High-dose vitamin E supplementation does not diminish ribavirin-associated haemolysis in hepatitis C treatment with combination standard α-interferon and ribavirin


Dr K. Saeian, Division of Gastroenterology and Hepatology, Medical College of Wisconsin, 9200 W. Wisconsin Avenue, Milwaukee, WI 53226, USA.
E-mail: ksaeian@mcw.edu


Background : Ribavirin is associated with haemolytic anaemia. Antioxidants have been reported to decrease severity of this anaemia.

Aim : To determine effect of vitamin E supplementation on ribavirin-associated haemolysis in chronic hepatitis C treated with standard α-interferon and ribavirin.

Methods : Fifty-one naive chronic hepatitis C patients were randomized to receive either α-interferon/ribavirin therapy (control) or therapy plus vitamin E 800 IU b.d. with 24-week follow-up. Alanine aminotransferase ALT, haemoglobin and reticulocyte percentage were monitored. Symptoms and health-related quality of life were also monitored at each visit.

Results : Forty-seven subjects were treated (27 vitamin E /20 controls). Thirteen withdrew because of adverse effects or non-compliance. Groups were similar in demographics, genotype and baseline lab indices. Comparison with baseline, treatment and follow-up values showed a significant haemoglobin and ALT reduction in both groups. There was no significant difference in haemoglobin and reticulocyte percentage between groups. Sustained viral response was not significantly different between vitamin E (11/18) and control (6/16) groups. Three patients required ribavirin dose-reduction in the vitamin E group compared with two controls. Health-related quality of life during and end-of-treatment was not different between groups.

Conclusions : Vitamin E supplementation alone during standard α-interferon and ribavirin therapy does not appear to diminish ribavirin-associated haemolysis.


Adherence to anti-viral therapy for hepatitis C is a major factor in predicting a sustained response to treatment.1 Ribavirin-associated haemolysis, which frequently results in anaemia, may exacerbate symptoms of fatigue and shortness of breath, resulting in reduced patient adherence. This common side-effect of ribavirin is dose-related and is usually reversed when the therapy is discontinued.2 Efforts directed at minimizing ribavirin-induced haemolytic anaemia include dose reduction and the use of erythropoietin. While both approaches may be effective, the former may compromise anti-viral response and the latter often requires weeks before taking effect and is costly.

Vitamin E is a fat-soluble vitamin, predominantly located in cell membranes and is believed to play an integral role in non-enzymatic protection of fatty acids against oxidative damage.3 Vitamin E supplementation has been reported to have protean benefits including reducing the risk of cardiovascular disease and preventing malignant disease.4 Vitamin E has also been used in renal dialysis to diminish haemolysis based on purported stabilization of the red blood cell membrane.5

Interestingly, vitamin E levels are consistently diminished in patients with chronic liver disease.6 A number of small studies have investigated the role of vitamin E in liver disease.7–9 Beneficial effects of supplementation have included an enhanced anti-viral response in chronic hepatitis B and diminished lipid peroxidation, stellate cell activity and collagen gene expression in chronic hepatitis C.10, 11 One report in abstract form suggested a marked delay and slight diminution of ribavirin-induced haemolysis in chronic hepatitis C virus (HCV) patients receiving combined vitamin C and vitamin E (800 IU/day) supplementation during combination standard α-interferon and ribavirin therapy.12 However, while many advocate the potential benefits of antioxidants, including vitamin E, on ribavirin-induced haemolysis, no systematic studies have been published.

The aim of the current study was to determine whether supplementation with high-dose vitamin E diminished the degree of ribavirin-associated haemolysis and improved patient tolerance to therapy in previously untreated patients with compensated chronic hepatitis C undergoing therapy with standard α-interferon and ribavirin.

Materials and methods

This was a randomized trial of treatment-naïve patients with compensated chronic hepatitis C. Subjects were randomized to receive standard α-interferon-2b (3 million units subcutaneously three times a week) and ribavirin (weight-based dosing according to package insert; 1000 mg/day for weight <75 kg, 1200 mg/day for weight >75 kg) alone (control group) or with 800 IU b.d. of vitamin E (vitamin E group). The study was not blinded or placebo-controlled. Patients with viral genotypes 1 or 4 were treated for up to 48 weeks and those with genotypes 2 or 3 were treated for up to 24 weeks. Inclusion criteria included age >18 years, a liver biopsy consistent with chronic hepatitis C, a positive serum viral assay for hepatitis C and an elevated ALT or Aspartate aminotransferase (AST) level within the prior year. Only those with haemoglobin (Hgb) values of ≥12 g/dL for females or ≥13 g/dL for males, WBC ≥ 3000/mm3, neutrophil count ≥ 1500/mm3, platelet count ≥ 80 000/mm3, albumin ≥ 3.0 g/dL, INR ≤ 1.4, serum creatinine ≤ 1.2, antinuclear antibody ≤ 1:160 and an α-foetoprotein within normal limits within the prior year were included. Participants were required not to have used vitamin E supplements for at least 30 days prior to enrollment.

Patients with chronic liver disease other than chronic hepatitis C, haemoglobinopathies (e.g. thalassaemia) or any other cause of haemolytic anaemia or evidence of decompensated liver disease such as a history of or presence of ascites, bleeding varices or encephalopathy were excluded.

The Institutional Review Boards at all participating institutions approved the study protocol and written informed consent was obtained from all subjects.

At the entry visit, a complete medical history and physical examination was performed. In the medical history, antioxidant and dietary herbal supplement use were specifically queried. Use of other antioxidants, including sylimarin and vitamin C was prohibited. Baseline laboratory evaluation was performed including studies for haemolysis (Hgb and reticulocyte count). Hepatitis C associated symptom assessment score (ASA score), previously validated against SF-36 in hepatitis C patients,13 was used to assess health-related quality of life (HRQoL). Safety and tolerance were evaluated at weeks 1, 2, 4, 8 and then every 4 weeks during treatment and at weeks 4, 8, 12 and 24 following the end of therapy. Additional safety precautions and monitoring were implemented as indicated for patients at risk for cardiovascular and neuropsychiatric complications. Serum HCV RNA was evaluated prior to initiating therapy, at week 24, at the end of therapy, and 6 months after completion of therapy. End-of-treatment response (ETR) was defined as loss of detectable HCV RNA at the end of treatment, sustained viral response (SVR) was defined as loss of detectable HCV RNA by polymerase chain reaction (PCR) 24 weeks following the completion of therapy and non-response (NR) defined as failure to lose HCV RNA by PCR at 24 weeks. Endpoints were assessed based on intention-to-treat analysis.

Haemolysis was defined as a drop in Hgb of 2 gm/dL accompanied by an increase in reticulocyte count to at least 2%. The use of growth factors was not permitted in the study.

The study was terminated because of the availability of pegylated interferon and it was felt by the investigators that standard α-interferon was no longer standard of care.

Statistical analysis

Unpaired t-test was used to compare data between vitamin E and control groups and paired t-test was used to compare patients’ lab data with baseline values. Chi-squared test was utilized to compare proportion of dose reduction, SVR, ETR and NR across groups.

Calculation of the sample size was based on degree of haemolysis. Using the degree of haemolysis, initial experience with ribavirin indicated that the difference in Hgb drop at 4 weeks to be 1.3, at 12 weeks to be 0.4 with an overall average difference of 0.9.14 The expected sample size (target power of 0.8 and α value of 0.05) was 21 for the overall average difference in Hgb.


Seventy-one patients were initially considered for enrollment, however 20 patients did not meet inclusion criteria. Fifty-one patients were prospectively randomized to the control or vitamin E group between July 2000 and July 2001. Four patients never received therapy and were excluded from analysis. The baseline demographic and virologic data of the 47 patients (27 vitamin E and 20 control group) who had sufficient data for analysis are presented in Table 1.

Table 1.  Baseline demographic and virological data
 Control (n = 20)Vitamin E (n = 27)
Age (years ± s.d.)43.7 ± 8.346.0 ± 5.5
Gender (M/F)14/624/3
Genotype I/IV1110
Genotype II/III 917

Thirteen patients (nine vitamin E and four control group) withdrew from the study before 24 weeks because of adverse effects (seven patients) or non-compliance (six patients) with therapy or follow-up. They were included in the analysis of the haematologic endpoints of the study. Sixteen controls and 18 vitamin E patients completed treatment and 24-week post-treatment follow-up.

Dose reduction

A total of five patients (10.6%) required dose reduction in accordance with protocol guidelines for anaemia. Three of these subjects were in the vitamin E group and two in the control group (P = N.S.).


The results of Hgb and reticulocyte count for 24 weeks are shown in Table 2. On paired t-test analysis, there was a significant decrease in haemoglobin and increase in reticulocyte count throughout the duration of the study within both groups, which resolved on follow-up. Ten controls and 11 patients in the vitamin E group had evidence of haemolysis at the end of treatment, which was not statistically significant (P = 0.9; chi-squared test). Unpaired t-test analysis did not show any significant differences between reticulocyte count and haemoglobin between the vitamin E group and the control group at baseline, week 12 and end-of-treatment.

Table 2.  Comparison of haematologic and biochemical indices with baseline (paired t-test)
Mean ± S.E.M.Baseline (n = 47)Week 1 (n = 44)Week 4 (n = 43)Week 12 (n = 31)Week 24 (n = 34)
  1. Hgb, haemoglobin; retic, reticulocyte.

  2. P < 0.05.

Hgb (vitamin E) g15.5 ± 0.215.0 ± 0.313.2 ± 0.312.9 ± 0.3*12.7 ± 0.3*
Hgb (control) g14.9 ± 0.314.6 ± 0.313.0 ± 0.412.8 ± 0.4*12.4 ± 0.5*
Retic % (vitamin E)1.3 ± 0.11.1 ± 0.13.5 ± 0.5*2.7 ± 0.5*2.5 ± 0.4*
Retic % (control)0.98 ± 0.10.88 ± 0.22.0 ± 0.4*2.0 ± 1.4*2.4 ± 0.3*
ALT (vitamin E) (IU/L)86.5 ± 1260.6 ± 68*37.8 ± 4.8*30.9 ± 3.8*26.9 ± 4.2*
ALT (control) (IU/L)111.4 ± 18.881.7 ± 12.1*67.8 ± 18.5*34.1 ± 4.9*37.9 ± 12.8

Virologic and biochemical response

There was significant ALT reduction throughout the duration of study in both groups as compared with baseline on paired t-test analysis as shown in Table 2. This reduction was maintained in the vitamin E group, but not the control group on post-treatment follow-up. Unpaired t-test analysis did not reveal a significant difference in ALT between the two groups at baseline, 12 weeks, 24 weeks or at end-of-treatment. There was no significant difference in ETR or SVR between the two groups (Table 3). ETR and SVR were not related to presence of haemolysis as defined by our criteria.

Table 3.  Virologic response across both groups
GroupNon-responderEnd-of-treatment responseSustained viral response
Vitamin E7/1811/1811/18
Control8/16 8/16 6/16

Health-related quality of life and adverse effects

Baseline ASA scores were similar across both groups (vitamin E 24 ± 17; control 28 ± 20: P = 0.5). Similarly there was no significant difference in ASA score between the two groups after 24 weeks as compared with baseline (P = 0.2). There was insufficient data for analysis at 48 weeks for this group. Rates of adverse effects to the combination therapy were similar across both groups. No vitamin E-associated adverse effects were observed. There was no difference in compliance with therapy between the two groups.


Our results indicate that high-dose vitamin E supplementation in chronic hepatitis C patients undergoing combination therapy does not prevent ribavirin-associated haemolysis. In addition, vitamin E also does not affect patient compliance, ETR or SVR. To our knowledge this is the first published randomized clinical trial using vitamin E in chronic hepatitis C combination therapy.

Our patient group had significant overall decrease in haemoglobin for the duration of the study, which was reversible on 24-week follow-up as is consistent with previous experience.15 This drop in haemoglobin was accompanied by reticulocytosis indicating haemolysis. Ribavirin-associated haemolysis was evident at 4 weeks and continued until end-of-therapy in both groups. Vitamin E did not affect these parameters.

Patient compliance was not affected by vitamin E. We did not encounter any untoward side-effects attributed to vitamin E supplementation in this trial. There was no significant difference in combination therapy-associated adverse effects or withdrawal between the two groups of patients. The majority of patients were apparently compliant with combination therapy as evidenced by the viral clearance in a significant percentage and treatment-associated adverse effects. Determination of vitamin E levels could have added an objective assessment of vitamin E compliance; however, doses of vitamin E used were supraphysiologic and would have been sufficient to replenish any pre-existing deficiency in vitamin E stores.

Both controls and patients given vitamin E supplementation had similar virologic and biochemical responses to therapy. ALT levels appropriately decreased in patients who responded to combination therapy, but this effect was apparent in both the control and vitamin E group. ETR and SVR also did not change significantly with the addition of vitamin E in the experimental group. This suggests that vitamin E therapy might not be effective in bolstering response to standard α-interferon and ribavirin in chronic hepatitis C patients as opposed to its beneficial effect along with interferon in chronic hepatitis B.10 The higher proportion of genotype 2 and 3 patients in the control group may have influenced the ETR and SVR data. The relatively low completion rate (68% of those randomized) may also limit conclusions about virologic outcomes.

Vitamin E is a lipophilic antioxidant and its preferred storage in the fat compartment could have potentially prevented its full desired antioxidant effect on the erythrocytes. Therefore, further studies evaluating water-soluble antioxidants such as N-Acetyl Cysteine, alone or in combination with vitamin E are warranted. Vitamin C has been used in combination with vitamin E to improve antioxidant potential and bioavailability of vitamin E.12 However, vitamin C increases gut absorption of iron and excess iron deposition in the liver has been shown to diminish the response to combination therapy in chronic hepatitis C.16 Furthermore, the intent of the study was to detect the effect of isolated vitamin E supplementation. By using a high-dose of vitamin E (800 IU twice a day), adequate bioavailability was ascertained and, therefore, supplementation with vitamin C was not part of our trial.

We conclude that vitamin E does not affect ribavirin-associated haemolysis, compliance or virologic clearance with combination therapy with standard α-interferon and ribavirin in patients with chronic hepatitis C. Concomitant supplementation with vitamin E and other antioxidants, particularly water soluble agents, should be considered in future trials to evaluate antioxidant effect on ribavirin-associated haemolysis.


This work was supported in part by Schering-Plough Corporation, Kenilworth, NJ, USA.