We investigated the prevalence of hepatitis C virus (HCV) in saliva because it is still controversial [1-3] and because lymphocytic sialadenitis of the Sjogren syndrome was associated with HCV infection in previous reports [4]. Serum and saliva samples were obtained on the same day from 28 patients with HCV infection and HCV viremia (13 of them were also infected with the human immunodeficiency virus [HIV]). Whole saliva was collected after the patients chewed gum for 10 minutes. Polymerase chain reaction using the Amplicor HCV test (Roche Diagnostic Systems, Branchburg, New Jersey) was used to detect HCV RNA. Genotyping of HCV was done using the HCV InnoLIPA assay (Innogenetics, Zwijndrecht, Belgium). Quantification of serum HCV RNA was done using the branched DNA assay (Chiron Corporation, Emeryville, California) or serial dilutions of polymerase chain reaction products with buffer before detection with the Amplicor HCV assay.

Seventeen patients (61%) had detectable HCV RNA in saliva. Six HCV genotypes were represented in saliva, including 1 (n = 1); 1a (n = 6); 1b (n = 5); 2a (n = 3); 3 (n = 1); and 4 (n = 1). Hepatitis C virus RNA was present in saliva with the same frequency among patients positive (7 of 13) and negative (10 of 15) for HIV infection. The best predictive factor for the presence of HCV in saliva was a high level of serum HCV RNA with the Chiron assay: Mean ±SD levels of serum genome equivalents were 75 x 10⁵ eq/mL ±62 x 10⁵ eq/mL and 25 x 10⁵ eq/mL ±29 x 10⁵ eq/mL in patients with and without HCV in saliva, respectively (P = 0.02). Likewise, a high titer of serum HCV RNA with the semi-quantitative assay of 10⁵.5 ±0.7 and 10⁴.1 ±1.1 in patients with and without HCV in saliva, respectively (P = 0.01) was linked to saliva infection. Moreover, three additional patients with antibodies to HCV antigen who did not have HCV viremia, had no HCV RNA in their saliva.

Hepatitis C virus in saliva might lead to lymphocytic sialadenitis in selected patients. However, whether HCV may infect salivary epithelial cells is still unknown. Because HCV cannot be detected in the saliva of patients without a high level of HCV viremia, the oropharynx is probably not a site of latency or of HCV replication. In fact, the presence of HCV in saliva may only reflect contamination of saliva with plasma, but it may also have a role in horizontal nonblood transmission of the virus, the mechanisms of which remain largely unknown. Indeed, if 27% of the spouses of HCV-infected patients have HCV-associated antibodies [5], sexual transmission of HCV remains debatable because HCV RNA is not detected in semen [2] and because HCV markers are not detected in spouses who have been married for less than 10 years [5].