Current guidelines for the screening and diagnosis of HCV denote using a relatively cheap screen with anti-HCV anti-body immunoassay, followed by the much costlier molecular test for HCV RNA levels using PCR assay to confirm active HCV infection. HCV Ag assays, which are easier to perform than real time-PCR also save time and are less expensive. Previous cost-effectiveness analysis showed that screening and evaluating HCV viremia using a strategy in which the two-step process is replaced by a one-step process could result in a net cost saving up to $44 per person screened [8, 9].
That is why, there has been a push for a one-step HCV antigen test to diagnose active HCV infection [10].
Detection of HCV Ag was initially described in late 90s. However, first generation antigen tests had soon become unpopular due to their lack of sensitivity [7]. The chemiluminescent immunoassay based new generation HCV Ag test used in this study has a sensitivity of ≤ 3 fmol/L, offering 16- to 25-fold sensitivity compared to preceding tests [11].
In our study, it was found that there was strong correlation between HCV RNA levels and HCV Ag levels. HCV Ag levels were detected, with a corresponding increase in HCV Ag levels to increased titer of viral RNA, with a high correlation coefficient of 0.958 (p < 0.001). Similar results were reported by various other studies [10,11,12]. Due to the excellent correlation between HCV Ag and HCV RNA concentrations, detection of HCV Ag in serum or plasma is useful as an indirect marker of HCV replication [13].
Sensitivity of the HCV Ag test, in the present study, was 95.05% compared to that of the RNA test. Previous studies which utilized the same HCV Ag test have reported a sensitivity of 75.8–99.5% for the HCV Ag test [11, 12, 14,15,16].
In our study, negative results were obtained with HCV Ag test for five samples with viral load of 38–813 IU/mL. In a study by Çetiner and colleagues [11], seven samples with positive HCV RNA results yielded negative HCV Ag results and viral load were lower than 10,000 IU/mL for these patients (six samples with viral loads of 17–178 IU/mL and one sample with a viral load of 2500 IU/mL). In another study by Ergünay and colleagues [15], the viral load was lower than 103 IU/mL in samples which were negative for the HCV Ag test.
In the present study, specificity was found to be 100%. Ross and colleagues [17], Kesli and colleagues [12], Park and colleagues [18], Çetiner and colleagues [11] and Chang and colleagues [10] have also found a specificity level of 100%.
In the present study, PPV was 100% and there is no false positivity with the HCV Ag test. Same result was found by Çetiner and colleagues [11].
However, our negative predictive value (61.54%), similar to result find by Abdelrazik and colleagues [13], was lower than those of the other studies [12, 19]. This result was probably due to the small number of RNA negative samples, that is, 8 out of 109.
Consequently, all positive results found by the HCV Ag assay were also positive with the HCV RNA assay. However, all negative results found by the HCV core Ag assay were not negative with the HCV RNA assay. Thus, it can be concluded that a positive test with the HCV Ag (almost) always represents a true positive since it is highly specific. However, when there is a serum sample showing anti-HCV positivity, the negative results found by the HCV core Ag assay should be also confirmed by a HCV RNA assay [13].
There is increasing, albeit insufficient, evidence that genotype may affect antigen and RNA assay detection results to different extents [20]. In our study, genotype 1 was the predominant detected genotype, corresponding to the genotypic distribution of HCV in Tunisia [21,22,23]. Specially, there was a strong correlation between HCV RNA levels and HCV Ag levels in genotype 1 (r = 0.966). Similar result was found by Chang and colleagues (r = 0.945) [10]. In contrast, the correlation coefficient was considerably higher than those reported in previous study [24, 25]. This discordance may be explained by the variability of studied population and the use of different assays.