Opravil, Milos MD*; Sasadeusz, Joe MD†; Cooper, David A MD‡; Rockstroh, Jürgen K MD§; Clumeck, Nathan MD∥; Clotet, Bonaventura MD¶; Montaner, Julio MD#; Torriani, Francesca J MD**; DePamphilis, Jean PhD††; Dieterich, Douglas T MD‡‡
Because of shared routes of transmission, coinfection with HIV and hepatitis C virus (HCV) is common. It is estimated that one third of all HIV-infected individuals,1,2 and an even higher proportion of those with blood exposures such as injection drug users or recipients of transfused blood products,3,4 are also infected with HCV. HCV-related liver disease is known to progress more rapidly in patients with HIV as compared with those infected with HCV alone,5,6 and the presence of HCV has been linked with increased mortality rate in HIV-infected individuals.2 Furthermore, concurrent HIV/HCV coinfection may be associated with higher serum HCV RNA levels, accelerated progression of hepatic fibrosis, hepatocellular carcinoma, and an increased risk of end-stage liver disease.7 As a result, the burden of HCV infection among those with HIV is substantial. Because advances in antiretroviral therapy have improved the life expectancy of patients infected with HIV,8,9 the negative impact of chronic liver disease caused by HCV infection has become more apparent, and has thus become an important focus of clinical investigation.7,10-14
The current standard therapy for patients with HCV monoinfection, pegylated interferon in combination with ribavirin (RBV), produced a sustained virologic response (SVR) in up to 66% of patients in randomized clinical trials.15-18 Recently, this combination has also been studied in patients with HIV/HCV coinfection.10,12-14 In the AIDS PEGASYS Ribavirin International Coinfection Trial (APRICOT),10 the largest international study of peginterferon therapy conducted to date in the HIV/HCV-coinfected population, peginterferon alfa-2a (40KD) (Peg-IFN alfa-2a [40KD]; PEGASYS; Roche, Nutley, NJ) plus RBV (COPEGUS; Roche) produced significantly higher SVR rates than Peg-IFN alfa-2a (40KD) plus placebo or conventional interferon alfa-2a (IFN) plus RBV (40% vs. 20% and 12%, respectively; P < 0.001). Although some adverse events (AEs), such as lactic acidosis and hepatic decompensation, not previously reported in HCV-monoinfected patients, were observed in coinfected patients, the incidence of these events was low and treatment was not associated with any worsening of HIV disease status.10
One of the most common side effects associated with interferon-based therapies is alteration of hematologic parameters, particularly white blood cell counts, including neutrophils and lymphocytes.15-17 Therefore, there has been some concern that treatment of coinfected patients with interferon may reduce CD4 cell count, thus potentially causing a worsening of HIV disease. Furthermore, early studies with conventional IFN in coinfected patients indicated that efficacy seemed to be dependent on the patient's immunologic status at baseline, with the response rate being directly proportional to the pretreatment CD4 cell count.19-21 As such, current HCV treatment guidelines for coinfected patients recommend that the CD4 count should be greater than 200 cells/μL before commencing anti-HCV therapy.22,23
The APRICOT is the largest randomized study of anti-HCV therapy to date that has included coinfected individuals with CD4 cell counts as low as 100 cells/μL. To determine whether anti-HCV therapy is efficacious and safe in HIV/HCV-coinfected patients, particularly those with low CD4 cell counts, we performed a retrospective analysis of efficacy and safety data from the APRICOT according to patients' baseline CD4 cell count. We also carried out the analyses for patients with HCV genotype 1 and genotype 2 or 3 to ascertain whether patterns of efficacy and safety data were influenced by genotype.
Detailed entry criteria for the APRICOT have been reported previously.10 In summary, eligible patients were ≥18 years old, naive to HCV treatment, and positive for HCV antibody, and they had detectable HCV RNA, elevated alanine aminotransferase levels, and compensated liver disease. Patients were also required to have detectable anti-HIV-1 antibodies or HIV-1 RNA in serum and stable HIV disease (no opportunistic infections or malignancies requiring systemic therapy and no recent changes in antiretroviral therapy). Those with a baseline CD4 cell count ≥200 cells/μL were eligible regardless of their HIV RNA level; those with a baseline CD4 cell count between 100 and 199 cells/μL were eligible only if their baseline HIV-1 RNA load was <5000 copies/mL. Individuals with a CD4 count <100 cells/μL were excluded from the study. All patients provided written informed consent before starting treatment.
Study Design and Interventions
The APRICOT was a 3-arm, partially blind, randomized, parallel-group, placebo- and active comparator-controlled study. It was conducted in accordance with the principles of the Declaration of Helsinki, Good Clinical Practice guidelines, or with local law if it afforded greater protection to the participant.
Patients were randomized to 3 million IU of IFN alfa-2a 3 times a week + 800 mg/d of RBV, 180 μg/wk of Peg-IFN alfa-2a (40KD) + placebo, or 180 μg/wk of Peg-IFN alfa-2a (40KD) + 800 mg/d of RBV. All patients were to be treated for 48 weeks and subsequently followed for 24 weeks (until week 72). The use of a stable active antiretroviral regimen was allowed without restriction during the study. Patients were to have been receiving stable antiretroviral therapy for at least 6 weeks before the study, with no changes expected during the first 8 weeks of the study, or not to have received any antiretroviral therapy for at least 8 weeks before randomization and to be able to delay the initiation of antiretroviral therapy for at least 6 weeks. Changes to any existing antiretroviral therapy were permitted during the remainder of the study, at the investigator's discretion. Stratification factors included HCV genotype (1 vs. 2/3), HIV treatment (antiretroviral therapy vs. no antiretroviral therapy), baseline CD4 cell count (<200 vs. ≥200 cells/μL), and cirrhosis versus no cirrhosis.
Endpoint Assessments and Analyses
For this analysis, only patients who had available data on baseline CD4 cell count were included. Efficacy and safety outcomes were analyzed retrospectively using data from the intent-to-treat (ITT) population (all patients who received at least 1 dose of study medication) and safety population (all treated patients who had at least 1 postbaseline safety evaluation).
The primary efficacy endpoint was the proportion of patients who achieved an SVR, defined as undetectable HCV RNA (<50 IU/mL) at week 72 using the COBAS AMPLICOR HCV Test v2.0 (Roche Diagnostics, Basel, Switzerland). SVR was analyzed according to CD4 cell count strata; 3 strata of CD4 cell counts based on those currently utilized for staging HIV disease were used to determine whether the efficacy and safety of the 3 treatments studied were influenced by patients' immunologic status at baseline: ≥100 to <200 CD4 cells/μL, ≥200 to <350 CD4 cells/μL, and ≥350 CD4 cells/μL. SVR rates were also analyzed according to CD4% (defined as the proportion of lymphocytes that are CD4 cells at baseline) for the overall study population and by HCV genotype (genotype 1 vs. genotype 2 or 3). Baseline CD4% could only be calculated if the numbers of total lymphocytes and CD4 cells were measured on the same day and at the same laboratory at baseline.
Safety was assessed by measuring laboratory parameters, withdrawal rates, the incidence of AEs and serious AEs, AIDS-defining events (according to 1993 Centers for Disease Control and Prevention definitions), and deaths. Dose reductions of Peg-IFN alfa-2a (40KD) or RBV, and the reasons for these, were also noted.
A further analysis was performed to compare the change in HIV RNA titer from baseline between all patients (ITT analysis) and those completing therapy (patients receiving 48 weeks of treatment) to control for effect of duration of treatment. The effect of treatment on HIV RNA titer over time according to baseline CD4 cell count and baseline CD4% was also assessed to determine if either parameter influenced HIV RNA titer. CD4 cell count quartiles and CD4% quartiles were calculated from baseline values across the 3 treatment groups for all patients and for patients with genotype 1 and genotype 2/3.
The effect of treatment on the maximal change in the absolute number of CD4 cells and CD4% was also determined according to baseline CD4 count for each of the treatment groups to identify any patterns.
The disposition of randomized patients in the study has been reported previously.10 In total, there were 860 treated patients included in the ITT population (IFN + RBV, n = 285; Peg-IFN alfa-2a [40KD] + placebo, n = 286; Peg-IFN alfa-2a [40KD] + RBV, n = 289). Three patients (2 in the IFN + RBV group and 1 in the Peg-IFN alfa-2a [40KD] + RBV group) were excluded from the present analysis because their baseline CD4 count was not available. One further patient (with ≥350 CD4 cells randomized to Peg-IFN alfa-2a [40KD] + RBV) had no postbaseline safety assessment, and was therefore excluded from the safety population.
Most patients in each treatment group (213 [75.3%] of 283 in the IFN + RBV arm, 218 [76.2%] of 286 in the Peg-IFN alfa-2a (40KD) + placebo arm, and 216 [75.0%] of 288 in the Peg-IFN alfa-2a (40KD) + RBV arm) had a baseline CD4 count ≥350 cells/μL, possibly an effect of the high proportion of patients receiving antiretroviral therapy during the study (75% to 95% of patients). By comparison, the proportion of patients with a baseline CD4 count <200 cells/μL was relatively small: 20 (7.1%) of 283 in the IFN + RBV arm, 14 (4.9%) of 286 in the Peg-IFN alfa-2a (40KD) + placebo arm, and 17 (5.9%) of 288 in the Peg-IFN alfa-2a (40KD) + RBV arm (Table 1).
Because the CD4 cell count and the total lymphocyte count were not carried out on the same day and at the same laboratory for all patients (a prerequisite for CD4% calculation), data by CD4% were only available for 705 of 856 patients.
Patient Characteristics at Baseline
Baseline characteristics of patients in the ITT population are presented in Table 1. The 3 treatment groups were well balanced with respect to baseline characteristics; most patients were white (79%) and male (81%), with a median age between 39 and 40 years old, a median weight of approximately 71 to 74 kg, and a median body mass index (BMI) of approximately 24.0. There were no significant differences between the 3 treatment arms and the 3 CD4 strata. The exceptions were distribution of HCV genotype and the proportions of patients with cirrhosis and undetectable HIV viral load. Among patients with <200 CD4 cells/μL, fewer Peg-IFN alfa-2a (40KD) + RBV recipients had HCV genotype 1 infection (47%) compared with recipients of Peg-IFN alfa-2a (40KD) + placebo (79%) or IFN + RBV (75%). Cirrhosis was more common in patients with CD4 counts <200 cells/μL. The proportion of patients with undetectable HIV RNA varied widely between treatment arms and CD4 strata.
HCV Virologic Response According to Baseline CD4 Cell Count
In the overall study population, SVR rates were 12% with IFN + RBV, 20% with Peg-IFN alfa-2a (40KD) + placebo, and 40% with Peg-IFN alfa-2a (40KD) + RBV (P < 0.001 Peg-IFN alfa-2a (40KD) + RBV vs. Peg-IFN alfa-2a (40KD) + placebo and IFN + RBV).10 Treatment with Peg-IFN alfa-2a (40KD) + RBV for 48 weeks produced greater SVR rates compared with the 2 other arms across all 3 CD4 strata (Fig. 1A). Differences between Peg-IFN alfa-2a (40KD) + RBV and IFN + RBV were significant for CD4 counts <200 cells/μL and ≥350 cells/μL (P = 0.03 and P < 0.0001, respectively). The SVR rate for Peg-IFN alfa-2a (40KD) + RBV was higher (47%) among those with <200 CD4 cells/μL at baseline compared with the other CD4 groups; however, 6 of the 8 patients with an SVR in this group were infected with HCV genotype 2 or 3.
Analysis by genotype showed that, as expected, SVR rates were higher among patients infected with HCV genotype 2 or 3 than in those infected with HCV genotype 1 across all treatment groups and CD4 strata (see Figs. 1B, C), with the exception of patients in the lowest CD4 strata treated with Peg-IFN alfa-2a (40KD) + placebo. Overall, Peg-IFN alfa-2a (40KD) + RBV was the most efficacious treatment across all genotypes (P < 0.0001; odds ratio [OR] = 5.30, 95% confidence interval [CI]: 2.42 to 11.61 vs. IFN + RBV for HCV genotype 1 and P < 0.0001; OR = 5.42, 95% CI: 2.50 to 11.75 vs. IFN + RBV for genotype 2/3).
Because of the low number of patients with baseline CD4 <350 cells/μL, an efficacy analysis stratified further by genotype has only limited power. For patients infected with HCV genotype 1, Peg-IFN alfa-2a (40KD) + RBV was more efficacious than the other 2 treatments across all CD4 strata (see Fig. 1B). Among the 136 patients with a CD4 count ≥350 cells/μL, this combination produced an SVR rate of 32%. Response rates were lower in patients with a CD4 cell count <350 cells/μL. Differences between Peg-IFN alfa-2a (40KD) + RBV and IFN + RBV were significant for CD4 count ≥350 cells/μL (P < 0.0001; OR = 5.59, 95% CI: 2.39 to 13.09 and P < 0.0001; OR = 6.55, 95% CI: 2.61 to 16.43 for genotypes 1 and 2/3, respectively) but not for the 2 lower CD4 strata (see Figs. 1B, C). SVR was achieved in 1 (13%) of 8 genotype 1 patients with a CD4 count <200 cells/μL treated with Peg-IFN alfa-2a (40KD) + RBV.
In genotype 2 or 3 patients, the efficacy of Peg-IFN alfa-2a (40KD) + RBV seemed to be independent of the baseline CD4 count. The number of patients with severe immunodeficiency was low, however; among those with a baseline CD4 count of <200 cells/μL, 6 (75%) of 8 treated with Peg-IFN alfa-2a (40KD) + RBV had an SVR, whereas in the other 2 treatment groups, no patients achieved an SVR (see Fig. 1C).
Analysis of on-treatment virologic response rates over time in patients treated with Peg-IFN alfa-2a (40KD) + RBV showed a CD4-dependent pattern in HCV genotype 1 (Fig. 2A). In HCV genotype 2 or 3, the response was rapid (starting at week 12) and independent of baseline CD4 count (see Fig. 2B).
HCV Virologic Response According to Baseline CD4%
When SVR rates were analyzed according to patients' baseline CD4% (the proportion of lymphocytes that were CD4 cells), a trend was seen in the Peg-IFN alfa-2a (40KD) + RBV group and, to a lesser extent, in the IFN + RBV group that HCV virologic response increased as the baseline CD4% increased (Fig. 3A). This trend was consistent across all CD4% quartiles; the difference in SVR rate between patients in quartiles 1 and 4 was 14% in the Peg-IFN alfa-2a (40KD) + RBV group. When SVR rates were analyzed by HCV genotype and baseline CD4%, genotype 1 patients treated with Peg-IFN alfa-2a (40KD) + RBV had a lower SVR if their CD4% was in the lowest quartile. The response among genotype 2 or 3 patients, however, presented no obvious trend (see Figs. 3B, C).
Analysis of HIV RNA Changes From Baseline
Comparison of Safety Population and Completers
In the safety population and in patients completing 48 weeks of therapy, treatment with Peg-IFN alfa-2a (40KD) + RBV and Peg-IFN alfa-2a (40KD) + placebo produced similar reductions in HIV RNA titers, whereas HIV RNA titers increased with IFN + RBV in both populations (Fig. 4). Because the 2 populations were similar in this regard, subsequent analyses by CD4 strata and CD4% quartiles were performed using the safety population data.
Influence of Baseline CD4 Count on Change in HIV RNA Titer
When analyzed by baseline CD4 count strata, HIV RNA titer at week 48 was reduced from baseline with peginterferon monotherapy or in combination with RBV, whereas HIV RNA titer increased with interferon plus RBV therapy (data not shown). The extent of HIV RNA reduction from baseline was inversely correlated with baseline CD4 count. This is not unexpected, because patients with a lower CD4 count had a higher baseline mean HIV RNA titer and a lower proportion of patients had undetectable HIV RNA (see Table 1). When HIV RNA titers were analyzed by HCV genotype and baseline CD4 count strata, genotype 1 patients seemed to show a similar pattern of change in HIV RNA titer as the overall population, with patients in the lowest CD4 strata demonstrating the greatest reduction in HIV RNA titer. The response among genotype 2 or 3 patients presented no obvious trend; this was probably attributable to the small patient numbers involved.
Influence of Baseline CD4% on Change in HIV RNA Titer
When the changes in HIV RNA titers from baseline at week 48 were analyzed according to patients' baseline CD4%, with the exception of CD4% quartile 1 to median, a trend was seen in the Peg-IFN alfa-2a (40KD) + RBV and Peg-IFN alfa-2a (40KD) + placebo groups that a greater reduction in HIV RNA titer was observed in the lower baseline CD4% quartiles (Fig. 5A). HIV RNA titers increased or remained the same with IFN + RBV treatment. Trends were similar when HIV RNA titers were analyzed by HCV genotype. As with CD4 counts, genotype 1 patients seemed to show a similar pattern of change in HIV RNA titer as the overall population, with patients in the lowest CD4% quartile demonstrating the greatest reduction in HIV RNA titer (see Fig. 5B). As with the data for baseline CD4 count, the response among genotype 2 or 3 patients by CD4% quartile presented no obvious trend, probably because of the small patient numbers involved (see Fig. 5C).
Changes in CD4 Cell Count and CD4% During Treatment
Generally, a decline in absolute CD4 count was observed during treatment, with the exceptions being those with a baseline CD4 count <200 cells/μL who received Peg-IFN alfa-2a (40KD) + placebo or IFN + RBV (Fig. 6A). The median maximal change from baseline in absolute CD4 count was smaller among patients with a baseline CD4 count <200 cells/μL compared with the other 2 CD4 groups; the decrease seemed to be proportional to the baseline CD4 count. CD4 counts generally recovered after cessation of therapy; at the end of follow-up (week 72), the median absolute CD4 count was higher than at baseline in all treatment arms in patients with baseline CD4 counts <200 cells/μL and ≥200 to <350 cells/μL (data not shown). In patients with a baseline CD4 count ≥350 cells/μL, however, the CD4 count at week 72 remained lower than at baseline [−27, −48, and −14 in the Peg-IFN alfa-2a (40KD) + RBV, Peg-IFN alfa-2a (40KD) + placebo, and IFN + RBV groups, respectively].
A slight increase in the proportion of CD4 cells relative to the total number of lymphocytes (CD4%) was seen across all CD4 groups during treatment with each of the 3 drug regimens under investigation. The increases tended to be smaller in those with CD4 counts ≥350 cells/μL (see Fig. 6B).
The proportion of patients prematurely withdrawing from study treatment was similar across each of the 3 treatment groups (Table 2). Withdrawal rates for all safety reasons were highest among patients in the <200 cells/μL CD4 group across all 3 treatment groups. This discrepancy was mainly attributable to differences in the frequency of laboratory abnormalities requiring discontinuation; there were no between-group differences in the proportions of patients withdrawing because of AEs.
Withdrawals for nonsafety reasons were least frequent in the Peg-IFN alfa-2a (40KD) + RBV arm (6% to 11% of patients) compared with the other 2 treatment arms (14% to 35% of patients). There was no discernible pattern for nonsafety withdrawal rates among Peg-IFN alfa-2a (40KD) + RBV recipients when assessed according to the 3 CD4 strata (6%, 9%, and 11%). In the Peg-IFN alfa-2a (40KD) + placebo and IFN + RBV arms, the nonsafety withdrawal rates were highest in patients with a CD4 count <200 cells/μL, and a substantial proportion of these were attributable to insufficient therapeutic response.
The proportion of patients reporting neutropenia (<0.5 × 109 neutrophils/L) and the mean maximum decrease in hemoglobin concentration during treatment were largely similar between the 3 CD4 strata in each treatment arm (Table 3). In both Peg-IFN alfa-2a (40KD) treatment arms, the occurrence of thrombocytopenia (<50 × 109 platelets/L) was notably higher in patients with CD4 counts <200 cells/μL compared with the other CD4 strata (see Table 3). This could be attributable, in part, to the higher incidence of cirrhosis in those with CD4 counts <200 cells/μL.
Dose Reductions of Peginterferon Alfa-2a (40KD), Conventional Interferon, or Ribavirin for Safety Reasons
In patients treated with Peg-IFN alfa-2a (40KD) + RBV, the occurrence of dose reductions for any reason was highest in the stratum with a CD4 count <200 cells/μL (Table 4). As with the withdrawal rates, the discrepancy in rates of dose reductions among the CD4 groups seemed to be attributable mainly to laboratory abnormalities; most Peg-IFN alfa-2a (40KD) or IFN dose reductions were attributable to neutropenia and thrombocytopenia, whereas most RBV dose reductions were attributable to anemia.
Adverse Events, Serious Adverse Events, and Deaths
The overall frequencies of AEs were similar across the 3 CD4 groups (Table 5). The most common individual AE was fatigue, which was reported in 30% to 59% of patients in each CD4 group across the 3 treatment arms. In general, patients with a baseline CD4 count <200 cells/μL, particularly those in the Peg-IFN alfa-2a (40KD) group, reported a higher incidence of certain AEs, including fatigue, nausea, headache, and pyrexia. AIDS-defining events were uncommon, occurring in only 10 patients in the entire study population (none of these occurred in the CD4 <200 cells/μL group). Lower respiratory tract infections or bacterial infections were also uncommon (≤2%) in each arm of the study.
The incidence of serious AEs was similar among CD4 groups across all treatment arms. In the Peg-IFN alfa-2a (40KD) + RBV group, serious AEs occurred in 18%, 18%, and 17% of patients with CD4 counts of <200 cells/μL, ≥200 to <350 cells/μL, and ≥350 cells/μL, respectively. In those patients with a CD4 count <200 cells/μL, serious AEs occurred with a similar frequency between the 3 treatment groups: 18% of recipients of Peg-IFN alfa-2a (40KD) + RBV, 21% of those treated with IFN + RBV, and 25% of IFN + RBV recipients.
A total of 12 deaths occurred during the study (see Table 5): 4 in the Peg-IFN alfa-2a (40KD) + RBV arm, 5 in the Peg-IFN alfa-2a (40KD) + placebo arm, and 3 in the IFN + RBV arm. Two of these deaths were judged by the investigator as being possibly related to a study drug: 1 in the Peg-IFN alfa-2a (40KD) + RBV group (suicide) and 1 in the IFN + RBV arm (respiratory failure). One additional death, in the IFN + RBV arm, was caused by pneumonia and was thought to be remotely related to a study drug.
Treatment of HCV among those infected with HIV has become a focus of recent clinical investigation because of the increased rate of HCV-related morbidity and mortality now apparent after the introduction of more effective life-prolonging HIV therapies. Data from several randomized clinical trials, of which the APRICOT is the largest to date, have shown that the highest SVR rates can be achieved with Peg-IFN + RBV,10,12,14 and guidelines providing recommendations for the treatment of HCV in coinfected individuals have since been published.22-24 Recent HCV treatment guidelines endorsed by the European Association for the Study of Liver Disease22 advocate the use of Peg-IFN/RBV combination therapy early in the course of HIV infection before antiretroviral therapy is started but state that for those with severe immunodeficiency (CD4 count <200 cells/μL), the CD4 count should be increased before commencing anti-HCV therapy. Practice guidelines in the United States23 follow a similar approach, suggesting that all HIV/HCV-coinfected patients should be considered candidates for treatment but that antiretroviral therapy should be optimized beforehand to ensure “maximal” CD4 benefit. These recommendations are largely reflective of the fact that early studies in coinfected patients showed a poorer SVR among those with a low CD4 count, and, historically, severely immunocompromised patients (ie, those with a CD4 count <200 cells/μL) have been excluded from most clinical trials; as a result, few data are available.19-21
In the overall population, which comprises results from one of the few clinical trials to include patients with CD4 counts <200 cells/μL, neither the efficacy nor the safety profile of Peg-IFN alfa-2a (40KD) + RBV therapy seems to be significantly compromised by a low baseline CD4 count, although efficacy differed between HCV genotypes. In keeping with results obtained in the overall APRICOT population reported elsewhere,10 treatment with Peg-IFN alfa-2a (40KD) + RBV was associated with a higher SVR rate than Peg-IFN alfa-2a (40KD) + placebo or conventional IFN + RBV. Analysis of the overall population showed that this effect was maintained across all 3 CD4 strata. The findings about patients with severe immunodeficiency should be interpreted with caution, however, because of the small number of patients with a CD4 count <200 cells/μL treated with Peg-IFN alfa-2a (40KD) + RBV (n = 17) and the relatively high proportion of patients with HCV genotype 2 or 3 infection (47%), who are known to be more responsive to Peg-IFN/RBV therapy than patients infected with HCV genotype 1.15-17
When data were analyzed according to HCV genotype and baseline CD4 count, Peg-IFN alfa-2a (40KD) + RBV was more efficacious than other treatment combinations studied for genotype 1 and genotype 2 or 3 patients across all CD4 strata. Among those infected with HCV genotype 2 or 3, there was no obvious trend to suggest that baseline CD4 count influenced the efficacy of any of the 3 treatments administered. Among those with a CD4 count <200 cells/μL, however, the response to Peg-IFN alfa-2a (40KD) + placebo and to IFN + RBV was poor, with no patients reporting an SVR. The small number of patients (3 in each treatment group) may explain this observation. For those infected with HCV genotype 1, there was an overall trend to suggest that response to treatment may be influenced by baseline CD4 count, particularly for those treated with Peg-IFN alfa-2a (40KD) + RBV. In this arm, the response rate ranged from 13% among those with <200 CD4 cells/μL at baseline to 32% in those in the highest CD4 category. These data indicate that for patients infected with HCV genotype 1, the ability to achieve an SVR is impaired if treatment is initiated when CD4 counts are less than 350 cells/μL. This observation warrants further study.
When SVR data were analyzed according to patients' baseline CD4%, overall, a trend was seen in the Peg-IFN alfa-2a (40KD) + RBV group and, to a lesser extent, in the IFN + RBV group, suggesting that SVR rates increase as the baseline CD4% increases.
When response data were analyzed according to HCV genotype (1 vs. 2 or 3), a less consistent trend was seen among genotype 2 or 3 patients compared with genotype 1 patients and the overall population. Of note, those genotype 1 patients in the fourth quartile of baseline CD4% values seemed to show a lower SVR rate in all 3 treatment groups than patients in quartiles 2 and 3, suggesting a potential blunted response in those with a high proportion of CD4 cells at baseline. No obvious trends were seen among the Peg-IFN alfa-2a (40KD) + placebo group, in the overall population and in the genotype-specific analyses.
In interpretation of these data, it is important to note that the number of patients enrolled in the APRICOT with a CD4 count <200 cells/μL was small, restricted to those with a CD4 count ≥100 cells/μL and a baseline HIV-1 viral load <5000 copies/mL; therefore, interpretations of the data regarding the influence of baseline CD4 count on response are conservative, and definitive conclusions cannot be made.
Because the study population was restricted to those with stable HIV, inevitably, patients with a low CD4 count, especially those with poorly controlled HIV infection, were not adequately represented in the present analysis. In addition, the distribution of HCV genotypes was not equal between the different CD4 strata analyzed. Among those treated with Peg-IFN alfa-2a (40KD) + RBV, the proportion of HCV genotype 2 or 3 patients was high in patients with a baseline CD4 count <200 cells/μL, which undoubtedly contributed to the good SVR rate (47%) in this group. Other imbalances in patient and disease characteristics at baseline, such as the proportion of patients with cirrhosis and the HIV viral load, may also have influenced the efficacy and safety outcomes reported.
There seemed to be no obvious relation between baseline CD4 count and the overall safety profile of the 3 treatment combinations studied. With all treatments, as expected, the incidences of some individual AEs and laboratory abnormalities were higher among patients with a baseline CD4 count <200 cells/μL, probably because of overall impaired general health status. It should be noted that the proportion of patients with cirrhosis was greater in those with a CD4 count <200 cells/μL than in those with a CD4 count ≥200 cells/μL, and this is likely to have resulted in greater hematologic toxicity in these patients. The higher withdrawal rate seen among patients treated with Peg-IFN alfa-2a (40KD) + RBV with a CD4 count <200 cells/μL was found to be attributable almost entirely to variation in the rate of laboratory abnormalities, and there were no between-group differences in the proportion of patients withdrawing because of symptomatic AEs. Dose reductions of study medication in the <200 CD4 cells/μL group were also higher in patients treated with Peg-IFN alfa-2a (40KD) + RBV than in the other CD4 groups, attributable mainly to an increased incidence of laboratory abnormalities, such an anemia, neutropenia, and thrombocytopenia. Consistent with previous findings in the HCV-monoinfected population,15-17 neutropenia and thrombocytopenia were the main reasons for Peg-IFN dose reductions, whereas most RBV dose reductions were attributable to anemia. In the other treatment arms, no differences among CD4 groups were seen. These findings suggest that although there is no apparent increased safety risk associated with Peg-IFN alfa-2a (40KD) + RBV therapy in those who are severely immunocompromised, clinicians should be aware of the increased potential for hematologic abnormalities in this group.
The primary analysis of APRICOT data revealed that the mean absolute CD4 count decreased uniformly in all treatment groups during the course of the study but that the mean percentage of CD4 lymphocytes increased slightly.10 There was no increase in the incidence of AIDS-defining events, however. To investigate this observation further, we analyzed maximal changes in absolute CD4 count and in the CD4% according to patients' baseline CD4 status. Although the absolute number of CD4 cells decreased during treatment, the magnitude of decrease was found to be proportional to baseline CD4 count, with the decrease being greater among those with a higher CD4 count at baseline than in those with a lower CD4 count. For patients with a baseline CD4 count <200 cells/μL, a small increase was seen in the Peg-IFN alfa-2a (40KD) + placebo group and in the IFN + RBV treatment group. Changes in the CD4% seemed to be inversely proportional to the baseline CD4 count, with the greatest increases being seen in those with the lowest baseline count. The maximum increase in CD4% (8%) was only marginal, however. The possible mechanisms by which interferon-based therapies affect CD4 dynamics and the clinical consequences of these changes have yet to be elucidated; however, because they seem to be related to patients' baseline immunologic status and seem to normalize after cessation of anti-HCV therapy, any significant positive or negative influence on patient outcome is not anticipated.
The APRICOT is the largest randomized study of a pegylated interferon to investigate the efficacy and safety of anti-HCV therapy among patients with a CD4 count as low as 100 cells/μL. Based on our findings, we suggest that Peg-IFN alfa-2a (40KD) + RBV treatment should be considered on a case-by-case basis in all patients with stable HIV and should not, as a rule, be withheld from HIV-infected patients with a low CD4 count (100 to 200 cells/μL). Because patients infected with HCV genotype 2 or 3 are known to be more responsive to Peg-IFN + RBV therapy, HCV genotyping before treatment is likely to be of benefit in determining those most likely to benefit from treatment. Although the safety profile of the interferon-based therapies studied in this population did not reveal any major cause for concern in those patients with the lowest baseline CD4 counts, because of the apparent increased frequency of laboratory abnormalities and dose reductions seen in this group, frequent monitoring, particularly during the early stages of therapy, seems prudent.
In summary, the findings from this analysis suggest that the benefits of treating concurrent HCV infection among immunocompromised individuals with HIV may outweigh the potential risks, particularly with Peg-IFN alfa-2a (40KD) + RBV therapy, where the SVR rate was substantially higher than with the other 2 treatment combinations studied across all HCV genotypes and CD4 strata. Although the number of patients with a baseline CD4 count <200 cells/μL was small, SVR rates with Peg-IFN alfa-2a (40KD) + RBV seemed to improve with higher baseline CD4 counts for genotype 1 but were independent of baseline CD4 counts for genotype 2/3. Additional prospective studies of Peg-IFN + RBV therapy in a larger number of coinfected patients with low CD4 counts are necessary to define better the optimal time point of treatment initiation in patients infected with HCV genotype 1 who are severely immunocompromised.
The authors thank Mary Hines for editorial support.
© 2008 Lippincott Williams & Wilkins, Inc.