Time to Detection of Hepatitis C Virus Infection With the Xpert HCV Viral Load Fingerstick Point-of-Care Assay: Facilitating a More Rapid Time to Diagnosis

Abstract

Background

Xpert HCV Viral Load Fingerstick assay (Xpert HCV VL FS) is a point-of-care test quantifying HCV RNA in <1 hour, enabling same-visit diagnosis and treatment.

Methods

This study evaluated time to HCV RNA detection using the Xpert HCV VL FS assay. Fingerstick whole-blood samples were collected from participants in an observational cohort in Australia.

Results

In May 2018–2019, 1468 participants were enrolled, 1426 had Xpert HCV VL FS testing performed, and 1386 had a valid result. HCV RNA was detected in 23% (325/1386). Among people with undetectable HCV RNA (n = 1061), median time to result was 57 minutes. Among people with detectable HCV RNA (n = 325), median time to HCV RNA detection was 32 minutes and 80% (261/325) had a detectable HCV RNA result in ≤40 minutes. Median time to HCV RNA detection was dependent on HCV RNA level.

Conclusions

A quicker HCV diagnosis could be achieved by monitoring the time when HCV RNA is first detected with the Xpert HCV VL FS test, rather than HCV RNA quantification, although the current platform does not allow for this. These findings could facilitate new strategies to reduce waiting times for an HCV diagnosis and improve linkage to treatment.

Viral hepatitis is the seventh leading cause of mortality globally, surpassing annual deaths due to HIV, malaria, and tuberculosis [1]. The development of simple therapies for hepatitis C virus (HCV) infection with cure rates >95% [2] is one of the greatest medical breakthroughs in decades. The World Health Organization (WHO) has set a goal to eliminate HCV as a major public health threat by 2030, with targets to increase HCV diagnoses and treatment, and reduce new infections and liver-related mortality [3]. Among the 71 million people with HCV worldwide, only 20% have been diagnosed and <10% treated [4]. Improving HCV testing and treatment to reduce disease burden is a key aim of HCV strategies globally [3, 5].

Low rates of HCV diagnosis and linkage to care are barriers to increasing treatment to achieve elimination. Current testing algorithms involve detection of HCV antibodies to confirm exposure, followed by HCV RNA or core antigen testing to detect active infection. This 2-step pathway requires multiple visits, leading to a drop-off in those receiving a diagnosis of active infection [6–12]. Barriers to testing include stigma and low HCV knowledge and awareness among providers and patients [13].

Point-of-care HCV tests increase testing uptake and linkage to care [13]. Most point-of-care HCV tests are limited in that they provide evidence of previous exposure (HCV antibodies) not active infection (HCV RNA or core antigen), which is a requirement for commencing treatment [13]. We undertook a series of studies that led to the first evaluation of the Xpert HCV Viral Load Fingerstick (Xpert HCV VL FS) point-of-care assay, enabling same-day diagnosis of active infection (HCV RNA) in 1 hour (sensitivity and specificity, 100%) [14, 15]. This research has led to a progressive development in HCV care, enabling same-visit HCV diagnosis and treatment [15].

Reducing the time from sample collection to diagnosis of active HCV infection during a single visit has the potential to improve patient acceptability of testing and linkage to treatment [16]. The Xpert HCV VL FS assay uses real-time polymerase chain reaction (RT-PCR) technology to quantify HCV RNA levels in under 1 hour. The time required to detect HCV RNA varies according to the level of virus present, with higher viral loads requiring a shorter time to detection and lower viral loads requiring a longer time to detection. There are no studies evaluating the time to HCV RNA detection for the Xpert HCV VL FS assay.

The aim of this study was to evaluate the time to HCV RNA detection using the Xpert HCV VL FS assay from capillary whole-blood collected by fingerstick among participants attending drug treatment and needle and syringe program (NSP) services in Australia.

METHODS

Study Participants

ETHOS Engage is an observational cohort study evaluating HCV testing, treatment, and HCV RNA prevalence among people who inject drugs (PWID) attending drug treatment clinics and NSPs. Between 28 May 2018 and 6 September 2019, participants were enrolled at 22 drug treatment clinics and 3 NSP services in Australia. The study protocol is provided in the Supplementary Material.

Inclusion criteria were age ≥18 years and lifetime injection drug use, with either recent injecting drug use (at least once in the last 6 months) or current receipt of opioid agonist therapy. Pregnant women were excluded. Participants received an AU$30 voucher for their participation. The study protocol was approved by St Vincent’s Hospital, Sydney Human Research Ethics Committee, and the Aboriginal Health and Medical Research Council.

Study Design, Intervention, and Study Assessments

Participants were provided information about the study while accessing services and consecutively enrolled into the study. Each clinic site held a series of campaign days (2 to 5 days). At enrollment, the following data and samples were collected: fingerstick capillary whole blood samples (100 µL for Xpert HCV VL FS assay), a self-administered survey on tablet computer (sociodemographic and drug use characteristics), liver stiffness measurement by transient elastography (FibroScan), and a clinical HCV assessment (performed by a nurse).

A capillary whole-blood sample was collected from participants via a fingerstick (Safety Lancet, Super Blade, No. 85.1018; Sarstedt) using procedures recommended by the WHO [17] and collected into a 100-µL minivette collection tube (Minivette POCT 100 µL K3E, No. 17.2113.101; Sarstedt).

Immediately following collection, 100 µL of capillary whole blood was placed directly into the Xpert HCV VL FS assay prototype cartridge (research use only; Cepheid; lower limit of detection 40 IU/mL, lower limit of quantification of 100 IU/mL, and upper limit of quantification 10⁸ log₁₀ IU/mL) for on-site HCV RNA testing. The cartridge was loaded into a GeneXpert instrument. The time to a quantitative HCV RNA result for Xpert HCV VL FS testing is 58 minutes [18].

All Xpert HCV VL FS assay testing was performed on a GeneXpert R2 6-colour, 4-module machine (GXIV-4-L System, 900-0513, GeneXpert Dx software v4.6a; Cepheid) operated by a trained member (Indika Jayasinghe, IJ) of the research team at the clinic site as per the manufacturer’s instructions (see Figure 1 for an example of the software output) [18]. As per the cartridge design, the high internal quantitative standard (IQS high, the high positive control) is 10⁶ copies (6 log₁₀) and the low internal quantitative standard (IQS low, the low positive control) is 10³ copies (3 log₁₀). Each kit lot of the Xpert HCV VL FS assay has a specific set of quantitation parameters (slope and intercept) built into the test cartridge. The kit lot specific quantitation parameters are determined through testing with the Cepheid HCV FS calibration panel (production calibrator, 5-member panel) using Xpert HCV VL FS assay determining delta cycle threshold (Ct) values, including the high internal quantitative standard (IQS high), the low internal quantitative standard (IQS low), and HCV Ct values. The calibration panel is value assigned to the Cepheid HCV secondary standard (working calibrator), which in turn is value assigned to the primary WHO fourth international standard [18]. Participants were not provided the result of their Xpert HCV test results, given that the Xpert HCV VL FS assay is not approved for clinical diagnostic use in Australia. Results were provided to clinic staff to inform subsequent clinical follow-up.

Statistical Analysis

The time to a detectable HCV RNA result using the Xpert HCV VL FS assay was evaluated by multiplying the Ct provided after the assay had completed (cycle at which the HCV RNA crosses the threshold and is considered detectable) by the cycle time (approximately 80 seconds). Information on the internal quantitative standards (high and low) was also assessed. Analyses were also performed to assess the time to HCV RNA detection in relation to the high and low internal standards (high and low internal controls), given that the positive standards always need to be positive (and reflects the minimum time to detection that is possible). When data were normally distributed, the mean was presented, while when data were not normally distributed, the median was presented.

The relationship between the time to a detectable HCV RNA result and the HCV RNA level (log IU/mL) was assessed using a scatter plot and stratified by categories of HCV RNA level (≥7 log, 6 to <7 log, 5 to <6 log, 4 to <5 log, 3 to <4 log, quantifiable 2 log to <3 log, and detectable but not quantifiable <2 log). The correlation of HCV RNA with time was assessed using Pearson correlation. Median HCV RNA levels were compared using the Wilcoxon-Mann-Whitney test. We also assessed the number and proportion of tests in this population sample that (1) could be detected at an earlier time (eg, equal to or above the positive high standard); (2) could be detected at an earlier or later time than low internal standards (eg, above or below lower standard); (3) fall below the lower internal standard but are still positive (eg, probably classified as detectable); and (4) need to wait the full assay run time to determine they are truly negative. Data were analyzed with Stata (version 14.0) and GraphPad Prism (version 7.03).

RESULTS

Participant Disposition

Among 1468 participants enrolled, 1426 had a fingerstick sample and 1386 had an available result. Reasons for not having a valid result (n = 40) included errors due to low sample volume (n = 21, 1.5%), invalid results due to the internal standards being out of range (n = 18, 1.3%), and other (n = 1, 0.1%).

Participant Characteristics

Among the 1426 participants with a fingerstick sample, the median age was 43 years, 35% (n = 503) were female, 74% (n = 1052) were currently receiving opioid agonist therapy, and 64% (n = 909) had injected drugs in the last month (Table 1). HCV RNA was detected by Xpert HCV VL FS assay in 325 of 1386 of participants with results (23%; 95% confidence interval [CI], 21–26). HCV RNA was <3 log₁₀ IU/mL (≥1000 IU/mL) in 2% (n = 32). Among people with HCV RNA levels 2 to <3 (log₁₀ IU/mL) (n = 17), 3 people were receiving HCV therapy.

Time to HCV RNA Detection

The overall median time to result was 32 minutes (range 24–55 minutes; interquartile range 30–37). The time to result was significantly faster among people with detectable HCV RNA (n = 325; median, 32 minutes) compared to people with undetectable HCV RNA (n = 1061; median, 57 minutes at completion of the assay; P < .001). Among people with detectable HCV RNA (n = 325), 80% (261/325) of participants had a detectable HCV RNA result in ≤40 minutes. As shown in Figure 2, there was a strong reverse linear relationship between HCV RNA levels (log₁₀ IU/mL) and elapsed time (r, −0.998; P < .001). Among people with detectable HCV RNA (n = 325), the median time to HCV RNA detection was longer in participants with lower HCV RNA levels (log₁₀ IU/mL): >7, 26 minutes (n = 3); 6 to <7, 29 minutes (n = 129); 5 to <6, 33 minutes (n = 100); 4 to <5, 38 minutes (n = 42); 3 to <4, 43 minutes (n = 19); 2 to <3, 49 minutes (n = 17); detectable but not quantifiable to <2, 53 minutes (n = 15) (Table 2 and Supplementary Figure 1). Among people with detectable HCV RNA (n = 325) 91% had a result ≥1000 IU/mL, with results available in ≤45 minutes.

Among participants with detectable HCV RNA (n = 325), the median time to result for the high internal quantitative standard (IQS high or the high positive control) was 31 minutes (range, 29–34 minutes) and the median time to result for the low internal quantitative standard (IQS low or the low positive control) was 45 minutes (range, 43–49 minutes) (Table 3).

Among participants with detectable HCV RNA, 36% (n = 118) had a result faster than the mean high internal quantitative standard (mean 31 minutes) and 80% (n = 261) had a result faster than the mean low internal quantitative standard (mean 45 minutes) (Figure 3).

Discussion

This is the first study to report the time to HCV detection with the Xpert Fingerstick HCV Viral Load assay in whole blood collected by fingerstick. The median time to HCV RNA detection was faster in people with detectable HCV RNA (32 minutes) as compared to people with undetectable HCV RNA (57 minutes). These findings suggest that an HCV diagnosis could be inferred by monitoring the time at which HCV RNA is first detected with the Xpert HCV VL FS test, although, in its current format, the assay must be completed before a diagnosis is made. These findings highlight opportunities for new HCV RNA formats, including an assay that focuses on HCV RNA detection rather than quantification, which may reduce times for an HCV diagnosis and improve linkage to treatment. A quicker time to HCV RNA detection and diagnosis would provide a major advance over the current fingerstick HCV RNA test and would impact on clinical HCV management, enabling the implementation of more realistic same-visit test and treat interventions.

The median time to result was faster among people with detectable HCV RNA (32 minutes) compared to those with undetectable HCV RNA (57 minutes). Further, 91% of participants with detectable HCV RNA in this cohort had an HCV RNA result ≥1000 IU/mL (clinically relevant lower limit of detection as determined by European Association for the Study of the Liver guidelines [19]) with HCV detection occurring in under 45 minutes. Similarly, 80% had a result faster than the mean low internal quantitative standard, which was 45 minutes. This is important, given that this quantitative assay requires both the high and low internal quantitative standards to be completed for a valid test result. Although these findings are novel in the setting of detection for HCV infection, the results are consistent with the ability of Xpert assays to detect and diagnose infectious diseases more quickly than the time required for quantification (eg, Trichomonas vaginalis, group B Streptococcus for intrapartum testing, norovirus, and influenza) [20–23]. Further work is needed to explore the potential of an Xpert HCV VL FS assay based on HCV RNA detection rather than HCV RNA quantification.

The Xpert HCV VL FS test was originally developed as a quantitative assay in an era when the quantification of HCV RNA had relevance for HCV clinical management [24]. With interferon-based therapy, treatment duration could be individualized based on changes in HCV RNA levels during therapy [24]. Further, with certain genotype-specific direct-acting antiviral therapies (eg, sofosbuvir-ledipasvir), pretreatment HCV RNA levels could be used to inform decisions about shortening the duration of HCV therapy [19]. However, pangenotypic HCV therapies with cure rates >95% are now available and HCV RNA levels prior to or during treatment do not provide any benefit for individualizing therapy. In the current era, it is sufficient to have an assay that can detect HCV RNA for diagnosis of infection, confirmation of cure, and monitoring of HCV reinfection following successful treatment [19].

Theoretically, it would be possible for an operator to use the current GeneXpert platform and Xpert HCV VL FS test to identify participants with detectable HCV RNA prior to assay completion. During the assay, it is possible to wait for the high internal quantitative standard to amplify first as a first-step quality control check (eg, using 30–50 fluorescence units as a cutoff with an expected Ct range). However, further extensive validation is required to determine a suitable fluorescence cutoff. With samples that are HCV RNA undetectable, there should be no/limited fluorescence for the HCV target. With samples that are HCV RNA detectable, the operator could use the same 30–50 fluorescence units as a cutoff with an expected Ct range.

However, there are a number of limitations with this approach and the results should not be provided until the assay has completed. First, and foremost, use of a detectable result prior to assay completion would be an off-label use of this diagnostic product in a manner not approved by regulatory authorities. Currently, the assay needs to be fully completed to provide a definitive diagnosis of HCV infection. The use of such a test off-label would require careful considerations around appropriate counseling for people who are suspected of having detectable HCV RNA. There are also concerns about providing someone with a diagnosis prior to the completion of the assay (particularly in the event of an invalid test result). Further, each cartridge is a standalone assay and there is a slight fluctuation (5–6 minutes) in the high and low internal quantitative standards between different participants. HCV RNA levels are assessed by quantitation parameters (slope and intercept) built into each test cartridge, calibrated within each lot using the Xpert HCV VL FS 5-member calibration panel. The exact Ct for HCV RNA detection is specific to each kit lot and is unknown to the operator. As such, there is no mechanism for the operator to determine the exact Ct at which HCV RNA is detected. This information is only available upon assay completion; therefore there is currently no visual notification or alarm to indicate when the sample is detectable. Further, visually inspecting the curves requires active monitoring of the system, particularly if multiple samples are being run at once. The assay must complete its run time before it is possible to download the result (PDF report or electronic export). As such, no result can be provided until the assay has completed. In addition to having a trained operator who is familiar with the software, it is also critical to have someone who can be trained in providing appropriate posttest counseling. There are laboratory quality control data management software available, allowing the ability to track internal quantitative standard fluctuations and develop normal Ct ranges to ensure optimal and accurate cartridge and platform performance. Moving forward, it is important to explore how the current Xpert HCV VL FS test could be modified to facilitate a more rapid time to diagnosis.

Xpert assays for other infectious diseases (eg, Trichomonas vaginalis, Group B Streptococcus, norovirus, and influenza) have an “early assay termination” feature in situations where quantitation is not required, enabling detectable results to be reported more quickly. This quicker turnaround time for the GeneXpert platform using the early assay termination feature is important in high-risk settings (eg, hospital emergency room) or in settings where a shorter time to result could improve patient management (eg, the same visit). Conformité Européene (CE) registration of a new Xpert HCV VL FS assay with an early assay termination feature would require substantial work and investment, including further verification and validation studies (development of another version of the cartridge from scratch, designed as a qualitative assay with appropriate internal controls, instead of a quantitative assay), but is something that the manufacturer, Cepheid, could explore moving forward. Alternatively, it should be technically possible from the back end of the software to note the time at which HCV RNA becomes detectable and provide this result to the operator in real time, with checks in place to ensure quality and accuracy is not compromised. In either case, negative samples will likely still require the assay to run to completion for confirmation and the assay is only terminated early in samples that are detectable. A qualitative assay with an early termination feature would require revalidation to confirm diagnostic performance (particularly at HCV RNA levels 2 to <3 logs [100–1000] to eliminate the possibility of false-positive results), but it is unlikely sensitivity and specificity would be affected. The potential challenges of adapting this retrospective analysis to a prospective test exploiting an early assay termination feature is feasible, demonstrated by assay kits already incorporating this feature [20–23].

Although the Xpert HCV VL FS assay represents a major advance in HCV management by enabling the diagnosis of active HCV infection in a single visit, a time to result in 1 hour is still too long in many settings. In practice, by the time the assay has completed and discussions around HCV treatment options with a practitioner have finished, it could be greater than 90 minutes. This may be suitable in some settings, but may not be ideal for the delivery of a test and treat model, such as settings providing services for more marginalized populations, as in PWID (eg, drug treatment clinics, community health centers, needle and syringe programs, and pharmacies). Strategies to reduce the time from testing to diagnosis may be an important component of a single visit test and treat model for HCV infection. As such, exploring options for an early assay termination version of the Xpert HCV VL FS assay should be explored. Other assays are also in development, such as the BLINK ONE cartridge (BLINK Diagnostics) HCV assay prototype, with a time to detectable HCV RNA result in under 20 minutes [25]. Shortening the time to HCV diagnosis and treatment would be a major clinical advance towards achieving a true test and treat model for HCV infection.

This study has several limitations. First, the number of samples with low HCV RNA levels was limited. Given the low prevalence of HIV infection among PWID in Australia [26], there were few people with HCV-HIV coinfection. Also, HCV genotype was not performed. Despite this being one of the first studies to investigate the time to HCV detection using this assay, further validation work is needed across larger populations with a broader range of HCV RNA levels, HIV coinfection, and known HCV genotypes. Also, all testing was performed by a trained operator with considerable experience in operating the GeneXpert R2 6-colour, 4-module machine. As such, the low proportion of errors and invalid results may not reflect what might be observed in a real-world clinical setting among operators without such experience.

Irrespective of these limitations, this is one of the first studies to demonstrate a proof of concept that the Xpert HCV VL FS assay could be modified to provide a faster time to diagnosis by focusing on HCV detection, rather than HCV quantification. Further work is required by manufacturers to explore the potential of HCV assays (including the Xpert HCV VL FS) based on HCV RNA detection rather than HCV RNA quantification. Regardless of the approach taken, by either converting the current quantitative assay or designing a new, custom qualitative cartridge, significant further verification and validation work would be required to fulfil CE in vitro diagnostics registration requirements. The availability of an HCV assay offering a faster time to HCV diagnosis would be a significant improvement over existing point-of-care HCV RNA assays by reducing wait times and improving linkage to HCV treatment, thereby facilitating efforts to eliminate HCV infection.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Author contributions. J. G., B. C., I. J., B. H., M. M., and T. L. A. contributed to the study design. J. G. and T. L. A. were the study investigators. I. J. and B. C. contributed to the laboratory work. All authors contributed to the study implementation and conduct, data interpretation, and writing and review of the article.

ETHOS II Protocol Steering Committee members: Nicky Bath (Chair, LGBTI Health Programming and Development), Jason Grebely (UNSW Sydney), Gregory Dore (UNSW Sydney), David Silk (UNSW Sydney), Carla Treloar (UNSW Sydney), Andrew Milat (NSW Health), Adrian Dunlop (Hunter New England Local Health District), Janaki Amin (Macquarie University), Jo Holden (NSW Health), Charles Henderson (NUAA), Kyle Leadbeatter (Hepatitis NSW), Emma Day (ASHM), Louisa Degenhardt (UNSW Sydney), and Phillipp Reid (KRC).

Acknowledgments. The authors thank the study participants for their contribution to the research, as well as researchers and staff who have participated in the project. The authors acknowledge Cepheid, including Marika Klemen, David Persing, Rayden Rivett, and Lazarela Vucinic, for technical assistance. The authors gratefully acknowledge the pivotal role played by the following partner organizations and key stakeholders in study planning and implementation: NSW Health; Hepatitis NSW; NSW Users and AIDS Association; Australasian Society for HIV, Viral Hepatitis, and Sexual Health Medicine.

Disclaimer. The sponsors had no role in the analysis and interpretation of the study results. The contents of the published material are solely the responsibility of the individual authors and not do not reflect the views of NHMRC or New South Wales Health. The Kirby Institute is funded by the Australian Government Department of Health and Ageing. The views expressed in this publication do not necessarily represent the position of the Australian Government.

Financial support. The work was supported by the Australian Government Department of Health and Ageing and New South Wales Health through a National Health and Medical Research Council (NHMRC) Partnership Project Grant (grant number APP1103165); NHMRC fellowships to G. D., J. G., B. H., and M. M.; and by Cepheid (provision of GeneXpert equipment and Xpert Fingerstick HCV Viral Load cartridges. The Xpert Fingerstick HCV VL FS cartridge was developed by Cepheid in collaboration with Foundation for Innovative New Diagnostics).

Potential conflicts of interest. J. G. reports grants from Cepheid during the conduct of the study; grants from Abbvie, Cepheid, and Hologic outside the submitted work; and grants and personal fees from Gilead Sciences and Merck outside the submitted work. M. M. reports personal fees from Abbvie outside the submitted work. G. J. D. reports grants from Cepheid during the conduct of the study and outside the submitted work; grants, personal fees, and nonfinancial support from Gilead, Abbvie, Merck, Bristol-Myers Squibb, Janssen, and Roche; and personal fees from GlaxoSmithKline and Abbott Diagnostics. T. L. A. reports grants and personal fees from Cepheid during the conduct of the study; and grants from Abbott Diagnostics outside the submitted work. All other authors report no potential conflicts.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References