Background Chronic hepatitis C virus HCV infection

Background Chronic hepatitis C virus (HCV) infection is a major public health concern, with 71 million people infected worldwide [1]. Treatment options have improved with the availability of interferon-free direct-acting antiviral (DAA) therapies with cure in >95% of people [2]. However, broadening the availability of DAA therapies will be vital in achieving the WHO target of eliminating HCV as a major global public health threat by 2030 [3]. One of the major challenges will be to address the low rates of HCV testing and diagnosis [4]. Simplifying models of HCV care by reducing the number of visits and time to result is necessary to increase HCV testing and treatment uptake [5]. Alternative sampling and diagnostic options, such as dried blood spots (DBS), have been shown to increase testing and linkage to care [6,7]. Sampling involves the finger-stick collection of a capillary whole-blood sample (50–100 μl, typically 70 μl), onto a filter paper. DBS samples allow for reflex viral testing, testing of other viral analytes (such as HIV), they are easy to biobank and store, they have good stability, and require no cold chain transport. DBS sampling also avoids the need for venepuncture, which is a barrier to sample collection and testing in settings where venepuncture is unavailable or presents a major barrier to testing, such as among people who inject drugs, who often have poor venous access [5,8]. The Aptima HCV Quant Dx assay (subsequently referred to as Aptima) for the detection of HCV RNA uses real-time TMA technology which has many advantages [9], such as integrated platform automation and random access with continuous sample and reagent loading. It is CE-marked and has previously demonstrated similar performance to other platforms, such as the COBAS TaqMan HCV2 assay or Abbott RealTime HCV assay [[9], [10], [11], [12]]. However, there is a gap in the literature for evaluating the performance of HCV RNA assays on DBS samples. HCV serological and nucleic allopurinol zyloprim testing from DBS samples has been successfully performed within clinical trials, epidemiological surveillance programs, screening programs and in limited clinical settings [[13], [14], [15], [16], [17], [18], [19], [20]]. However, there are currently no registered assays for HCV RNA testing from DBS samples. As such, there is a growing need for validated, standardized manufacturers protocols and registration of HCV quantitative and qualitative molecular diagnostic assays with DBS as a sample type [1].
Objectives
Study design
Results The sensitivity of the Aptima assay for HCV RNA detection in DBS was 96.4% (95% CI 89.8–99.3%) and specificity was 95.8% (95% CI 78.9–99.9%) when compared to the CAP/CTM assay (Table 1). Sensitivity for HCV RNA detection in DBS above a quantifiable threshold (≥15 IU/mL in plasma) was 95.1% (95% CI 88% to 98.7%) and specificity was 96.0% (95% CI 79.7% to 99.9%) (Table 2). The sensitivity of HCV RNA detection ≥1000 IU/mL in DBS (based on a clinically relevant threshold obtained from EASL guidelines) was 100% (95% CI 95.3–100%) and specificity was 100% (95% CI 88.4–100%) (Table 2). A similar sensitivity and specificity of the Aptima assay for HCV RNA detection in DBS compared to Aptima HCV RNA detection in plasma was observed across all three of the above thresholds (Supplementary Tables 1–2). CAP/CTM assay results in plasma were in agreement for 102/107 (95.3%) with Aptima DBS (23 undetectable, 80 detectable) (Table 1). Three samples were ‘undetectable’ in DBS and were detectable at very low HCV RNA concentrations in plasma (27 IU/mL, 22 IU/mL, 20 IU/mL); one DBS sample was detectable, but below the LLOQ, while the paired plasma sample result was quantified at 33 IU/ mL and one DBS sample was detectable (13 IU/mL) while the paired plasma sample result was ‘undetectable’. Correlation between paired plasma and DBS tested on the Aptima HCV Dx Quant assay (HCV RNA detectable in plasma (n = 86)) was assessed using Deming regression analysis (Fig. 1). This demonstrated a strong, positive correlation (R2 = 0.929) that was stastically significant (p < 0.001). Bland-Altman plot analysis (Fig. 2) showed a bias of +0.51 log10 between plasma and DBS HCV RNA concentrations (n = 86). The limits of agreement indicate that 95% of the difference between plasma and DBS on the Aptima are between -0.48 and 1.50 log10. Both Deming and Bland-Altman anlysis showed that HCV RNA tended to be higher in plasma than DBS for samples with HCV RNA above 4.5 log10 IU/mL. Importantly, HCV RNA in DBS could be quantitated as low as 525 IU/mL (2.72 log10 IU/mL), with agreement between sample types.