Antibody screening tests variably overestimate the prevalence of hepatitis C virus infection among HIV-infected adults in Ghana

Summary HIV coinfection with HCV has been poorly studied in sub-Saharan Africa, and the reliability of available seroprevalence estimates remains uncertain. The study aim was to determine HCV RNA prevalence in HIV-infected subjects receiving care in Kumasi, Ghana, and relate the findings to HCV antibody detection. From a population of 1520 HIV-infected adults, all HBsAg-positive subjects (n = 236) and a random subset of HBsAg-negative subject (n = 172) were screened for HCV RNA using pooled plasma; positive samples were genotyped by core and NS5B sequencing. HCV antibodies were detected by three commercial screening assays and confirmed by the line immunoassay. HCV RNA was detected in 4/408 subjects (1.0%, 95% confidence interval 0.0-1.9%), comprising 3/236 (1.3%; 0.0-2.8%) HBsAg-positive and 1/172 (0.6%; 0.0-1.8%) HBsAg-negative subjects. HCV RNA-positive subjects showed reactivity in all three antibody screening assays. Among HCV RNA-negative subjects, 5/67 (7.5%), 5/67 (7.5%) and 19/67 (28.4%) showed antibody reactivity by each screening assay, respectively, including two (3.0%) with reactivity by all three assays. Only one sample (1.5%) had confirmed antibody reactivity by line immunoassay indicating past HCV infection. HCV-positive subjects (three males, two females) were aged 30-46 years, by questionnaire-based interview reported surgical procedures and blood transfusion as risk factors for infection. HCV genotypes were 2 (subtypes 2j, 2l, 2k/unassigned) and 1 (subtype unassigned). Without further testing, HCV antibody screening assays variably overestimated HCV prevalence among HIV-infected subjects in Ghana. These findings inform the interpretation of previous seroprevalence estimates based upon screening assays alone.