A Multicenter Controlled, Randomized, Open Trial of Interferon α2b Treatment of Anti–Human Immunodeficiency Virus-Negative Hemophilic Patients With Chronic Hepatitis C

MOST HEMOPHILIC PATIENTS, who in the past have been treated with large-pool plasma-derived clotting factor concentrates, are persistently infected with the hepatitis C virus (HCV), a condition that entails a high risk of morbidity and mortality.1,2 In 15% to 25% of nonhemophilic patients with chronic hepatitis C, interferon α therapy has been shown to suppress HCV replication and normalize serum alanine aminotransferase (ALT) long-term.3-5 Whether this also happens in hemophilic patients is unclear. The few, small pilot studies in hemophilia performed thus far have given disappointingly low rates (0%-8%) of sustained virologic responses to interferon.6-8 Treatment failures were thought to result from specific epidemiologic and virologic characteristics of hemophilic patients, as well as from coinfection with the human immunodeficiency virus (HIV-1). HIV-1 is present in 30% to 50% of the patients with hepatitis C, and this might weaken the anti-HCV activity of interferon α by impairing host immunity, thus facilitating multiplication of HCV and hastening the patients' deaths.7,9-11 

A demonstration that interferon α can stop hepatitis C in the hemophilic patients who are not coinfected with HIV-1 would have important clinical implications, since in many hemophilia centers patients with hepatitis C uninfected with HIV-1 outnumber those who are concurrently infected with HCV and HIV-1.12-14 We have carried out a multicenter, randomized, controlled trial to assess the efficacy and tolerability of interferon α2b in anti-HIV seronegative hemophilic patients with chronic hepatitis C. We chose to treat patients for 12 months instead of the standard 6-month period, because several studies in nonhemophilic patients have suggested that prolonging interferon therapy should increase the rates of responses to treatment.5,15 A subsidiary aim of this study was to assess whether HCV genotype and/or pretreatment serum HCV-RNA might predict response to interferon.

Study design.This multicenter trial was designed as a randomized, controlled, open study. Nine Italian Centers participated: “Angelo Bianchi Bonomi” Hemophilia and Thrombosis Center, University of Milan; Department of Hematology and Hemophilia Center, Careggi Hospital, Florence; Department of Biopathology, La Sapienza University, Rome; Department of Hematology, Nuovo Pellegrini Hospital, Naples; Hemophilia and Thrombosis Center, Policlinico Hospital, Bari; Department of Internal Medicine, University of Pavia; Hemophilia Center, Hematology Division, Pugliese Hospital, Catanzaro; “E. Tosatti” Hemophilia Center, Castelfranco Veneto Hospital; Pediatric Division, G. Gaslini Hospital, Genoa. The study conformed to international standards of good clinical practice and was approved by the Institutional Review Board of each hospital. All patients gave their written informed consent according to the Declaration of Helsinki.

Between December 1992 and February 1994, 107 patients were enrolled. Random assignment was conducted by a computer-generated list kept by the principal investigator (MC) blind to the clinical data. Patients were randomly assigned to 3 million units of recombinant interferon α2b (Intron A, Schering-Plough, Milan, Italy) by subcutaneous injections thrice weekly for 6 months or to no therapy. Patients showing a biochemical response to treatment continued to take the same dose of interferon for 6 additional months (total duration of treatment = 12 months). Patients who did not respond to therapy at month 6 or who had a permanent breakthrough during therapy discontinued interferon but continued with follow-up until the end of the study. All patients, treated and untreated, were followed up for at least 12 months after cessation of treatment (total duration of follow-up: 24 months).

Patient selection.Patients with congenital bleeding disorders (Hemophilia A and B, von Willebrand disease) were included if they were older than 14 years and had serum anti-HCV, HCV-RNA detected by RT-PCR method and ALT values greater than two times the upper limit of normal range, on three consecutive occasions 6 months apart before enrollment. Patients were not included if they had serum HBsAg or HBV-DNA in the presence of isolated anti-HBc. Additional criteria for exclusion were serum anti-HIV, serum organ and nonorgan specific autoantibodies, white blood cell counts less than 3,000/mm³, platelet counts less than 100,000/mm³, abnormal tests of thyroid function, decompensated cirrhosis, liver cell cancer, an ethanol intake greater than 80 g/die, or previous treatments with either interferon or steroids.

Patient monitoring.The following parameters were assessed before beginning treatment: anti-HCV, HCV-RNA, HBsAg and HBV-DNA, liver function tests, urea, creatinine, complete blood counts, and abdominal ultrasound. In the same pretreatment sera, quantitative serum levels of HCV-RNA, virus genotypes, and HGV-RNA were assessed retrospectively. RT-PCR for HCV-RNA and abdominal ultrasound were repeated at the end of treatment (month 12) and follow-up (month 24). Liver function and routine chemistry tests were performed at monthly intervals during the 12 months of therapy, then at 3-month intervals until the end of follow-up.

Definition of response.A primary response to treatment was complete normalization of serum ALT and aspartate aminotransferase (AST) values during the first 6 months of interferon administration. Cases with partial reduction in transaminases were considered nonresponders. In the primary responders, a breakthrough was any increase in serum transaminases occurring during interferon administration. Nonresponse was persistence of serum transaminase values above the normal limits during the first 6 months of therapy. A complete response was a negative RT-PCR result for serum HCV-RNA and normal transaminase values at the end of the 12-month treatment period. Patients showing a complete response at the end of the treatment period and at the end of the follow-up (month 24) were considered to have sustained complete response. A relapse was defined as an increase in serum transaminase values above the upper limit of normal occurring at any time after treatment had been completed.

Measurements.ALT and AST were measured by automated optimized colorimetry at 37°C. Commercially available enzyme immunoassays were used to determine serum HBsAg, anti-HBs, anti-HBc, anti–HIV-1 (Abbott Laboratories, Chicago, IL), anti-HCV (EIA-2 Ortho Diagnostic System, Raritan, NJ), and levels of thyrotropin (Boehringer Mannheim, Mannheim, Germany). Factor VIII inhibitor was assayed by the Bethesda method and expressed as BU/mL. Antibodies to nuclear, smooth muscle, mitochondrial, and liver-kidney microsomal antigens were detected on rat liver and kidney cryostat sections by immunofluorescence. HBV-DNA was detected by dot-blot hybridization, with a previously described method.16 HCV-RNA was detected by nested RT-PCR, using specific primers from the 5′ noncoding region as previously described.17 The minimum detectable level of HCV-RNA was calculated to be approximately 50 virus particles/mL. HCV-RNA was quantitated in the pretreatment serum specimens with a branched DNA signal amplification assay (bDNA, Quantiplex, HCV-RNA 2.0 Assay; Chiron Co, Emeryville, CA).18 The sensitivity threshold for bDNA was calculated to be 0.2 million equivalents per mL (MEq/mL). HCV genotypes were identified by nested PCR amplification with type-specific primers of the core and NS5b regions.19 Newly designed type-specific primers in both core and NS5b regions were used in the second round of PCR. HGV-RNA was detected in serum by RT-PCR amplification with nested primers from the highly conserved 5′ untranslated region.20 First-strand cDNA synthesis was primed with G362 antisense primer (5′-GTC AAG AGA GAC ATT GAA GGG-3′, nt 362 to 342). The resulting cDNA was amplified by PCR with primers G97 (5′-GAC CGG CCA AAA GGT GGT GG-3′, nt 97 to 116) and G362. The product from the first amplification was subjected to a second round of nested PCR amplification using primers G120 (5′-GGT GAT GAC AGG GTT GGT AG-3′, nt 120 to 139) and G340 (5′-GAC GTG GAC CGT ACG TGG G-3′, nt 340 to 322). Analysis was performed in duplicate for each specimen. Specimens that gave positive results were confirmed by RT-PCR analysis with two sets of nested primers derived from the putative NS3 helicase region.

Cirrhosis was established on the basis of clinical signs of portal hypertension (platelets fewer than 150,000 mm³, albumin less than 3.5 g/L, serum cholinesterase activity less than 4,500 U/L), endoscopic signs (presence of esophageal varices, portal hypertensive gastropathy), and/or by abdominal ultrasound (irregular margins of the liver, dilated portal vein axis, splenomegaly).

Statistical analysis.Continuous variables were expressed as mean values and standard deviations. Categorical variables were expressed as frequency and percent values. Univariate analysis was applied to each variable to assess any association with response to treatment. Chi-square analysis or Fisher's exact tests were used for categorical variables. Wilcoxon rank-sum test was used for continuous variables. The median values were compared by the median test for two samples.

Baseline characteristics of the patients.Two thousand four hundred eleven patients attended the nine participating centers. Fifty-eight percent (1,398) were serum anti-HCV positive but anti-HIV negative. Of these, 435 (18% of total) had ALT values greater than two times the upper limit of normal range on three consecutive checkups (Table 1). One hundred seven patients were eligible for the study; 101 accepted to be enrolled. The patients' characteristics at entry are summarized in Table 2. Fifty patients were allocated to treatment, and 51 to the untreated group. At the start of therapy, the two groups were comparable in terms of age, severity of clotting defect, serum levels of ALT, prevalence of cirrhosis, genotype distribution, and prevalence of HGV-RNA. By converse, treated patients were found to have higher pretreatment viremia than control patients (1.5 vs 1.1 MEq/mL, P < 0.03). Of the 101 patients included, five in the treatment group and one untreated control dropped out. Reasons for dropping were onset of severe flu-like symptoms on day 15 (1 case); refusal to continue therapy (3 cases) at months 1, 3, and 4, and interrupted follow-up at months 6 and 15. Only one of these patients showed a primary response. The characteristics of the 6 patients who dropped out from the study were similar to those of the 95 patients (45 treated, 50 controls) who completed the study.

Response to treatment.Twenty (44%) treated patients had a primary response to interferon; however, five (11%) of them had breakthroughs 4 to 10 months after the onset of treatment. Thus, only 15 (33%) patients had a biochemical persistent response to interferon. Considering the five patients with breakthrough, 30 (67%) patients were nonresponders. The corresponding figures for the 50 untreated patients were 2 (4%) primary responses, 0 breakthroughs, and 48 (96%) nonresponses. At the end of the treatment, 12 (27%) treated patients and none of the untreated controls had a complete biochemical and virologic response. At the end of the follow-up, six (13%) treated patients and none of the controls had sustained complete responses. All the differences in the response rates between treated and untreated patients were statistically significant (Table 3).

Adverse effects.During the study, interferon-related adverse events were sought for by both open inquiry and by inquiry about specific effects. One (2%) patient dropped out of the study following grade 3 fever at month 1 and four (8%) required dose reduction for thrombocytopenia, fatigue, and leukopenia, between month 3 and month 6. An influenza-like syndrome was reported by 82% of the patients accompanied by headache in one, myalgias in four, and arthralgias in one. One patient had mild psychological depression and one developed alopecia.

Correlation studies.The characteristics of the six patients who showed sustained complete responses to treatment and those of the 39 who transiently responded or did not respond at all are summarized in Table 4. Univariate analysis revealed a correlation between pretreatment viremia and response to therapy (0.5 MEq/mL v 2.0 MEq/mL, P < .01).

Anti-HIV negative hemophilic patients with chronic hepatitis C are at risk of significant liver-related morbidity and mortality. Hepatomegaly or clinical signs of advanced liver disease were detected in at least one third of all anti-HIV negative hemophiliacs with HCV who were followed-up for longer than 10 years.9,21 In a cohort study of 138 patients, 22 years after HCV infection, 19% developed cirrhosis and 9% developed liver failure.22 Hepatitis C progressed to end-stage liver disease in both the 36 HIV-infected patients and the 102 HIV-negative ones.22 Liver disease was a major cause of morbidity and mortality in HIV-uninfected hemophilic patients.

This is the first randomized controlled study of interferon therapy that has enrolled exclusively hemophilic patients with chronic hepatitis C not coinfected with HIV-1. We chose these patients to avoid potential confounding with HIV-related treatment failures. Among other factors, HIV-1 is thought to be implicated in the poor treatment outcomes observed in the previous studies in hemophiliacs.6-8 Moreover, in this study for the first time, interferon treatment was extended to 12 months instead of the standard period of 6 months, since previous studies in nonhemophilic patients showed that increasing the duration of treatment might also increase the chances of a sustained response.5,15 

In our study, interferon treatment of anti-HIV seronegative hemophiliacs with chronic hepatitis C led to sustained clearance of serum HCV-RNA and normalization of ALT values in six (13%) patients, compared with none in the control group, a difference that was statistically significant (P < .01). However, these figures are lower than those reported in studies of nonhemophilic patients treated with a similar schedule of interferon. A recent study of 61 Italian nonhemophilic patients treated for 1 year with 3 million U of interferon-α showed 25% rates of sustained complete responses.5 

The low rates of sustained responses to interferon that we observed suggest that factors other than HIV-1 may interfere with treatment outcome in hemophilia. In our patients, poor responsiveness to interferon could be due to excess virus load as is suggested by the pretreatment levels of viremia that we detected with the bDNA assay. In agreement with a previous study, we found that hemophilic patients had fairly high levels of HCV-RNA (1.3 MEq/mL), which may result from several factors, including impaired patient immunocompetence due to repeated exposures to allogenic serum proteins23 and accumulation of multiple virus strains (quasispecies), resulting from repeat infusions with infected clotting factor concentrates.24,25 Unfortunately, the small number (six) of patients showing sustained responses to interferon prevents us from assessing with multivariate analysis the potential predictive power of viremia. However, in at least three studies in nonhemophilic patients with chronic hepatitis C, multivariate analysis showed a clear-cut correlation between high pretreatment viremia and increased risk of nonresponse to interferon.26-28 

Poor responsiveness to interferon in hemophiliacs could be due also to the long duration of HCV infection, another well-recognized predictor of poor response in nonhemophilic patients.29 Our patients were 33 years old, on the average, suggesting a corresponding long duration of their HCV infections, because it is very likely that HCV was first transmitted at the time of the first infusion with unsterilized factor concentrates.30 In these patients, the high prevalence (63%) of genotype 1 infection could be another factor accounting for hyporesponsiveness to interferon. This HCV strain is less sensitive to interferon than types 2 and 3.31-33 At present, it is not clear whether genotype 1 differs in terms of intrinsic sensitivity to interferon from other HCV genotypes, or whether it preferentially segregates with more prolonged infections and/or more advanced liver disease, another factor known to be associated with poor responsiveness to interferon.34 

There was no correlation between response to interferon and coinfection with HGV, a novel member of the Flaviridae family, which can be cotransmitted with clotting factor concentrates.20 In our patients, the overall prevalence of serum HGV-RNA was low (8%), suggesting that this virus plays a marginal role in the hepatitis of hemophilic patients.

Finally, the low rates of primary responses (30%) observed in our patients could have been due to inadequate initial doses of interferon therapy, which in turn could be another factor accounting for unsatisfactory long-term therapeutic results. However, results of studies in nonhemophilic patients suggest that the advantage of administering initial high doses is to increase durability of the responses, not to recruit more primary responses. In a prospective study, patients treated with initial doses of 6 million U of interferon α for 4 months followed by a 3 million U dose for 8 months had similar rates (73%) of primary responses but higher rates of sustained responses (42% vs 25%) than patients treated with the same treatment schedule used for our patients.5 

In our study, treatment was well tolerated; none of the patients developed inhibitors to the deficient clotting factors and only one (2%) treatment had to be interrupted because of severe flu-like symptoms occurring during the first 2 weeks of the study. For four patients (8%) the dose had to be reduced by 50% between months 3 and 6 because of the onset of moderate leukopenia, thrombocytopenia, or fatigue. These figures overlap with those for nonhemophilic patients treated by the same schedule of therapy.35 

In conclusion, this study demonstrates that only 13% of anti-HIV seronegative hemophilic patients benefited long-term from prolonged interferon therapy. Since hemophilic patients with hepatitis C who escaped HIV infection have a long prospect of survival, they should be offered all chances for suppressing HCV. One approach for obtaining this could be to increase the durability of the response by administering high initial doses of interferon and treating with more cycles of interferon those patients in whom hepatitis relapsed after discontinuation of treatment.

Members of the Hepatitis Study Group are as follows: A. Alberti, Clinica Medica 2, University of Padova; N. Ciavarella, M. Schiavoni, Hemophilia and Thrombosis Center, Policlinico Hospital, Bari; R. de Biasi, A. Rocino, Department of Hematology, Nuovo Pellegrini Hospital, Naples; G. Gamba, Department of Internal Medicine, University of Pavia; P.G. Mori, Pediatric Division, G Gaslini Hospital, Genoa; G. Iannaccaro, G. Muleo, Hemophilia Center, Pugliese Hospital, Catanzaro; A. Russo, Epidemiology Unit, Aviano Cancer Center; R. Soffredini, Institute of Internal Medicine, University of Milan, IRCCS Maggiore Hospital, Milan.