Extrahepatic Immunologic Manifestations in Chronic Hepatitis C and Hepatitis C Virus Serotypes
- Jean-Michel Pawlotsky, MD;
- Francoise Roudot-Thoraval, MD;
- Peter Simmonds, PhD;
- Janet Mellor, PhD;
- Mustapha Ben Yahia, MD;
- Chantal Andre, MD;
- Marie-Catherine Voisin, MD;
- Liliane Intrator, MD;
- Elie-Serge Zafrani, MD;
- Jean Duval, MD; and
- Daniel Dhumeaux, MD
- From the Universite de Paris XII, Creteil, France; and the University of Edinburgh, Edinburgh, United Kingdom. Requests for Reprints: Jean-Michel Pawlotsky, MD, Service de Bacteriologie-Virologie, Hopital Henri Mondor, F-94010 Creteil Cedex, France.
Abstract
Objective: To determine, using a serotyping assay, whether the occurrence of extrahepatic immunologic disorders in patients with chronic hepatitis C is dependent on hepatitis C virus serotype.
Design: Prospective study.
Setting: Liver unit and virology laboratory of a university hospital.
Patients: 59 consecutive patients with chronic hepatitis C.
Measurements: Hepatitis C virus serotype was determined using a recently developed immunoenzymatic assay that detects antibodies directed to serotype-specific immunodominant epitopes. Cryoglobulin, rheumatoid factor, and numerous antitissue antibodies were sought. Biopsies of labial salivary glands were done in 49 of the 59 patients.
Results: Prevalence was 59% for serotype 1, 10% for serotype 2, 12% for serotype 3, and 3% for mixed infection. Fifteen percent of patients could not be serotyped. Cryoglobulinemia was found in 36% of patients and rheumatoid factor was found in the serum of 71%. At least one antitissue antibody was found in the serum of 41% of patients; salivary gland biopsy showed lymphocytic capillaritis in 49% of patients. These immunologic abnormalities were seen in patients infected with any of the three serotypes, and prevalences of the abnormalities did not differ significantly among patients infected with different serotypes.
Conclusions: We confirm that the prevalence of extrahepatic immunologic abnormalities is high in patients with chronic hepatitis C. These abnormalities may occur in patients infected with any of the three major hepatitis C virus serotypes now present in developed countries.
Extrahepatic immunologic abnormalities have been shown to occur frequently in patients with chronic hepatitis C virus (HCV) infection. Hepatitis C virus now appears to cause those cases of mixed cryoglobulinemia that were previously considered essential [1-5]. Indeed, HCV RNA has been detected in the serum specimens of about 90% of patients with essential mixed cryoglobulinemia [1, 2, 4]. In addition, cryoglobulin is found, usually at low levels, in the serum specimens of one third to one half of patients with chronic hepatitis C [6, 7]; rheumatoid factor, which may play a role in cryoglobulinemia, is present in the serum specimens of about 70% of patients [6]. Various autoantibodies have been seen in the serum of 40% to 50% of patients with chronic HCV infection [6, 8], and HCV has been associated with cases of autoimmune thyroiditis [9]. Salivary gland lesions, characterized by lymphocytic capillaritis, are seen in about half of patients and are sometimes associated with lymphocytic sialadenitis resembling that of the Sjogren syndrome [6]. Finally, HCV may cause the chronic liver disease frequently associated with lichen planus [10].
Recently, sequences of different HCV variants were classified into different genotypes on the basis of overall sequence similarity [11-22]. A consensus nomenclature for HCV genotypes has been proposed [23], in which the six HCV genotypes identified so far are numbered in the order of their discovery. Within each genotype, subtypes have been identified by lower case letters, which are also given in order of discovery [23]. Correspondence among the classifications reported so far is presented in Table 1. Different techniques for determining HCV genotype have been developed in recent months. Currently, in addition to sequencing the genome, investigators can use three techniques based on the polymerase chain reaction (PCR). The technique described by Okamoto and colleagues [24, 25] is based on a “nested” PCR amplification of the HCV genome and uses primers located in the core region: The first round of PCR uses a pair of universal (non-type-specific) primers and the second uses a pair of type-specific primers. The method of McOmish and colleagues [17, 18] is based on PCR amplification of the 5′ noncoding region of the genome done with a pair of universal primers, followed by enzymatic digestion of the amplified products and analysis of their restriction fragment length polymorphism. Stuyver and colleagues described a line probe assay for the determination of HCV genotypes [22], in which a PCR amplification is done using universal primers located in the 5′ noncoding region of the genome. This is followed by hybridization of the amplified products to oligonucleotide probes attached as parallel bands on nitrocellulose strips. On the other hand, a serotyping immunoenzymatic assay to detect genotype-specific antibodies directed to epitopes encoded by the NS4 region of the HCV genome has been developed [26]. This technique, in its present form, allows the differentiation of HCV serotypes 1, 2, and 3, which correspond to HCV genotypes 1, 2, and 3 in the consensus nomenclature [23].
Several studies indicate that particular HCV genotypes are associated with more severe liver disease and poorer response to interferon-α therapy [27-30]. The factors determining immunologic abnormalities in patients with chronic hepatitis C are largely unknown. We used a serotyping assay to study whether the occurrence of extrahepatic immunologic abnormalities in patients with chronic hepatitis C is serotype dependent.
Methods
Patients
Fifty-nine consecutive patients with chronic hepatitis C were prospectively studied. Thirty-four were men and 25 were women; their mean age was 52 years (range, 18 to 77 years). In all cases, the diagnosis of chronic hepatitis C was based on long-term elevation of serum alanine aminotransferase levels in the blood, positive serologic markers of HCV infection (found using second-generation enzyme-linked immunosorbent assay and recombinant immunoblot assay, Ortho Diagnostic Systems, Raritan, New Jersey), and the absence of any other cause of chronic liver disease. Specimens obtained by percutaneous liver biopsy showed chronic active hepatitis in all 56 patients tested and associated cirrhosis in 15 of the 56 (27%). Before any treatment was given, serum specimens were tested for cryoglobulin, rheumatoid factor, and many antitissue antibodies, and biopsy of labial salivary glands was done. Hepatitis C virus serotype was determined in all patients by immunoenzymatic assay.
Study Methods
Detection of Cryoglobulinemia
Venous blood (20 mL) was taken from fasting patients in a room at 37 °C, allowed to clot at this temperature, and then separated by centrifugation. After centrifugation, the supernatant was removed from the serum, incubated at 4 °C for 8 days, and examined daily for cryoprecipitation.
Detection of Rheumatoid Factor
Rheumatoid factor was measured using a nephelometer analyzer (BNA, Behring, Marburg, Germany); polystyrene particles coated with human γ globulin were agglutinated when mixed with samples containing rheumatoid factor. Normal values were those less than 18 IU/mL.
Detection of Autoantibodies
Antinuclear, anti-smooth muscle, type 1 anti-liver-kidney microsomal (anti-LKM1), and antimitochondrial antibodies were detected by indirect immunofluorescence using air-dried cryostat sections from rat or mouse livers and kidneys and HEp-2 cells (Kallestad, Chaska, Minnesota) as substrates. Antithyroid microsomal antibodies were detected by indirect immunofluorescence using surgical specimens of human thyrotoxic thyroid as substrate. In all cases, the classic Weller and Coon indirect immunofluorescence method was used with fluorescein-labeled goat immunoglobulin directed to IgG, IgA, and IgM (Pasteur Diagnostics, Marnes la Coquette, France) as a second layer [31]. The serum specimens were tested undiluted for anti-DNA antibodies, at a 1/5 dilution for antimicrosomal antibodies, and at a 1/10 dilution for other antibodies. The titers were established using increasing dilutions up to 1/2560. Antithyroglobulin antibodies were detected using an hemagglutination kit (Thymune-T, Wellcome Diagnostics, Dartford, United Kingdom).
Labial Salivary Gland Examination
All biopsies were done in macroscopically normal mucosa. The samples were fixed in Bouin fluid, embedded in paraffin, and stained with hematoxylin-eosin-safranin. All sections were examined blind by two pathologists and graded according to the Chisholm and Mason classification system [32].
Determination of Serotypes
The 59 serum specimens in our study were tested for the presence of serotype-specific antibodies using the recently developed enzyme immunoassay [26]. A series of eight branched peptides, synthesized from two antigenic regions of HCV genotypes 1, 2, and 3, were used to coat polypropylene microtiter wells overnight at 4 °C. After washing, the wells were blocked with 150 µL of blocking solution (phosphate-buffered saline, 0.1% Tween 20, and 2% bovine serum albumin) for 1 hour at room temperature. Blocking assays were done using mixes of type-specific peptides at a final concentration of 1 mg/mL (for example, 100:1 excess over that used to coat the wells).
Plasma specimens from the 59 patients were diluted in the blocking solution and 100 µL were added to antigen-coated and blocked wells. The first incubation was done overnight at 4 °C. Plates were washed four times in phosphate-buffered saline and 0.1% Tween 20 and then incubated with horseradish peroxidase-conjugated anti-human IgG (1/20 000 in phosphate-buffered saline and 0.1% Tween 20 for 1 hour at room temperature). The plates were finally washed four times in phosphate-buffered saline and 0.1% Tween 20 and incubated with substrate (50 µg of ο-phenylenediamine per milliliter and 0.1% H2O2 [30 volumes] for 30 minutes in the dark at room temperature). Optical densities were read at 490 nm; values ranged from 100 to 2000 mU.
Results
Prevalence of Immunologic Abnormalities
Our results are presented in Table 2. Cryoglobulin was found in the serum specimens of 20 of the 56 patients tested (36%), and rheumatoid factor was present at abnormal levels in 42 of the 59 patients (71%).
At least one type of antitissue antibody was detected in the serum specimens of 24 of the 59 patients (41%). Thirteen patients (22%) had serum antinuclear antibodies and 13 (22%) had anti-smooth muscle antibodies at a significant titer (greater than 1/40). Anti-LKM1 antibodies were found in 3 patients (5%) and antithyroid antibodies were found in 5 (8%); 4 of these 5 had antithyroglobulin and 1 had antithyroid microsomes. No antimitochondrial antibodies were found.
Labial salivary gland biopsies were done in those 49 of the 59 patients who had no contraindication and who gave informed consent; lesions were found in 24 of them (49%). In all patients, these lesions were characterized by lymphocytic capillaritis, as previously described [6]. In 7 patients (14%), they were associated with more severe lesions, grades 3 and 4 by the Chisholm and Mason classification (lymphocytic sialadenitis) [32], and resembled the lymphocytic sialadenitis of the Sjogren syndrome (14%). Only one of the patients with salivary gland lesions had a mild case of the ocular sicca syndrome shown by the Schirmer test.
The prevalences of the different immunologic abnormalities did not vary significantly according to the presence of cirrhotic findings in liver specimens.
Hepatitis C Virus Serotypes
Thirty-five of the 59 patients (59%) were infected with HCV serotype 1, 6 (10%) were infected with serotype 2, and 7 (12%) were infected with serotype 3 (serotypes are here described using the proposed consensus nomenclature [23]). Two patients (3%) had mixed infections (serotypes 1 and 2 in one case and 1 and 3 in the other) and 9 (15%) had nonreactive sera that could not be serotyped.
Prevalence of Immunologic Abnormalities according to HCV Serotype
Table 2 shows the prevalences of immunologic abnormalities according to HCV serotype. Cryoglobulin and rheumatoid factor were found in the serum specimens of patients infected with any of the three serotypes studied. Although cryoglobulinemia was less prevalent in patients infected with HCV serotype 1 than in those infected with other serotypes, the difference was not significant. Similarly, antitissue antibodies were detected in the serum specimens of patients infected with any of the three serotypes. Although antinuclear, anti-smooth muscle, and antithyroid antibodies were never found in the few patients with HCV serotype 3, no significant difference among the three groups could be shown. Finally, lymphocytic capillaritis and lymphocytic sialadenitis were seen on salivary gland biopsy in similar proportions of patients infected with each of the three HCV serotypes.
Discussion
Since the identification of HCV [33], reported sequences of different HCV variants have been classified into at least 12 different genotypes on the basis of overall sequence similarity in both coding and noncoding regions of the viral genome [11-17, 20]. Analysis of the more variable coding regions of the genome indicates that many of the major types are composed of two or more distinct subtypes [15, 19, 21, 22]. A consensus nomenclature for HCV genotypes has recently been proposed (23; (Table 1). The methods currently used to determine HCV genotypes are based on PCR amplification [17, 18, 22, 24, 25]. However, these methods are expensive and time-consuming and can be done only in highly specialized laboratories. The serotyping method that we used was recently developed [26] to detect genotype-specific antibodies directed to epitopes encoded by the NS4 region of the HCV genome. Little is known about the concordance of results produced by the different methods of determining HCV genotypes. Although these methods are based on the study of different regions of the genome, including the core [24, 25], 5′ noncoding [17, 18, 22], and NS4 [26] regions, they can be considered reliable on the basis of the following arguments: All were designed using sequence analysis of different regions of the HCV genome as a reference for genotype identification; the serotyping assay used in our study was extensively compared (using approximately 150 specimens from blood donors infected with HCV) with genotyping by restriction fragment length polymorphism in the 5′ noncoding region of the genome [26]; and we recently compared this serotyping assay with the HCV line probe assay (Inno-LiPA HCV, Innogenetics, Gent, Belgium) [22] in 30 patients with chronic hepatitis C and found a 93% concordance between these two techniques (unpublished data). In addition, because serotyping is rapid and easy to do on a large number of samples in any laboratory with enzyme immunoassay equipment, it can be considered a useful tool for studies of the relation between HCV genotypes or serotypes and pathogenicity or response to antiviral therapy [30].
Most of the HCV strains present in Western Europe, the United States, and Japan are HCV genotypes 1, 2, or 3 [17, 18, 30, 34-37]. We confirmed, by using the serotyping assay, that most of the HCV strains present in France belong to one of these three serotypes. Nine patients in our study could not be classified into any of these three groups, but this could be due to the absence in their serum of detectable antibodies directed to the epitopes included in the assay. It is of interest that our results accord with the prevalences found using the same serotyping assay in patients from Northern Italy who have chronic hepatitis C [30]. The only difference was that patients in the Italian series who were infected with serotype 1 were more often coinfected with another serotype; this could be due to differences in the mode of infection, for example, multiple transfusion or intravenous drug abuse.
Frequent and various immunologic abnormalities have been reported in patients with chronic hepatitis C. In a recent study, we classified them into the following four categories, according to their putative mechanisms [6]: 1) immune complex-mediated disease, mainly represented by mixed cryoglobulinemia, which is present in one third to one half of patients [1-7]; 2) autoimmune abnormalities, primarily antitissue antibodies in the serum and autoimmune thyroiditis [6, 9, 38-42]; 3) salivary gland lesions, which are present in about half of patients and are characterized by lymphocytic capillaritis, which is sometimes associated with lymphocytic sialadenitis resembling that of the Sjogren syndrome [6]; and 4) the chronic liver disease that is frequently associated with lichen planus and that may be caused by HCV [10]. Hepatitis C virus genotypes have been shown to have different pathogenicities, most likely because of differences in nucleotide sequence that may play a crucial role in viral replication and protein synthesis. Indeed, the severity of disease and the rates of response to interferon-α therapy have been reported to differ according to HCV genotype [27-30]. Similarly, it could be hypothesized that genotype-specific sequences may play a role in the occurrence of HCV-associated extrahepatic immunologic manifestations, either alone or in combination with HCV-specific virologic features, including the ability to induce perturbations of the immune system.
We show that the extrahepatic immunologic abnormalities associated with chronic HCV infection can be seen in patients infected with any of the three main HCV serotypes now present in France. Thus, it seems unlikely that genotype-specific sequences are responsible for the occurrence of these abnormalities. Cryoglobulinemia is usually present at low levels and does not cause symptoms in patients with chronic hepatitis C [6]. However, cases of severe cryoglobulinemia, characterized by cutaneous involvement, or by renal or neurologic involvement or both, have been reported in these patients [5, 6, 43, 44]. Because some HCV genotypes have been shown to induce more severe liver disease [28, 29], whether these rare complications of HCV-associated cryoglobulinemia could be due to peculiar HCV serotypes or subtypes remains unknown.
Anti-LKM1 antibodies are usually detected in fewer than 5% of patients with chronic hepatitis C [6, 8, 45]. Nevertheless, HCV has been shown to be responsible for about half of the cases of anti-LKM1-antibody-associated chronic active hepatitis in adults, so that this entity can now be distinguished from the so-called type 2 autoimmune chronic active hepatitis [39, 40]. A mechanism of molecular mimicry has been proposed to explain this association. Indeed, the HCV core region has been shown to share a common epitope with cytochrome P450 of the 2D6 family, a protein to which anti-LKM1 antibodies are directed [46, 47]. The fact that anti-LKM1 antibodies can be seen in patients infected with different HCV serotypes suggests that this common epitope is not encoded by a genotype-specific sequence of the core region of the HCV genome. However, genetic variability could be involved to a greater extent. Indeed, the low prevalence of anti-LKM1 antibodies in patients with HCV infection suggests that this epitope could be randomly produced as a result of nucleotide substitutions in the involved region. The additional role of immunologic perturbations remains to be determined, but the overall 40% to 50% prevalence of various types of autoantibodies in patients with chronic hepatitis C [6, 8] suggests that these abnormalities occur frequently. Indeed, the occurrence of antinuclear, anti-smooth muscle, and antithyroid antibodies in serum specimens appears to be a nonspecific event that could be related to HCV-induced immunologic perturbations favoring the synthesis of autoantibodies by B lymphocytes. These antibodies were seen in patients infected with HCV serotypes 1 and 2, suggesting that their occurrence is not serotype dependent. However, these antibodies were not seen in any patients infected with HCV serotype 3. Our series had too few patients with this rare serotype to allow us to draw any conclusions; studies with larger series are needed to determine whether these autoantibodies may be seen in all HCV serotypes.
Salivary gland lesions and more severe lesions characterized by lymphocytic sialadenitis could be seen in patients infected by any of the three main serotypes. This result was not surprising because these lesions are caused by pericapillary lymphocyte infiltration of salivary glands, a pattern that favors a local inflammatory process [6]. These lesions could be due to the presence of immune complex deposits in capillaries, because circulating immune complexes are usually found in patients with chronic hepatitis C [48]. They could also be due to HCV replication into salivary glands or lymphocytes infiltrating the glands or both. These two putative mechanisms appear unlikely to be genotype dependent.
We have confirmed the high prevalence of extrahepatic immunologic abnormalities in patients with chronic HCV infection. We have also showed that the occurrence of these abnormalities is not dependent on HCV serotype. Whether some serotypes or genotypes could be associated with different features of these abnormalities remains to be determined.
- Copyright ©2004 by the American College of Physicians
RSS Feeds









