HCV core antigen as an alternative to HCV RNA testing in the era of direct-acting antivirals: retrospective screening and diagnostic cohort studies
The Lancet Gastroenterology and Hepatology , Volume 3 - Issue 12 p. 856- 864
Background: Direct-acting antivirals for chronic hepatitis C (HCV) infection have reduced the need for on-treatment HCV RNA monitoring. We assessed the accuracy and cost implications of using HCV core antigen testing to replace HCV RNA testing for confirmation of diagnosis, on-treatment monitoring, and determination of sustained virological response (SVR). Methods: In a retrospective screening cohort study, de-identified residual serum from unselected samples were obtained from commercial laboratories in Ontario, Canada. Samples from each 5-year age-sex band from birth years 1945–74 collected from Aug 1, 2014, to Feb 28, 2015, were included. All samples that tested positive for HCV antibodies, and 10% of samples that tested negative for HCV antibodies, were tested for HCV core antigen and HCV RNA. A retrospective clinical cohort study was also done using blood samples from patients with confirmed HCV infection collected at four tertiary academic centres: one in Canada, two in Germany, and one in the USA. For assessment of SVR, we included samples from patients who started direct-acting antiviral-based treatment (excluding telaprevir and boceprevir) with or without peginterferon, ribavirin, or both, from Jan 1, 2014, to March 31, 2015. To ensure inclusion of adequate numbers for analysis, patients who relapsed after any treatment regimen were included. Serum samples included in the study were from baseline, week 4 on-treatment (only for patients treated with direct-acting antivirals), end of treatment, and week 12 or 24 of follow-up. The sensitivity and specificity of core antigen testing as a diagnostic tool was assessed in the screening cohort, using HCV RNA as a reference. The sensitivity and specificity of core antigen testing as well as its concordance with HCV RNA testing in the clinical cohort was assessed at baseline, week 4 on-treatment, and at weeks 12 or 24 after the end of treatment in patients undergoing therapy with direct-acting antivirals. The cost-effectiveness of core antigen testing with and without confirmatory HCV RNA testing for negative samples was also assessed. Findings: From 10 006 samples in the screening cohort, 75 of 80 viraemic (HCV RNA-positive) samples tested positive for HCV core antigen (sensitivity 94%, 95% CI 86–98), and none of the 993 HCV RNA-negative samples tested positive for HCV core antigen (specificity 100%, 95% CI 94–100). The five viraemic samples that tested negative for HCV core antigen had low corresponding HCV RNA concentrations. In the clinical cohort, two (1%) of 202 baseline samples tested negative for HCV core antigen; one had a low HCV RNA concentration (468 IU/mL), the other had a high HCV RNA concentration (>2 000 000 IU/mL). By week 4 of treatment, HCV core antigen concentrations decreased in all patients but were not predictive of SVR. Although there was good concordance between HCV RNA and HCV core antigen results at 12 weeks after the end of treatment (r=0·97; p<0·0001), three of the 148 patients who achieved SVR at 12 weeks tested HCV core antigen positive. 12 weeks after the end of treatment, HCV core antigen was undetectable in one (1%) of 71 samples from patients who were identified as having relapsed according to HCV RNA detection. On-treatment and end-of-treatment testing of core antigen or HCV RNA provided little clinical value. The use of HCV core antigen testing as a confirmatory diagnostic strategy was cost saving relative to HCV RNA testing, with a reduction of CAD$0·29–3·70 per patient screened depending on whether HCV RNA testing was used to confirm HCV core antigen-negative results. Interpretation: These data support the use of HCV core antigen testing to document HCV viraemia in a cost-saving diagnostic algorithm. In a treatment setting, HCV core antigen testing can be used instead of HCV RNA testing for diagnosis and documentation of treatment adherence, but it might not be adequate to determine SVR. This approach might improve access to care, particularly in low-income and middle-income countries. Funding: Abbott Diagnostics and Toronto Centre for Liver Disease.