Interferon-α and ribavirin treatment in patients with hepatitis C virus–related systemic vasculitis

Authors


Abstract

Objective

Hepatitis C virus (HCV)–related vasculitis may involve multiple organs, including the skin, kidneys, and nervous system, and may be life-threatening. Although HCV is increasingly recognized as a cause of systemic vasculitis, limited data are available regarding the optimal treatment of this potentially serious condition. Therefore, we retrospectively analyzed the response to treatment in patients with chronic hepatitis C complicated by systemic vasculitis who had received antiviral therapy with interferon-α (IFNα) and ribavirin.

Methods

This retrospective study included 27 patients with systemic vasculitis and chronic HCV infection. Each patient had received treatment with IFNα and ribavirin for at least 6 months. The response to antiviral treatment was analyzed by comparing clinical, immunologic, and virologic data at the time of entry and during followup. Clinical response was defined according to the evolution of weight, arthralgia, nervous system, renal system, and cutaneous involvement. The virologic and immunologic responses were defined by the absence of HCV RNA and the absence of cryoglobulinemia, respectively, both 6 months after stopping antiviral therapy and at the end of followup.

Results

Patients received IFNα for a mean ± SD of 20 ± 14 months and ribavirin (at a mean ± SD dosage of 895 ± 250 mg/day) for 14 ± 12 months. Other treatments included low-dose corticosteroids and plasma exchange. After a mean ± SD followup of 57 ± 29 months, 25 of 27 patients are alive and are being followed up as outpatients. Because of the heterogeneity of anti-HCV treatments received, the main results were stratified according to patients with 6 months of followup after stopping antiviral treatment (group 1, n = 14) and those who were still undergoing antiviral therapy at the time of analysis (group 2, n = 13). Nine patients in group 1 had a sustained virologic response and were clinical and immunologic complete responders. Four patients in group 1 were virologic nonresponders, and 3 of these patients had partial clinical and immunologic responses. Overall, 10 patients in group 1 had a complete clinical and immunologic response of their vasculitis (all 9 of the sustained virologic responders and 1 of the 5 patients who remained viremic). At the end of followup, 7 patients in group 2 were negative for HCV RNA; 6 were complete clinical responders. Among the other 6 patients in group 2, who had persistent viremia, 4 had a partial clinical response. Among the patients in group 1, HCV RNA was more often undetectable and genotype 1 was less frequent in complete clinical responders compared with partial/nonresponders. Age, sex, clinical vasculitic involvement, mean duration or total cumulative dose of IFNα or ribavirin, and use of steroids or plasmapheresis did not differ significantly according to clinical response.

Conclusion

Treatment with IFNα and ribavirin can achieve a complete clinical response in most patients with HCV-related systemic vasculitis. Complete clinical response correlates with the eradication of HCV.

Although the cause of most systemic vasculitides was unknown at the beginning of the 1950s, the prognosis for patients with systemic vasculitis improved dramatically with the introduction of corticosteroids (1). Later, other studies demonstrated that immunosuppressive treatment, usually with cyclophosphamide, improved the prognosis even further (2, 3). Although more efficacious, combination therapy with corticosteroids plus cyclophosphamide also increased the frequency of adverse events (3). A more rational therapeutic strategy in patients with systemic vasculitis related to an underlying infection (4, 5) is therapy aimed at the associated infection in combination with short-term, low-dose corticosteroids (6, 7). Such an approach has been used successfully in patients with vasculitis related to hepatitis B virus (HBV) and the human immunodeficiency virus (HIV) (6).

The hepatitis C virus (HCV) is a major problem worldwide and is the main cause of what was previously termed non A, non B hepatitis (8). Numerous extrahepatic manifestations have been associated with HCV infection, including mixed cryoglobulinemia, glomerulonephritis, porphyria cutanea tarda, sicca syndrome, and systemic necrotizing vasculitis (9–17). Chronic HCV infection may be associated with different types of vasculitis, including small-vessel vasculitis such as mixed cryoglobulinemia (MC), or vasculitis of larger-caliber vessels such as classic polyarteritis nodosa (PAN) (18).

Limited data are available regarding the treatment of patients with HCV-related systemic vasculitis. Interferon-α (IFNα), corticosteroids, immunosuppressive agents, and plasmapheresis have been proposed as treatments, but the optimal therapeutic regimen remains controversial (19–23). In patients who have chronic HCV infection without extrahepatic manifestations, combination therapy with IFNα plus ribavirin for 6–12 months has been shown to have much greater efficacy than IFNα monotherapy (24, 25). The effect of IFNα plus ribavirin in patients with HCV-related vasculitis has been reported in only a small series of 9 patients (26). After short-duration treatment and followup, only 2 of these patients had a virologic response. A clinical response was noted in all patients with skin involvement but in only half of the patients with nerve or renal involvement.

We now report the clinical and virologic features of 27 patients with chronic HCV infection complicated by systemic vasculitis, all of whom received antiviral therapy for HCV with IFNα plus ribavirin for at least 6 months. Clinical, immunologic, and virologic responses, therapeutic tolerance, and long-term followup are analyzed.

PATIENTS AND METHODS

We retrospectively studied the files of 50 patients at the departments of internal medicine and neurology of the Pitié-Salpêtrière Hospital, all of whom presented with signs of systemic vasculitis and chronic HCV infection. HCV infection was defined by the presence of anti-HCV antibodies (detected by third-generation enzyme-linked immunosorbent assay; Ortho Diagnostics, Cardiff, UK) and positive serum HCV RNA (detected by polymerase chain reaction [PCR]). Other causes of liver dysfunction (e.g., chronic infection with HBV, autoimmune hepatitis, primary biliary cirrhosis) were excluded using standard tests. The diagnosis of vasculitis was based on previously defined clinical and biologic criteria (27, 28). Inclusion criteria for the present study were as follows: 1) chronic HCV infection, 2) signs of systemic vasculitis in the absence of any alternative condition known to cause vasculitis, 3) treatment with IFNα and ribavirin for a minimum of 6 months, and 4) a minimum of 6 months of followup.

Among the 50 eligible patients, 23 were excluded for the following reasons: followup <6 months (n = 3), underlying malignant hematologic disorder representing an alternative possible cause of systemic vasculitis (lymphoma, n = 3; acute leukemia, n = 1), duration of antiviral treatment <6 months (n = 7), treatment with IFNα only (n = 3), no treatment (n = 3), and no treatment of HCV infection (n = 3). The remaining 27 patients, 22 of whom had histologically confirmed systemic vasculitis, were included in the current study. The 5 patients without histologic confirmation of system vasculitis presented with typical signs of “essential” MC vasculitis (29), including purpuric skin lesions of the lower extremities, arthralgia, and asthenia.

For each patient, clinical and biologic data were recorded at the time of the initial evaluation, at the end of antiviral treatment and 6 months after stopping antiviral treatment (for those who completed antiviral therapy), and at the end of followup. The clinical evaluation included age, sex, recent weight loss, neurologic involvement (central, peripheral), recent-onset hypertension (systolic blood pressure >160 mm Hg and/or diastolic blood pressure >95 mm Hg), cutaneous involvement (Raynaud's phenomenon, purpura, livedo, distal ulcers, or gangrenous changes), arthralgia, myalgia, and clinical signs of hepatic involvement (i.e., hepatomegaly, splenomegaly, collateral venous circulation, spider angiomata, ascites, jaundice). The biochemical evaluation included a determination of the complete blood cell count, the erythrocyte sedimentation rate (ESR), and levels of C-reactive protein (CRP), alanine aminotransferase (ALT), aspartate aminotransferase, alkaline phosphatase (AP), albumin, creatinine, rheumatoid factor (RF), cryoglobulin, and complement components (C3, C4, and total hemolytic complement activity). A urinalysis was also completed to screen for hematuria, and a 24-hour urine collection was performed to quantify daily excretion of protein.

The type of vasculitis for each patient was classified according to the Chapel Hill criteria (27). According to these criteria, patients with vasculitis of medium-sized vessels without evidence of small-vessel involvement were considered to have PAN vasculitis. Patients with purpura were considered to have small-vessel vasculitis, and in the context of HCV and mixed cryoglobulinemia, they were classified as having MC vasculitis.

Treatment of HCV-related vasculitis.

Because of the heterogeneity of anti-HCV treatments administered to the patients, we calculated the total cumulative dose and duration of only IFNα and ribavirin therapy. The total cumulative dose of IFNα (in millions of IU [MIU]) and the number of courses of antiviral therapy were recorded. For ribavirin, the number of courses, mean dosage (mg/day), and total cumulative dose were calculated. When used, corticosteroids were given in a short-term, low-dose regimen (prednisone 1 mg/kg/day for 2 weeks, with a rapid decrease to 10 mg/day within 6 weeks). In patients treated with plasma exchanges (PEs), the regimen consisted of 3 PEs per week for 2 weeks, 2 PEs per week for 2 weeks, and 1 PE per week for 2 weeks, for a total of 12 PEs, when possible. The amount of plasma scheduled to be exchanged was 60 ml/kg of body weight per session. The replacement fluid consisted of 500 ml of fluid gelatin and 4% albumin. The duration of followup after the end of antiviral treatment was calculated, and the clinical status of each patient (dead or alive) was recorded.

The response to treatment was analyzed by comparing clinical, immunologic, and virologic parameters at the initial evaluation, at the end of antiviral treatment and 6 months after stopping antiviral treatment (for those who completed therapy), and at the end of followup. Clinical response was defined by analyzing the evolution of the following main clinical signs: general condition (weight increase), peripheral neuropathy (electrophysiologic improvement at 2 successive examinations), renal involvement (a decrease in the serum creatinine level of >100 μmoles/liter), skin involvement (absence of purpura), and absence of arthralgia. Because asthenia and myalgia are frequently reported during IFNα treatment, these symptoms were not assessed.

A complete clinical response was defined by an improvement in all baseline clinical manifestations. Patients were classified as having a partial clinical response when at least half of the baseline clinical manifestations had improved, and all other patients were classified as clinical nonresponders. In patients with a complete or partial clinical response, the type, number, and timing of vasculitis relapses were analyzed. A sustained virologic response was defined by the absence of detectable serum HCV RNA 6 months after stopping antiviral treatment; the remaining patients were classified as virologic nonresponders. Because of the heterogeneity of anti-HCV treatments received, the main results of treatment were separated into 2 groups: those for patients who underwent 6 months of followup after stopping antiviral treatment (group 1), and those for patients who were still undergoing antiviral therapy at the time of analysis (group 2). A complete immunologic response was defined by the absence of serum cryoglobulin, and a partial immunologic response was defined by a decrease of >50% in the baseline cryoglobulin level.

Virologic and immunologic serum markers.

PCR testing for HCV RNA was performed on sera that had been stored at −80°C and was previously unthawed. Serum RNA was extracted, reverse transcribed, and amplified as previously described (18). The sensitivity of the HCV RNA assay was 100 copies/ml. HCV genotyping was performed using a second-generation line probe assay (LiPA; Innogenetics, Brussels, Belgium) (30). Sera were also tested for the presence of hepatitis B surface antigen (HBsAg), antibodies to HBsAg, and anti–hepatitis B core antibody using commercial immunoassays (Abbott, Abbott Park, IL; Sanofi Diagnostics Pasteur, Paris, France). In all sera, HBV DNA was assessed by molecular hybridization (Genostics; Abbott). All sera were negative for anti-HIV antibodies using commercial immunoassays (from Abbott and Sanofi Diagnostics Pasteur).

Cryoglobulins were isolated from patients' sera, purified, and characterized by immunoblotting at 37°C, as previously described (31). A previous evaluation using this method showed that only 5 (3.8%) of 131 healthy blood donors had MC, with a cryoglobulin level lower than 0.03 gm/liter in all 5 samples (16). In the present study, patients were considered to have a significant cryoglobulin level when it was ≥0.05 gm/liter on 2 determinations. Cryoglobulins were classified, according to the method described by Brouet et al (32), as either type II MC, which includes a monoclonal component, or type III MC, defined by the association of polyclonal immunoglobulins.

Electrophysiologic studies.

Electrophysiologic studies were performed with a Nicolet Viking electromyograph (Nicolet Biomedical Instruments, Madison, WI) using previously described techniques (33). Based on the clinical and electrophysiologic findings, each peripheral neuropathy observed was classified as either a polyneuropathy or a multifocal mononeuropathy. The peripheral neuropathies were then classified as either axonal or demyelinating, using the criteria for chronic inflammatory polyneuropathies of the American Academy of Neurology (34). For patients with peripheral neuropathy who had undergone serial electrophysiologic studies, detailed clinical and electrophysiologic parameters obtained at the time of the initial examination (before antiviral treatment) until the time of the last electrophysiologic examination (after treatment) were analyzed.

Histologic studies.

When indicated, neuromuscular biopsy of the superficial peroneal nerve and peroneus brevis muscle was performed in the most affected limb, using previously described techniques (18, 35). Necrotizing vasculitis was diagnosed when vessel walls were interrupted and destroyed by a transmural inflammatory process with segmental necrosis of the vessel wall. The vasculitis was classified as PAN or cryoglobulin type, as previously described (18, 27). Histologic examination of liver biopsy specimens included qualitative and quantitative analyses of the inflammatory activity and severity of fibrosis, classified using the scoring system described by Knodell et al (36).

Statistical analysis.

All data were expressed as the mean ± SD. A chi-square test or Fisher's exact test was used to compare proportions, and Wilcoxon's rank sum test was computed for ordinal variables. A 95% confidence interval (95% CI) was computed for each odds ratio (OR), when applicable. All calculated P values were 2-tailed, and the 5% significance level was used. All analyses were performed using SAS statistical software (SAS Institute, Cary, NC).

RESULTS

The study included 14 men and 13 women with a mean age of 59.5 ± 13 years (range 33–78 years). The mode of contamination with HCV was intravenous drug use in 5 patients, blood transfusion in 9, and unknown in 13. The distribution of HCV genotypes was as follows: genotype 1 (n = 13), 2 (n = 6), 3 (n = 3), 4 (n = 3), and 5 (n = 2). Weight loss (5–10 kg) was noted in 11 patients. Skin manifestations included purpura in 22 patients (81%), Raynaud's phenomenon in 4 patients (15%), livedo and leg ulcers in 3 patients each (11%), and erythema nodosa in 2 patients (7%). Neurologic involvement included a sensorimotor multifocal mononeuropathy in 8 patients (30%), a sensory polyneuropathy in 8 patients (30%), and a sensorimotor polyneuropathy in 3 patients (11%). Renal insufficiency was observed in 7 patients (26%); the mean creatinine concentration was 164 ± 67 μmoles/liter (range 104–280). Hypertension was noted in 9 patients and in 5 of them was associated with renal insufficiency. Microaneurysms and kidney infarction were detected by arteriography in 3 patients and 1 patient, respectively. Arthralgia was noted in 16 patients (59%) and myalgia in 13 patients (48%). Five patients (19%) presented with abdominal pain, assumed to be ischemic in nature, and severe weight loss. Hepatomegaly and splenomegaly were observed in 7 (26%) and 5 patients (19%), respectively. Five patients (19%) had sicca syndrome, and 1 patient (4%) experienced orchitis.

The mean ALT concentration was 2.9-fold the upper limit of normal (range normal to 14-fold); 21 patients (78%) had elevated ALT levels. Thirteen patients (48%) had an elevated serum AP or gammaglutamyl transpeptidase concentration. The mean serum albumin concentration was 37 ± 13 gm/liter (range 27–49). Glomerular proteinuria was noted in 5 patients (19%) and was always associated with an abnormal urinary sediment. An increased serum ESR or elevated CRP levels, potentially indicative of systemic inflammation, were observed in 19 patients (70%). MC was detected in 25 (93%) of 27 patients, with a mean cryoglobulin level of 1.58 gm/liter (range 0.05–5.5). MC was of the type II IgMκ variety in 21 of 25 patients and was type III in 4 patients. RF activity was observed in 20 patients (74%).

Renal biopsy specimens showed membranoproliferative glomerulonephritis in 5 patients and ischemic cortical necrosis in 1 patient. Skin biopsy specimens revealed leukocytoclastic vasculitis in 6 patients. Neuromuscular biopsy specimens obtained from 16 patients showed severe axonal degeneration in 9 (56%) and moderate axonal degeneration in 7 (44%), and an inflammatory process involving nerve in 8 patients (50%) and involving both muscle and nerve in the other 8 patients. When considering the clinical and pathologic features of vasculitis, we identified 7 cases of PAN-type vasculitis and 22 cases of MC-type vasculitis; 2 patients presented with both. Liver biopsy was performed in 25 of 27 patients and revealed chronic active hepatitis in all, with a mean Knodell score of 6.5 (range 1–12); cirrhosis was observed in 3 (12%) of 25 patients.

The main treatment-related data, including virologic outcomes, are summarized in Tables 1 and 2. All patients received combination treatment with IFNα and ribavirin. The median cumulative dose of IFNα was 648 MIU (range 180–1,872), and the mean duration of therapy was 20 ± 14 months (range 6–78). Twelve patients (44%) received at least 2 courses of IFNα. The mean duration of ribavirin therapy was 14 ± 12 months (range 6–56), with a mean daily dosage of 895 ± 250 mg (range 400–1,000) and a median cumulative dose of 293 gm (range 125–1,464). Other treatments included steroids (10 patients) and PE (10 patients). After a mean followup of 57 ± 29 months (range 6–98), 25 (93%) of 27 patients are alive and continue to be followed up as outpatients. Two patients (patients 10 and 11) with MC-type vasculitis and liver cirrhosis died, due to septicemia and massive digestive hemorrhage at 4.5 and 5.5 years of followup, respectively. Both patients were clinical, immunologic, and virologic nonresponders.

Table 1. Data for 14 patients followed up for at least 6 months after completing therapy (group 1)*
PatientNo. of months after EOTInterferonRibavirinClinical symptomsViremiaCryoglobulinemia, gm/liter
No. of coursesNo. of monthsNo. of coursesNo. of monthsEntry6 months after EOTEntry6 months after EOTEntry6 months after EOT
  • *

    EOT = end of therapy; W = weight loss; N = nerve involvement; S = skin disease; A = arthralgia; R = renal disease; ND = not done.

214118118W,N,S,ANonePos.Neg.1.750.05
311122115W,N,SNonePos.Neg.30
4261616S,ANonePos.Neg.5.50.04
8621816S,ANonePos.Neg.50
96129124W,R,SNonePos.Neg.0.030.15
1238213110N,S,ASPos.ND0.32ND
15151818R,SRPos.Pos.00
1714224119N,S,ANonePos.Neg.20
1818218112S,ASPos.Pos.1.610.8
196246130N,S,ANonePos.Pos.0.190.36
2224256156W,N,S,ANonePos.Neg.0.920
2319122112W,SNonePos.Neg.0.60
243312416NNPos.Pos.2Pos.
2713111111N,ANonePos.Neg.1.450
Table 2. Data for 13 patients still receiving therapy at the time of analysis (group 2)*
PatientInterferonRibavirinClinical symptomsViremiaCryoglobulinemia, gm/liter
No. of coursesNo. of monthsNo. of coursesNo. of monthsEntryEnd of followupEntryEnd of followupEntryEnd of followup
  • *

    N = nerve involvement; S = skin disease; A = arthralgia; R = renal disease; W = weight loss; ND = not done.

  • Died.

1113113N,S,ANPos.Neg.0.490.24
5112112R,ARPos.Pos.00
6222122S,ASPos.Neg.1.50.57
7178153N,R,S,AN,SPos.Pos.1.330.22
10324113N,S,ANPos.Pos.4.50.75
1129113N,SNPos.Pos.0.953.5
1311816N,R,SRPos.Pos.0.150.01
1421216N,R,SNPos.Neg.1.18ND
16335123W,N,SW,NPos.Neg.3.50.2
20334122N,R,S,ANonePos.Neg.0.50.03
21112112NNonePos.Neg.0ND
251616N,R,SRPos.Neg.Pos.0.02
261717W,N,S,ANPos.Pos.0.260.38

Considering the definition of a sustained virologic response, main followup data were stratified according to patients with at least 6 months of followup after stopping antiviral treatment (group 1) (Table 1) and those who were still receiving antiviral therapy at the time of the analysis (group 2) (Table 2). Patients in group 1 (n = 14) were followed up for an average of 17 months after discontinuation of therapy (range 6–38). They received a median cumulative dose of IFNα of 828 MIU (range 216–1,872) for a mean duration of 24 ± 15 months (range 6–56). The duration of ribavirin therapy was 17 ± 12 months (range 6–56), with a mean daily dosage of 780 ± 250 mg (range 400–1,000), and a median cumulative dose of 298 gm (range 110–1,460). Twelve of these patients did not receive any additional treatment, and 2 received low-dose corticosteroids (<10 mg/day).

Nine (64%) of the 14 patients in group 1 had a sustained virologic response and were clinical and immunologic complete responders. Four patients (29%) in group 1 were virologic nonresponders (i.e., they had a persistent viremia during treatment); 3 of these patients had a partial clinical and immunologic response. One patient (patient 22) had a virologic response during treatment (negative viremia at the end of therapy) associated with a complete clinical and immunologic response but relapsed virologically and clinically 2 months after stopping antiviral therapy. Overall, 10 patients in group 1 had a complete clinical and immunologic response of their vasculitis (9 [100%] of 9 of the sustained virologic responders compared with 1 [20%] of 5 patients who remained viremic [P = 0.01]). Among the patients in group 1, HCV RNA was more often undetectable in complete clinical responders (9 [90%] of 10 versus 0 of 4 partial responders and nonresponders; P = 0.001); in addition, the mean HCV RNA level was lower (P = 0.001), and genotype 1 was less frequent in complete clinical responders compared with partial responders and nonresponders (1 [10%] of 10 versus 3 [75%] of 4; P = 0.04) (Table 3). All other factors, including age, sex, main clinical vasculitic involvement (weight loss, arthralgia, skin, nerve, or renal involvement), total cumulative dose and duration of IFNα and ribavirin, and use of corticosteroids or PE, were similar in responders and partial responders and nonresponders.

Table 3. Characteristics of 14 patients with HCV-related systemic vasculitis with at least 6 months of followup after completing antiviral therapy with interferon-α plus ribavirin*
 Complete clinical response (n = 10)Nonresponse or partial response (n = 4)P
  • *

    Values are the no. (%) unless otherwise indicated. HCV = hepatitis C virus.

  • Detectable in only 1 patient.

  • Not determined in 1 patient.

Age, mean ± SD years59.9 ± 11.951.3 ± 13.20.26
Male sex3 (30)1 (25)1.0
Weight loss5 (50)0 (0)0.08
Cutaneous involvement9 (90)3 (75)0.51
Arthralgia7 (70)2 (50)1.0
Nerve involvement6 (60)2 (50)1.0
Renal involvement1 (10)1 (25)0.51
Genotype 11 (10)3 (75)0.04
Negative viremia at the end of followup9 (90)0 (0)0.001
HCV RNA level at the end of followup, millions of copies/ml0.22.0 ± 2.50.001
Negative cryoglobulin at the end of followup4 (40)1 (33)0.56
Cryoglobulin level at the end of followup, mean ± SD gm/liter0.12 ± 0.140.26 ± 0.210.36
Cumulative dose of interferon, mean ± SD millions IU987 ± 558603 ± 1980.23
Cumulative dose of ribavirin, mean ± SD gm509 ± 409240 ± 1160.29
Corticosteroid use5 (50)2 (50)0.58
Plasmapheresis6 (60)1 (25)0.56

HCV genotype 1, which has a major negative impact on the response to IFNα and ribavirin therapy, was identified in 4 (29%) of 14 patients in group 1. Compared with patients infected with non–genotype 1 HCV, those with genotype 1 were more likely to be older (59 versus 53 years), male (3 [30%] of 10 versus 1 [25%] of 4), and to have a lower response to treatment (i.e., complete clinical response, 1 [25%] of 4 versus 9 [90%] of 10; complete virologic response, 1 [25%] of 4 versus 8 [80%] of 10; and complete immunologic response, 1 [11%] of 4 versus 4 [40%] of 10). None of these differences, however, reached statistical significance.

In general, IFNα was well tolerated, although therapy was discontinued in 6 (22%) of the 27 patients because of depression (n = 2), severe thrombocytopenia (n = 2), and an exacerbation of peripheral neuropathy (n = 2). IFNα was rechallenged in 4 of these 6 patients, without further problems. Other transient side effects included diarrhea (n = 2), alopecia (n = 2), and dysgeusia (n = 1). The dosage of ribavirin was reduced in 8 (30%) of 27 patients because of hemolytic anemia; nevertheless, all patients received a daily dose higher than 11 mg/kg of body weight. During the initial phase of vasculitis treatment, 10 patients were treated with steroids, and 10 patients were treated with PE (range 8–14 sessions). In general, PE was well tolerated. However, 1 patient developed septicemia at the time of the tenth PE, and another had a decreased hemoglobin level (82 gm/liter), and PE in this patient was stopped after the tenth session. At the end of followup, the mean decrease in the cryoglobulin level was lower in patients who had undergone PE compared with that in patients who did not receive PE (−4% [−100% to 268%] versus −74% [−99% to 0%]; P not significant), but there was no correlation with clinical response.

Four patients who were complete clinical responders had a subsequent clinical relapse. All of these patients had negative HCV RNA levels at the end of the first course of antiviral treatment. Features of relapse included purpura (n = 4), peripheral neuropathy (n = 2), ischemic abdominal pain (n = 1), and arthralgia (n = 2). The mean delay between the end of antiviral treatment and clinical relapse was 4 months (range 2–7). Clinical relapse was always associated with reappearance of HCV RNA. All clinical relapsers went back into complete clinical remission following another course of antiviral treatment.

For 12 of the 19 patients with peripheral neuropathy, results of serial electrophysiologic studies obtained before and after antiviral treatment were analyzed (Table 4). Seven patients had a dramatic improvement, and 3 were stable, whereas 2 patients had an increase in nerve damage. After a mean delay of 36 ± 16 months (range 12–48) between the initial and last electrophysiologic examinations, motor conduction studies showed an increase of distal compound motor amplitude potential (mV) of the peroneal (165 ± 236%) and median (170 ± 326%) nerves. Sensory conduction studies showed an increase in potential amplitude (μV) on sural (180 ± 268%), superficial peroneal (260 ± 360%), and median (90 ± 182%) nerves. Comparison of the electrophysiologic course between patients with (n = 6) and without (n = 6) sustained virologic response showed no significant difference.

Table 4. Course of electrophysiologic parameters in 12 patients with HCV-related systemic vasculitis and peripheral neuropathy
PatientNo. of months between first and last examinationsMotor conduction studies, distal CMAP amplitudeSensory conduction studies, potential amplitude
Peroneal nerveMedian nerveSural nerveSuperficial peroneal nerveMedian nerve
Initial, mVLast, %Initial, mVLast, %Initial, μVLast, %Initial, μVLast, %Initial, μVLast, %
  1. * HCV = hepatitis C virus; CMAP = compound motor amplitude potential; initial = examination just before beginning treatment; last = last examination after treatment during followup (values are the percentage of the initial-visit value); NO = not obtained.

2120+201.4+1712.30NO015+10
3480.6+3009+15NO+400NO+30014+25
5122+406.7+410+505+30380
10120.8−5011−30NO0NO0110
11480.1−105−20NO0NO018−38
12481.3+847+216+45NO+90026+34
16322.906.408−3730250
17360.106+28NO0NO012+50
19480.5+606.4+313+66NO0460
21145.2+134.4+2010+255.2+16342+22
22480.2+6500.05+900NO+700NO+600NO+500
27140.5+406.8+293.6+503.6+228+10

Thirteen patients were still undergoing antiviral treatment at the time of analysis (group 2) (Table 2) and were followed up for an average of 26 months (range 6–78). At the end of followup, 7 of these 13 patients had negative HCV RNA, and 6 were complete clinical and immunologic responders. Six patients in group 2 had persistent viremia, and 4 of these patients showed a partial clinical response despite the lack of a virologic response, prompting ongoing therapy. In the 2 remaining patients who were also virologic nonresponders, IFNα was continued for its antifibrotic properties.

DISCUSSION

Treatment of HCV-related cryoglobulinemia with IFNα is associated with a relatively poor response and a high relapse rate, especially in severe cases (19, 20). In previous studies (19–22), IFNα monotherapy was effective in 50–100% of patients with purpuric skin lesions but did not clearly demonstrate efficacy for neural or renal involvement. In both our study and a study by Zuckerman et al (26), combination therapy with IFNα and ribavirin demonstrated enhanced efficacy for the main HCV-related vasculitic manifestations (cutaneous, 100%; renal, 50%; neural, 25–75%).

In the present study, 93% of patients are alive and continue to be followed up as outpatients, with good control of their vasculitis. This outcome was seen despite the fact that most patients in this series presented with more severe disease than that described in patients in other series. As previously reported by Misiani et al (19), clinical improvement of patients with HCV-related vasculitis correlates with virologic response. In previous series of HCV-related vasculitis, a virologic response at the end of treatment was reported in 15–60% of patients receiving IFNα monotherapy (19–22). However, when followup was sufficient, it was revealed that most of the responders developed virologic and clinical relapses following IFNα withdrawal (Table 5). Such results are quite similar to those reported in patients without extrahepatic manifestations, in whom a 12-month course of IFNα monotherapy leads to a sustained virologic response in only 15–20% of patients compared with 35–80% with IFNα plus ribavirin (24, 25, 37).

Table 5. Overview of treatments in HCV-related systemic vasculitis*
StudyNo. of patientsNo. with visceral involvementTreatmentTreatment duration, monthsFollowup duration, monthsResponse, no. of patients
Authors, year (ref.); designSKN
  • *

    S = skin; K = kidney; N = nerve; IFN = interferon; MIU = millions of international units; ND = not determined. Clinical = clinical improvement of manifestations induced by hepatitis C virus–related (HCV).

  • Values for clinical response and negative (neg.) viremia apply to the groups that received IFN 3 MIU 3 times/week, IFN 3 MIU 3 times/week plus prednisone 16 mg/day, prednisone 16 mg/day, or no treatment, respectively.

Ferri et al, 1993 (20); randomized, crossover2020120IFN 2 MIU/day for 1 month, then 2 MIU every other day for 5 months66Clinical: S 20/20, K 0/1, N 0/20; death: 0/20; neg. viremia: 2/13
Misiani et al, 1994 (19) prospective, randomized, controlled53314015IFN 3 MIU 3 times/week for 6 months; or steroids <0.2 mg/kg/day66–12Clinical: IFN, 56% vs. steroids, 0%; death: 4/53; neg. viremia: IFN, 15/25 vs. controls, 0/24
Dammaco et al, 1994 (21) prospective, randomized, controlled52522820IFN 3 MIU 3 times/week; or IFN 3 MIU 3 times/week plus prednisone 16 mg/day; or prednisone 16 mg/day; or no treatment128–17Clinical: 8/15, 9/17, 3/18, 1/15; death: 0/52; neg. viremia: 5/12, 7/14, 0/17, 0/15
Casato et al, 1997 (22) retrospective31311124IFN 3 MIU/day for 3 months, then 3 MIU every other day for 9 months1633Clinical: S 17/31, K ND, N ND; death: 5/31; neg. viremia: 11/24
Durand et al, 1998 (23) open, uncontrolled5411Ribavirin, 100–1,200 mg/day10–36NDClinical: S 4/4, K 1/1, N 0/1; death: 0/5; viremia decrease: 3/5
Zuckerman et al, 2000 (26) open, uncontrolled9724IFN 3 MIU 3 times/week plus ribavirin 400–1,000 mg/day68Clinical: S 7/7, K 1/2, N 1/4; death: 0/9; neg. viremia: 2/9
Naarendorp et al, 2001 (38) open, uncontrolled11847IFN 3 MIU 3 times/week (plus ribavirin in 4 patients)5–60Clinical: S 3/8, K 1/4, N 3/7; death: 1/11; neg. viremia: 2/9
Cacoub et al (present study); retrospective2722719IFN MIU 3 times/week plus ribavirin 400–1,000 mg/day20, IFN; 14, ribavirin57Clinical: S 20/23, K 4/8, N 11/19; death: 2/27; neg. viremia: 16/27

In our series, a complete clinical response was observed in 16 (59%) of 27 patients, and a partial clinical response in 9 (33%) of 27. A sustained virologic response was observed in 9 (64%) of 14 patients who had sufficient followup after stopping antiviral treatment, and it was associated with sustained clinical remission in all but 1 patient, confirming the findings of other investigators (19). The high rate of virologic response compared with that reported in previous trials in patients with chronic HCV infection but no extrahepatic manifestations is somewhat surprising (24, 25). There are several possible explanations. Our patients were treated for much longer compared with patients in other trials (an average of 20 months with IFNα and 14 months with ribavirin versus only 12 months on average in previous trials). Furthermore, the patients in our series had favorable pretreatment characteristics: only 48% had genotype 1 compared with 60–70% in previous trials, 12% were cirrhotic compared with 7–33% in previous trials, and only 50% had high viral load (>2 million copies/ml) at baseline compared with 60–70% in previous studies.

In 4 patients, clinical relapses of HCV-related vasculitis were associated with relapsing HCV viremia. These results are much better than those reported in studies of IFNα monotherapy, in which vasculitis relapses were described in 40–100% of patients (19, 21, 22). In our experience, IFNα and ribavirin therapy must be prolonged at least 18–24 months, particularly in patients with peripheral nerve or renal involvement, to avoid such relapses (38). In our series, all patients who relapsed went into clinical remission following another course of therapy. Despite the successes with combination antiviral treatment, HCV-related vasculitis remains a severe disease. In most series for which the effects of treatment were reported, the mortality rate after sufficient followup was 8–15% (19, 22). In our series, 7% of patients died during a mean followup period of nearly 5 years. Usually, death occurs in nonresponders after a prolonged course of vasculitis and is often attributable to sepsis, as was observed in 1 of our patients.

We observed only a trend toward lower rates of clinical, immunologic, and virologic responses in patients infected with HCV genotype 1. These results must be analyzed cautiously, however, because patients with HCV genotype 1 infection had several clinical features (older age and male sex) that suggested a poor prognosis for antiviral treatment (24, 25). There was no clear association between the presence of abnormal levels of liver enzymes and HCV-related vasculitis; 22% of patients (6 of 27) had normal aminotransferase levels before treatment. Therefore, as in the routine care of patients with chronic HCV infection, the efficacy of antiviral treatment in patients with HCV-related vasculitis should not be monitored on the basis of ALT levels but rather on the basis of the disappearance of HCV RNA from serum.

In rare cases of HCV-related vasculitis (22), complete clinical responders had viral clearance long after clinical remission. In addition, some patients (3 reported by Casato et al [22] and 1 in our series) may remain in clinical remission despite the persistence of viremia. Monotherapy with ribavirin, although ineffective in eradicating HCV when used in isolation, may be effective in patients with symptomatic HCV-related MC, as suggested by Durand et al (23). These findings suggest that IFNα and ribavirin may have both antiviral and immunomodulatory effects. For example, IFNα may exert its effects via inhibition of viral replication rather than by complete eradication of the virus, or by an antiinflammatory action mediated by an altered balance between proinflammatory (Th1-like) and antiinflammatory (Th2-like) cytokines (39, 40). The ability of ribavirin to lead to a reduction in serum aminotransferase levels despite having no effect on HCV viremia suggests that it has antiinflammatory properties independent of any antiviral effect.

In comparison with IFNα monotherapy or combination therapy with IFNα and ribavirin, other treatment strategies have not shown superior efficacy. Corticosteroids, used alone or in addition to IFNα, do not improve the response of HCV-related vasculitic manifestations (19, 21). However, high-dose intravenous corticosteroids may be useful initially for the control of life-threatening organ involvement while awaiting the generally slow response to antiviral treatments. One potential concern regarding the use of corticosteroids and immunosuppressives is the propensity of these agents to worsen HCV viremia, as observed in the posttransplantation setting. We did not document this phenomenon; however, this was not a specific interest of our study. Use of PE may permit the rapid control of life-threatening symptoms of vasculitis without using high-dose intravenous corticosteroids or immunosuppressive treatments (7).

In general, a 12-month course of IFNα at a dose of 3 MIU 3 times weekly, alone or in combination with ribavirin, is well tolerated, with side effects necessitating treatment discontinuation in 10–20% of patients (19–22). In our study, 21% of patients required discontinuation of therapy despite a mean treatment duration of 20 months with IFNα and 14 months with ribavirin. Careful monitoring for adverse effects is mandatory, because some manifestations of HCV-related vasculitis, such as peripheral neuropathy, may worsen during interferon therapy, as observed in 2 patients in our study (41, 42). We did not observe a worsening of skin ulcers at the time of initiation of IFNα therapy, as has been reported by Cid et al (43).

In conclusion, IFNα and ribavirin are effective in the treatment of HCV-related vasculitis. Clinical response mirrors virologic response, which generally requires a prolonged period (at least 18–24 months) to achieve and avoid vasculitis relapse. Therapy is well tolerated in patients with HCV-related vasculitis, and tolerance appears to be no different from that observed in patients without vasculitis. Further large, prospective, multicenter studies are warranted in patients with HCV-related vasculitis who are followed up for prolonged periods in order to establish the optimal dosage, duration, and formulation (i.e., classic or pegylated IFNα) of combination therapy.

Ancillary