Pegylated Interferon Alfa-2b vs Standard Interferon Alfa-2b, Plus Ribavirin, for Chronic Hepatitis C in HIV-Infected Patients:  A Randomized Controlled Trial FREE

About one third of human immunodeficiency virus (HIV)–infected patients in Europe and the United States are also infected by hepatitis C virus (HCV) while 5% to 10% of HCV-infected patients are also infected by HIV.^1- 2 Human immunodeficiency virus coinfection accelerates the progression of HCV infection, which is now a leading cause of morbidity and mortality among HIV-infected individuals.^3- 7

The treatment of chronic HCV infection was transformed in the 1990s by the advent of the interferon-ribavirin combination and was further improved with the use of pegylated interferon (peginterferon), in which a polyethylene glycol molecule is added to standard interferon, yielding a longer half-life and more favorable pharmacokinetics.^8- 9 Two forms of peginterferon (alfa-2a and alfa-2b) have been approved for use with or without ribavirin. In combination with ribavirin, both peginterferons provide sustained virologic responses in 54% to 63% of HIV-seronegative HCV-infected patients (7% to 12% higher than with standard interferon-ribavirin).^10- 12

In vitro antagonism between ribavirin and some antiretrovirals (especially zidovudine and stavudine) delayed the use of ribavirin in HIV-coinfected patients, but a recent randomized trial of the stavudine-ribavirin combination showed no negative impact on antiretroviral efficacy.¹³

Recently, 2 randomized controlled studies showed the efficacy and safety of peginterferon alfa-2a plus ribavirin in HIV-HCV coinfected patients.^14- 15 The few available data on the virologic efficacy and tolerability of peginterferon alfa-2b plus ribavirin in HIV-HCV coinfected patients come mainly from small uncontrolled or single-center trials.^16- 19 The aim of our prospective randomized study of initial treatment of chronic HCV in HIV-infected patients was to compare the efficacy of a 48-week course of ribavirin combined with either standard interferon alfa-2b or peginterferon alfa-2b.

Adults who had never received interferon and who had the following characteristics were eligible for the study: second-generation enzyme-linked immunosorbent assay positivity for anti-HCV antibodies and polymerase chain reaction–based assay positivity for HCV-RNA in serum; interpretable results of liver biopsy performed within the previous 18 months, showing at least mild activity or fibrosis; anti-HIV antibody positivity and a stable plasma HIV-1 RNA level (variation of less than 1 log₁₀ copies × 10⁶/L during the 3 months before randomization); stable antiretroviral treatment during the preceding 3 months (or no antiretroviral treatment); and a CD4 cell count higher than 200 × 10⁶/L. Patients were not eligible if they had neutropenia (<1.5 × 10⁹/L neutrophils); thrombocytopenia (<100 × 10³/μL platelets); anemia (<11.0 g/dL hemoglobin); a serum creatinine level higher than 1.70 mg/dL (150 μmol/L); circulating hepatitis B surface antigen positivity; decompensated cirrhosis (defined as biopsy-proved cirrhosis with serum albumin below the lower limit of normal: a prothrombin level <60%; a total bilirubin level higher than the upper limit of normal; or a history of ascites, hepatic encephalopathy or esophageal varices); biliary, tumoral, or vascular liver disease; psychiatric disorders (history of major depression, suicide attempts, suicidal ideation, or other severe psychiatric disorders; psychosis); a history of seizures; cardiovascular disease; poorly controlled diabetes mellitus; or autoimmune disorders; or if they had actively injected illicit drugs 3 months before enrollment or reported daily alcohol intake greater than 40 g (women) or 50 g (men). Women were not eligible if they were unwilling to use effective contraception.

This randomized, phase 3, open-label, parallel-group study was conducted in 71 French centers. The study was approved by the ethics committee of Saint-Germain en Laye hospital and by the sponsor's institutional review board (Agence Nationale de Recherches sur le SIDA [ANRS]). All the patients gave their written informed consent. The study was designed by the Groupe d'Etude et de Recherche en Médecine Interne et Maladies Infectieuses sur le Virus de l'Hépatite C (GERMIVIC) joint study group, which comprised experts in internal medicine, infectious diseases, and hepatology. Data analysis was performed by the sponsor and the authors, both of whom were independent of the drug manufacturers. The study followed the Helsinki Declaration and Good Clinical Practices.

Patients were randomly assigned to receive subcutaneous injections of 1.5 μg/kg peginterferon alfa-2b (Peg Intron, Schering-Plough, Kenilworth, NJ) once a week or subcutaneous injections of 3 million units of interferon alfa-2b (Intron A, Schering-Plough) 3 times a week for 48 weeks. All patients also received 400 mg of ribavirin twice a day (Rebetol, Schering-Plough), orally. Randomization was managed by the central data center (INSERM U444, Paris, France). Randomization was balanced within centers, with blocking within strata. The randomization code was developed using a computerized random number generator to select random permuted block sizes of 2, 4, 6, and 8. The randomization list was concealed from the medical monitor (located in the data center), who assigned participants to the treatment groups after reviewing the eligibility criteria. Allocated treatments were communicated to the investigator during the week preceding the visit at which the first treatment prescription was planned.

The patients were evaluated after 2 and 4 weeks of treatment, every 4 weeks thereafter during treatment, and 4, 12, and 24 weeks after treatment was completed. Patients were followed up until week 72 to assess sustained responses.

Biochemical and hematologic tests were performed in local laboratories. Hepatitis C virus–RNA tests, viral genotyping, and histological evaluation of biopsy specimens were performed in central laboratories. Liver biopsy was performed at the end of follow-up.

Assessment and End Points. Hepatitis C virus–RNA was detected with a polymerase chain reaction assay (Amplicor 2.0 HCV Monitor, Roche Diagnostics Systems, Basel, Switzerland) with a detection limit of 50 IU (100 copies) × 10³/L. Hepatitis C virus–RNA levels were measured with a branched-chain DNA assay (bDNA3.0, Bayer Diagnostics, Tarrytown, NY) with a detection limit of 615 IU (3200 copies) × 10³/L. Hepatitis C virus genotyping was performed by sequence analysis of the 5′ untranslated region. The primary end point for efficacy was a sustained virologic response, defined as undetectable serum HCV-RNA at week 72. The secondary end point was histological improvement. Virologic and histologic end points were evaluated by individuals blinded to the treatment assignments. Pretreatment biopsy specimens were examined locally before randomization and were then coded and evaluated in parallel with those obtained at week 72, by 2 experienced pathologists (A.B. and C.D.). Hepatic inflammation and fibrosis were graded with the Metavir scoring system²⁰ (scores ranging from 0 [none] to 3 for severe necroinflammatory activity, and 0 [none] to 4 for cirrhosis) and Ishak’s classification²¹ (scores ranging from 0 [none] to 12 for severe inflammation, and 0 [none] to 6 for cirrhosis). Histological improvement in disease activity or fibrosis was defined as a decrease of 1 point or more between the relevant pretreatment and posttreatment Metavir and Ishak subscores.²² Histological aggravation was defined as a score increase of 1 point or more.

Special attention was paid to the possible effect of treatment on HIV infection (CD4 cell count, HIV viral load). Adverse events were graded 1 (mild) to 4 (life-threatening), using the ANRS grading system.²³ Stepwise reductions in the peginterferon alfa-2b dose to 1 and 0.5 μg/kg per week, and reductions in the interferon alfa-2b dose to 1.5 million units 3 times a week or the ribavirin dose to 600 mg/d, were allowed to manage adverse events or laboratory abnormalities that had reached predefined thresholds. Patients who discontinued therapy prematurely because of adverse effects were encouraged to remain in the study.

The study was designed to have a power of 80% to detect a 15% difference (chosen for its clinical relevance) between the rates of sustained virologic response (from 20% vs 35% to 40% vs 55%) at a 5% level of significance (2-tailed test). Intention-to-treat analysis was used as the primary analysis for all measures of efficacy. Patients violating major eligibility criteria were excluded from the analyses. Patients who missed the final examination (week 72) were included as nonresponders. Histological responses were only analyzed in patients who underwent both a pretreatment and a posttreatment biopsy. Patients who received at least 1 dose of study medication were included in the safety analysis.

The Cochran-Mantel-Haenszel test (or Fisher exact test) was used to compare categorical variables, with stratification by center for comparisons between treatment groups and stratification by center and treatment group for other categorical variables. The Wilcoxon rank-sum test was used to compare quantitative variables between the groups. Logistic regression analyses were used to explore the influence of treatment and pretreatment characteristics on the response. Characteristics with P values below 0.20 in univariate analysis were included in multivariate models based on a backward elimination procedure. Adjusted odds ratios were transformed into approximated risk ratios (RRs) to correct for overestimation due to common events.²⁴ The Mac-Nemar χ² test or the Wilcoxon signed-rank test was used to compare pretreatment and posttreatment characteristics. All statistical tests were 2-tailed; P<.05 was considered statistically significant. We used SAS, version 8.2 (SAS Institute Inc, Cary, NC).

Patients were enrolled from February 2000 to February 2002. The trial ended in October 2003. A total of 442 patients were screened for eligibility, of whom 416 met the entry criteria and were randomly assigned to a treatment group (Figure 1). Four patients were excluded after randomization: 3 patients tested HCV-RNA–negative in the central laboratory, and one patient had previously received interferon alfa. Twenty-nine patients withdrew before receiving the study treatment because they refused or did not come for treatment or because a serious medical event occurred between randomization and the first treatment visit. During the follow-up period 4 patients in the peginterferon group and 1 patient in the standard interferon group withdrew. The pretreatment characteristics of the patients were similar in the 2 groups (Table 1).

A sustained virologic response (main end point) was obtained in 56 patients (27%) in the peginterferon group and 41 patients (20%) in the standard interferon group (P = .047). End-of-treatment virologic responses (at 48 weeks) were obtained in 72 (35%) and 44 (21%) of patients, respectively (P = .001). Undetectable serum HCV-RNA was obtained in 30 patients (15%) in the peginterferon group and 15 patients (7%) in the standard interferon group (P = .04) at week 4, 67 (33%) and 51 (25%) at week 12 (P = .09), and 83 (40%) and 57 (28%) at week 24 (P = .004).

A total of the 67 (99%) of the 68 patients in the peginterferon group and all 91 (100%) of patients in the standard interferon group who had detectable serum HCV-RNA and a viral load decline of less than 2 log₁₀ IU × 10³/L from baseline at week 12 failed to achieve a sustained virologic response at week 72 (P = .43, Fisher exact test).

Twenty-one patients (17%) with HCV genotype 1 or 4 infection who received peginterferon alfa-2b plus ribavirin had a sustained virologic response compared with 8 patients (6%) who received interferon-alfa2b plus ribavirin (P = .006, Figure 2). Among patients with HCV genotype 2, 3, or 5 infection, the rates of sustained virologic response were not different between the 2 treatment groups (35 [44%] vs 33 [43%], respectively, P = .88). Among patients who took at least 80% of the planned total dose, sustained virologic responses were achieved in 44 (40%) of 111 patients in the peginterferon group and 33 (29%) of 115 patients in the standard interferon group (P = .30).

Table 2 shows the variables that were included in the multiple logistic regression models on the basis of univariate analysis. Preliminary analysis showed that the sustained virologic response rates were influenced by the HCV genotype. A first multiple logistic regression model confirmed this result and showed that HCV genotypes 2, 3, or 5 were the main predictors of response (adjusted risk ratio [RR], 3.77; 95% confidence interval [CI], 2.69-4.93; P<.001). Other predictors were no protease inhibitor treatment, age 40 years or younger, and baseline alanine aminotransferase greater than 3 upper limits of normal. Because preliminary findings also indicated that the treatment effect differed according to the HCV genotype, further exploratory analyses were stratified according to the genotype. In patients with HCV genotype 1 or 4 infection, peginterferon treatment (RR, 2.43; 95% CI, 1.12 to 3.79; P = .03) and HCV viral load ≤5.7 log₁₀ IU x 10³/L (RR, 2.07; 95% CI, 1.04 to 3.74; P = .04) were independently associated with a sustained virologic response. In patients with HCV genotypes 2, 3, or 5, the absence of protease inhibitor treatment (RR, 1.80; 95% CI, 1.17-2.42; P = .01), age 40 years or younger (RR, 1.60; 95% CI, 1.08-2.10; P = .02), and baseline alanine aminotransferase more than 3 times the upper limits of normal (RR, 1.57; 95% CI, 1.07-2.02; P = .03) were independently associated with a sustained virologic response. Liver histology, the per kilogram body weight ribavirin dose, and the CD4 cell count were not independently associated with a sustained virologic response.

Paired pretreatment and posttreatment histological results were available for 205 patients (50%, Table 3). The reasons for missing posttreatment results were refusal of biopsy in 153 cases (74%), failure to return in 46 cases (22%), and clotting disorders in 8 cases (4%). The only significant difference in baseline characteristics between patients with and without posttreatment biopsy was a higher mean (SD) Metavir fibrosis score in patients with available posttreatment results (2.4 [1.0] vs 2.2 [1.0]), P = .03).

For disease activity, the Metavir score change was –0.19 in the peginterferon group and 0.01 in the standard interferon group (P = .02); and the mean changes in the Ishak grade were –0.57 and –0.26 (P = .24), respectively. The decline in both subscores was significant among sustained virologic responders while the subscores were stable among nonresponders. Changes in fibrosis did not differ between the 2 groups, but fibrosis worsened in patients who did not have a sustained virologic response. Steatosis improved significantly in patients infected by HCV genotype 3 who had a sustained virologic response (–13%, P<.001).

Similar proportions of patients in the 2 groups withdrew from the study because of clinical adverse events or laboratory abnormalities (Table 4). The doses of the study treatments were modified in 31 (16%) of patients in the peginterferon group and 13 (7%) of patients in the standard interferon group because of clinical adverse events (P = .004), and in 38 (20%) and 13 (7%) of patients respectively because of laboratory abnormalities (P = .004).

The incidence of most clinical adverse events was similar between the 2 treatment groups. The following parameters were not significantly different by week 12 in the peginterferon vs standard interferon groups: hemoglobinemia (–18 and –14 g/L, respectively, P = .01, at week 12), platelets (–34.3 vs –21.3 × 10³/μL, P = .004), neutrophils (–1.10 vs –0.70 × 10⁹/L, P = .009), lymphocytes (–0.66 vs –0.54 × 10⁹/L, P = .06), and CD4 cells (–121 vs –110 × 10⁶/L, P = .21). These parameters generally remained stable after week 12, then returned to near baseline values shortly after treatment cessation. When viral loads less than 400 copies/mL were attributed a value of 2.6 log₁₀, mean (SD) HIV viral load (log₁₀ copies/mL) was 2.96 (0.66) at baseline and 3.12 (0.86) at the end of treatment in the peginterferon group (P<.001), 3.04 (0.73) and 3.09 (0.78) respectively, in the standard interferon group (P = .12).

Seven deaths occurred among randomized patients: there were 5 deaths in the peginterferon group (2 from liver failure, 1 from metastatic neuroendocrine carcinoma, 1 from metastatic vulvar cancer—the patient who tested negative for HCV infection in the central laboratory—and 1 from accidental nitrate propyl overdose before the first dose of study treatment), and 2 deaths in the standard interferon group (1 from liver failure and 1 from liver cancer). One of the deaths from liver failure was considered possibly related to peginterferon-ribavirin therapy. This patient had fibrosis (F2) on a biopsy performed 9 months before the first dose of study treatment, was at Centers of Disease Control and Prevention stage A, and received didanosine, stavudine, and abacavir. He discontinued the study treatment after 8 weeks because of severe thrombocytopenia. He was hospitalized at week 12 with symptoms of decompensated cirrhosis and ascites infection. Liver biopsy showed significant cirrhosis (F4). Death occurred at week 32 from sepsis and liver failure. The proportion of patients reporting serious adverse events was generally similar among the groups. Symptomatic mitochondrial toxicity (symptomatic hyperlactatemia, lactic acidosis, or acute pancreatitis) occurred in 11 patients, 9 patients receiving peginterferon and 2 patients receiving standard interferon. All these patients were receiving didanosine.

In this population of HIV-infected patients with chronic HCV, peginterferon alfa-2b plus ribavirin yielded more sustained virologic responses than standard interferon alfa-2b plus ribavirin. Efficacy and tolerability in the peginterferon group were in keeping with the results of 3 uncontrolled trials, in which the sustained virologic response rates to peginterferon alfa-2b plus ribavirin were 28% to 31%.^16- 18 However, the absolute difference between the 2 groups (7%) was smaller than initially expected (15%); and was also smaller than in recent trials comparing peginterferon alfa-2a with standard interferon alfa-2a^14- 15 and in a small single-center trial comparing peginterferon alfa-2b with standard interferon alfa-2b.¹⁹ A high frequency of severe HIV- and/or HCV-related disease (62% of serious adverse events were unrelated to the study treatments, and 40% of our patients had bridging fibrosis or cirrhosis) and a high baseline prevalence of characteristics associated with poorer adherence to treatment (injection drug use, 79%; psychiatric disorders, 21%)²⁵ may have contributed to the smaller than expected difference between the 2 treatment groups in our study. However, the fact that we did not exclude such patients may make our results relevant to the general population of patients living with HIV and HCV coinfection in the United States and Europe.^26- 27

The benefit of peginterferon alfa-2b relative to standard interferon alfa-2b was most apparent in patients with genotype 1 or 4 infection. No significant differences were found between the treatment groups in patients with genotype 2 or 3 infection. Similar findings have been reported for HIV-seronegative and HIV-seropositive patients in studies comparing either peginterferon alfa-2b^11,19 or peginterferon alfa-2a¹² plus ribavirin with standard interferon plus ribavirin.

The HCV genotype was the predominant predictor of a sustained virologic efficacy, consistent with most previous studies of both HIV-seropositive^16,18 and HIV-seronegative patients.^{10- 11,28- 29} A young age, a lower HCV-RNA level, or a higher pretreatment alanine amino transferase level are also reported to be predictive of sustained virologic responses.^{14- 16,18,29- 30} The higher virologic failure rates that we observed among patients treated with protease inhibitors may be related to drug hepatotoxicity,³¹ increased HCV replication,³² restoration of anti–HCV immune responses³³ or cytochrome P450–mediated drug interactions.³⁴ Liver histology and the CD4 cell count were not independent predictors of the virologic response in our study, contrary to some previous reports^16,18 but in line with recently published clinical trials.^14- 15

The effect of treatment on histologic activity was similar to that observed in the HIV-seronegative population.¹¹ Interestingly, the progression of fibrosis was slowed when a virologic response was achieved.

Overall, tolerability was similar in the peginterferon and standard interferon groups. However, serious adverse events were far more frequent (35%) than what has been reported among HIV-seronegative patients (10%-15%). The incidence of opportunistic infections did not appear to be affected by anti-HCV therapy. Mitochondrial toxicity was particularly frequent in patients receiving didanosine, as previously reported,^{16,18,35- 36} possibly owing to increased intracellular concentrations of active triphosphorylated didanosine metabolites. As a result, a warning was added to the didanosine product information in September 2002, stating that ribavirin should be used cautiously in patients also receiving didanosine.³⁷ It is noteworthy that the patient who died from liver failure that was possibly related to the treatment had cirrhosis at the time of death and received didanosine; both these factors were recently reported as independent risk factors for hepatic decompensation.³⁸

In combination with ribavirin, peginterferon alfa-2b is more effective than standard interferon alfa-2b for initial treatment of HCV infection with genotypes 1 and 4 in HIV-infected patients. Treatment can be discontinued after 3 months if a satisfactory virologic response is not obtained at this time. Coadministration of ribavirin with didanosine should be avoided. However, the poor outcome of HIV-coinfected patients and particularly that of patients with HCV genotype 1 infections calls for new therapeutic approaches in this setting.