Article first published online: 9 MAY 2008
Copyright © 2008 American Association for the Study of Liver Diseases
Volume 48, Issue 2, pages 407–417, August 2008
How to Cite
Zeuzem, S., Yoshida, E. M., Benhamou, Y., Pianko, S., Bain, V. G., Shouval, D., Flisiak, R., Rehak, V., Grigorescu, M., Kaita, K., Cronin, P. W., Pulkstenis, E., Subramanian, G. M. and McHutchison, J. G. (2008), Albinterferon alfa-2b dosed every two or four weeks in interferon-naïve patients with genotype 1 chronic hepatitis C. Hepatology, 48: 407–417. doi: 10.1002/hep.22403
This study is registered on ClinicalTrials.gov (identifier NCT00115908; http://clinicaltrials.gov/ct2/show/NCT00115908?term=human+genome+sciences&rank=2)
Potential conflict of interest: Dr. McHutchison is a consultant for and received grants from Anadys, Colsy Pharmaceutical Group, First Circle Medical, GlaxoSmithKline, Human Genome Sciences, Idenix Pharmaceuticals, Intarcia, Novartis, Pharmasset, Pfizer, Roche, Schering-Plough, Valeant Pharmaceuticals, Vertex Pharmaceuticals, and Wyeth. He received grants from Globe Immune, Medtronics, Salix Pharmaceuticals, and Sanofi-Aventis. He is also a consultant for Biolex, Epiphany Biosciences, InterMune Pharmaceuticals, National Genetics Institute, Peregrine, and United Therapeutics. Drs. Subramanian, Cronin and Pulkstenis own stock in Human Genome Sciences, Inc. Dr. Kaita is a consultant for, is on the speakers' bureau of, and received grants from Schering-Plough, Roche, and Gilead. Dr. Zeuzem is a consultant for, is on the speakers' bureau of, and received grants from Human Genome Sciences, Novartis, Roche, and Schering-Plough. Dr. Yoshida received grants from and advises Hoffmann-La Roche. He also received grants from Schering-Plough, GlaxoSmithKline, Ortho Janssen, Pfizer, Gilead Sciences, and Cangene. He advises Novartis. Dr. Benhamou is a consultant for and is on the speakers' bureau of Human Genome Science and Roche. Dr. Pianko is a consultant for, advises, and is on the speakers' bureau of Human Genome Sciences, Novartis, and Roche. Dr. Bain is a consultant for Human Genome Sciences. Dr. Shouval received grants from Human Genome Sciences. Dr. Flisiak received grants from Debiopharm. He advises and received grants from Human Genome Sciences and Roche.
- Issue published online: 29 JUL 2008
- Article first published online: 9 MAY 2008
- Accepted manuscript online: 9 MAY 2008 12:00AM EST
- Manuscript Accepted: 25 APR 2008
- Manuscript Received: 21 FEB 2008
- Human Genome Sciences Inc., Rockville, MD
- Novartis Pharma AG, Basel, Switzerland
The efficacy and safety of albinterferon alfa-2b (alb-IFN), a novel recombinant protein consisting of interferon alfa-2b genetically fused to human albumin, was evaluated in a phase 2b, open-label study of patients with genotype 1, chronic hepatitis C. In all, 458 IFN-alfa treatment-naïve patients were randomized to 48-week treatment with peginterferon alfa (PEG-IFNα)-2a 180 μg one time per week (qwk), or alb-IFN 900 or 1,200 μg once every two weeks (q2wk), or 1,200 μg once every four weeks (q4wk), administered subcutaneously, plus weight-based oral ribavirin 1,000 or 1,200 mg/day. Hepatitis C virus RNA was measured by real-time polymerase chain reaction (limit of detection: 10 IU/mL). The primary efficacy endpoint was sustained virologic response (hepatitis C virus RNA <10 IU/mL 24 weeks after the end of treatment). By intention-to-treat analysis, sustained virologic response rates were 58.5% (69/118) with alb-IFN 900 μg q2wk, 55.5% (61/110) with 1,200 μg q2wk, and 50.9% (59/116) with 1,200 μg q4wk, and 57.9% (66/114) with PEG-IFNα-2a (P = 0.64 for overall test). Discontinuation rates due to adverse events were 9.3% with alb-IFN 900 μg q2wk, 18.2% with 1,200 μg q2wk and 12.1% with 1,200 μg q4wk, and 6.1% with PEG-IFNα-2a (P = 0.04). Hematologic reductions were lowest in the q4wk group and comparable across other groups. At week 12, mean treatment-associated missed workdays were significantly lower with alb-IFN 900 μg q2wk versus PEG-IFNα-2a (1.1 versus 4.3 days; P = 0.006). Conclusion: Alb-IFN administered q2wk or q4wk may offer comparable efficacy, with an improved dosing schedule, compared with PEG-IFNα-2a. (HEPATOLOGY 2008;48:407–417.)
Chronic hepatitis C (CHC) infection is one of the most common causes of chronic liver disease, with approximately 3% of the world's population (or as many as 170 million people) infected.1, 2 Peginterferon alfa (PEG-IFNα)-2a or PEG-IFNα-2b dosed once weekly in combination with daily ribavirin (RBV) represents the current standard of care for the treatment of CHC, and has demonstrated superior sustained virologic response (SVR) rates and health-related quality of life (HRQOL) compared with standard interferon-alfa (IFNα) therapy.3-8 Approximately 70% of all patients with CHC are infected with genotype 1, hepatitis C virus (HCV), 42%-52% of whom achieve SVR with PEG-IFNα therapy.4, 6, 9, 10 New treatment options for patients with genotype 1 CHC that offer improved efficacy, tolerability, and convenience are clearly needed.
Albinterferon alfa-2b (alb-IFN) is an 85.7-kDa protein consisting of recombinant human IFNα-2b genetically fused to recombinant human albumin. In vitro assays have shown that alb-IFN retains the antiviral properties of IFNα.11 Dose-ranging phase 1/2 studies of alb-IFN in patients with CHC demonstrated acceptable tolerability at doses up to 1,200 μg, an extended half-life of ∼144 hours, and evidence of dose-dependent antiviral activity in IFN treatment-experienced and IFN treatment-naïve patients.12, 13 The extended pharmacokinetics of alb-IFN provides detectable drug throughout the entire dosing interval, corresponding to viral dynamic changes observed with 900-μg and 1,200-μg doses in IFNα treatment-naïve patients with genotype 1 CHC.12 These data provided the rationale for exploring doses of alb-IFN at 2-week intervals or 4-week intervals in patients with CHC.
This is the first report of the efficacy, safety, and patient-reported HRQOL of alb-IFN administered every 2 or 4 weeks compared with PEG-IFNα-2a, all in combination with RBV, in previously untreated patients with genotype 1 CHC.
Patients and Methods
Adult patients with genotype 1 CHC who had not previously received IFNα therapy were invited to participate in the study. Patients were excluded if they had decompensated liver disease or other possible etiologies for chronic liver disease; thrombocytopenia (<100,000 platelets/mm3); neutropenia (<1,800 neutrophils/mm3); history of moderate–severe psychiatric disease; immunologically mediated disease; uncontrolled thyroid disease; coinfection with hepatitis B virus or human immunodeficiency virus; a significant coexisting medical condition; or alcohol or drug dependence.
Alb-IFN is prepared as a sterile, single-use, lyophilized product.12 PEG-IFNα-2a (PEGASYS; Hoffmann-La Roche Inc., Nutley, NJ) was supplied in sterile, single-use, graduated, clear-glass, prefilled syringes delivering 180 μg of drug product. Ribavirin (COPEGUS; Roche Laboratories, Nutley, NJ) was supplied as 200-mg tablets.
This phase 2b, randomized, multicenter, active-controlled, open-label, dose-ranging study was conducted at 82 centers in eight countries (Australia, Canada, Czech Republic, France, Germany, Israel, Poland, and Romania) between May 2005 and May 2007. The median number of patients enrolled per site was four (range, 1-29). A centralized randomization using an interactive voice-response system assigned patients in blocks of eight, in an equal allocation (2:2:2:2) ratio, to one of four treatment groups, including three different alb-IFN groups (900 μg once every two weeks [q2wk], 1,200 μg q2wk, and 1,200 μg once every four weeks [q4wk]) and the active-control group PEG-IFNα-2a 180 μg once per week (qwk), with both agents administered subcutaneously. All patients were to receive oral RBV 1,000 mg/day or 1,200 mg/day based on body weight (<75 kg or ≥75 kg, respectively) in two divided doses. Randomization was stratified by pretreatment serum HCV-RNA levels (<800,000 IU/mL or ≥800,000 IU/mL) and body mass index (<25 kg/m2 or ≥25 kg/m2). Treatment duration was 48 weeks, with a 24-week follow-up. The study protocol recommended stepwise (≥1 level) dose reductions of alb-IFN to 900, 700, and 500 μg, and of PEG-IFNα-2a to 135 μg, 90 μg, and 45 μg, to manage adverse events (AEs) and hematologic reductions. Criteria for dose reductions for hematologic values included reducing the alb-IFN or PEG-IFNα-2a dose for absolute neutrophil counts (ANC) of 500-750/mm3 or platelet counts of 30,000-50,000/mm3, and holding or delaying the dose for ANC <500/mm3 or platelet counts <30,000/mm3. The RBV dose was to be reduced to 600 mg/day for hemoglobin of >8.5-10 g/dL and held steady for ≤8.5 g/dL. Likewise, dose reductions (1-3 levels or holding the dose) were recommended for managing moderate–severe AEs, based on the discretion of the clinical investigator.
The institutional review boards of participating centers approved the protocol. All patients provided written informed consent. Human Genome Sciences Inc. (Rockville, MD) sponsored the study, and was responsible for collection and statistical analysis of the data. The conduct of the study was coordinated centrally by the Duke Clinical Research Institute (Durham, NC). The authors had full access to the data. The study was conducted according to the guideline provisions of “Good Clinical Practices,” as defined in the International Conference on Harmonisation guidelines.
Serum HCV RNA and HCV Genotype.
HCV-RNA levels were assessed by real-time polymerase chain reaction (CE-marked COBAS Ampliprep/COBAS Taqman HCV test; Roche Diagnostics, Basel, Switzerland). The dynamic range of this assay was 43 IU/mL-69 million IU/mL. The limit of quantitation (LOQ) was 43 IU/mL, while the lower limit of detection (LOD) was 10 IU/mL. HCV genotyping was based on hybridization of the amplified segment of the 5′ nontranslated region of the HCV genome, with oligonucleotide probes representing genotype 1 (Abbott Diagnostics assay).
The primary efficacy endpoint was SVR, defined as HCV RNA < LOD at 24 weeks after the end of therapy. The major secondary efficacy endpoints (on-treatment virologic response rates) were rapid virologic response at week 4 (RVR4), defined as HCV RNA < LOQ; early virologic response at week 12 (EVR12), defined as HCV RNA < LOQ or a ≥2-log reduction in HCV RNA; undetectable HCV RNA (<LOD) at weeks 24 and 48; and end-of-treatment response, defined as undetectable HCV RNA (<LOD) at the end of treatment.
Evaluation of safety included AE monitoring, physical examination, and clinical laboratory assessments (hematology, serum chemistry, and urinalysis). Safety was assessed from the first dose of study agent to completion of treatment (up to 48 weeks) and through at least 12 weeks after completion of treatment. AEs were graded using the Division of Microbiology and Infectious Diseases toxicity tables to assess severity. The development of antibody response (antibodies to alb-IFN/PEG-IFNα-2a and human albumin, as well as neutralizing response to alb-IFN/PEG-IFNα-2a) was also assessed prior to dosing and periodically up to 24 weeks posttreatment using immunogenicity assays based on enzyme-linked immunosorbent assay methodology and standard neutralization assays.12
Health-Related Quality of Life (SF-36 and Disability Days).
The SF-36v2 Health Survey is a validated questionnaire that measures patient health status using a 4-week recall period.14, 15 The eight multi-item health domains include physical functioning, role-physical, bodily pain, general health, vitality, social functioning, role-emotional, and mental health. Two aggregated scores—physical and mental component summaries—were also derived. Lower scores indicate worse HRQOL. Patient disability days were defined as the number of missed workdays in the previous 4 weeks. Questionnaires were administered to patients prior to any procedures/evaluations at each visit on day 0 (pretreatment), and at weeks 4, 12, 24, and 48 during treatment, and weeks 4, 12, and 24 posttreatment.
A target sample size of 110 patients per treatment group was selected to provide estimation, via 95% confidence interval (CI), of the SVR rate for alb-IFN with a precision of ±10 percentage points. All analyses were performed in the intention-to-treat population, defined as the subset of all randomized patients who received ≥1 dose of study agent. Adherence to IFNα and RBV therapy was calculated as the total dose received/planned (based on 48 weeks of planned full-dose treatment) and expressed as a percentage. All statistical tests were two-sided and performed at the 5% level of significance. All analyses were performed using SAS 9 (SAS Institute Inc., Cary, NC) and R (version 1.9.1) statistical software.
The SVR and on-treatment response rates for each treatment group, and the differences between each alb-IFN group and the PEG-IFNα-2a group were estimated with 95% CIs. Statistical testing was performed using the likelihood ratio test (or Fisher's exact test when >20% of expected contingency table cell counts were <5). As a secondary analysis of the primary efficacy endpoint, a logistic-regression model was used to examine the association of SVR with treatment, as well as with other covariates of interest including treatment adherence (≥80% adherence to both IFN and RBV), gender, age, weight ≥75 kg, body mass index, alanine aminotransferase >3× the upper limit of normal (ULN), γ-glutamyl transpeptidase ≤ULN, and HCV RNA. A logistic-regression model was fit including all main effects and two-way interactions with treatment-group indicator variables. The model was reduced one term at a time, beginning with all nonsignificant two-way interactions and followed by main effects in order of the highest P value, with the model being refit at each step. This process resulted in a final model for inference. Analysis of variance was used to determine treatment imbalances at baseline for continuous variables; the likelihood ratio test (or Fisher's exact test) was used to determine treatment imbalances at baseline for categorical variables.
Scores for SF-36 domains, and aggregate physical and mental components were summarized. Scores for the alb-IFN and PEG-IFNα-2a groups were compared using analysis of covariance controlling for baseline score. Similarly, mean/median numbers of disability days and percentages of patients missing various levels of work were compared using the Wilcoxon test and the likelihood ratio test (or Fisher's exact test), respectively. Missing data for SF-36 and disability days were handled by using the last (postbaseline)-observation-carried-forward method.
In all, 585 patients were screened; of 478 who were randomized, 458 received ≥1 dose of study drug (Fig. 1). The most common reasons for screening failure were the violation of inclusion/exclusion criteria (n = 92) and patient request (n = 14). The 20 patients who did not receive treatment after randomization were evenly distributed among treatment groups. Alb-IFN was administered to 344 patients, whereas PEG-IFNα-2a was given to 114. In all, 369 patients (80.6%) completed treatment per protocol, defined as completing the 48-week treatment period (or discontinuation due to lack of efficacy). Demographics and baseline characteristics are summarized in Table 1 and were generally comparable across treatment groups. This genotype 1 population was predominately Caucasian (92.4%) and male (59.8%), with a median age of 44 years.
|PEG-IFNα-2a 180 μg qwk (n = 114)||Alb-IFN 900 μg q2wk (n = 118)||Alb-IFN 1,200 μg q2wk (n = 110)||Alb-IFN 1,200 μg q4wk (n = 116)||P Value|
|Male (%)||66 (57.9)||66 (55.9)||64 (58.2)||78 (67.2)||–|
|Female (%)||48 (42.1)||52 (44.1)||46 (41.8)||38 (32.8)||–|
|All patients, n||114||118||110||116||0.51|
|Caucasian, n (%)||105 (92.1)||109 (92.4)||101 (91.8)||108 (93.1)||0.99|
|Mean age ± SD, year||41.9 ± 10.3||42.5 ± 10.5||41.3 ± 11.4||42.7 ± 10.7||0.75|
|Mean weight ± SD, kg||73.4 ± 17.0||74.6 ± 15.3||76.4 ± 15.2||77.3 ± 17.2||0.27|
|≥75 kg, n (%)||46 (40.4)||59 (50.0)||59 (53.6)||62 (53.4)||0.15|
|Mean BMI ± SD, kg/m2||25.1 ± 4.3||25.4 ± 4.8||25.6 ± 4.6||26.1 ± 5.0||0.48|
|≥25 kg/m2, n (%)||59 (51.8)||58 (49.2)||53 (48.2)||59 (50.9)||0.95|
|Mean HCV RNA ± SD, log10 IU/mL||6.0 ± 0.8||6.0 ± 0.7||6.0 ± 0.8||6.0 ± 0.7||0.92|
|≥800,000 IU/mL, n (%)||72 (63.2)||67 (56.8)||60 (54.5)||73 (62.9)||0.44|
|<800,000 IU/mL, n (%)||42 (36.8)||51 (43.2)||50 (45.5)||43 (37.1)||0.44|
|<400,000 IU/mL, n (%)||27 (23.7)||35 (29.7)||36 (32.7)||27 (23.3)||0.30|
|Mean ALT ± SD, IU/L||92.3 ± 103.7||85.4 ± 64.7||77.4 ± 52.9||88.0 ± 74.5||0.52|
|>1.5× ULN, n (%)||42 (36.8)||49 (41.5)||45 (40.9)||47 (40.5)||0.89|
|F3-4, n (%)||9/73 (12.3)||12/73 (16.4)||10/69 (14.5)||8/67 (11.9)||0.86|
Antiviral Response and Efficacy
Table 2 details the intention to treat analysis of antiviral response rates observed in the study. The SVR rates were comparable (55%-58%) across the alb-IFN q2wk and PEG-IFNα-2a treatment groups, but slightly lower (51%) in the alb-IFN 1,200-μg q4wk group (P = 0.28 versus PEG-IFNα-2a). Relapse rates (that is, proportions of patients who became HCV-RNA positive after having achieved an end-of-treatment response) were comparable across the alb-IFN 1,200-μg and PEG-IFNα-2a groups (28%-31%), and lowest in the alb-IFN 900-μg q2wk group (21%; P = .22 versus PEG-IFNα-2a). Rates of virologic breakthrough (that is, proportions of patients who became HCV-RNA positive on treatment after having previously achieved HCV RNA < LOD) were also comparable across treatment groups, although lowest in the alb-IFN 1,200-μg q2wk arm. Early antiviral response rates are shown in Fig. 2. The alb-IFN 1,200-μg q2wk arm showed greater antiviral activity, with a higher proportion of patients achieving HCV RNA < LOQ at weeks 4 (33.6% [+7.3%]; P = 0.23) and 12 (74.6% [+8.8%]; P = 0.15) compared with PEG-IFNα-2a (26.3% and 65.8%, respectively). At week 2 following a single dose of alb-IFN, the proportion of patients who achieved a ≥2-log HCV-RNA reduction was significantly higher in the alb-IFN 1,200-μg arms than in the PEG-IFNα-2a arm (P = 0.03). The alb-IFN 900-μg q2wk arm had an overall antiviral response profile that was similar to PEG-IFNα-2a. The alb-IFN 1200-μg q4wk arm had the lowest antiviral response rates from weeks 4 to 24.
|PEG-IFNα-2a 180 μg qwk (n = 114)||Alb-IFN 900 μg q2wk (n = 118)||Alb-IFN 1,200 μg q2wk (n = 110)||Alb-IFN 1,200 μg q4wk (n = 116)||P Value|
|SVR (ITT) (%) [95% CI]||66 (57.9) [48.3, 67.1]||69 (58.5) [49.0, 67.5]||61 (55.5) [45.7, 64.9]||59 (50.9) [41.4, 60.3]||0.64|
|SVR and IFN+RBV treatment adherence|
|≥80% (%)||48/72 (66.7)||47/65 (72.3)||36/51 (70.6)||49/79 (62.0)||0.57|
|<80% (%)||18/42 (42.9)||22/53 (41.5)||25/59 (42.4)||10/37 (27.0)||0.39|
|Relapse (%) [95% CI]||26/90 (28.9) [19.8, 39.4]||18/86 (20.9) [12.9, 31.0]||27/88 (30.7) [21.3, 41.4]||23/81 (28.4) [18.9, 39.5]||0.34|
|Relapse and IFN+RBV treatment-adherence|
|≥80% (%)||16/63 (25.4)||7/53 (13.2)||12/48 (25.0)||16/65 (24.6)||0.31|
|<80% (%)||10/27 (37.0)||11/33 (33.3)||15/40 (37.5)||7/16 (43.8)||0.92|
|Breakthrough (%) [95% CI]||4/96 (4.2) [1.1, 10.3]||5/95 (5.3) [1.7, 11.9]||3/95 (3.2) [0.7, 9.0]||5/88 (5.7) [1.9, 12.8]||0.86|
|Breakthrough and IFN+RBV adherence|
|≥80% (%)||1/65 (1.5)||2/58 (3.4)||1/50 (2.0)||2/69 (2.9)||0.95|
|<80% (%)||3/31 (9.7)||3/37 (8.1)||2/45 (4.4)||3/19 (15.8)||0.46|
Predictors of SVR.
The predictive value for SVR of early antiviral response was consistent among all treatment groups (Table 3). RVR at week 4 had a high positive predictive value for SVR (83%-100%), whereas lack of EVR12 and detectable HCV RNA at week 24 both had a high negative predictive value for SVR (85%-100%). The factors predictive of SVR or relapse were consistent across all four treatment groups. Adherence to therapy (≥80%), baseline viral load <400,000 IU/mL, serum alanine aminotransferase levels >3× ULN, and baseline serum γ-glutamyl transpeptidase levels ≤ULN were strongly associated with a greater likelihood of SVR (P < 0.001). Additional modeling was performed omitting adherence as a potential covariate to allow identification of baseline factors that may be associated with adherence to impact SVR. This resulted in the same final model, as no additional factors or interactions were found to be significant.
|PEG-IFNα-2a 180 μg qwk (n = 114)||Alb-IFN 900 μg q2wk (n = 118)||Alb-IFN 1,200 μg q2wk (n = 110)||Alb-IFN 1,200 μg q4wk (n = 116)|
|SVR, n (%)||66 (57.9)||69 (58.5)||61 (55.5)||59 (50.9)|
|RVR4, n (%)*||30 (26.3)||29 (24.6)||37 (33.6)||21 (18.1)|
|EVR12, n (%)†||101 (88.6)||99 (83.9)||99 (90.0)||90 (77.6)|
|Week-24 response, n (%)‡||87 (76.3)||85 (72.0)||87 (79.1)||72 (62.1)|
Adherence to Therapy.
The SVR rates in patients who received ≥80% of the prescribed IFN and RBV therapy were higher in the alb-IFN 900-μg (72.3%) and alb-IFN 1,200-μg (70.6%) q2wk groups than in the PEG-IFNα-2a (66.7%) and alb-IFN 1,200-μg q4wk (62.0%) groups (P = 0.57 for overall test; Table 2). Of note, in heavier patients (≥75 kg) adherent to therapy, SVR rates declined to 53.3% with PEG-IFNα-2a, but were maintained with alb-IFN 900 μg q2wk (74.2%), alb-IFN 1,200 μg q2wk (67.9%), and alb-IFN 1,200 μg q4wk (61.0%); this was reflective of the higher observed relapse rate in these patients with PEG-IFNα-2a (37.5%) than with those in the alb-IFN regimens (11.5%, 29.6%, and 21.9%, respectively). The overall reduced rate of virologic breakthrough in the alb-IFN 1200-μg q2wk group was driven by the fact that patients in this group who were <80% adherent to RBV had a lower breakthrough rate than did those in the other treatment groups (Table 2).
Safety and HRQOL
The alb-IFN 900-μg q2wk and alb-IFN 1,200-μg q4wk treatment groups had overall safety profiles comparable to that of the PEG-IFNα-2a group (Tables 4 and 5). In the alb-IFN 1,200-μg q2wk group, the rates of certain AEs, as well as severe/serious AEs, were increased, although the types of AEs reported were as expected with any IFN therapy (Table 4). Rates of discontinuation due to AEs were comparable between the alb-IFN 900-μg q2wk, alb-IFN 1,200-μg q4wk, and PEG-IFNα-2a groups (9%, 12%, and 6%, respectively), and were higher in the alb-IFN 1,200-μg q2wk group (18%; P = 0.04 for overall test; Table 5). One death occurred during the study, due to cardiac arrest secondary to purulent bronchitis, in a patient receiving PEG-IFNα-2a. The most frequent AEs observed were common and expected side effects of IFN therapy (Table 4). Most AEs were mild to moderate in severity and resolved or stabilized after the completion of treatment. Of the common AEs, alopecia (hair thinning) occurred at a higher rate in the alb-IFN treatment groups (46.8%) than in the PEG-IFNα-2a group (23.7%). The rates of cough and dyspnea were also higher with alb-IFN 1,200 μg q2wk, and were reported primarily in the first 12 weeks of treatment, with the rates declining thereafter. The majority of AEs involving cough, dyspnea, and alopecia were mild in severity, resolved during or after the study, and did not lead to discontinuations. There was no increased risk of clinically significant infection in the alb-IFN groups; all groups had comparable rates, including for severe infections.
|Most frequent AEs (>15% of patients)||PEG-IFNα-2a 180 μg qwk (n = 114) (%)||Alb-IFN 900 μg q2wk (n = 118) (%)||Alb-IFN 1,200 μg q2wk (n = 110) (%)||Alb-IFN 1,200 μg q4wk (n = 116) (%)|
|Headache||64 (56.1)||68 (57.6)||58 (52.7)||64 (55.2)|
|Fatigue||57 (50.0)||57 (48.3)||66 (60.0)||57 (49.1)|
|Pyrexia||47 (41.2)||55 (46.6)||55 (50.0)||54 (46.6)|
|Myalgia||43 (37.7)||47 (39.8)||49 (44.5)||61 (52.6)|
|Alopecia||27 (23.7)||61 (51.7)||53 (48.2)||47 (40.5)|
|Insomnia||44 (38.6)||44 (37.3)||31 (28.2)||37 (31.9)|
|Nausea||39 (34.2)||39 (33.1)||39 (35.5)||39 (33.6)|
|Arthralgia||42 (36.8)||30 (25.4)||37 (33.6)||40 (34.5)|
|Cough||31 (27.2)||37 (31.4)||45 (40.9)||31 (26.7)|
|Weight decreased||31 (27.2)||34 (28.8)||45 (40.9)||32 (27.6)|
|Pruritus||39 (34.2)||37 (31.4)||18 (16.4)||38 (32.8)|
|Dry skin||23 (20.2)||30 (25.4)||31 (28.2)||26 (22.4)|
|Chills||16 (14.0)||28 (23.7)||31 (28.2)||33 (28.4)|
|Anorexia||23 (20.2)||29 (24.6)||29 (26.4)||26 (22.4)|
|Asthenia||28 (24.6)||28 (23.7)||22 (20.0)||26 (22.4)|
|Mood altered||22 (19.3)||26 (22.0)||24 (21.8)||26 (22.4)|
|Diarrhea||27 (23.7)||33 (28.0)||13 (11.8)||22 (19.0)|
|Back pain||16 (14.0)||24 (20.3)||23 (20.9)||19 (16.4)|
|Depression||26 (22.8)||16 (13.6)||19 (17.3)||17 (14.7)|
|Dyspnea||11 (9.6)||18 (15.3)||25 (22.7)||15 (12.9)|
|Dry mouth||16 (14.0)||23 (19.5)||15 (13.6)||14 (12.1)|
|PEG-IFNα-2a 180 μg qwk (n = 114) (%)||Alb-IFN 900 μg q2wk (n = 118) (%)||Alb-IFN 1,200 μg q2wk (n = 110) (%)||Alb-IFN 1,200 μg q4wk (n = 116) (%)||P Value|
|Completed treatment per protocol, n||99 (86.8)||96 (81.4)||78 (70.9)||96 (82.8)||0.02|
|Discontinued due to AE||7 (6.1)||11 (9.3)||20 (18.2)||14 (12.1)||0.04|
|≥1 severe AE||34 (29.8)||38 (32.2)||44 (40.0)||35 (30.2)||0.34|
|≥1 serious AE||7 (6.1)||14 (11.9)||17 (15.5)||14 (12.1)||0.15|
|IFN dose reduction||33 (28.9)||37 (31.4)||38 (34.5)||12 (10.3)||<0.001|
|IFN dose reduction due to AE||10 (8.8)||4 (3.4)||8 (7.3)||4 (3.4)||0.19|
|IFN dose reduction due to laboratory abnormality||27 (23.7)||32 (27.1)||31 (28.2)||7 (6.0)||<0.001|
|≤750/mm3||23 (20.2)||28 (23.7)||25 (22.7)||12 (10.3)||0.03|
|≤500/mm3||2 (1.8)||4 (3.4)||5 (4.5)||1 (0.9)||0.29|
|Platelets ≤50,000/mm3||4 (3.5)||2 (1.7)||3 (2.7)||1 (0.9)||0.50|
|Hemoglobin <10 g/dL||22 (19.3)||25 (21.2)||31 (28.2)||14 (12.1)||0.02|
|Treatment-emergent antibodies||24/114 (21.1)||1/117 (0.9)||4/108 (3.7)||4/115 (3.5)||<0.001|
Hematologic reductions were generally comparable across the alb-IFN q2wk and PEG-IFNα-2a groups, with lower rates observed in the alb-IFN q4wk group (Table 5). Nadirs in ANC or platelet count generally occurred by weeks 4-8 and remained stable during treatment, returning to baseline levels by 12 weeks after completion of therapy in all groups. Changes in ANC did not appear to be associated with severe infections, and no patient experienced an ANC <250/mm3. The alb-IFN q4wk group had significantly fewer IFN and RBV dose reductions due to laboratory abnormalities than did the other three groups. No patient discontinued treatment due to a reduction in ANC or platelet count, whereas one in the PEG-IFNα-2a group failed to complete treatment due to anemia. The ANC of the patient in the PEG-IFNα-2a group who died during the study was normal (>2,000 mm3).
The rate of immunogenicity was 2.6% (nine of 340) in the alb-IFN groups compared with 21.1% (24/114) in the PEG-IFNα-2a group (Table 5). In general, no correlation between immunogenicity and safety parameters (including AEs and laboratory changes) or antiviral response was observed.
SF-36 and Missed Workdays.
Changes in HRQOL as assessed by SF-36 during the study period are shown in Fig. 3. The week-48 time point is less relevant, since the variable discontinuation rates across the four treatment groups and the last-observation-carried-forward analysis may tend to bias the conclusions. SF-36 data were available for 94% of patients through week 12, for 89% through week 24, and for 78% through week 48. Overall, alb-IFN had less impact on both the physical and mental domains of SF-36 during treatment than did PEG-IFNα-2a. Reductions in SF-36 scores occurred by weeks 4-12, which is consistent with the occurrence of most treatment-associated AEs. At week 12, there was significantly less worsening in the aggregated physical (P = 0.04) and mental (P = 0.02) component scores, as well as in the role-physical (P = 0.02), bodily pain (P = 0.003), vitality (P = 0.02), social-functioning (P <0.001), and mental-health (P = 0.003) domains, of the SF-36 with alb-IFN 900 μg q2wk than with PEG-IFNα-2a.
Overall, disability-day assessment was significantly more favorable for the alb-IFN 900-μg q2wk arm than for the PEG-IFNα-2a arm (Table 6). The mean number of missed workdays was significantly lower with alb-IFN 900 μg than with PEG-IFNα-2a in the initial 4 weeks of treatment (1.4 versus 4.0 days; P = 0.005) and at week 12 (1.1 versus 4.3; P = 0.006). Likewise, the proportions of patients missing ≥7 workdays were significantly lower with alb-IFN 900 μg versus PEG-IFNα-2a at weeks 4 (4.5% versus 21.9%; P = 0.002), 12 (4.2% versus 18.1%; P = 0.006), and 24 (5.3% versus 20.3%; P = 0.005). The disability-day rates for the alb-IFN 1,200-μg q2wk and q4wk groups were slightly better than those for the PEG-IFNα-2a group.
|Disability Days (n = no. working)||Week 4||Week 12||Week 24|
|PEG-IFNα-2a 180 μg qwk (n = 64)||Alb-IFN||PEG-IFNα-2a 180 μg qwk(n = 72)||Alb-IFN||PEG-IFNα-2a 180 μg qwk (n = 69)||Alb-IFN|
|900 μg q2wk (n = 66)||1,200 μg q2wk (n = 61)||1,200 μg q4wk (n = 62)||900 μg q2wk (n = 72)||1,200 μg q2wk (n = 62)||1,200 μg q4wk (n = 72)||900 μg q2wk (n = 76)||1,200 μg q2wk (n = 62)||1,200 μg q4wk (n = 66)|
|Mean ± SE||4.0 ± 0.91||1.4 ± 0.58*||3.3 ± 0.93||2.6 ± 0.69||4.3 ± 1.00||1.1 ± 0.45*||2.8 ± 0.78||3.1 ± 0.88||4.7 ± 1.10||1.6 ± 0.52||3.8 ± 0.98||3.3 ± 0.92|
|Missed ≥3 day (%)||21 (32.8)||5 (7.6)*||17 (27.9)||15 (24.2)||22 (30.6)||7 (9.7)*||15 (24.2)||17 (23.6)||20 (29.0)||10 (13.2)*||15 (24.2)||15 (22.7)|
|Missed ≥7 day (%)||14 (21.9)||3 (4.5)*||8 (13.1)||7 (11.3)||13 (18.1)||3 (4.2)*||9 (14.5)||8 (11.1)||14 (20.3)||4 (5.3)*||11 (17.7)||9 (13.6)|
|Missed ≥14 day (%)||7 (10.9)||3 (4.5)||7 (11.5)||4 (6.5)||8 (11.1)||1 (1.4)*||5 (8.1)||7 (9.7)||10 (14.5)||2 (2.6)*||9 (14.5)||6 (9.1)|
The pharmacologic rationale for the development of alb-IFN was to optimize drug exposure to maximize sustained viral suppression, with the enhanced convenience of an improved 2-week to 4-week dosing schedule. This study represents the first evaluation of alb-IFN in combination with RBV in IFN treatment-naïve patients with genotype 1 CHC.
The SVR rates with alb-IFN administered every 2 weeks were comparable to that observed with PEG-IFNα-2a administered weekly, while the alb-IFN 1,200-μg q4wk regimen resulted in a lower SVR rate (50.9%; P = 0.28). Given the wide CIs (∼45%-∼67%) around the observed SVR rates, caution should be exercised in the interpretation of results from this moderately sized phase-2 study. It is, however, reassuring that compared with historical results (SVR rates of 46%-51% in genotype 1 CHC),4, 9 the PEG-IFNα-2a control arm in this study performed at least as well (SVR rate of 57.9%), and thus provides confidence in the ability of the results to guide further development of alb-IFN. This study also highlighted the similarities of alb-IFN with PEG-IFNα-2a in the context of factors that are predictive of SVR. Consistent with the literature,16–19 the positive predictive value for SVR of RVR4 and the negative predictive value of lack of EVR12 or viral clearance at week 24 were similar for alb-IFN and PEG-IFNα-2a in this study. Although the proportion of patients achieving RVR4 was lower in the alb-IFN q4wk arm, patients in that arm who achieved RVR4 had a 100% positive predictive value for SVR. Furthermore, the low viral breakthrough rate and the favorable hematologic toxicity profile associated with the alb-IFN 1,200-μg q4wk regimen provide the rationale for further dose finding with alb-IFN dosed every 4 weeks. The absence of a higher SVR rate in the alb-IFN 1,200-μg q2wk group in this study, despite the improved early antiviral response profile at weeks 4 and 12, was in part driven by higher discontinuation rates. Given the comparable SVR rates across both alb-IFN q2wk groups and the PEG-IFNα-2a group, both the alb-IFN 900-μg and alb-IFN 1,200-μg q2wk regimens warrant further evaluation in definitive phase 3 trials.
Overall, alb-IFN demonstrated an acceptable safety and tolerability profile. This study showed that the types of AEs observed with alb-IFN therapy were consistent with those expected during standard IFN or PEG-IFNα therapy. Some differences in AE rates, for example, cough and dyspnea, were observed in the alb-IFN 1,200-μg q2wk group compared with the lower-dose alb-IFN arms, as well as with PEG-IFNα. The majority of these respiratory events resolved following completion of treatment and none led to discontinuations. The AEs of cough and dyspnea have been commonly reported in previous PEG-IFN studies,4, 6 although the mechanisms and relationship to IFN or RBV dose remain unclear. Hematologic reductions were significantly lower in the alb-IFN q4wk treatment group, and were comparable across the alb-IFN q2wk and PEG-IFNα-2a groups, with rapid recovery of hematologic parameters after cessation of treatment. It is also worth noting the low rates of immunogenicity seen with alb-IFN, which provide evidence that the genetic fusion to albumin with a resultant novel polypeptide molecule may be associated with reduced rates of immunogenicity.
This study also showed that HRQOL assessments were more favorable for the alb-IFN 900-μg q2wk arm in terms of both SF-36 scores and missed workdays. Although the potential for bias in open-label studies of a novel agent with a less-frequent dosing schedule should be considered, the SF-36 results were consistent with previous studies comparing both PEG-IFNs with standard IFNα.3, 5, 7, 8, 20 Whereas patients experiencing AEs consistently reported worse HRQOL, there remained a range of worsening HRQOL observed in patients experiencing (or not experiencing) specific AEs. This resulted in more favorable SF-36 scores and fewer missed workdays for the alb-IFN 900-μg q2wk arm, although the rates of AEs and discontinuations due to AEs were comparable between the alb-IFN 900-μg q2wk and PEG-IFNα-2a groups. Likewise, while the rates of AEs and discontinuations due to AEs were higher with alb-IFN 1200-μg q2wk, HRQOL assessments in these patients were comparable to those receiving PEG-IFNα-2a.
In summary, the alb-IFN 900-μg q2wk regimen demonstrated an antiviral response and safety profile comparable to that of PEG-IFNα-2a, but with less impaired HRQOL, whereas the alb-IFN 1,200-μg q2wk regimen resulted in the greatest early antiviral response rate. The efficacy of the alb-IFN 900-μg q2wk regimen provides support for total dose reduction in the higher dose 1,200-μg q2wk regimen to enable optimal dose titration to individual patient tolerability in the phase-3 program. This study supports evaluation of the alb-IFN 900-μg and alb-IFN 1,200-μg q2wk regimens in phase-3 trials currently being conducted in patients with genotypes 1 and 2/3 CHC.
We thank Jessie Wolfe for her contribution to the authorship of the clinical report for this study and for editorial assistance on the manuscript, and Bob Abelson for statistical programming. BioScience Communications, New York, NY, provided editorial support funded by Human Genome Sciences and Novartis Pharma AG. Principal investigators (and locations) participating in the study were as follows: Australia (AU): P. Angus, W. Cheng, G. Cooksley, D. Crawford, P. Desmond, J. George, H. Harley, G. Jeffrey, I. Kronborg, A. Lee, L. Mollison, S. Pianko, J. Sasadeusz, W. Sievert, S. Strasser, J. Watson; Canada (CA): R. Bailey, V. Bain, J. Heathcote, K. Kaita, P. Marotta, M. Sherman, M. Swain, L. Worobetz, E. Yoshida; Czech Republic (CZ): P. Dlouhy, J. Galský, P. Husa, P. Kumpel, S. Plisek, V. Rehak, P. Urbanek; France (FR): K. Barange, Y. Benhamou, M. Bourliere, J. Bronowicki, P. Cales, X. Causse, P. Couzigou, P. Marcellin, P. Mathurin, J. Pawlotsky, S. Pol, R. Poupon, A. Tran, C. Trepo, J. Zarski; Germany (DE): T. Berg, P. Buggisch, J. Encke, P. Galle, G. Gerken, M. Gregor, D. Häussinger, M. Manns, J. Rasenack, W. Schmiegel, G. Teuber, J. Wiegand, S. Zeuzem; Israel (IL): R. Tur-Kaspa, Y. Lurie, R. Safadi, D. Shouval, E. Veitsman, E. Zuckerman; Poland (PO): M. Beniowski, A. Boron-Kaczmarska, J. Cianciara, R. Flisiak, A. Gladysz, W. Halota, A. Horban, M. Jablkowski, J. Juszczyk, W. Kryczka, T. Mach; Romania (RO): P. Calistru, M. Grigorescu, M. Manuc, C. Stanciu, C. Tanasescu.
- 14The SF-36 Health Survey (version 2.0) Technical Note. Boston, MA: Health Assessment Lab, September 20, 1996 (updates September 27, 1997)., .
- 15SF-36 Health Survey manual and interpretation guide. Boston, MA: New England Medical Center, The Health Institute, 1993., , , .