Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003), pp. 126–139 ( C 2003)
Effect of Ribavirin and Amantadine on Early
Hepatitis C Virus RNA Rebound and Clearance
in Serum During Daily High-Dose Interferon
GEROND LAKE-BAKAAR, MD,* LYNDA RUFFINI,* and PETR KUZMI ˇ
C, PhD†
The early rebound in serum HCV RNA during HCV dynamic studies with high-dose interferon maybe due to de novo infection with interferon escape quasispecies. We simultaneously measured serumalanine aminotransferase (ALT) and HCV RNA at rapid intervals in chronic HCV liver diseasepatients during interferon therapy alone or in combination with ribavirin and amantadine. HCVRNA declined rapidly between 0 and 48 hr in all patients (phase 1). Ribavirin and amantadinesignificantly increased this phase 1 decline. In all four monotherapy patients with viral rebound, theincreasing levels of HCV RNA were associated with a parallel increase in serum ALT, consistentwith a hepatitis flare or de novo infection. By contrast, in the four monotherapy patients withoutviral rebound, and all eight patients receiving combination therapy, the slow progressive phasetwo decay was associated with declining serum ALT levels. Ribavirin or ribavirin and amantadinesignificantly and incrementally increased the phase two HCV RNA clearance. Dynamic sequencingin the HVR1 region in one rebound patient confirmed the potential for rapid evolutionary changesduring interferon therapy. These preliminary data suggest that early viral rebound might be associatedwith de novo infection with interferon escape HCV variants, which in turn are attenuated by ribavirinand amantadine. KEY WORDS: hepatitis C viral dynamics; interferon; ribavirin; amantadine; HCV clearance; HCV molecular evolutionary genetics.
Mathematical modeling of changes in plasma viral load
dynamic parameters are based assumes the rapid initial
after administration of antiviral compounds (1), has pro-
decline in plasma HCV RNA levels (phase 1) that con-
vided significant insights into the kinetics of host–virus
sistently follows high-dose daily interferon monotherapy
interactions and the action of drugs used in the treatment
reflects rapid inhibition of viral replication or release and
of chronic human immunodeficiency virus, HIV (2, 3),
that the variable and much slower phase 2 decay can be
hepatitis B virus, HBV (4), and hepatitis C virus (HCV)
attributed to a progressive loss of infected cells (5). How-
infections (5–9). The simple model of chronic viral infec-
ever, in a significant number of patients, an early rebound
tion on which mathematical models for establishing HCV
in serum HCV RNA that is not predicted by this determin-istic simple model follows the phase 1 decline, replacing
Manuscript received May 10, 2002; revised manuscript received
the expected phase 2 decay (5, 8, 10, 11). The model as-
September 16, 2002; accepted September 20, 2002.
sumes constant growth and other virus–host characteris-
From the *Department of Medicine, VA Medical Center Northport,
tics during interferon treatment. However, HCV exists as a
New York, and †Biokin Ltd., Pullman, Washington, USA.
Supported in part by an unrestricted Investigator initiated study grant
complex mixture of quasispecies with marked heterogene-
from Roche Pharmaceutical, Nutley, New Jersey, USA.
ity. As has been observed during nevirapine monotherapy
Address for reprint requests: Dr. Gerond Lake-Bakaar, Department
for HIV (3), early emergence of drug-tolerant quasis-
of Medicine 111E, Building 200, Room A1-1 Veterans AdministrationMedical Center 79, Middleville Road, Northport, New York 11768, USA.
pecies that are capable of continuing de novo hepatocyte
Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003)
0163-2116/03/0100-0126/0 C 2003 Plenum Publishing Corporation
infection and replication could explain the HCV rebound.
Flu-like symptoms induced by interferon were treated with
In that event, a parallel increase in serum ALT, as well as
acetaminophen, 500–650 mg orally, 1 hr before interferon and
structural or molecular evolutionary changes in HCV qua-
repeated at a dose of 325–500 mg every 4–6 hr as needed.
The physician instructed patients on the technique of self-
sispecies might be expected to accompany the rebound.
administration of interferon. After discharge, the patients re-
Combining interferon with other antivirals should inhibit
turned to the hospital for daily collection of blood samples until
the emergence of these variants, attenuating the rebound
day 14. To ensure compliance, they were asked to return empty
and increasing the estimated serum clearance rates of
vials, unused drugs, and used needles and syringes.
At the conclusion of this 14-day study period, patients were
transferred to a standard regimen of 3 million units of inter-
In this study, we provide evidence for the first time that
feron, three times weekly. Dual- and triple-combination ther-
a parallel increase in the serum ALT levels accompanies
apy patients also continued with daily ribavirin, and ribavirin
HCV RNA rebound, consistent with de novo hepatocyte
plus amantadine, respectively. They were evaluated at the end of
infection. We also demonstrate that combining interferon
three months for the presence or absence of HCV RNA in serum.
with other antivirals eliminates the rebound, and increases
Ribavirin with or without amantadine, was added to the treat-ment regimen of all monotherapy patients who had failed to clear
both phases of serum HCV RNA clearance, consistent
with attenuation of escape variants. Finally, we present
Materials. Roche Pharmaceutical (Nutley, New Jersey, USA)
preliminary data from a single patient study that treat-
kindly provided IFN-α-2a. Ribavirin was obtained from Scher-
ment with interferon is associated with rapid evolution
ing Pharmaceutical (Kenilworth, New Jersey, USA) and aman-
and structural change within the HCV genome.
tadine from Endo Pharmaceuticals (Chadds Ford, Pennsylvania,USA).
The measurements of HCV RNA were performed using ei-
MATERIALS AND METHODS
ther the branched DNA, (bDNA) assay (Chiron, Emeryville,California, USA) for patients with high pretreatment copy num-
Patient Population. The diagnosis of chronic hepatitis
bers of HCV RNA or the COBAS Amplicor (Roche Diagnostics
C virus infection was based on the simultaneous presence of
Systems, Branchburg, New Jersey, USA) for patients with low
anti-HCV and HCV RNA in serum. All patients were hepatitis
pretreatment copy numbers. Samples from individual patients
B surface antigen negative and had been tested for HIV infection.
were assayed in the same batch to avoid batch to batch variation.
None had received previous treatment with interferon, ribavirin,
Using conversion tables for bDNA (1 IU = 6.3 eq) and for qPCR
or amantadine. All but two of the patients were infected with
(100 copies = 60 IU) provided by the manufacturers, the results
from either assay were recorded in international units (IU) per
Patients with other forms of liver disease and those actively
milliliter. In preliminary studies, we assayed eight serum sam-
abusing alcohol and intravenous drugs were excluded. None of
ples using both techniques and recorded the results. The results
the serum samples collected during the study contained signifi-
varied by 11.4% ± 12.2 (mean ± SE).
Serum liver function tests (LFTs) were measured through the
For clinical diagnostic staging of the fibrosis scores, a routine
routine chemistry laboratory. All the serum samples from each
pretreatment liver biopsy was performed within six months prior
individual patient were assayed in the same batch to avoid sig-
to the study and assessed by a single pathologist. Informed con-
nificant interassay variation. HCV genotyping was performed
sent was obtained from all patients prior to entry into the study,
using a DNA line probe assay InnoLiPA (Innogenetics Inc., Nor-
which was approved by the Human Studies Subcommittee at the
cross, Georgia, USA). Serum ethanol levels were assayed using
Veterans Administration Medical Center, Northport, New York,
a blood alcohol kit (Sigma Diagnostics, St. Louis, Missouri,
Study Design. Patients were hospitalized during the first two Evaluation of Changes in E2 HVR1: Sequencing and
days of the study. The weights of the patients were recorded on
Cloning PCR Products. We used sera obtained from patient 1
admission. Venous blood samples, 5–10 ml, were collected 1 hr
at 0, h 12, and 24 hr and days 5 and 10. We prepared PCR
before and then at 0, 2, 4, 6, 8, 12, 16, 20, 24, 32, and 48 hr and
products from the HVR1 region according to the methods de-
then daily for up to 14 days, after the first injection of 10 million
scribed by Gonzalez-Peralta et al (12), using eLONGASE, a
units subcutaneously of interferon-α-2a (IFN-α-2a). In a second
high-fidelity polymerase with low error rates. Following incuba-
group of patients assigned to receive dual combination therapy,
tion with Taq polymerase (0.7–1 units) at 72◦C for 8–10 min, the
the first morning dose of ribavirin, 400 mg, was given orally at
PCR products were purified using Qiagen columns. This purifi-
the same time as the initial interferon injection. Ribavirin was
cation step was found to improve cloning efficiency. The purified
repeated orally, 600 mg in the evening. In a third group of patients
PCR products were cloned into the plasmid vector pCR4-TOPO
assigned to triple therapy, amantadine, 100 mg orally, was given
and transformed into TOP10 E. coli cells using the TOPO TA
twice daily in combination with interferon and ribavirin. The
Cloning kit for Sequencing, version E (Invitrogen Corporation,
same dosing regimen for all three drugs was continued daily for
Carlsbad, California, USA) according to the manufacturers pro-
tocol. Transformants were grown on LB plates containing either
All blood samples were kept on ice, centrifuged, and separated
100 µg/ml ampicillin or 50 µg/ml kanamycin, which typically
within 6 hr of collection. Sera were frozen in aliquots at −20◦C
initially, and transferred within seven days to −80◦C for storage.
Up to 10 colonies per time point were selected at random
HCV RNA assays were performed within two weeks of sample
and grown overnight in 2 ml LB or SOB medium in 15 ml
loosely capped tubes at 37◦C, containing 50 µg/ml ampicillin
Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003)
or kanamycin. Plasmid DNA were isolated and purified using
considered only those ALT data points that were obtained after
the SNAP miniprep kit, version E (Invitrogen) according to the
the nadir or point of inflection on the HCV RNA curve. These
manufacturers instructions. The concentration of the DNA were
data points are represented by the filled circles in Figure 4 below.
determined using a DNA Dipstick kit (Invitrogen).
All linear regression analyses were performed using the soft-
DNA (500 ng) and universal T3 primers (3.2 pmol) in a to-
ware program SigmaPlot 2.0 (SPSS, Chicago, Illinois, USA;
tal volume of 12 µl were provided to GeneWiz, a commercial
www.spss.com). All nonlinear regression analyses were per-
laboratory for sequencing using the ABI PRISM model 373.
formed by using the computer software ViraFit 1.05 (Biokin
The nucleotide sequences from the clones were aligned by
Ltd., Pullman, Washington, USA; www.biokin.com). Statisti-
using Clustal W (1.8; EMBL European Bioinformatics Insti-
cal comparison between groups was performed by the Student’s
tute). The 196-bp HVR1 sequence is readily identified within
the 3957-bp pCR4-TOPO plasmid, bordered by the known plas-mid sequences at the PCR product insertion site. Data Analyses and Statistical Methods. For quantitative
analysis of observed changes in serum viral load after inter-feron therapy, we followed the model and equations described
Patient Demographics. Sixteen male patients were en-
by Neumann et al (5), in which V0 is the initial steady-state vi-
rolled and completed the study. Details regarding age,
ral load; ε is the efficacy of drug treatment in reducing viral
genotype and degree of histologically assessed fibrosis
production and release, and ranges from zero (no effect) to one
(13) are shown in Table 1. All but two of the patients were
(complete removal of the virus in the steady state); c is the clear-ance rate constant and directly reflects the initial phase 1 decay
infected with HCV genotype 1. One patient was positive
for HIV. There were three African-Americans in the study.
0 is the initial delay that elapses before interferon can
express its therapeutic action. Additionally, δ is the death rate
Two were genotype 1 and received monotherapy. A third
constant for infected hepatocytes and directly reflects the phase
was infected with genotype 3a and received triple therapy.
Pretreatment serum HCV RNA levels varied by
The rate of serum ALT increase or decrease, associated with
the HCV rebound or decay, was calculated by a linear least-
less than one logarithmic unit, indicating steady state
squares fit of a portion of the ALT data. In this linear fit we have
TABLE 1. HCV RNA KINETICS AND ALT KINETICS DURING TREATMENT WITH INTERFERON, INTERFERON PLUS RIBAVIRIN (DUAL THERAPY B)
AND INTERFERON PLUS RIBAVIRIN AND AMANTADINE (TRIPLE THERAPY, C)*
7.17 ± 1.45 3.33 ± 2.95 0.90 ± 0.10 0.28 ± 0.12 0.30 ± 0.34 0.05 ± 0.027.87 ± 2.69 4.10 ± 2.53 0.93 ± 0.08 0.50 ± 0.24 0.69 ± 0.39 0.15 ± 0.068.29 ± 3.4 1.76 ± 1.17 0.91 ± 0.01
*Explanation of symbols and corresponding units: t0 (hours), delay time; V0 (106×IU/ml), initial viral load; ε, efficacy; c (per hour), rate constant
for viral clearance during phase 1 of therapy (0–48 hours); rv (1012×virions/day), virion production rate; δ (per day), rate constant for death ofinfected cells during phase 2 of therapy (2–14 days); rALT (IU/day), daily rate of ALT level changes (increase or decrease) during Phase 2; P,probability (0–1, computed from Student’s t statistic) that the mean values of each parameter are identical for treatment regimens. Parameters t0,V0, ε and c were obtained by fitting HCV data in Figure 1a, 1b to equation 1. Virion production rate rv was computed as V0 × c, defined in equation1, times a factor, based on body weight, corresponding to the extracellular fluid volume. The infected cell death rate constant δ is determined bynonlinear least squares fit of phase 1 + 2 data (Figure 2a,b) to equation 2. The rate of ALT changes was computed by linear regression of phase2 data shown in Figure 3a,b. AR = interferon monotherapy patients with rebound; ANR = interferon monotherapy patients without rebound. (a),African-American; (b), HIV-positive; (c), rebound; (d), below detection limit during phase 2 (2–14 days); (e), fixed in nonlinear regression; rv,units are 1012 per day; V0 units are 106 IU/ml. Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003)Fig 1. Changes in the HCV viral load (106 IU/ml) for (a) patients treated with interferon alone and (b) patients treated by the interferon plus ribavirin combination therapy during phase 1 (0–48 hr) of therapy. The solid curves were obtained by nonlinear least-squares fit to equation 1. HCV Viral Dynamics and Serum ALT Kinetics. t0, the initial viral load V0, the efficacy ε, and the
HCV RNA viral dynamic data observed during phase 1
clearance rate constant c for each patient. The re-
(0–48 hr) for monotherapy patients and patients treated
sults, including those from patients on triple therapy
with interferon and ribavirin are shown graphically in
with interferon, ribavirin, and amantadine, are summa-
Figure 1a, b. The data were fit to the Neumann equa-
rized in appropriate columns of Table 1, along with
tion in order to obtain estimates of the delay time
P values computed from Student’s t statistic for the
Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003)Fig 1. (Continued).
treatment. The changes in serum ALT levels during
significant HCV rebound was observed in any of the
phase 1 of therapy (0–48 hr) followed no consistent
Serum ALT rose in parallel with the rise in HCV RNA
Phase 2 viral dynamic data (2–14 days) for interferon
in the rebound patients (Figure 4a,b), as signified by the
monotherapy and, dual and triple therapy are shown in
positive slope of rALT in Table 1. By contrast, serum ALT
Figures 2a,b and 3, 4 respectively. In four of eight patients
consistently decreased in parallel with declining levels of
(patients 1–4) who received interferon monotherapy,
HCV RNA in both the nonrebound patients on monother-
the rapid initial decline during phase 1 was followed by
apy and in all of the patients on combination treatment. In
a marked increase or rebound in HCV RNA to levels
all of these patients, rALT was negative.
significantly above the nadir. This rebound typically
The phase 2 data were fit to the Neumann equation in
started around 24–48 hr after the initial injection, with
order to determine the infected cell death rate constant, δ,
HCV RNA levels remaining over twice the nadir at
from the slope of the decay. Since the derived equation
day 14. No HCV rebound was observed in the other
anticipates progressively decreasing levels of HCV RNA,
four patients on monotherapy (patients 5–8). Although
it was not possible to fit the data from any of the four
there were significant oscillations above the baseline
patients on monotherapy with rebounding HCV RNA lev-
in patient 7, serum HCV RNA by day 14 was only
els to the curve. In patinet 6 on monotherapy, HCV RNA
marginally (less than double) greater than the nadir. No
levels during phase 2 were below detection limit. Thus,
Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003)Fig 2. Changes in the HCV viral load (106 IU/ml) for (a) patients treated with interferon alone and (b) patients treated by the interferon plus ribavirin combination therapy during phase 2 (2–14 days) of therapy. The dashed curves are interpolated from the data for those patients where the phase 2 viral dynamics included a rebound or oscillations. The solid curves were obtained by nonlinear least-squares fit to equation 2. Data points identified by triangles were excluded from the analysis.
the infected cell death rate constant, δ, could not be deter-
tial viral load V0, the delay time t0, the mean clearance rate
constant c, and the daily virus production rates rv, are sim-
The HCV dynamic parameters for the interferon
ilar for both patients demonstrating a rebound ( AR) and
monotherapy patients (regimen A), which include the ini-
those without a significant rebound ( ANR). The rebound
Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003)Fig 2. (Continued).
monotherapy group included two African American (AA)
Combination therapy (groups B and C combined) also
patients with genotype 1 HCV infection. One was also
significantly increased the phase 2 slope and the derived
coinfected with HIV. These two AA patients had the low-
estimate for the infected cell death rate constant, δ. These
est efficacies estimated in any of the groups, 0.74 and 0.75
results are significant even after patients 12 (genotype 2b)
and 14 (genotype 3a) have been excluded from the anal-
The addition of ribavirin and amantadine to interferon
ysis. The combination of interferon with ribavirin and
made no significant difference to the overall efficacy. This
amantadine increased the calculated infected cell death
is consistent with their known weak antiviral activity to-
rate constnat δ, almost 10-fold above that of interferon.
wards HCV. By contrast, combination therapy (groups
In almost all of the patients treated with combination
B and C combined) significantly increased the phase 1
interferon plus ribavirin therapy, we observed a transient
slope and the derived estimate for viral clearance, c. The
increase in serum HCV RNA levels around 50–200 hr after
results are significant even when patients 12 (genotype
initiation of therapy (Figure 2b). The hump was attenuated
2b) and 14 (genotype 3a) are excluded from the analy-
and occurred earlier, between 24 and 125 hr, in patients
sis. The difference between combination group B (0.5 ±
0.24) and group C (0.6 ± 0.2) did not reach statisti-
Preliminary Virological Response at 3 Months. A
cally significant levels, presumably because of the small
preliminary assessment of virological response to treat-
ment was made based on the presence or absence of HCV
Digestive Diseases and Sciences, Vol. 48, No. 1 (January 2003)Cloning, Sequencing, and Phylogenetic Analysis of E2 Gene PCR Products. We analyzed samples of sera from patient 1 with clearly demonstrated HCV RNA re- bound. PCR products from five time points were cloned and 10 clones from each time point were sequenced, for a total of 50 sequences. Prior to analyses, plasmid se- quences promixal and distal to either end of the insertion site were removed. Sequences containing gaps, deletions, or artifacts due to mispriming were excluded. A total of 15 sequences were discarded, providing 35 analyz- able sequences, each 196 nucleotides in length. Double- stranded PCR products inserted into the plasmid in the 3 –5 position were reversed and the complementary se- quence determined using a reverse-complement trans- lator (http://arep.med.harvard.edu/labgc/adnan/projects/ Utilities/revcomp.html) as necessary. Data from the in- dependent clones from 0 hr (N = 8), 12 hr (N = 8), 24 hr (N = 6), day 5 (N = 6), and day 10 (N = 7) were analyzed. Diversity and Evolutionary distances in E2 HVR1.
The deduced amino acid sequences translated and alignedusing molecular evolution genetics analysis (MEGA soft-ware) (14) are shown, with the published HCV-1 proto-type, HCV J1 sequence (Table 2). The table also shows thededuced amino acid sequences 359–422, which includethe HVR1 sequences 384–410. The HVR1 sequences con-tain the amino acids threonine, T 385; glycine, G 406;glycine G 389 and glutamine, Q 409, which have recentlybeen noted to be 100% conserved among HCV subtypes(15).
We estimated the average genetic distance within each
time point (group) for both the nucleotide (Table 3) andthe deduced amino acid sequences (Table 4). The averagegenetic distance within both nucleotide and amino acidsequences before treatment (0 hr) was small, suggestinga homogeneous, stable viral population (16, 17, 18). Theamino acid variability decreased to complete homogeneity
Fig 3. Changes in serum HCV RNA levels during treatment with the
by 12 hr, in line with a marked reduction in viral load. This
triple combination of interferon, ribavirin, and amantadine.
presumed haplotype-independent culling was followed bya gradual increase back to baseline levels by day 5 and to
at the end of three months. Serum HCV RNA was still
above baseline by day 10 (Figure 5).
present in all eight monotherapy patients, compared totwo of five (40%) on dual therapy and only one of three
DISCUSSION
(33 percent) on triple therapy. In one patient on triple ther-apy in whom HCV RNA reappeared in serum, increasing
An analysis of the rapid changes that take place imme-
the freuqency of interferon injections from three times to
diately after the initiation of antiviral therapy has provided
six times a week resulted in loss of HCV RNA from serum.
valuable information regarding the pathogenesis and treat-
The patient subsequently had a sustained virological re-
ment of chronic viral infections and has helped to elucidate
sponse. In a second patient, end treatment HCV RNA was
mechanisms of antiviral drug action (3, 5, 19).
negative in serum, but relapsed one month after treatment
The reduction in serum HCV RNA within less than
was discontinued. The third patient discontinued therapy
8 hrs that follows the initial subcutaneous injection ofSource: http://www.biokin.com/papers/Lake0326.pdf
Eur. J. Biochem. 250 , 242Ϫ248 (1997) Modulation of platelet-activating-factor production by incorporation of naturally occurring 1- O -alkylglycerols in phospholipids of human leukemic monocyte-like THP-1 cells Aziz HICHAMI 1, Vale´rie DUROUDIER 1, Ve´ronique LEBLAIS1, Laurent VERNHET 1, Franc¸ois LE GOFFIC2, Ewa NINIO 3 and Alain LEGRAND 11 Laboratoire de Pharmacologie Mole´culaire,