| |
Black patients with chronic hepatitis C have a lower sustained viral response rate than non-Blacks with genotype 1, but the same with genotypes 2/3, and this is not explained by more frequent dose reductions of interferon and ribavirin
|
| |
| |
Journal of Viral Hepatitis
Volume 13 Page 242 - April 2006
N. Brau1,23, E. J. Bini2, S. Currie3,24, H. Shen3,24, W. N. Schmidt4, P. D. King5, S. B. Ho6, R. C. Cheung7, K.-Q. Hu8, B. S. Anand9, F. R. Simon10, A. Aytaman11, D. P. Johnson12, J. A. Awad13, J. Ahmad14, C. L. Mendenhall15, M. C. Pedrosa16, R. H. Moseley17, C. H. Hagedorn18, B. Waters19, K.-M. Chang20, T. R. Morgan21, S. J. Rossi3, L. J. Jeffers22, T. L. Wright, 3,24 The VA-HCV-001 Study Group
1Veteran Affairs Medical Centers Bronx, NY; 2New York, NY; 3San Francisco, CA; 4Iowa City, IA; 5Columbia, MO; 6Minneapolis, MN; 7Palo Alto, CA; 8Loma Linda, CA; 9Houston, TX; 10Denver, CO; 11Brooklyn, NY; 12Bay Pines, FL; 13Nashville, TN; 14Pittsburgh, PA; 15Cincinnati, OH; 16Boston, MA; 17Ann Arbor, MI; 18Atlanta, GA; 19Memphis, TN; 20Philadelphia, PA; 21Long Beach, CA; 22Miami, FL; 23Mount Sinai School of Medicine, New York, NY; and 24University of California, San Francisco, CA, USA
Summary. In previous hepatitis C virus (HCV) treatment studies, Black patients not only had a lower sustained viral response (SVR) rate to interferon and ribavirin (RBV) than non-Black patients but also a higher frequency of HCV genotype 1 (GT-1) infection. The aim of this community-based study was to determine whether Black patients have a lower SVR rate independent of genotype. We prospectively enrolled 785 patients (24.8% Black, 71.5% White, 3.7% others) who received interferon alpha-2b 3 MU three times weekly + RBV 1000-1200 mg/day for 24 weeks (GT-2/3) or 48 weeks (GT-1). Black patients were more commonly infected with GT-1 (86.8%vs 64.8%, P < 0.001) and less frequently had an SVR compared with non-Black patients (8.4%vs 21.6%, P < 0.001). Within GT-1, Black patients had a lower SVR rate than non-Black patients (6.1%vs 14.1%, P = 0.004) but not within GT-2/3 (50.0%vs 36.5%, P = 0.47). Black patients had lower baseline haemoglobin levels (14.8 vs 15.3 g/dL, P < 0.001) and neutrophil counts (2900 vs 4100/mm3, P < 0.001) and required more frequent dose reductions of RBV (29.8%vs 18.5%, P < 0.001) and interferon (4.7%vs 1.6%, P = 0.012). However, dose reductions were not associated with lower SVR rates while early treatment discontinuations were (2.9%vs 25.7%, P < 0.001). Independent predictors of SVR were GT-1 [odds ratio (OR) 0.33; 95% confidence interval (CI) 0.20-0.55; P < 0.001], Black race (OR 0.45; 95% CI 0.22-0.93; P = 0.030), and advanced fibrosis, stages 3 + 4 (OR 0.53; 95% CI 0.31-0.92; P = 0.023). In conclusion, Black patients infected with HCV GT-1 (but not GT-2/3) have a lower SVR rate than non-Black patients. This is not explained by their lower baseline haemoglobin levels and neutrophil counts that lead to higher rates of ribavirin and interferon dose reductions.
Introduction
In chronic hepatitis C, treatment with interferon (standard or pegylated) and ribavirin (RBV) leads to a viral cure or sustained viral response (SVR) in a limited number of patients. Even with the best current therapy of pegylated interferon (PEG-IFN) plus RBV, 49% of patients infected with hepatitis C virus (HCV) genotype 1 (GT-1) and 18% of patients with GT-2/3 do not permanently clear the virus [1,2]. Furthermore, the SVR rate is lower in Black than in White patients [3]. One hepatitis C treatment trial with 3% Black subjects not only found a lower SVR rate among Black patients but also a higher frequency of HCV GT-1 infection. Within GT-1, the study found no difference in SVR rate between Black and White individuals [4]. However, the study was limited by the small number of Black patients. The present study was performed with a significant number of Black patients to examine the influence of HCV genotype and other factors on the SVR rate of Black vs non-Black patients.
Results
Patient characteristics
The baseline characteristics are displayed in Table 1. The study included a sizable proportion of Black patients (24.8%). Among non-Black individuals, 95% were White. Among 743 study subjects with genotype results, 70% were infected with HCV GT-1 and 24% with GT-2/3, a distribution similar to other US-based hepatitis C studies [7] and the US population as a whole [10]. Among 569 patients who had a liver biopsy, cirrhosis was seen in 13.9% of patients. Black patients differed significantly from non-Black patients in that they were more commonly infected with HCV GT-1 (86.8%vs 64.8%, P < 0.001). At baseline, they also had lower mean levels of ALT (86 IU/L vs 101 IU/L, P = 0.001) and haemoglobin (14.8 g/dL vs 15.3 g/dL, P < 0.001) and lower mean counts of white blood cells (6000/mm3vs 7100/mm3, P < 0.001) and of neutrophils (2900/mm3vs 4000/mm3, P < 0.001).
Viral response
An ETR was achieved in 29.7% of patients and a SVR in 18.2%. The SVR rate in GT-1 was lower (11.7%) than in GT-2/3 (36.5%, P < 0.001). The SVR rate in GT-4 (44.4%) was similar to that in GT-2/3 (P = 0.73) as shown in Fig. 1. Overall, Black patients had a significantly lower SVR rate than non-Black subjects (8.4%vs 21.6%, P < 0.001). Within GT-1, the SVR rate of Black patients was still lower than that of non-Black patients (6.1%vs 14.1%, P = 0.017) but in GT-2/3, the rates were similar (50.0%vs 36.5%, P = 0.47). The results are displayed in Fig. 2. Univariate logistic regression analysis found the following baseline factors to be significantly associated with SVR rate: HCV GT-1, Black race, HCV-RNA level ≥1 000 000 copies/mL, alanine aminotransferase (ALT) level >90 IU/L (twice upper limit of normal), advanced fibrosis (stages 3 + 4), and neutrophil count <2000/mm3. Multivariable logistic regression analysis identified only GT-1, Black race, and advanced fibrosis as independently associated with SVR rate (Table 2).
Safety
Early treatment discontinuations occurred in 30.5% of patients with no difference between Black and non-Black individuals (30.9%vs 29.9%, P = 0.80). Treatment was most commonly discontinued because of adverse events (18.7% of subjects) of which neuropsychiatric side-effects dominated (4.1%) followed by influenza-like symptoms (2.8%), fatigue (1.3%), rash (0.6%), and anaemia (0.5%). Treatment was discontinued prematurely because of patient non-adherence in 3.7%, loss to follow-up in 3.2%, and for unspecified reasons in 4.8%. In multivariable logistic regression analysis, two factors were independently associated with early treatment discontinuation: ≥3 alcoholic drinks per day in the last 12 months prior to treatment and annual household income <$25 000 (Table 3).
Dose reductions of RBV occurred in 21.1% of patients. The RBV dose was reduced more frequently in Black than in non-Black subjects (29.8%vs 18.5%, P < 0.001). Multivariable logistic regression analysis found only baseline haemoglobin concentration <14.0 g/dL to be independently associated with RBV dose reductions (Table 4).
The dose of interferon was reduced in 2.3% of individuals, again with more frequent dose reductions in Black than in non-Black patients (4.7%vs 1.6%, P = 0.012). In logistic regression analysis, baseline weight ≥75.0 kg, and neutrophil count <2000/mm3 were independently associated with interferon dose reductions (Table 5).
There was no difference in SVR rates (all patients) between those who completed the full treatment course and had no dose reductions (25.7%) and those with IFN dose reductions [40% (2/5), P = 0.61], those with RBV dose reductions (22.2%, P = 0.46), and with dose reductions of both IFN and RBV [0% (0/5), P = 0.34]. However, patients who discontinued therapy prematurely did have a lower SVR rate (2.9%, P < 0.001) than those with a full treatment course. The results are displayed in Fig. 3. When this analysis was performed separately for Black and non-Black patients, the same pattern was seen.
When a study was published that found that RBV dose reductions are associated with a lower SVR rate when the dose reduction occurs in weeks 1 to 20, but not in weeks 21 to 48 [11], a post hoc analysis was performed with the same differentiation by early vs late dose reductions. This analysis showed that patients with RBV dose reductions until week 20 had a similar SVR rate (25.0%) as patients with no dose reductions (25.7%, P = 0.92), and the same applied to RBV dose reductions in weeks 21 to 48 (SVR, 19.6%vs 25.7%, P = 0.33). Dose reductions of IFN did not influence the SVR rate in any of the sub-analyses.
In patients who received a full course of therapy with full doses of both RBV of IFN, Black subjects still had a lower SVR rate than White subjects (11.1%vs 29.8%, P = 0.003). This difference occurred in GT-1 (7.6%vs 18.8%, P = 0.033), but not in GT-2/3 (75%vs 49%, P = 0.62).
Side-effects were similar to the ones that are expected with combination therapy of interferon and RBV. The most common adverse events (≥5% incidence) were fatigue (50%), myalgia (38%), headache (36%), dyspnoea (26%), nausea or vomiting (25%), depression (19%), rash (19%), fever (14%), injection-site reactions (12%), and alopecia (8%). Two of 785 patients (0.25%) died during the study. One patient died of pneumonia, and the cause of death is unknown in the other.
Discussion
In treatment with interferon monotherapy, several studies have shown that Black individuals with chronic hepatitis C have a lower SVR rate and a slower reduction of HCV viral load than White patients [4,12-16].
In combination therapy with IFN and RBV, a study by McHutchison et al. [4] with pooled data from two phase III trials found that Black patients (n = 53, 3.0% of study population) had a lower overall SVR rate than White patients (11%vs 27%, P = 0.01), but Black patients also had a higher frequency of HCV GT-1 infection (96%vs 65%). Among patients with GT-1, there was no difference in the SVR rate between Black and White patients (22%vs 23%). On multivariable analysis, Black race was not independently associated with SVR. In this study, only six Black patients who received combination therapy achieved a SVR [4].
Jeffers et al. [17] performed a study of 78 Black and 28 White subjects with chronic hepatitis C GT-1 who were treated with PEG-IFN plus RBV. The study did not show a difference in the SVR rate between Black and White subjects (26%vs 39%, P = 0.17) but the study was not intended to show a difference but rather to determine the SVR rate in Black patients with a 95% confidence interval of <20%. The findings of these two studies are in contrast with our study where Black patients with GT-1 did have a significantly lower SVR rate than non-Black patients (most of them White) with GT-1 (6.1%vs 14.1%, P = 0.017) when treated with HCV combination therapy. This discrepancy is likely explained by the fact that neither of the previous two studies had a large enough sample size to detect a significant difference in SVR rates between Black and White individuals.
In a recent study, Muir et al. [18] compared 100 Black with 100 White patients (GT-1 in 98%) who were treated with PEG-IFN and RBV, and they found a lower SVR rate in Black than in White patients (19%vs 52%, P < 0.001). In the multivariable logistic regression analysis, only Black race was independently associated with SVR rate (odds ratio, 0.22; P < 0.001) [18]. In our study, multivariable analysis of SVR found GT-1, Black race, and advanced fibrosis to be independently associated with lower SVR rate. Our study thus confirms the findings by Muir et al. that Black patients have a lower SVR rate than White patients in HCV GT-1 infection. However, in HCV infection with GT-2/3 we found no difference in the SVR rate between Black and non-Black patients. Studies published to date have not been able to address differences in viral response in infection with GT-2/3 as they have either focused exclusively on GT-1 [17,18] or have included insufficient numbers of Black patients to comment on the issue[4].
In our analysis of haematological differences that may explain the lower SVR rate in Black patients, several differences in baseline and safety characteristics became apparent between the two racial groups. Interferon dose reductions were more frequent in Black compared with non-Black patients (4.7%vs 1.6%). The main reason for this observation was the significantly lower baseline neutrophil count of Black individuals compared with non-Black patients (2900/mm3vs 4100/mm3). When controlled for a neutrophil count of <2000/mm3, Black race was no longer associated with IFN dose reductions. Of note, higher weight (≥75.0 kg) was also independently associated with less frequent IFN dose reductions. This reflects the fact that standard IFN has a small molecular weight and a large volume of distribution (30 L), which, in heavier patients, leads to lower plasma IFN levels and less myelosuppression. There was also a higher rate of RBV dose reductions among Black vs non-Black patients (30.0%vs 18.4%). One possible reason for this finding was the slightly, but significantly, lower baseline haemoglobin level in Black vs non-Black patients (-0.5 g/dL), a difference that had been observed in another study [19]. Another possibility may be a genetically higher susceptibility of red blood cells to the haemolytic properties of RBV, although glucose-6-phosphate dehydrogenase deficiency (which is more frequent in Black than in White individuals) does not seem to be a factor as shown in a recent study [20]. In univariate analysis, both a low baseline haemoglobin level (<14.0 g/dL) and Black race were associated with more frequent RBV dose reductions. However, when controlled for low haemoglobin concentration, Black race was no longer associated with RBV dose reductions. Thus, the lower baseline haemoglobin concentration alone explains the more common rate of RBV dose reductions in Black patients. By comparison, Muir et al. [18] found no difference between Black and White patients in the rate of dose reduction of PEG-IFN and RBV combined (22%vs 24%) as well as the combined dose reduction rate because of anaemia (4% in both groups), but the study did not differentiate between PEG-IFN and RBV dose reductions. There was no difference in early treatment discontinuation between Black and non-Black patients.
When patients with dose reductions of either IFN or RBV or both, who completed a full course of treatment, were compared with those without any dose reductions, there was no significant difference in SVR rates. This applied to all patients, as well as to Black and non-Black patients separately. It can therefore be concluded that while Black patients had more frequent dose reductions of either IFN or RBV than non-Black patients, this difference does not explain the lower SVR rate of Black patients. Early dose discontinuations were associated with a lower SVR rate, but they were equally frequent between Black and non-Black patients.
A decrease in SVR rate with early (weeks 1-20), but not late (weeks 21-48) RBV dose reduction was found in a study by Shiffman et al. [11]. Patients with chronic hepatitis C who had failed prior therapy with IFN and RBV were retreated with PEG-IFN plus RBV. The SVR rate decreased in patients with early RBV dose reductions from 21% (>80% RBV dose received) to 17% (61-80% RBV dose) to 11% (<60% RBV dose, P = 0.031). In patients with late RBV dose reductions, there was no difference in SVR rates: 55%, 47%, 54%, respectively for above RBV doses (P = 0.32). Dose reductions of PEG-IFN at any stage did not influence SVR rates either [11]. Our study did not confirm that early RBV dose reductions led to lower SVR rates. The reasons for this discrepancy are not apparent, and it is also unclear whether the use of PEG-IFN vs standard IFN can explain the different findings.
The SVR rates in our study using standard IFN and RBV were visibly lower than those in the previously published US-based phase III trial by McHutchison et al. [7] using the same regimen. Our study patients had a lower SVR rates, both with GT-1 (12%vs 28%) and with GT-2/3 (37%vs 66%). These difference in viral response rates may be due to the fact that compared with subjects in the McHutchison trial, our study population had a higher frequency of most factors that have been associated with lower SVR rates. Our subjects were older (49 vs 44 years) and heavier (89 vs 81 kg), more commonly male (97%vs 67%) and Black (25%vs 3%), and they had a higher frequency of cirrhosis (14%vs 5%) and advanced fibrosis (stage 3 + 4, 37%vs 25%). Of note, in our study, the rate of early treatment discontinuation because of adverse events was similar to that in the phase III trial (19%vs 21%). The rate of discontinuation because of non-safety events was not reported in that study. Given the many differences in baseline characteristics between the two studies, it cannot be concluded that the SVR rate to IFN plus RBV is genuinely lower in one population than the other. However, the comparison shows that a community-based patient population can expect a lower SVR rate than that published from phase III trials when they are more commonly Black or male and when they have more advanced liver disease than that in the published studies.
Nowadays, in hepatitis C therapy, standard interferon has been replaced by once weekly PEG-IFN, a longer-lasting and more effective interferon that is combined with RBV. However, the main outcomes of our study, i.e. Black people have lower SVR rates than non-Black people in GT-1 but the same in GT-2/3, and that dose reductions do not play a role in differences of SVR rates between the races, are equally applicable to therapy with PEG-IFN plus RBV. Indeed, our study confirms the different GT-1 outcome of Muir's study [18]. The study also confirms the lack of IFN dose reductions on SVR rate seen in Shiffman's trial [11] with PEG-IFN, but the discrepancy between the two studies in the effect of early RBV dose reductions on SVR is unexplained.
We conclude that a reduced SVR rate in Black vs non-Black patients with chronic hepatitis C GT-1 is consistently seen in our and other studies. Our data further show that there is no racial difference in SVR rates in HCV infection with GT-2/3, a finding not reported previously. We were unable to shed new light on the question why such a different outcome is seen in GT-1. However, we showed that differences in IFN or RBV dose reductions do not explain the difference in SVR rates between the races. In community-based practices, SVR rates to HCV combination therapy can be substantially lower than that in published phase-III trials when their patients are more commonly Black and male, have more cirrhosis and other factors associated with a lower SVR rate.
An ongoing prospective trial sponsored by the National Institutes of Health called VIRAHEP-C (http://www.edc.gsph.pitt.edu/virahepc) is intended to treat 400 HCV-infected patients (50% Black and White each) with PEG-IFN and RBV and is designed to examine factors associated with viral nonresponse. It is the purpose of that trial to elucidate why Black patients have a lower SVR rate than White patients.
Acknowledgements
This study was supported in part by a grant from the Schering-Plough Corporation and the VA National HIV/Hepatitis C Program.
Methods
Patient selection
In a prospective study from November 1999 to November 2001, 4364 chronically HCV-infected patients were evaluated for treatment in 24 Veterans Affairs Medical Centers. The characteristics and epidemiology of these 4364 subjects are reported elsewhere [5]. In 813 patients (18.6%), the physician recommended treatment with standard interferon and RBV, and the patient agreed to start therapy and consented to be followed up prospectively throughout this study. We excluded 28 patients with human immunodeficiency virus (HIV)/HCV-coinfection leaving 785 patients with HCV-monoinfection for this analysis. Inclusion and exclusion criteria were similar to those in previously reported phase III trials of interferon and RBV [6,7]. Patients were eligible for the study if they had compensated liver disease, were hepatitis C treatment-naive, were seronegative for hepatitis B virus surface antigen, had normal values of creatinine, and met predefined laboratory parameters (total bilirubin <2.0 g/dL, haemoglobin ≥12 g/dL, neutrophils ≥1500/mm3, platelets ≥85 000/mm3). Patients with active ischaemic heart disease or other significant medical or psychiatric disease were excluded from the study. The study was approved by the Human Studies Subcommittee of each participating institution, and all patients signed an informed consent.
Study design
All patients were prospectively followed up during treatment and a 24-week follow-up period. Therapy consisted of the standard anti-HCV regimen at the time of interferon alpha-2b (IFN) 3 million units (MU) three times per week (TIW) plus weight-based RBV 1000-1200 mg/day (cut-off at 75 kg). Duration of treatment was 48 weeks for GT-1 and GT-4 with discontinuation of therapy at week 24 in patients who still had detectable plasma HCV-RNA. Patients with GT-2 or GT-3 were treated for 24 weeks. As recommended in the product information [8], the dose of IFN was reduced from 3.0 to 1.5 MU TIW for a neutrophil count of <750/mm3 or a platelet count of <50 000/mm3, and combination therapy was discontinued for a neutrophil count of <500/mm3 or a platelet count of <20 000/mm3. Similarly, the RBV dose was reduced from 1000-1200 mg/day to 600 mg/day for a haemoglobin level of <10.0 g/dL, and RBV was stopped with a haemoglobin concentration of <8.5 g/dL. The same schedule was applied when other clinical events required dose reductions.
The primary endpoint was the rate of sustained viral response (SVR), i.e. undetectable plasma HCV-RNA (<100 copies/mL by qualitative polymerase chain reaction assay, Amplicor , Roche Diagnostics, Nutley, NJ, USA) 24 weeks after the end of treatment comparing Black and non-Black patients. An analysis of causes of a possible difference in SVR rates focused on differences in haematological toxicity and early treatment discontinuations and dose reductions between the two groups. Viral response was analysed on an intention-to-treat basis, i.e. lacking HCV-RNA data was counted as viral non-response. A secondary efficacy endpoint was the rate of end-of-treatment response (ETR) at treatment week 24 (GT-2/3) or week 48 (GT-1) or at early treatment discontinuation. Safety endpoints included the incidence of clinical adverse events and the rates of dose reduction and early discontinuation of therapy. Early treatment discontinuation was defined as stopping interferon and RBV prior to the scheduled end of therapy. Dose reduction was defined as reducing the dose of either RBV or interferon or both but completing 24 or 48 weeks of therapy.
Baseline HCV-RNA testing was performed locally in each facility. The dynamic range varied between sites, but all tests measured HCV-RNA quantitatively at least until 1 000 000 copies/mL. Thus, no mean or median value of HCV viral load was determined for the study population but instead for the proportion of patients with HCV viral load <1 000 000 copies/mL. Liver biopsies were interpreted by the local pathologists, and necroinflammation and fibrosis were each rated on a scale from 0 to 4 with a fibrosis score of 3 indicating bridging fibrosis and 4 indicating cirrhosis [9]. The estimated time of HCV infection was determined by the first time a risk factor occurred (injection drug use, transfusion, high-risk sex, etc.). Duration of HCV infection was the time from this estimated initial infection to enrollment in the study. Race was recorded by self-identification of each patient.
Statistical analysis
Categorical variables were compared by chi-square analysis or Fisher's exact test where appropriate, and continuous variables were compared using Student's t-test. Logistic regression analyses were performed to determine association of baseline variables with SVR, early discontinuation, and dose reductions of interferon or RBV. Variables that had an association on univariate analysis with P < 0.10 and no collinearity with other factors were also examined for independent association by multivariable logistic regression analysis.
|
|
| |
| |
|
|
|
