010 BUTI ET AL.1203
ratio 3.991; 95% CI 1.043-15.277; P ¼ 0.017). Other
variables including weight, treatment arm (group
A versus group B), and week 12 HCV RNA levels
(<50 versus >5,000 IU/mL or 50-5,000 versus
>5,000 IU/mL) did not achieve any statistical signifi-
cance. There was a trend toward a significant associa-
tion between week 8 viral load (<2-log versus ?2-log
drop from baseline), 72 weeks of treatment, and SVR
among slow responders (odds ratio 2.504; 95% CI
0.948-6.613; P ¼ 0.064). The negative predictive
value for a <2-log decline at week 8 was 81% among
patients treated for 48 weeks and 62% among those
Table 1. Patient Demographics and Disposition
Group A (n 5 86):
Slow Responder, 48 Weeks
Group B (n 5 73):
Slow Responder, 72 Weeks
Group C (n 5 816):
cEVR, 48 Weeks
(n 5 452*)
Age in years, mean (SD)
White, n (%)
Female, n (%)
Weight in kg, mean (SD)
Baseline HCV RNA in IU/mL, median log
Alanine aminotransferase level in U/L, mean (SD)
Baseline viral load, n (%)
Change in HCV RNA between screening and baseline, n (%)
Decreased by ?1 log
Increased by ?1 log
*Includes 371 nonresponders and 81 patients who withdrew from therapy prior to week 12.
Fig. 1. Study design. *One patient with partial early virologic response and detectable HCV RNA at week 24 was randomized to group A.
†One administrative, two deaths, and five ineligible.‡One administrative and one treatment failure. AE, adverse event.
1204 BUTI ET AL.HEPATOLOGY, October 2010
treated for 72 weeks (Fig. 3). Thus, absence of a
?2-log decline in HCV RNA at week 8 was a strong
predictor for treatment failure with a 48-week treat-
ment duration, but was less predictive if the treatment
duration was 72 weeks. In addition, among the slow
responders with a <2-log decline in HCV RNA at
week 8, SVR was attained by 19% of patients treated
for 48 weeks and 39% of those treated for 72 weeks.
Among slow responders with a ?2-log decline in
HCV RNA at week 8, treatment outcomes were simi-
lar regardless of treatment duration.
Safety and Tolerability. Safety and tolerability were
generally similar across all treatment groups (Table 2).
Serious adverse events were similar across the treat-
ment arms; however, adverse events leading to early
withdrawal from therapy appeared slightly higher in
group B compared with group A.
This was the largest prospective, randomized study
of patients with hepatitis C G1 infection and a slow
virologic response. These data show that a weight-
based regimen of PEG-IFN alfa-2b plus RBV for 72
weeks resulted in a similar rate of SVR compared with
the same regimen administered for 48 weeks. Although
there was a numerical trend for improved SVR in the
72-week treatment arm, this failed to achieve statistical
significance. This observation has important implica-
tions for clinical practice because of the increasing
tendency, as recommended by some guidelines,12to
extend treatment duration beyond 48 weeks for slow
virologic responders and, occasionally, for G1-infected
patients with detectable HCV RNA at week 4. This
practice results in an increase in adverse events and cost
of therapy without a clear benefit in increasing SVR.
The results of two studies suggest that treatment
with PEG-IFN alfa 2a plus RBV for 72 weeks
increases SVR rates in patients with varying definitions
of slow response compared with the standard 48-week
treatment. However, in these studies (one prospective
study in patients with detectable HCV RNA at week
4, and one retrospective analysis of patients with HCV
RNA ?50 IU/mL at week 12 and <50 IU/mL at
week 24),6,7patients were treated with a fixed dose of
RBV (800 mg), resulting in SVR rates of 17% and
Fig. 2. Virologic response rates.
24.9%-49%.b95% CI ?3.3%-32.1%.
Fig. 3. Positive predictive value (PPV) and negative predictive value
(NPV) of change in HCV RNA levels at week 8 in slow responders
treated for 48 or 72 weeks. Week 8 HCV RNA analyses were not avail-
able for all slow responders.
HEPATOLOGY, Vol. 52, No. 4, 2010 BUTI ET AL. 1205