Because of the variability of the HCV genome, one might speculate that particular viral variants are responsible for mixed cryoglobulinemia. Thus, we assessed the prevalence of different genotypes in HCV-positive cryoglobulinemic patients and in patients with chronic HCV infection who did not have cryoglobulinemia.

Methods

Top Methods Results Discussion Author & Article Info References

From March 1994 to September 1994, we recruited (at ambulatory visits) 29 consecutive HCV-positive (anti-HCV antibody-positive and HCV RNA-positive) patients with type II (IgM ) mixed cryoglobulinemia (9 men, 20 women; mean age ±SD, 60 ± 7.5 years; age range, 46 to 72 years) and 61 patients with chronic HCV infection who did not have mixed cryoglobulinemia (control group). All patients studied were Italian-born, were heterosexual, and had no history of blood transfusion or drug or alcohol abuse. Cryoglobulinemic and control patients were followed at the rheumatology and hepatology units of the University of Pisa and University of Florence. The mean (±SD) duration of follow-up was 8.4 ± 5.5 years (range, 1 to 25 years) for cryoglobulinemic patients and 6.8 ± 4.5 years (range, 2 to 14 years) for controls. A diagnosis of mixed cryoglobulinemia was made if a patient had the typical syndrome (purpura, arthralgias, weakness, and circulating mixed cryoglobulins) and if other well-known systemic disorders could be ruled out. Eight cryoglobulinemic patients developed B-cell non-Hodgkin lymphoma 4.3 ± 2.7 years (range, 1.5 to 8 years) after diagnosis.

After informed consent was obtained, percutaneous liver and renal biopsies were done as previously described [4, 7]. Cryocrit determinations were done and cryoglobulin composition was evaluated as previously described [4, 8]. Antinuclear, anti-smooth muscle, anti-liver-kidney microsomal 1, anti-soluble liver antigen, and antimitochondrial autoantibodies were assayed by current techniques [9]. A titer greater than 1:40 was considered positive. Anti-extractable nuclear antigen antibody determinations were done using the method of Bunn and colleagues [10].

Serum samples and aliquots of peripheral blood mononuclear cells with the last washing liquid (phosphate-buffered saline) for HCV polymerase chain reaction (PCR) analysis were collected as previously described [11, 12]. In addition, to ascertain the presence of a latent HCV infection, peripheral blood mononuclear cell samples were cultured for 72 hours in the presence of mitogens (phytohemagglutinin-phorbol myristate acetate) as previously described [11, 12].

Different samples were tested by "one-tube" nested reverse transcriptase PCR with primers derived from the 5' noncoding region [13]. Several precautions were taken to prevent false-positive results [14], including the incorporation of deoxyuridane-triphosphate instead of deoxythymidine-triphosphate during amplification steps followed by incubation of PCR mixtures for 3 minutes at 50 °C in the presence of uracil-N-glycosilase (UNG; Perkin Elmer Cetus, Norwalk, Connecticut).

In nine unselected patients with mixed cryoglobulinemia, aliquots of peripheral blood mononuclear cells were also available for HCV genotyping. Hepatitis C virus genotyping was done using two different methods, both based on amplification by PCR. The first technique used type-specific primers localized in the core region, as described by Okamoto and colleagues [15], with the difference that PCR was done without mixing genotype-specific antisense primers. Moreover, for the detection of genotype III, we used a new primer that, in a previous study, made it possible to classify most previously unclassified HCV isolates as genotype 2a/III [16]: This primer was CRIIIa antisense 5'-TTCCCCAGGAYT TGCCAGTGG-3' (Y equals C or T). The second one employed biotinyled, universal primers localized in the 5' noncoding region of HCV RNA; amplification products were then hybridized to genotype-specific probes (Line Probe Assay, LiPA, Innogenetics, Brussels, Belgium).

Statistical analysis was done using the chi-square test and the Fisher exact test whenever the z approximation was inadequate.

Results

Top Methods Results Discussion Author & Article Info References

Table 1 shows the values for the main clinicoepidemiologic and laboratory variables in patients with mixed cryoglobulinemia. The following complications of mixed cryoglobulinemia were recorded: peripheral neuropathy in 15 of 29 patients (52%); mild sicca syndrome in 11 of 28 (39%); glomerulonephritis in 4 of 29 (13%); Raynaud phenomenon in 1 of 28 (4%); and skin ulcers in 3 of 28 (11%). One or more serum autoantibodies were detected in 8 of 28 (29%) patients with mixed cryoglobulinemia and in 19 of 61 (31%) controls.

Hepatitis C virus RNA sequences were shown in uncultured peripheral blood mononuclear cells from 23 of 29 (75%) patients with mixed cryoglobulinemia and in cultured cells from 3 other patients (total, 90%) (Table 1). In the control group, viral sequences were detected in uncultured or mitogenstimulated peripheral blood mononuclear cells from 46 of 61 (75%) and 49 of 61 persons (total, 80%), respectively (Table 1).

Among the 29 patients with mixed cryoglobulinemia, serum specimens showed a single infection with type 1a/I in 1 patient (3 %), with type 1b/II in 14 patients (48%), and with type 2a/III in 12 patients (41%). Two patients (6.6%) had mixed infection (1a/I plus 1b/II and 1b/II plus 2a/III, respectively) (Table 1). Among the 61 controls, genotypes 1a/I, 1b/II, 2a/III, 3a/V, and 4a were observed in 7 (11%), 37 (61%), 9 (15%), 4 (7%), and 1 (1%) patient, respectively, whereas mixed infection (1a/I plus 1b/II; 1b/II plus 2a/III; 1b/II plus 3a/V) was observed in 3 (5%) patients.

When HCV genotypes detected in peripheral blood mononuclear cells were also considered, type 2a/III was found in 15 of the 29 (52%) patients with mixed cryoglobulinemia and in most autoantibody-positive patients (6 of 8; 75%) (Table 1).

The prevalence of 2a/III genotype was significantly higher in patients with mixed cryoglobulinemia (12 of 29; 41%) than in controls (9 of 61; 15%), a difference of 27 percentage points (95% CI, 6.6% to 46.6%; P = 0.009). No other significant differences were observed between the two groups.

Sixteen of the 29 patients with mixed cryoglobulinemia had chronic aminotransferase elevations. Analysis of serum samples showed that 12 of these patients (75%) were infected with HCV genotype 1b/II and that 4 (25%) were infected with HCV genotype 2a/III (Table 1). Of the remaining 13 patients who showed no clinical evidence of liver damage, 8 (61%) had infection with genotype 2a/III, 3 (23%) had infection with genotype 1b/II, 1 (7%) had infection with genotype 1a/I, and 1 had coinfection with types 1b/II and 2a/III.

Liver biopsy, done in 14 patients, showed chronic hepatitis in 13 patients and liver cirrhosis in 1 patient; 2 patients with chronic hepatitis and 1 patient with cirrhosis had persistently normal aminotransferase levels (Table 1). Among the 61 controls, 27 (44%) had chronic hepatitis, 21 (34%) had liver cirrhosis, and 13 (21%) had hepatocellular carcinoma; none had normal aminotransferase values.

Discussion

Top Methods Results Discussion Author & Article Info References

In our study, HCV genotype 2a/III had a significantly higher prevalence in HCV-positive patients with mixed cryoglobulinemia than in patients with chronic hepatitis who did not have cryoglobulinemia. Among cryoglobulinemic patients, this genotype was more frequent in those without a symptomatic liver disease or with circulating autoantibodies.

Recently, several reports have suggested different clinical outcomes for the HCV genotypes. Type 1b/II infection, for example, has been associated with a more severe liver disease and a lower response to interferon treatment, whereas type 2a/III infection has been considered relatively benign [17-19]. This hypothesis is consistent with the observation that genotype 2a/III is more prevalent in cryoglobulinemic patients without symptomatic liver disease than in those with chronic hepatitis. On the other hand, the higher prevalence of genotype 2a/III in patients with mixed cryoglobulinemia than in controls, especially in cryoglobulinemic patients with circulating autoantibodies, suggests that type 2a/III might be involved in the pathogenesis of autoimmune-lymphoproliferative disorders. The recent observation that type 2a/III is particularly frequent in Italian patients with anti-liver-kidney microsomal 1 autoantibody-positive type 2 autoimmune hepatitis further supports the possibility of a peculiar pathogenetic role for this genotype [20].

A recent study [8] showed that patients with mixed cryoglobulinemia have a high prevalence (81%) of HCV infection in peripheral blood mononuclear cells, suggesting that HCV lymphotropism may play a key role in determining the lymphoproliferative disorder underlying the disease. Our study confirms these data and also shows the frequent infection of lymphatic cells in HCV-positive patients with chronic hepatitis who do not have cryoglobulinemia. We can thus hypothesize that different viral, genetic, or environmental factors, in addition to the infection of lymphatic cells, may be involved in the pathogenesis of this disorder.

The exact role of HCV variants, namely 2a/III, which are possibly related to different host immune reactivity or to a greater lymphotropism, should be clarified through deeper virologic analysis, including examination of lymph-node and bone marrow specimens from a larger series of patients. In addition, because the clinical expression of chronic HCV infection varies according to geographical area [20], it would be interesting to evaluate the prevalence of HCV genotypes in patients from different countries.

Our findings suggest that viral and possibly environmental or host factors may be involved in the pathogenesis of mixed cryoglobulinemia in patients with HCV infection. A different combination of the above factors may lead to differing degrees of liver or lymphatic damage, ranging from chronic hepatitis without evidence of immune alterations to mixed cryoglobulinemia without liver involvement.

Drs. Ferri, La Civita, Longombardo, Lombardini, and Bombardieri: Istituto di Patologia Medica I, Via Roma 67, Pisa, Italy