Misiani and colleagues [1] recently reported a high frequency of GB virus C infection [2] in patients with type II mixed cryoglobulinemia. In 1996 we studied the prevalence of plasma hepatitis G virus (HGV) RNA (a different isolate of GB virus C) [3] in 35 patients who were positive for hepatitis C virus (HCV). Nine patients were men and 26 were women; the mean age ±SD was 66.2 ± 9.0 years. Twenty-three patients had type II and 12 had type III mixed cryoglobulinemia. Twenty patients (9 men and 11 women, with a mean age of 58.8 ± 8.9 years) who had noncryoglobulinemic chronic hepatitis C served as controls. All of the patients with mixed cryoglobulinemia and all controls had histologically proven chronic hepatitis (mean serum alanine aminotransferase levels, 81.1 ± 78.7 U/L [1.35 ± 1.31 µkat/L] compared with 112.4 ± 104.3 U/L [1.87 ± 1.74 µkat/L]; P = 0.2), none had received interferon in the previous 24 months, and none reported any history of blood transfusion or injection drug use.

We detected HGV RNA by reverse transcriptase polymerase chain reaction (PCR) amplification of the NS5 region of the genome and by reverse transcriptase PCR of the putative 5'-noncoding region, as described elsewhere [4]. Plasma HGV RNA was present in one patient with type II mixed cryoglobulinemia (2.9% of all patients and 4.3% of those with type II mixed cryoglobulinemia) and two of the controls (10%). The patient with type II mixed cryoglobulinemia who was positive for HGV RNA was a 72-year-old woman with cirrhosis.

The discordance between our data and the higher prevalence of HGV RNA reported by Misiani and colleagues [1] may have been due to the different sensitivity of HGV and GB virus C detection, which resulted from the amplification of different genomic regions (NS3/helicase for Misiani and colleagues and putative 5'-noncoding region and NS5 in our study), or to differences in the epidemiology of HGV and GB virus C (or their variants) in our patient samples. On the other hand, the cryoglobulinemic syndrome is known to affect mainly women after their fifth decade of life and probably after a long history of chronic HCV infection. Our recent study [4] of the prevalence of HGV RNA in Italian HIV-1-positive drug addicts showed that the prevalence of HGV infection was 23% in the sample as a whole and was higher in patients who had been exposed more recently to parenterally transmitted infection (HIV-1-positive, asymptomatic persons infected with both HIV-1 and HCV type 3a); this suggests that HGV infection is often self-limiting. It is difficult to interpret the significance of the prevalence observed by Misiani and colleagues in a sample of persons with no recent risk for parenterally transmitted infection. Nevertheless, we completely agree with the conclusion of these authors that HGV plays no role in the pathogenesis of mixed cryoglobulinemia.