Impact of weight-based ribavirin with peginterferon alfa-2b in african americans with hepatitis C virus genotype 1§

Authors


  • Trial registration number NCT00299936, available at: http://clinicaltrials.gov

  • Potential conflict of interest: Ira M. Jacobson, Schering-Plough consultant, speaker bureau, research support; Merck consultant, GlobeImmune consultant, research support, Human Genome Sciences consultant, research support; Coley consultant, research support; Gilead consultant, speaker bureau, research support; Vertex consultant, research support; Intermune consultant, research support; Intarcia consultant; Valeant consultant, research support; GlaxoSmithKline consultant; Idenix consultant, speaker bureau, research support; Novartis consultant, speaker bureau, research support; Bristol-Myers Squibb consultant, speaker bureau; Boehringer Ingelheim consultant; XTL consultant. Robert S. Brown, Schering-Plough consultant, research support, honoraria; Jonathan McCone, Schering-Plough speaker bureau, grant support; Roche speaker bureau; Martin Black, Schering-Plough speaker bureau, Roche speaker bureau, Valeant speaker bureau, Bristol-Myers Squibb speaker bureau, Gilead speaker bureau, GlaxoSmithKline speaker bureau, Axcan speaker bureau; Clive Albert, Wyeth speaker bureau, AstraZeneca speaker bureau; Paul Y. Kwo, Schering-Plough research support; Greg W. Galler, Schering-Plough consultant, speakers bureau, advisor, grant support; Roche consultant, speakers bureau, advisor, grant support; Victor Araya, Schering-Plough speakers bureau, grant support, Roche speakers bureau, grant support, Gilead speakers bureau; Bradley Freilich, Schering-Plough speaker bureau, research support; Joann Harvey, Schering-Plough consultant; Louis H. Griffel, Schering-Plough employee; Clifford A. Brass, Schering-Plough employee.

  • §

    See Editorial on Page 953

Abstract

WIN-R (Weight-based dosing of pegINterferon alfa-2b and Ribavirin) was a multicenter, randomized, open-label, investigator-initiated trial involving 236 community and academic sites in the United States, comparing response to pegylated interferon (PEG-IFN) alfa-2b plus a flat or weight-based dose of ribavirin (RBV) in treatment-naive patients with chronic hepatitis C and compensated liver disease. Patients were randomized to receive PEG-IFN alfa-2b at 1.5 μg/kg/week plus flat-dose (800 mg/day) or weight-based-dose RBV (800 mg/day for weight <65 kg, 1000 mg/day for 65-85 kg, 1200 mg/day for >85-105 kg, or 1400 mg/day for >105-<125 kg). Sustained virologic response (SVR; undetectable [<125 IU/mL] hepatitis C virus [HCV] RNA at end of follow-up) in patients ≥65 kg was the primary end point. Low SVR rates have been reported among African American individuals, in whom there is a preponderance of HCV genotype 1. This subanalysis of WIN-R was conducted to evaluate the efficacy of weight-based dosing among African American individuals with genotype 1 infection enrolled in the trial. Of 362 African American patients in the primary efficacy analysis, 188 received RBV flat dosing and 174 received weight-based dosing. SVR rates were higher (21% versus 10%; P = 0.0006) and relapse rates were lower (22% versus 30%) in the weight-based-dose group than in the flat-dose group. Safety and rates of drug discontinuation were similar between the 2 groups. Conclusion: Weight-based dosing of RBV is more effective than flat dosing in combination with PEG-IFN alfa-2b in African American individuals with HCV genotype 1. Even with weight-based dosing, response rates in African American individuals are lower than reported in other ethnic groups. (HEPATOLOGY 2007.)

Combination therapy with pegylated interferon (PEG-IFN) alfa-2b (PegIntron; Schering Corp., Kenilworth, NJ) and ribavirin (RBV) has been available for the treatment of chronic hepatitis C since 2001. The superior efficacy of this regimen over standard interferon (IFN) alfa-2b and RBV was established by Manns et al. in an international multicenter study that compared administration of IFN alfa-2b at 3 million units (MU) 3 times weekly, PEG-IFN alfa-2b at 0.5 μg/kg/week (after an initial 4 weeks of dosing with 1.5 μg/kg/week), and PEG-IFN alfa-2b at 1.5 μg/kg/week, all in combination with RBV.1 In the first 2 dosing groups, RBV was administered at a dose of 1000 or 1200 mg/day based on body weight < 75 kg or ≥ 75 kg. RBV in a dose of 800 mg was given to all patients receiving the higher dose of PEG-IFN because of the perceived potential for hematological toxicity. Despite this lower, flat dose (FD) of RBV, the higher dose of PEG-IFN was associated with a greater sustained virologic response (SVR) rate (54%) than was the lower PEG-IFN dose (47%) or IFN (47%; P < 0.01).

The large volume of distribution of RBV2 and its narrow therapeutic index suggest that dosing for RBV should be adjusted by patient body weight. Logistic regression analysis of the pivotal study by Manns et al. showed a direct linear relationship between RBV dose, measured in milligrams per kilogram of body weight, and SVR.1 These observations warranted a prospective study to confirm the greater therapeutic efficacy of a weight-based dose (WBD) of RBV over the FD RBV of 800 mg licensed in the United States.

The WIN-R (Weight-based dosing of pegINterferon alfa-2b and Ribavirin) study was a large, multicenter, randomized, prospective, open-label study conducted in the United States between December 2000 and June 2005.3 Patients with chronic hepatitis C received PEG-IFN alfa-2b in combination with FD or WBD RBV. We hypothesized that SVR would be greater in the group receiving weight-based dosing and that response and tolerability in that group would be similar across all body weights, given that patients of different weights would be receiving equivalent doses of RBV per kilogram of body weight. The results of this study showed that overall SVR rates were significantly higher for patients receiving PEG-IFN alfa-2b and WBD RBV than for those receiving FD RBV (P = 0.008). This was also true for the difficult-to-treat genotype 1 (G1) patients (P = 0.005). Besides greater hemoglobin reductions with WBD RBV, safety profiles were similar across RBV dosing groups, including the group receiving 1400 mg/day.3

The prevalence of hepatitis C virus (HCV) infection is higher in African American individuals than in other ethnic groups in the United States,4 and up to 90% of infected African American individuals have HCV G1a or G1b (both of which are associated with lower response to therapy), compared with approximately 70% of non–African American persons.5–8 Although the African American population has been underrepresented in pivotal trials of IFN-based therapies,1, 4, 9–11 it is clear that this population has an impaired response to HCV therapy.5, 6, 12–18 Given the disproportionate prevalence of this disease in African American individuals and the lower rates of response to IFN-based therapy, it is important to evaluate ways of maximizing treatment response to HCV therapy. Tailoring therapy according to body weight may accomplish this, particularly in light of the finding that the prevalence of obesity is higher in African American persons, particularly women, than in other ethnic groups.19, 20

The WIN-R study enrolled more than 400 African American patients, constituting the largest database available on treatment of African American individuals with HCV infection. Because of the importance of acquiring and analyzing additional information in this population, we present here the results of the WIN-R study in African American individuals with HCV G1, which is overwhelmingly the most common HCV genotype in African American individuals infected with HCV.5–8

Abbreviations

ANC, absolute neutrophil count; FD, flat dose; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IFN, interferon; PCR, polymerase chain reaction; PEG-IFN, pegylated interferon; RBV, ribavirin; SVR, sustained virologic response; WBD, weight-based dose; WIN-R, Weight-Based Dosing of Peginterferon Alfa-2b and Ribavirin trial.

Patients and Methods

The WIN-R study was a multicenter, randomized, open-label, investigator-initiated study conducted at 236 practices at community and academic sites in the United States between December 2000 and June 2005. The study was carried out in accordance with Good Clinical Practices and the Declaration of Helsinki. All patients provided written, informed consent, and the study protocol was approved by the local or central institutional review boards for each site.

Patient Selection.

Treatment-naive adult patients were eligible to enroll in the study if HCV RNA was detectable by polymerase chain reaction (PCR) or branched DNA assay and the patient had compensated liver disease. Patients were also required to have undergone liver biopsy, with findings consistent with HCV infection, as determined by a local pathologist, within 36 months prior to screening, and to have had elevated alanine aminotransferase levels at least once during the 6 months prior to screening.

Exclusion criteria included positivity for serum hepatitis B surface antigen or human immunodeficiency virus (HIV) infection, decompensated liver disease, neutrophil count less than 1500 cells/mm3, platelet count less than 70,000 cells/mm3, hemoglobin levels less than 12 g/dL for females and less than 13 g/dL for males, poorly controlled psychiatric disease, pregnancy or lactation, history of substance abuse (such as alcohol, intravenous drugs, and inhaled drugs) within 6 months of study entry, history of significant retinal abnormalities, or preexisting or coexisting medical conditions that could interfere with the subject's participation in, and completion of, the study.

Study Design and Conduct.

Patients participating in the WIN-R study were randomized to 1 of 2 treatment arms: once-weekly PEG-IFN alfa-2b at 1.5 μg/kg via subcutaneous injection at approximately the same time each week plus RBV (Rebetol; Schering Corp.) at a FD of 800 mg/day or RBV at 800 to 1400 mg/day (weight-based dosing) administered as 800 mg for patients with a body weight <65 kg; 1000 mg for body weight 65 kg to 85 kg; 1200 mg for body weight >85 kg to 105 kg, and 1400 mg for body weight >105 kg to <125 kg. RBV was administered orally, twice daily with food. Randomization was stratified by genotype (2/3 versus non-2/3) and degree of fibrosis (METAVIR score F3-F4 versus F0-F2), as interpreted by a local pathologist. Central randomization (by computer) was implemented using a block size of 10; sequence was concealed until treatment was allocated. Randomization was generated by a central statistician and assigned by a third-party database manager (Therapeias Health Management Inc., Claremont, CA). Treatment duration for patients with G1 was 48 weeks. For patients with G2 or G3, treatment was randomized to 24 versus 48 weeks; these patients are not included in this analysis.

Assessment and End Points.

Quantitative testing of serum viral load was performed at baseline and at weeks 24, 48, and 72. Testing for HCV RNA was performed at a central laboratory (Schering-Plough Research Institute [SPRI], Union, NJ) using quantitative real-time PCR with a lower limit of detection of 125 IU/mL. With this assay, 2 million copies/mL is equivalent to 600,000 IU/mL. The HCV reverse transcriptase PCR (RT-PCR) assay used was developed by the central laboratory and is based on a published method.21 In cases in which end-of-therapy or posttreatment serum samples were not sent to the central laboratory, PCR was accepted from local laboratories. A subset of the samples tested at the SPRI laboratory was sent for confirmatory testing to Quest Nichols Laboratory (San Juan Capistrano, CA). In the WIN-R trial, 65% of the population had available 24-week follow-up PCR data: 85% was obtained from the central laboratory and 15% was obtained from local laboratories. Similarly, for the African American patients in this study who were infected with HCV G1, 54.6% had available 24-week follow-up PCR data: 83% was performed at the central laboratory and 17% was performed at local laboratories. Other laboratory tests, including HCV genotyping, were performed at local laboratories.

Physical and liver examinations were performed at weeks 4, 12, 24, 36, 48, 60, and 72. At every treatment and follow-up visit, vital signs and adverse events were recorded, and pregnancy tests were performed on females with reproductive potential. Hematological laboratory parameters (hemoglobin, hematocrit, red blood cells, white blood cells, and platelet count) were measured at each treatment and follow-up visit, and complete blood chemistry (including measurement of alanine aminotransferase, aspartate aminotransferase, and creatinine) was performed at weeks 2, 4, 8, 12, 24, 36, 48, and 72.

Discontinuation criteria included serious or life-threatening adverse events, pregnancy, and failure to comply with dosing or other requirements. Patients were free to withdraw from the study at any time. A case report form that focused on key safety parameters and known adverse events relating to use of PEG-IFN and RBV was filled out at each site. Adverse events were graded as mild, moderate, severe, or life-threatening according to the World Health Organization grading system and were assessed regarding relationship with the study drug.

Management of non–life-threatening adverse events was achieved by dose reduction of PEG-IFN alfa-2b or RBV, or both. In the case of neutropenia, PEG-IFN alfa-2b given at 1.5 μg/kg/week was reduced to 1.0 μg/kg/week in patients who had an absolute neutrophil count (ANC) <750 cells/mm3, and PEG-IFN alfa-2b was discontinued in patients with an ANC <400 cells/mm3. In conjunction with a dose reduction of PEG-IFN alfa-2b, use of filgrastim was permitted at the discretion of the investigator for patients with an ANC <750 cells/mm3. Resumption of the initial PEG-IFN alfa-2b dose was permitted for patients in whom ANC increased to >1250 cells/mm3. If hemoglobin levels decreased to <10 g/dL, RBV dose was reduced by 200 or 400 mg in patients receiving doses of 800 to 1200 mg/day or 1400 mg/day, respectively. In patients in whom hemoglobin concentration did not increase to ≥10 g/dL after the first dose reduction, a second 200-mg dose reduction (to 600 mg or 800 mg) was performed, except in initial recipients of 800 mg/day in whom treatment was discontinued if hemoglobin remained <10 g/dL with RBV given at 600 mg/day. Patients with hemoglobin concentrations <8.5 g/dL were required to discontinue therapy. Erythropoietin could be used at the discretion of the investigator, along with concomitant dose reduction of RBV, only if hemoglobin levels dropped to ≤10 g/dL. Resumption of initial RBV dosing was permitted for patients with hemoglobin levels ≥11.5 g/dL.

Statistical Analysis.

The primary efficacy analysis as defined in the original protocol was a comparison of SVR rates among all patients with a body weight ≥65 kg because patients weighing <65 kg received RBV at 800 mg in both treatment arms and were included to fulfill the intent of the study to serve as an access trial. End-of-treatment response was defined as the absence of detectable serum HCV RNA at 48 weeks, and SVR was defined as the absence of detectable serum HCV RNA 24 weeks after treatment completion. Relapsers were those individuals having undetectable HCV RNA at the end of treatment and reappearance of HCV RNA during the 24 weeks after treatment. All patients who received at least 1 dose of the study medication were included in the efficacy analysis. Patients for whom there were no follow-up data were considered nonresponders. The main study, including all patients, had an 80% power to detect a 5.5% difference in SVR rates. The analysis of African American patients was a secondary evaluation and was not powered for statistical analysis.

The effect of treatment on SVR (and on end-of-treatment response) was tested using logistic regression, controlling for pretreatment fibrosis, which was 1 stratification factor. The impact of body weight on treatment with respect to its effect on SVR, including patients of all body weights, was also tested using logistic regression (controlling for treatment, pretreatment fibrosis, and body weight). In addition, 2 other tests were performed to confirm the results of the logistic regression. A 2-way comparison between treatment and body weight was conducted using the Cochran-Mantel-Haenszel test. Also, a Breslow-Day test of homogeneity of odds ratios was performed. Tests of treatment effects within subgroups were performed using a chi-squared test. SAS version 8.2 (SAS Institute Inc., Cary, NC) was used for all statistical analyses.

Results

Of the 4913 patients enrolled in the WIN-R study who received at least 1 dose of study drug, 387 were African American individuals infected with HCV G1; 227 completed therapy (Fig. 1). Of these 387 individuals, 362 weighed ≥65 kg and were included in the primary efficacy analysis: 188 received FD RBV and 174 received WBD RBV. Baseline demographics are shown in Table 1 for all 387 African American patients with HCV G1 infection.

Figure 1.

Patient flow.

Table 1. Baseline Demographics of African American Patients with HCV Genotype 1 Infection
CharacteristicPEG-IFN alfa-2b + FD RBV (n = 202)PEG-IFN alfa-2b + WBD RBV (n = 185)
  1. Abbreviations: BMI, body mass index; FD, flat dose; HCV, hepatitis C virus; PEG-IFN, pegylated interferon; RBV, ribavirin; WBD, weight-based dose.

Male (%)6464
Mean age (years)47.848.5
Mean weight (kg)87.889.6
Mean BMI (kg/m2)29.529.9
HCV viral load >600,000 IU/mL (%)5963
METAVIR stage F3-F4 (%)3031

Among patients weighing ≥65 kg, SVR was achieved in 21% of WBD recipients and in 10% of FD recipients (P = 0.006; Table 2). If patients weighing <65 kg were also included in the efficacy analysis, despite the identical RBV doses in both treatment arms, similar results were obtained: an SVR of 19% (36 of 185) in the WBD group and of 11% (23 of 201) in the FD group (P = 0.03). End-of-treatment response rates were also significantly higher in the WBD group than in the FD group (29% versus 17% of patients, respectively; P = 0.008). Of the patients in each group who were confirmed to have undetectable HCV RNA by PCR assay at end of treatment and for whom 24-week follow-up PCR analysis was available, relapse rates were 30% and 22% in the FD and WBD groups, respectively.

Table 2. Virologic Response to Therapy in African American Patients with HCV Genotype 1 Infection and Body Weight ≥65 kg (Efficacy Analysis)
CharacteristicPEG-IFN alfa-2b + FD RBVPEG-IFN alfa-2b + WBD RBVP Value*
  • Abbreviations: EOT, end-of-therapy response; FD, flat dose; HCV, hepatitis C virus; PEG-IFN, peginterferon; RBV, ribavirin; SVR, sustained virologic response; WBD, weight-based dose.

  • *

    P values for treatment comparison of SVR and EOT response calculated using logistic regression; those for treatment comparison of SVR by subgroup were calculated using the chi-squared test.

  • Analysis performed in patients in each group who were confirmed to be negative for HCV RNA at end of treatment and in whom 24-week follow-up HCV RNA levels were available.

SVR, % (n)10 (19/188)21 (36/174)0.006
 ≤600,000 IU/mL18 (10/56) P = 0.03 vs >600,000 IU/mL33 (16/48) P = 0.02 vs >600,000 IU/mL0.069
 >600,000 IU/mL7 (8/112)16 (18/111)0.035
 METAVIR stage F0-F29 (12/132)21 (25/120)0.009
 METAVIR stage F3-F413 (7/56)20 (11/54)0.265
 Male10 (13/126)23 (27/117)0.007
 Female10 (6/62)16 (9/57)0.316
EOT response, % (n)17 (32/188)29 (50/174)0.008
Relapse, % (n)30 (8/27)22 (10/46)

Univariate logistic regression analyses and chi-squared tests were performed to evaluate patient (for example, age, sex, weight, fibrosis stage), treatment (for example, treatment duration, RBV dose), and viral (for example, baseline viral load) factors influencing SVR rates within and between RBV dosing strategies. There were no statistically significant differences in SVR rates between males and females in either treatment group, and fibrosis had no effect on response in patients in either treatment group (Table 2). However, there were significant differences between FD and WBD RBV groups for males (P = 0.007) and those with METAVIR stage F0-F2 (P = 0.009). In contrast, baseline viral load affected SVR in both treatment groups. Of patients in the FD group, 7% of those with high baseline viral loads (>600,000 IU/mL) and 18% of those with low baseline viral loads (≤600,000 IU/mL) achieved SVR (P = 0.03; Table 2); and, in WBD recipients, the corresponding figures were 16% and 33%, respectively (P = 0.02). When comparing the 2 RBV dosing regimens, significantly more patients in the WBD group attained SVR than those in the FD group for those with high baseline viral load (>600,000 IU/mL, 16% versus 7%; P = 0.035). Although there was an even larger numerical difference for the subset of patients with low baseline viral load (≤600,000 IU/mL, 33% versus 18%; P = 0.069), this did not reach statistical significance.

When the impact of body weight on SVR in the 2 treatment arms was assessed, results indicated that there was a negative effect on SVR rates of increasing body weight among those in the FD arm: as body weight increased, SVR rates decreased. In contrast, there was a positive effect on SVR rates of increasing body weight among those in the WBD arm: SVR rates increased as body weight increased (Fig. 2). This observation was also investigated using 2 more conventional tests. A Cochran-Mantel-Haenszel test, which assesses treatment effect after adjusting for body weight (categorized into 4 groups as in Fig. 2), showed that the SVR rates were higher among patients in the WBD arm than in those in the FD arm (P = 0.02). A Breslow-Day test showed that the magnitude of the treatment effects (that is, the differences in SVR rates between the WBD and the FD arm) were different among the 4 weight categories (P = 0.01). Logistic regression analysis, controlling for treatment, pretreatment fibrosis (a stratification factor), and body weight (as a continuous variable) were also used to determine whether there was a significantly different effect of body weight on SVR in each of the 2 treatment arms. Results from this analysis showed that there was a statistically significant difference between the effect of body weight on SVR in the FD arm and the WBD arm (P = 0.01).

Figure 2.

Sustained virologic response (SVR) rate according to body weight. (All African American patients with genotype 1.) PEG-IFN, peginterferon.

In a multivariate regression analysis, RBV dosing (WBD versus FD) had a significant impact on SVR, even after adjusting for baseline viral load (P = 0.0002), duration of treatment (P < 0.0001), and RBV dose (P < 0.0001), which were shown to be prognostic factors in the univariate analysis.

Overall, the incidence and type of adverse events reported in this trial were consistent with those generally associated with PEG-IFN alfa and RBV treatment. Mean hemoglobin levels during therapy were slightly higher in the FD group than in the WBD group (Fig. 3A). However, there were no differences in mean hemoglobin concentrations among the 4 WBD arms receiving between 800 and 1400 mg RBV per day (Fig. 3B). Decreases in hemoglobin levels to <10 g/dL occurred in 15% of the FD and in 20% of the WBD recipients (Table 3). Dose reductions occurred at the following rates for the WBD and FD arms, respectively: overall, 32% and 24%; PEG-IFN alfa-2b, 7% and 8%; RBV, 11% and 8%; and PEG-IFN alfa-2b plus RBV, 14% and 9%. Reported use of erythropoietin was also greater in the WBD group (12%) than in the FD group (2%). In contrast, nadir ANC and platelet counts were similar between the 2 RBV dosing groups (Table 3).

Figure 3.

Mean hemoglobin levels over time in the (A) ribavirin flat-dose and weight-based-dose treatment groups and (B) 4 ribavirin dose arms in the weight-based dose treatment group. (All African American patients with genotype 1). PEG-IFN, peginterferon.

Table 3. Hematological Toxicities in African American Patients with HCV Genotype 1 Infection
CharacteristicPEG-IFN alfa-2b + FD RBV (n = 202)PEG-IFN alfa-2b + WBD RBV (n = 185)
  1. Abbreviations: ANC, absolute neutrophil count; FD, flat dose; PEG-IFN, pegylated interferon; RBV, ribavirin; WBD, weight-based dose.

Nadir hemoglobin (%)  
 <10 g/dL1520
 <8.5 g/dL14
RBV dose-reduction (%)2637
Nadir ANC (%)  
 <750 cells/mm32824
 <500 cells/mm358
Nadir platelets (%)  
 <100 × 103cells/mm31511
 <50 × 103cells/mm311

Overall study discontinuation rates for all reasons, including failure to respond to therapy, were 42% for the FD arm and 41% in the WBD arm. Discontinuation rates were similar for adverse events and serious adverse events in each arm. The combined discontinuation rate related to the categories of patient withdrew plus noncompliance was as great (18% FD, 18% WBD) as that of withdrawal for all adverse events (17% FD, 18% WBD).

Discussion

This prospective, community-based study establishes the superiority of weight-based dosing over flat dosing of RBV in previously untreated African American individuals with HCV G1 infection. Although lower compliance rates, high patient withdrawal rates, and missing data in this large community-based study likely contributed to a lower than expected SVR rate, the magnitude of the difference in SVR rates from the primarily Caucasian patients with G1 infection in the pivotal trial of Manns et al.1 (42% in the overall cohort receiving PEG-IFN alfa-2b at 1.5 μg/kg/week and RBV at 800 mg/day and 48% in patients receiving RBV >10.6 mg/kg) and the SVR rates in African American individuals in the present study are consistent with the finding of the reported lower viral response rates in African American individuals in smaller studies.5, 6, 12–18 Consistent with this, analysis of the WIN-R trial showed an SVR rate of 37% for Caucasian patients with HCV G1 infection receiving WBD RBV and 33% of those receiving FD RBV.

Enhanced efficacy of RBV administered at 1000 to 1200 mg versus RBV at 800 mg in G1 HCV infection was reported in a previous study of PEG-IFN alfa-2a at 180 μg/week in combination with RBV.22 In that report, an SVR rate of 52% was reported after 48 weeks of treatment with RBV at 1000 to 1200 mg, whereas SVR occurred in 41% of those receiving RBV at 800 mg for the same duration. However, the magnitude of the increment in efficacy with weight-based dosing was notably smaller in that study than in our observations of African American individuals, in whom a 2-fold increase in SVR occurred.

An unexpected finding of our study was the increase in efficacy with an increase in RBV dose in heavier patients. This was most evident in the 1400-mg dosing arm. In contrast, we observed the expected modest decrease in efficacy observed in heavier patients receiving the standard 800-mg dose. Our hypothesis had been that SVR rate would be equivalent across all body weights, provided patients received equivalent doses of RBV adjusted for weight. There are potential explanations for these observations: heavier patients may respond better to this therapy, for unknown reasons, and this observation could have been masked by previous underdosing of these patients with FD regimens of IFN with or without RBV. Alternatively, RBV distribution may be more complex than realized and body weight may only approximate the marker for size required to dose RBV consistently. Therefore, tissue exposure may increase with increased RBV dose to a degree disproportionate to the increase in body weight. However, the similar degrees of anemia in the 4 weight strata in the WBD RBV group speak against this alternative.

Although the effect of WBD RBV in this study was predicted, the magnitude of increase in efficacy with weight-based dosing of RBV (2-fold) in this population was not foreseen. However, this result is consistent with results of recent viral kinetic studies, which show that the maximum increment in efficacy conferred by RBV occurs in patients with the lowest IFN-induced phase 1 decrease in HCV RNA.23 This is the early phase of decline in HCV RNA that reflects inhibition of viral production. Prior viral kinetic studies indicated that African American individuals have particularly poor phase 1 responses to IFN-based therapy.24 Proposed antiviral and immunomodulatory mechanisms of action of RBV25 include a mutagenic effect, which leads to “error catastrophe”26 and the creation of noninfective particles. For groups in which IFN greatly reduces viral production, RBV may have less of an effect because viral particle production is already low. In groups such as the African American population, in which IFN is less effective in shutting down viral particle production, RBV may have a substantial effect if it renders the majority of produced particles noninfectious. The differential increment in efficacy conferred by RBV suggests that the dose used, in addition to the use of the drug itself, may be of particular importance in IFN-refractory subgroups. Thus, the 2-fold increase in SVR with WBD in African American individuals in our study is similar to, but quantitatively much greater than, differences in SVR rates in other HCV G1-infected patients in the WIN-R study.3

Limitations of the study include high patient dropout rates and missing patient data. The high patient dropout rate could reflect, in part, the participation of a large number of community sites in this study. Nearly half the discontinuations in both arms were classified by the local investigator as “patient withdrawal” or “noncompliance,” which might be attributable to a decrease in interest of the investigator or patient when this investigational therapy became licensed during the course of this study.1 In contrast, the rate of discontinuation due to adverse events was similar in this study (17%) to that reported in the pivotal trial of PEG-IFN alfa-2b and RBV (14%).1 Other limitations of the study include the absence of rigorous compliance monitoring more characteristic of pivotal registration trials. RBV dose reductions were reported on a yes-no basis, without details about when the dose reduction occurred or when normal dosing was resumed. Growth factor use was reported in this trial on an as-used basis, without details about the circumstances surrounding their use. Many of the case report forms querying investigators about growth factor use were not returned, limiting the scope of this analysis.

Whether the findings of this study can be extrapolated to other relatively treatment-resistant populations requires further investigation. However, it is tempting to speculate that there are groups of patients for whom strategies that facilitate maximum possible RBV doses may enhance response rates relative to those reported in studies that have used relatively low RBV doses. These would include not only African Americans, but also other subgroups, such as nonresponders or relapsers to previous HCV therapy, patients with HIV-HCV coinfection, patients who have undergone liver transplantation, and, potentially, the group that is the majority in the United States, Europe, and Japan: those with G1 and high viral load. This possibility is supported by a recent report on 10 patients with HCV G1 and high viral load who were given doses of RBV titrated to achieve serum levels of 15 mmol/L, a higher concentration than that obtained with usual therapeutic doses. This resulted in a mean RBV dosage of 2400 mg/day at 24 weeks. Although hemolytic anemia occurred in most patients and transfusions were required in 2, SVR occurred in 90%.27

The greater degree of anemia seen with weight-based dosing in this study is consistent with other data indicating that the hemolysis associated with RBV is a dose-related toxicity.22 However, the similarity of hemoglobin levels in the different weight groups in the WBD arm serves as a clinical affirmation of the scientific observation of the large volume of distribution of RBV and, thereby, suggests the appropriateness of WBD RBV from a safety as well as an efficacy perspective. Underscoring the latter point, we found that drug discontinuation owing to adverse events, as well as the incidence of serious side effects, was equivalent between the FD and WBD groups. We acknowledge, however, that our incomplete data regarding erythropoietin use creates the possibility that the use of this growth factor made continued RBV dosing feasible when it otherwise might not have been.

In conclusion, when combined with PEG-IFN alfa-2b, WBD RBV offers a significant advantage in efficacy, with acceptable tolerability, over FD of RBV in the treatment of African American patients infected with HCV G1. However, even with WBD RBV, the rate of SVR in African American individuals is low. Studies designed to explain this phenomenon have been conducted,18, 28, 29 but further studies are needed to elucidate the fundamental basis for the impaired responsiveness in this population.

Acknowledgements

We thank Chris Rozanitis and Peter Savino for programming assistance, Nicole Stauffer for assistance with data collection, Weiping Deng for statistical support; Maribeth Bogush, Ph.D., for editorial assistance in the preparation of this manuscript; and Alan Perelson, Ph.D., for his review of the manuscript.

Ancillary