COMPARE

A randomised trial to compare the pharmacokinetic,pharmacodynamic,and antiviral effects of peginterferon alfa-2band peginterferon alfa-2a in patients with chronic hepatitis C (COMPARE)Marcelo Silva 1,*,Jorge Poo 2,Frank Wagner 3,Mary Jackson 4,David Cutler 4,Michael Grace 4,Ronald Bordens 4,Connie Cullen 4,Joann Harvey 4,Mark Laughlin 41Hospital Universitario Austral,Pilar,Argentina2CIF-BIOTEC/Hospital Medical Sur,Mexico City,Mexico3Clinical Research,Berlin,Germany4Schering-Plough Research Institute,Kenilworth,NJ,USASee Editorial,pages 172–173Background /Aims :To compare the pharmacokinetics,pharmacodynamics,and antiviral activity of peginterferon alfa-2b and peginterferon alfa-2a in patients with chronic hepatitis C virus genotype 1.Methods :Thirty-six patients were randomised to peginterferon alfa-2b (1.5l g /kg/week)or peginterferon alfa-2a (180l g/week)for 4weeks,then in combination with ribavirin (13mg/kg/day)for a further 4weeks.The pharmacokinetic profile of both peginterferons,mRNA expression of a selected group of interferon-induced gene transcripts,and serum HCV-RNA levels were assessed.Results :Patients receiving peginterferon alfa-2b had significantly greater up-regulation of interferon-alfa response genes compared with those receiving peginterferon alfa-2a.Correspondingly,patients treated with peginterferon alfa-2b also had a significantly greater log 10maximum and log 10time-weighted average decrease in serum HCV-RNA.A greater propor-tion of peginterferon alfa-2b patients achieved a P 2.0log 10reduction in serum HCV-RNA levels by week 8(72%vs 44%of peginterferon alfa-2a patients,P =0.09).There was an approximately 16-fold greater exposure to peginterferon in the serum of patients treated with peginterferon alfa-2a.Conclusions :These findings suggest that the biological activity,measured by early interferon-induced gene transcripts and early antiviral responsiveness,may have been greater in patients treated with peginterferon alfa-2b despite their lower exposure to the drug compared with patients treated with peginterferon alfa-2a.Ó2006European Association for the Study of the Liver.Published by Elsevier B.V.All rights reserved.Keywords :Chronic hepatitis C infection;Pegylated interferon;Peginterferon alfa-2b;Peginterferon alfa-2a;Ribavirin1.IntroductionHepatitis C virus (HCV)is the leading cause of liver transplantation in the US and Europe and is associatedwith an increased risk of hepatocellular carcinoma [1,2].Owing to the chronic nature of the disease,estimates suggest that its healthcare burden will increase dramat-ically [3].The current standard of treatment is based on interferon alfa therapy and consists of the combina-tion of peginterferon alfa-2plus ribavirin [4].Type I interferons do not interact directly with the virus but rather exert their effects through interaction with their specific receptor.The interaction initiates a0168-8278/$32.00Ó2006European Association for the Study of the Liver.Published by Elsevier B.V.All rights reserved.doi:10.1016/j.jhep.2006.03.008Received 15August 2005;received in revised form 7March 2006;accepted 9March 2006;available online 18April 2006*Corresponding author.Tel.:+542322482624.E-mail address:msilva@.ar (M.Silva)./locate/jhepJournal of Hepatology 45(2006)204–213cascade of intracellular Janus-kinase-mediated events, leading to changes in gene-transcript expression.Some of these up-regulated genes[e.g.,RNA-dependent pro-tein kinase(PKR),2050oligoadenylate synthetase (OAS)]are believed to have an important role in the interruption of HCV replication within infected hepato-cytes[5–7].Recent advances in pegylation chemistry have led to the development of interferon alfas with longer half-lives.The covalent attachment of a polyethylene glycol (PEG)molecule to the interferon alfa protein results in decreased renal clearance and a longer half-life in the plasma[8–10].The size and position of the PEG mole-cule used in the two currently licensed peginterferons: peginterferon alfa-2b(PegIntronÒ,Schering Corp., Kenilworth,NJ,USA)and peginterferon alfa-2a(Pega-sysÒ,Hoffmann-La Roche,Basel,Switzerland)differ significantly in their respective physical-chemical charac-teristics[11–14].Although pegylation improves the pharmacokinetic properties of the core protein,it also results in loss of in vitro biological activity[8,15].The antiviral activity of peginterferon alfa-2b is approxi-mately28%of the interferon alfa-2b core protein[12]. The antiviral activity of peginterferon alfa-2a ranges between1%and7%of the antiviral activity of the inter-feron alfa-2a core protein[16].Recent work has demon-strated that the size and site of attachment of the PEG moiety on the interferon alfa molecule markedly affects its specific activity[12,17].The purpose of this trial was to determine if these in vitro differences between peginterferon alfa-2b and peginterferon alfa-2a would translate to observable dif-ferences in their in vivo biological activity.This was done by assessing the effect of treatment on early viro-logic response as well as on expression of interferon response genes as markers of their mechanism of action [12,17].2.Materials and methods2.1.Study designThis was an8-week,double-blind,randomised,multicentre,paral-lel-group study comparing the pharmacokinetics,pharmacodynamics, and antiviral activity of peginterferon alfa-2b and peginterferon alfa-2a in patients infected with HCV genotype1.Study drug was prepared by a site pharmacist and administered by a qualified,independent third party who was blinded to protocol assignments.Neither the pharma-cist nor the individual administering the injection had further involve-ment in the study.Patients eligible for the study were randomised to receive once-weekly subcutaneous injections of either peginterferon alfa-2b1.5 l g/kg or peginterferon alfa-2a180l g for8weeks.After the fourth week of treatment,oral ribavirin therapy was added to the regimen at a dose of13mg/kg,in a divided BID dose.At the end of the study period,patients were offered a full course of weight-based peginterfer-on alfa-2b and ribavirin.The study protocol and patient-informed con-sent were approved by independent Ethics Committees of the participating institutions and regulatory agencies of their respective countries,and the study was conducted according to the Declaration of Helsinki and the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use(ICH)Guidance for Good Clinical Practice.Written informed consent was obtained from all patients prior to conducting study-related procedures.2.2.PatientsThirty-six treatment-naı¨ve patients between the ages of18and65 years who were infected with HCV genotype1a or1b[with a minimum of6.0·105HCV-RNA IU/mL,determined by quantitative polymer-ase chain reaction(PCR)]were eligible for enrollment into the study. Additional inclusion criteria were alanine aminotransferase(ALT)/as-partate aminotransferase(AST)levels610times the upper limit of normal(ULN),normal haemoglobin,white-blood-cell count P4000cells/l L,neutrophil count P1500cells/l L,and platelet count P100,000/l L.Subjects were excluded from participation if they had evidence of liver disease due to causes other than chronic HCV infec-tion,HIV positivity,haemoglobinopathy,haemophilia,severe pre-ex-isting psychiatric disease,poorly controlled diabetes mellitus, significant ischaemic heart disease,chronic obstructive lung disease, or active autoimmune disease.2.3.Study endpointsThe primary outcome was to assess the differential impact on inter-feron response genes and early viral kinetics of the two peginterferon treatments.Secondary outcome measures included the proportion of virologic responders(i.e.,patients with a P2.0log10decrease in HCV-RNA by8weeks of peginterferon treatment)as well as the safety and tolerability of treatment.2.4.Study proceduresWhen subjects met inclusion/exclusion criteria and qualified to enter the study,the investigator requested randomisation by faxing the central randomisation service.Only the site pharmacist responsible for medication preparation received confirmation of both the assigned treatment and the subject number.The investigator received the sub-ject number only.Screening evaluations were performed at three independent,research out-patient clinics.Patients were confined to pharmacology in-patient units on day1of week1(day1)and day1of week4(day22)for approximately72h while pharmacokinetic,pharmacodynamic,and antiviral samples were boratory safety tests were performed at the sites,but viral kinetics,interferon pharmacokinetics,and pharmacodynamics were centrally analysed in a blinded fashion.2.5.Interferon serum concentrationsSerum interferon samples were drawn at days1and22immediately before peginterferon administration and at6,10,12,24,48,72,and 120h after the dose.The peginterferon serum concentrations were determined in a blinded manner,using a validated immunological (electrochemiluminescent)assay[18,19]and standards for both alfa peg-interferons.Once the study was unblinded,the values from the appro-priate standard were reported in pg/mL as per the assigned treatment.2.6.Interferon-alfa response-gene-transcript analysisBlood samples for determination of interferon-stimulated gene reg-ulation were collected into PAX-Gene tubes(Becton–Dickinson, Franklin Lakes,NJ,USA)prior to thefirst dose and at6,24,48, and72h after thefirst dose of peginterferon.Samples were analysed for mRNA using quantitative real-time PCR(qRT-PCR)(TaqManÒ, Applied Biosystems,Foster City,CA,USA)[19,20].Interferon-modu-lated genes,including signal transducers and activators of transcrip-tion(STAT)-1,STAT-2,cyclin D,interferon-gamma,inducibleM.Silva et al./Journal of Hepatology45(2006)204–213205protein10(IP10),interferon-stimulated gene(ISG)15,ISG54,perforin, PKR,and2050OAS,were tested.Data were expressed as fold increase in mRNA over baseline.In order to investigate the cumulative expres-sion or sustained activity of treatment on gene transcription,changes in mRNA observed over the sampling period were expressed as maxi-mal fold increase from baseline(C max)and area under the curve (AUC),derived from the time-weighted averages of the mRNA fold-increase data.2.7.HCV-RNA serum determinationsHCV-RNA was determined prior to thefirst administration of study drug to establish the baseline and at0,24,48,72,120,and 168h after thefirst and fourth dose of peginterferon alfa.HCV-RNA was also assessed immediately prior to the weekly dose of pegin-terferon alfa at weeks2,3,5,6,7,and8.Data were expressed as log10 decrease in HCV-RNA from baseline.Maximal decrease and time-weighted average decrease over the sampling period were calculated from the data at weeks1and4.HCV-RNA was quantified by qRT-PCR(TaqManÒ;Schering-Plough Research Institute,Kenilworth, NJ,USA),with a lower limit of detection of29IU/mL.HCV genotyp-ing was performed by direct sequencing of the50-noncoding region.2.8.Adverse events and laboratory testingStandard safety laboratories were collected weekly.Adverse events were recorded at each visit and throughout the confinement.Patients were withdrawn from the study for a white-blood-cell count <1500cells/l L,neutrophil count<750cells/l L,platelet count <80,000/l L,creatinine clearance<50mL/min,ALT/AST levels >10·ULN,or indirect bilirubin>5mg/dL.No modification of the peginterferon alfa or ribavirin dose was allowed.2.9.Statistical analysisThe assessment of the viral concentration kinetics during weeks1 and4was made by determining the time-weighted mean decrease in viral concentration and the mean maximal decrease in viral concentra-tion over the week.For the weekly pre-dosing determinations,patients were defined as responders if a P2.0log10reduction in viral concentra-tion was achieved at the last determination.The viral kinetic differences between treatment groups were tested using Wilcoxon rank-sum tests of the average and maximum decreases from weekly baseline.The difference between treatment groups with respect to the once-weekly measurements of viral load change from baseline was tested using a model that tested for equality of slopes via SAS PROC MIXED.The model includedfixed effects terms for treatment,day,and treatment by day interaction(slope).Random effects included in the model were the patient and the addition of ribavirin.Using the observed standard deviation in the sample size calcula-tion,a sample size of18in each group would have80%power to detect a difference in means of log viral change of1.65(the difference between a group1mean ofÀ3.65and a group2mean ofÀ2.00)at week8, assuming that the common standard deviation is1.700using a two-group t-test with a0.050two-sided significance level.The association between the mRNA changes and the attainment of a P2.0log10drop in viral load was tested using Wilcoxon rank-sum tests of AUC of the fold increase in mRNA over baseline.The change from baseline of a patient’s last virologic observation,carried forward to week8if necessary,was used to ascertain whether a patient had attained a P2.0log10decrease in viral load during the treatment period.A post hoc multivariate correlation analysis using Spearman correlation values was performed in order to further explore the rela-tionship between treatment,gene expression,and viral response.To complement thesefindings,a secondary post hoc analysis using princi-pal components was conducted.Principal component analysis(PCA)is a variable reduction technique in which a number of[possibly]corre-lated variables are transformed into an equal number of uncorrelated variables called principal components[21].PCA was used to confirm the findings of thefirst post hoc analysis by determining whether the inter-feron-response genes identified in thefirst analysis remained signifi-cantly correlated with treatment and viral response.Table1Baseline characteristicsBaseline parameter Peginterferonalfa-2a(n=18)Peginterferonalfa-2b(n=18) Age(years)45.6(±11.8)a48.3(±9.7)a Males(n)910Ethnicity(n)Caucasian1614Hispanic24Weight(kg)71.0(±11.9)a69.6(±15.4)a Baseline viral load(·106IU/mL)1.8(±0.1)a 1.8(±0.2)aProthrombin time(%of control)109.5(±9.4)a111.4(±13.1)a Platelets(·103/l L)233(±75.8)a223(±80.7)a Neutrophils(·103/l L) 3.96(±1.26)a 3.34(±1.36)a ALT(IU/L)82.9(±59.8)a70.5(±51.5)a AST(IU/L)60.0(±42.8)a46.3(±25.5)aa Data are reported as means(±SD).Fig. 1.Individual time–concentration profiles following subcutaneous administration of(a)peginterferon alfa-2a180l g or(b)peginterferon alfa-2b1.5l g/kg.There was a$16-fold greater exposure in the serum of patients treated with peginterferon alfa-2a compared with peginterferon alfa-2b,but there was also a larger variability in patient exposure(38% vs20%,respectively).206M.Silva et al./Journal of Hepatology45(2006)204–2133.ResultsThe study was conducted in three centres,in Argenti-na,Mexico,and Germany,between November 2002and November 2003.3.1.Baseline characteristicsThe baseline characteristics in the two peginterferon treatment groups were comparable with respect to race,gender,weight,and viral load (Table 1).3.2.Pharmacokinetics resultsThe comparative exposure of patients to peginterfer-on alfa-2b and peginterferon alfa-2a during the first week of dosing is shown in Fig.1.There was a $16-fold greater exposure in the serum of patients treated withpeginterferon alfa-2a;however,there was also a larger variability in their exposure,compared with peginterfer-on alfa-2b (38%vs 20%,respectively).The week 1AUC 0–168h for each treatment group plotted against bodyweight is shown in Fig.2.There was no effect of weight on exposure when peginterferon alfa-2b was dosed according to patient weight.In contrast,when peginterferon alfa-2a was administered as a fixed dose,there was a decrease in exposure with increasing bodyweight.3.3.Interferon-alfa response-gene transcript results Thirty-one patients had sufficient,recoverable mRNA at baseline (as acquired via sampling prior to and immediately after the first treatment dose).Based on the results of a preliminary Wilcoxon rank-sum test that suggested a correlation with treat-ment response,transcripts were analysed for the fol-lowing interferon-response genes:IP10,ISG15,PKR,2050OAS,and ISG54[no statistically significant differ-ences in mRNA expression were observed for patients with or without a virologic response (P 2.0log 10decline in HCV-RNA after 8weeks of treatment)for the other genes tested].Patients who achieved a virologic response by 8weeks of treatment had significantly greater up-regula-tion of transcripts for the interferon-alfa response genes IP10,ISG15,PKR,2050OAS,and ISG54as assessed by mRNA C max and AUC levels (P <0.05;Table 2).There was consistently greater up-regulation of tran-scripts in patients treated with peginterferon alfa-2b (n =14)compared with patients treated with peginter-feron alfa-2a (n =17)(Table 2),with thedifferencesFig.2.Week 1exposure to peginterferon by weight.Unlike peginterferon alfa-2b (b),in which there was no effect of weight on exposure when dosed by patient weight,peginterferon alfa-2a (a),administered as a fixed dose,showed a trend towards decrease in exposure with increasing bodyweight.Table 2Mean AUC and C max of interferon-alfa response-gene transcripts analysed by virologic response and treatment groupResponders (n =17)Nonresponders (n =14)Peginterferon alfa-2a (n =17)Peginterferon alfa-2b(n =14)C max (fold increase)IP10253.2*83.4126.4237.3ISG1591.0*39.741.999.3PKR 11.3* 6.57.511.1*2050OAS 21.0*13.313.322.6*ISG5441.1*17.018.044.9*Mean AUC (fold increase Æh)IP104654.1*1703.72427.04408.1*ISG153589.8*1854.81886.73922.8*PKR 502.3*314.9345.2505.72050OAS 1099.6*669.0682.11176.1*ISG541368.7*630.6678.91468.1**P <0.05by Wilcoxon rank-sum test for responders vs nonre-sponders and peginterferon alfa-2b vs peginterferon alfa-2a.(Respond-ers were defined as patients with a P 2.0log 10decrease in HCV-RNAby week 8.)M.Silva et al./Journal of Hepatology 45(2006)204–213207reaching statistical significance (P <0.05)for the majority of the interferon-response genes investigated,as illus-trated in Fig.3.Within each treatment group,respond-ers demonstrated greater up-regulation of transcripts (as measured by AUC)than did nonresponders,although statistical significance was generally not demonstrated (Fig.4).The proportion of responders with available mRNA in the peginterferon alfa-2b group was 71%compared with 41%in the peginterferon alfa-2a group.3.4.Antiviral resultsThere was a significantly greater decline in HCV-RNA in patients treated with peginterferon alfa-2b com-pared with patients treated with peginterferon alfa-2a during weeks 1and 4(P <0.05)(Table 3),as illustrated by the mean maximum log 10reductions from baseline (Fig.5).At week 4(day 29),the mean log 10decrease from baseline was À1.891(95%CI:À2.468,À1.313)for peginterferon alfa-2b (n =16)and À1.331(95%CI:À2.159,À0.502)for peginterferon alfa-2a (n =17).The mean reduction from baseline in HCV-RNA over the course of 8weeks,by treatment group,is shown in Fig.6.The viral concentration curves begin to diverge at week 2.The curve for patients treated with peginterferon alfa-2b showed a trend towards a greater mean decrease in HCV-RNA through the 4weeks of monotherapy,a trend that persisted throughout the study.The rate of decline (95%CI)in HCV-RNA during the 8weeks of treatment was significantly greater in patients treated with peginter-feron alfa-2b compared with the rate for patients treated with peginterferon alfa-2a,À0.346(À0.396,À0.296)log 10/week vs À0.233(À0.281,À0.185)log 10/week,respectively (P <0.002).At the end of the study,the mean reduction in HCV-RNA was 3.13log 10in patients treated with peginterferon alfa-2b and 2.44log 10in patients treated with peginterferon alfa-2a (Fig.6).There was a larger number of virologic responders (defined as a P 2.0log 10reduction in HCV-RNA)to peginterferon alfa-2b than peginterferon alfa-2a at week 4(50%vs 28%,respectively;P =0.17)and week 8(72%vs 44%,respectively;P =0.09).Although sustained virologic response (SVR)was not measured as part of the study,such data were available for one of the centres (Argentina).Eleven patients were treated with peginterferon alfa-2b for 48weeks,of which 9patients achieved SVR,1patient was a non-responder,and 1patientdiscontinued.Fig. 3.Mean AUC of interferon-alfa response-gene transcripts by treatment group.There was consistently greater up-regulation of transcripts in patients treated with peginterferon alfa-2b compared with patients treated with peginterferon alfa-2a,with the differences reaching statistical significance (P <0.05)for most of the interferon-response genesinvestigated.Fig.4.Mean AUC of interferon-alfa response-gene transcripts analysed by within-treatment responders vs nonresponders.In general,responders demonstrated more up-regulation of transcripts than nonresponders,although statistical significance was only reached for one AUC group in each treatment arm.The proportion of responders with available mRNA in the peginterferon alfa-2b group (b)was 71%compared with 41%in the peginterferon alfa-2a group (a).(Responders were defined as patients with a P 2.0log 10decrease in HCV-RNA by week 8.)208M.Silva et al./Journal of Hepatology 45(2006)204–2133.5.Post hoc multivariate correlation analysisPairwise correlations(Spearman correlation values) between mRNA AUC values(gene expression),treat-ment response,and treatment arm(peginterferon alfa-2b or peginterferon alfa-2a)were determined to explore the relationship between these variables.IP10,ISG15, ISG54,PKR,and2050OAS were significantly correlat-ed with treatment group and virologic response at week 8(Table4).A secondary correlation analysis using the combined set of mRNA AUC data for thesefive genes and their derived principal components showed that thefirst principal component,which accounts for 63%of the variability in each individual mRNA AUC value,was significantly correlated with both treatment group and treatment response(Table5),thus providing further evidence of a link between these variables.3.6.Adverse eventsAdverse events were typical of those reported previ-ously with administration of pegylated interferons: fever,myalgia,headache,flu-like symptoms,fatigue, anaemia,and leucopenia.Treatment-emergent adverse events are shown in Table6(classified by body system). Patients treated with peginterferon alfa-2a showed sta-tistically significantly greater reductions in total white-blood-cell count over the8weeks of treatment compared with patients treated with peginterferon alfa-2b(P=0.04,Fig.7).This was due to statistically significantly greater reductions in the absolute neutro-phil count(P=0.03),as shown in Fig.8.There were no differences in monocyte or lymphocyte counts between the two groups(data not shown).Six patients discontinued treatment due to adverse events prior to completion of the8-week study,four receiving treatment with peginterferon alfa-2b and two with peginterferon alfa-2a.Three patients in the peginterferon alfa-2b group discontinued due to reduced platelet counts(<80,000/l L);however,all three began treatment with platelet counts close to the exclusion threshold,ranging from96,000to100,000/l L.The patient in the peginterferon alfa-2a group who discontinued due to thrombocytopenia had a baseline level of125,000/l L that dropped to71,000/l L.The remaining reasons for subject discontinuations were a case of anaemia in the peginterferon alfa-2b group and one of erythaema multiforme in the peginterferon alfa-2a group.4.DiscussionIn this study,we used changes in the mRNA abun-dance for genes known to be regulated by interferons as a marker of biological activity.The data presented are consistent with the in vitro data[12,17,20]and sug-gest that biological activity,as indicated by gene response,is greater with peginterferon alfa-2b than with peginterferon alfa-2a.Beginning as early as week2, when the viral concentration curves began to diverge, the rate of decline in HCV-RNA during the8weeks of treatment was significantly greater in patients treated with peginterferon alfa-2b compared to patients treated with peginterferon alfa-2a.The kinetic studies,conduct-ed at weeks1and4during the monotherapy phase, revealed that treatment with peginterferon alfa-2b resulted in significantly greater reduction in the maximal and time-weighted mean viral concentration during each week when compared with peginterfesron alfa-2a.Cor-respondingly,compared with peginterferon alfa-2a, peginterferon alfa-2b treatment also stimulated greater increases in many interferon-alfa response-gene tran-scripts for proteins believed to be associated with antiviral activity.The apparently greater biologicalTable3HCV viral load reductions during weeks1and4Week1Week4Peginterferon alfa-2a(n=18)Peginterferon alfa-2b(n=18)Peginterferon alfa-2a(n=18)Peginterferon alfa-2b(n=18) Log10time-weighted average decrease in HCV-RNAMean(±SD)0.58(0.76) 1.21(0.52)*0.04(0.29)0.73(0.48)***Median0.43 1.240.040.64Max 2.36 2.320.44 1.18MinÀ0.290.20À0.45À0.01Maximum log10decrease from baseline in HCV-RNAMean(±SD) 1.08(0.93) 2.11(0.79)**0.38(0.33) 1.46(0.83)***Median0.64 2.080.29 1.44Max 3.23 3.310.96 2.92Min0.000.590.000.33*P<0.05by Wilcoxon rank-sum test for peginterferon alfa-2b vs peginterferon alfa-2a.**P<0.01by Wilcoxon rank-sum test for peginterferon alfa-2b vs peginterferon alfa-2a.***P<0.001by Wilcoxon rank-sum test for peginterferon alfa-2b vs peginterferon alfa-2a.M.Silva et al./Journal of Hepatology45(2006)204–213209activity associated with peginterferon alfa-2b treatment was observed despite an approximately16-fold greater exposure in the serum of patients treated with peginter-feron alfa-2a.Thefindings of the post hoc analysis(Tables4and5) and the data shown in Fig.4both demonstrated that treatment response was correlated with a significantly greater increase in mRNA expression.They suggest that the higher antiviral activity of peginterferon alfa-2b is directly related to its higher associated interferon-mediated gene expression.However, because of small patient numbers,these analyses, and their implications for other therapies,require con-firmation in larger trials.Adverse events and impact on haematological param-eters were not unexpected for peginterferon alfa treat-ment.However,there was a significantly greater reduction in white-blood-cell and neutrophil counts associated with peginterferon alfa-2a treatment,even though the biological activity in these patients was low-er,afinding recently reported elsewhere[22].Peginter-feron alfa-2b treatment was associated with a greater incidence of fever compared with peginterferon alfa-2a (10patients vs1,respectively).Becauseflu-like symp-toms and fever are associated with the natural response induced by endogenous interferon,this may reflect the greater biological activity of peginterferon alfa-2b. These observations,however,need to be confirmed in a larger study before making any conclusions regarding their clinical implications.This study provides a better understanding of the in vivo impact that the size and site of an attached PEG molecule can have on interferon-alfa-dependent biological activity and subsequent anti-viral activity. Because peginterferon alfa-2b appears to induce inter-feron-response genes more effectively than peginterferon alfa-2a,despite a lower drug exposure,this implies that peginterferon alfa-2b has a better interaction with the interferon-alfa receptor.However,due todifferential Fig.6.Mean HCV viral load reduction at the end of each weekly dosing interval.Baseline viral loads were equivalent in both treatment groups, 1.8·106IU/mL;however,the rate of decline(95%CI)in HCV-RNA during the8weeks of treatment was significantly greater in patients treated with peginterferon alfa-2b than in patients treated with pegin-terferon alfa-2a,À0.346(À0.396,À0.296)log10/week vsÀ0.233 (À0.281,À0.185)log10/week,respectively(P<0.002).[The estimated slopes(i.e.,HCV-RNA reduction rates,95%CIs,and subsequent P-value)were derived from the mixed-model analysis.]Each data point is the mean HCV concentration determined at the pre-dosing weekly samplinginterval.Fig.5.Mean maximum HCV viral load reductions from baseline duringweeks1and4.The mean maximum log10decrease from baseline(day22for week4)in HCV-RNA was significantly greater in patients treatedwith peginterferon alfa-2b than in patients treated with peginterferonalfa-2a during both weeks1(a)and4(b)(P<0.01).210M.Silva et al./Journal of Hepatology45(2006)204–213。