Antiviral therapy for chronic hepatitis C in patients with inherited bleeding disorders: an international, multicenter cohort study

Authors


Eveline P. Mauser-Bunschoten, Van Creveldkliniek, University Medical Center Utrecht, C.01.425, Heidelberglaan 100, 3508 GA, Utrecht, the Netherlands.
Tel.: +31 30 250 8450; fax: +31 30 250 5438; e-mail: e.mauser-bunschoten@umcutrecht.nl

Abstract

Summary. Background: Hepatitis C is a major co-morbidity in patients with hemophilia. However, there is little information on the efficacy of antiviral therapy and long-term follow-up after treatment.Objectives: To assess the effect of interferon-based (IFN-based) therapy on hepatitis C virus (HCV) eradication, to identify determinants associated with treatment response, and to assess the occurrence of end-stage liver disease (ESLD) after completing antiviral therapy.Patients and methods: In a multicenter cohort study, 295 treatment-naïve hemophilia patients chronically infected with HCV were included. The effect of therapy was expressed as sustained virological response (SVR). Determinants associated with treatment response were expressed as odds ratios (ORs). Cumulative incidence of ESLD was assessed using a Kaplan–Meier survival table.Results: Among human immunodeficiency virus (HIV) negative patients (n = 235), SVR was 29% (29/101) for IFN monotherapy, 44% (32/72) for IFN with ribavirin, and 63% (39/62) for pegylated IFN (PegIFN) with ribavirin. In patients co-infected with HIV (= 60), IFN monotherapy, IFN with ribavirin, and PegIFN with ribavirin eradicated HCV in 7/35 (20%), 1/2 (50%), and 11/23 (48%), respectively. SVR increased with genotype 2 and 3 [OR 11.0, 95% CI: 5.8–20.5], and combination therapy (IFN and ribavirin OR 3.7, 95% CI: 1.7–8.4), PegIFN and ribavirin (OR 4.2, 95% CI: 1.8–9.5). Up to 15 years after antiviral treatment, none of the patients with a SVR relapsed and none developed ESLD. In contrast, among unsuccessfully treated patients the cumulative incidence of ESLD after 15 years was 13.0%.Conclusions: Successful antiviral therapy appears to have a durable effect and reduces the risk of ESLD considerably.

Introduction

Prior to the introduction of adequate hepatitis C virus (HCV) elimination techniques during the manufacture of clotting factor products (concentrates and cryoprecipitates), almost 100% of patients with inherited bleeding disorders were infected with HCV [1,2]. Approximately 80% of HCV-infected patients develop chronic hepatitis C, which is a major cause of morbidity and mortality in patients with inherited bleeding disorders [3–6].

In 1987, interferon monotherapy became available, resulting in normalization of aminotransferases or eradication of HCV in 10–20% of patients treated [7–9]. From 1995, the virological response markedly improved with the addition of ribavirin to IFN; 30–40% of patients became HCV RNA negative after 24–48 weeks of therapy [8,9]. The introduction of pegylated IFN (PegIFN) in the year 2000 in combination with ribavirin resulted in further improved HCV eradication in 50–60% of patients treated [10,11].

The reported efficacy of IFN is lower in patients coinfected with human immunodeficiency virus (HIV) with responses ranging from 16% for IFN monotherapy to 25–40% for combination therapy with (Peg-) IFN and ribavirin [12,13].

Only a few trials with IFN have been conducted in patients with inherited bleeding disorders and results appeared to be similar to those in the general population [14–17]. However, most of these studies have been limited by a small number of patients treated, a short follow-up, and experience with PegIFN and ribavirin has not yet been reported in peer reviewed publications. Data on response to IFN-based therapies are of importance for both physicians and patients, as the majority of hemophilia patients with chronic hepatitis C have received no or suboptimal treatment [18]. In the present multicenter study, the response to initial IFN-based therapy in a large cohort of patients with inherited bleeding disorders was assessed and determinants associated with response were identified. In addition, patients were followed post-treatment in order to assess the long-term response of antiviral therapy and its impact on the development of end-stage liver disease (ESLD).

Patients and methods

Study design and participants

All HCV RNA positive patients with inherited bleeding disorders who have ever been treated with IFN monotherapy, IFN and ribavirin, or PegIFN and ribavirin at three large hemophilia treatment centers were included in this analysis. The first anti-HCV treatment of the present study started in May 1987 and the treatment of the last patient ended in December 2005. The patients are registered at the hemophilia treatment centers of Sheffield (Royal Hallamshire Hospital, UK), London (Royal Free Hospital, UK), and Utrecht (Van Creveldkliniek, the Netherlands). Patients were seen at least annually. The study protocol was in accordance with the Helsinki Declaration and was approved by the institutional review boards of all three hospitals. Informed consent was obtained from all participants.

Laboratory methods

HIV antibodies, HCV RNA, and HCV genotype testing were performed in the local laboratories as described previously [4,5,19].

Date of HCV infection

Onset of infection was estimated as time of first treatment with inadequately (viral) or non-virus inactivated clotting factor products (concentrates or cryoprecipitate) and has been shown to be very reliable in other studies [19–21]. For 53 patients (18%) the precise date of first exposure was unknown and the date of July 1977 (London), January 1972 (Sheffield), or January 1970 (Utrecht) was taken. These different dates correspond to the median date of introduction of clotting factor in the different centers. For patients born after these dates it was set at the date of their first birthday.

Outcomes

The main outcome was sustained virological response (SVR), defined as the absence of HCV RNA in serum at the end of treatment and six months after completing therapy, in treatment-naïve patients. Before HCV RNA testing became available, the response to IFN was defined as normalization of aminotransferases: alanine aminotransferase (ALT) and aspartate aminotransferase (AST). In the present study, 18 patients were classified as sustained virological responders initially based on normalization of aminotransferases. SVR in all 18 patients was later confirmed by negative HCV RNA tests. A relapse was defined as the absence of HCV RNA at the end of treatment, but positive six months later. Non-responders were defined as HCV RNA positive, both at the end of treatment and six months later. Discontinuation was defined as prematurely ending of treatment, because of adverse events.

ESLD was defined as the occurrence of liver failure (ascites, bleeding esophageal varices, hepatic encephalopathy), hepatocellular carcinoma, or liver-related death.

Data analysis

The analyses were performed for responses to initial IFN-based therapy. In order to identify determinants associated with SVR after antiviral therapy in treatment-naïve patients, multivariate logistic regression was used. Coefficients were expressed in odds ratios (ORs) and 95% CIs. All determinants yielding an OR significant at the 0.20 level in the univariate analysis were included in the subsequent multivariate analysis. Determinants considered were as follows: gender, age at HCV infection, genotype (1, 4, and 5 vs. 2 and 3), duration of infection, age at start first antiviral therapy, history of alcohol abuse, treatment regimen (IFN monotherapy, IFN and ribavirin, and PegIFN and ribavirin), HIV co-infection, and baseline ALT and AST levels. Alcohol abuse was defined as intake of more than 20 units per week. Genotype was not available for 15% of patients. Therefore, we used imputation techniques (missing value analysis) to predict the missing genotypes using regression models. We performed the analyses both without those patients and with imputation of missing genotype. Because the results were similar, further analyses were performed with imputed genotype. HCV load was only available in 60% of patients and was therefore not used in the analyses.

In order to estimate the cumulative incidence of ESLD since the end of treatment, the Kaplan–Meier survival table method was used. For this analysis we compared 147 patients with a SVR after antiviral therapy (i.e. 119 with a SVR after first treatment, and an additional 28 with a SVR after re-treatment) with 148 who failed to antiviral therapy.

All statistical analyses were performed by using spss version 12.0 (SPSS Inc., Chicago, IL, USA).

Results

Patient characteristics

The cohort consisted of 295 patients with chronic hepatitis C, predominantly males with hemophilia (Table 1). The median age at first antiviral therapy was 36 years (range: 12–68) and 43 years (range: 19–84) at the end of follow-up. The median duration of infection at the start of initial IFN-based treatment was 22 years (range: 2–41). Coinfection with HIV was present in 60 patients (20%) and coinfection with hepatitis B in 10 (3%).

Table 1.   Patient characteristics
Patients295
  1. IFN, interferon; HCV, hepatitis C virus; HIV, human immunodeficiency virus; ALT, alanine aminotransferase; AST, aspartate aminotransferase.

  2. *End of follow-up is last clinical evaluation or death. At start of initial IFN-based treatment.

  3. Note that values are medians (ranges) or numbers (%)

Diagnosis
 Hemophilia A224 (76%)
 Hemophilia B50 (17%)
 von Willebrand disease15 (5%)
 Other6 (2%)
Male283 (96%)
Age (years)
 At infection13 (<1–56)
 At initial IFN treatment36 (12–68)
 At end of follow-up*43 (19–84)
Duration of infection (years)22 (2–41)
HCV genotype
 1146 (50%)
 242 (14%)
 358 (20%)
 42 (1%)
 54 (1%)
 Unknown43 (15%)
HIV coinfection60 (20%)
ALT (U L−1)82 (16–1091)
AST (U L−1)48 (12–94)

Response to initial antiviral therapy: role of HIV coinfection and HCV genotype

In total, 136 patients have been treated with IFN monotherapy, 74 patients with IFN and ribavirin, and 85 with PegIFN and ribavirin (Table 2). The median duration of antiviral therapy was 0.5 years (range: 0.1–7.8). 119 of 295 patients (40%) achieved a SVR after their initial antiviral therapy. In HIV-negative patients, response to antiviral therapy ranged from 29% for IFN monotherapy to 63% for PegIFN and ribavirin. In patients coinfected with HIV, the response was lower: 20% for IFN monotherapy and 48% for PegIFN and ribavirin.

Table 2.   Proportion of patients with sustained virological response to initial antiviral therapy
 HIV negativeHIV positiveTotalFollow-up (years)*
  1. IFN, interferon; HCV, hepatitis C virus; HIV, human immunodeficiency virus; PegIFN, pegylated IFN.

  2. *Median follow-up (range) after end of first treatment.

IFN29/101 (29%)7/35 (20%)36/136 (26%)9 (1–15)
IFN and ribavirin32/72 (44%)1/2 (50%)33/74 (45%)5 (1–8)
PegIFN and ribavirin39/62 (63%)11/23 (48%)50/85 (59%)1 (1–5)
Total100/235 (42%)19/60 (32%)119/295 (40%)5 (1–15)

Among patients with genotype 1, 4 or 5, the SVR was 7% for IFN monotherapy, 29% for combination therapy of IFN and ribavirin, and 33% for PegIFN and ribavirin. In contrast, antiviral therapy was more effective in patients with genotype 2 and 3: SVR of 49% for IFN monotherapy, 76% for combination therapy of IFN and ribavirin, and 84% for PegIFN and ribavirin.

All patients who achieved a SVR remained HCV RNA negative up to 15 years post-treatment (median follow-up of five years; range: 1–15). Discontinuation of therapy because of adverse events was similar for both IFN monotherapy and combination therapies at 11%, irrespective of HIV status (10% for HIV-negative patients and 13% for HIV-positive patients).

Determinants associated with response to initial antiviral therapy

In the univariate analysis, a younger age at HCV infection, longer duration of infection, presence of genotype 2 and 3, and combination therapy were associated with a higher SVR (Table 3). In the multivariate analysis (Table 4) determinants independently associated with SVR were the presence of genotype 2 and 3 (OR 11.0, 95% CI: 5.8–20.5) and the use of combination therapy [IFN vs. IFN and ribavirin (OR 3.7, 95% CI: 1.7–8.4) and IFN vs. PegIFN and ribavirin (OR 4.2, 95% CI: 1.8–9.5)]. An older age at infection, presence of HIV co-infection, and history of alcohol abuse appeared to be associated with a lower likelihood of SVR; however, this was not statistically significant at P = 0.05.

Table 3.   Univariate analysis of sustained virological response after initial antiviral therapy
DeterminantOdds ratio (95% CI)P-value
  1. IFN, interferon; HIV, human immunodeficiency virus; ALT, alanine aminotransferase; AST, aspartate aminotransferase.

  2. *At start of therapy.

Gender
 Male1.0 
 Female1.1 (0.3–3.4)0.92
Age at infection (per 10 years)0.7 (0.6–0.9)0.01
Duration of infection at start therapy (per 10 years)2.1 (1.5–3.1)<0.01
Age at start therapy (per 10 years)1.0 (0.8–1.2)0.71
Genotype
 Type 1, 4 and 51.0 
 Type 2 and 37.6 (4.5–12.8)<0.01
Treatment regimen
 IFN monotherapy 1.0 
 IFN and ribavirin2.2 (1.2–4.1)0.01
 PegIFN and ribavirin4.0 (2.2–7.1)<0.01
History of alcohol abuse
 No1.0 
 Yes0.5 (0.2–1.3)0.15
HIV
 No1.0 
 Yes0.6 (0.3–1.1)0.13
ALT* (per 100, U L−1)1.1 (0.9–1.4)0.43
AST* (per 100, U L−1)0.8 (0.6–1.3)0.41
Table 4.   Multivariate analysis of sustained virological response (SVR) after initial antiviral therapy
DeterminantOdds ratio (95% CI)P-value
  1. Interpretation: after adjustment for all other determinants in the multivariate model, treatment with interferon (IFN) and ribavirin resulted in a 3.7 higher chance of SVR than IFN monotherapy.

  2. PegIFN, pegylated IFN; HIV, human immunodeficiency virus.

  3. *At start of therapy.

Age at infection (per 10 years)0.8 (0.6–1.0)0.07
Duration of infection (per 10 years)1.3 (0.7–2.2)0.32
Genotype
 Type 1, 4 and 5 1.0 
 Type 2 and 311.0 (5.8–20.5)<0.01
Treatment regimen
 IFN monotherapy 1.0 
 IFN and ribavirin3.7 (1.7–8.4)<0.01
 PegIFN and ribavirin4.2 (1.8–9.5)<0.01
History of alcohol abuse
 No1.0 
 Yes0.3 (0.1–1.1)0.08
HIV
 No1.0 
 Yes0.6 (0.3–1.2)0.15

In other words, after adjustment for age at infection, presence of HIV, history of alcohol abuse, duration of infection, and genotype, treatment with IFN and ribavirin resulted in a 3.7 higher chance of a SVR than treatment with IFN monotherapy.

End-stage liver disease at the end of follow-up

Up to 15 years after completing antiviral therapy, none of the patients with a SVR developed ESLD. However, 12 patients, unsuccessfully treated, developed ESLD (10 liver failure, two hepatocellular carcinoma) a median of four years (range: < 1–15) after treatment and after a median duration of infection of 24 years (range: 18–36). The cumulative incidence of ESLD in these non-responders and relapsers was 13.0% 15 years after initial antiviral therapy and 35 years after infection (Fig. 1). After stratification for HIV among these patients without a SVR, the cumulative incidence of ESLD was 5.4% for HIV-negative patients and 23.7% for HIV-positive patients (P < 0.0001, log rank test).

Figure 1.

 Cumulative incidence of end-stage liver disease (ESLD) after antiviral therapy. 119 patients achieved a sustained virological response after first treatment and an additional 28 patients were successfully re-treated. None of these 147 patients developed ESLD. Of the 148 non-responders, 12 patients developed ESLD: all of these were treated with interferon (IFN) monotherapy and three of these were re-treated with combination therapy.

Re-treatment of previous non-responders and relapsers

One hundred seventy-six of 295 patients did not respond to or relapsed after initial IFN-based therapy. Of these, 71 (40%) were re-treated at least once with IFN monotherapy, or combination therapy of (Peg-)IFN with ribavirin. Eventually, re-treatment resulted in an overall SVR in 28 patients (39%).

Re-treatment with IFN and ribavirin after unsuccessful IFN monotherapy resulted in a SVR of 35% (12/34, 27% genotype 1, 4 or 5 vs. 50% genotype 2 or 3) and re-treatment with PegIFN and ribavirin in a SVR of 36% (5/14, 22% genotype 1, 4 or 5 vs. 60% genotype 2 or 3). Patients unsuccessfully treated with combination therapy of IFN and ribavirin achieved a SVR of 32% (6/19, 21% genotype 1, 4 or 5 vs. 60% genotype 2 or 3) when re-treated with PegIFN and ribavirin.

In total, 147 of 295 patients (49%) showed a sustained response after at least one IFN-based therapy.

Discussion

This is the first study describing the long-term responses to three IFN-based therapies in a cohort consisting of almost 300 hemophilic patients with chronic hepatitis C. Sustained response was similar to that seen in the non-hemophilic population and mainly associated with HCV genotype (2 and 3) and combination therapy (IFN/PegIFN and ribavirin) [8,11]. None of the sustained responders became HCV RNA positive or developed ESLD up to 15 years after antiviral therapy. In contrast, 15 years post-treatment, 13% of patients unsuccessfully treated developed ESLD.

To appreciate these findings, some limitations of the study have to be considered. First, direct comparisons between IFN monotherapy and combination therapy have to be interpreted with care. Initially, mainly patients with elevated aminotransferases were treated with IFN monotherapy. This might have resulted in a selection bias: patients who were treated with IFN monotherapy might have had more advanced liver disease with possibly lower response to IFN [22]. However, the duration of infection is probably shorter in these patients because IFN monotherapy was introduced in the late 1980s and combination therapy several years later [4,8]. Moreover, elevated ALT and AST levels were not associated with sustained response in our study and in the multivariate analysis we included duration of infection and age at infection as well, rendering bias less likely.

The response to antiviral therapy in HCV–HIV coinfected patients may also have been overestimated. The use of IFN in this population started in the 1990s and, at that time, many of the HIV–HCV coinfected patients had died, resulting in a group of ‘survivors’ [23]. Almost all of these patients started highly active antiretroviral therapy (HAART) with concomitant lower HIV viral load and slower fibrosis progression rates, and possibly better response to IFN [24,25].

Interestingly, we found that an older age at infection appeared to be associated with a lower response to antiviral therapy (although not statistically significant at P = 0.05). However, neither age at start of therapy nor duration of infection appeared to be associated with SVR to IFN. The effect of age at infection has been previously described in a small study of 20 hemophilia patients treated with IFN and ribavirin [26]. The mechanisms underlying the reduced response to treatment in patients infected at an older age remain elusive, but are probably multifactorial. From studies on the natural history of HCV infection, it is known that patients infected at an older age developed ESLD more frequently [4–6]. Possibly, these patients had more severe fibrosis or cirrhosis at the start of therapy, resulting in less effective treatment compared to those infected at a younger age.

In the present study we found that all patients with a SVR remained HCV RNA negative up to 15 years post-treatment. This corroborates the results of others in the non-hemophilic population and indicates the beneficial and long-lasting effects of successful antiviral therapy [27,28]. Furthermore, ESLD did not occur in patients with a SVR, whereas the cumulative incidence was 13% in those unsuccessfully treated 15 years after completing therapy, underscoring the importance of antiviral therapy. However, there are still many hemophilic patients with chronic hepatitis C who have not been treated [18]. Current guidelines recommend treatment in all patients with HCV genotype 2 and 3 [29]. In contrast, because of lower SVR and consequently a longer duration of treatment, the assessment of liver fibrosis is recommended in patients with genotype 1, 4, 5, and 6 before considering treatment. Patients with at least portal fibrosis and septa on liver biopsy (F ≥ 2) are eligible for treatment. However, biopsies are rarely performed in the hemophilic population because of possible bleeding complications and high costs of clotting factor replacement therapy, resulting in an unknown grade of fibrosis in most patients. Recently, a new device (Fibroscan®) has been developed that measures stiffness of the liver, showing high correlation between liver stiffness and stage of fibrosis in biopsies, as well as excellent receiver operating curves [30,31]. This technique may replace liver biopsy in the hemophilic population, and may offer an alternative way of selecting which patients should be treated with antiviral therapy [32].

There is no consensus on re-treatment of patients in whom previous IFN-based therapy has failed. However, our data show that re-treatment was successful in 39% of patients. This is in accordance with others, who found that re-treatment of non-hemophiliacs with PegIFN and ribavirin was effective in 8–21% of previous non-responders and in 42–55% of previous relapsers [33–35]. These data suggest that re-treatment of patients with inherited bleeding disorders is worthwhile because of a reasonable chance of response and improved natural history if they eradicate HCV.

Currently, there are no treatment options for those who failed after PegIFN and ribavirin. New drugs are being developed (e.g. new IFN molecules, HCV RNA protease and polymerase inhibitors, and immune modulators), but it will take some years before these will become available [36].

In conclusion, IFN-based therapies for chronic hepatitis C are effective in a significant proportion of patients with hemophilia, especially when treated with PegIFN and ribavirin. Treatment responses appear to be similar to those seen in the general population. Improvements in antiviral therapy have led to better virological response over the last decade and have resulted in long-term clearance of HCV. Furthermore, it will prevent ESLD in a considerable proportion of patients.

Disclosure of Conflict of Interests

The authors state that they have no conflict of interest.

Ancillary