Advertisement

Reducing Read Time of Point-of-Care Test Does Not Affect Detection of Hepatitis C Virus and Reduces Need for Reflex RNA

  • Author Footnotes
    a Authors share co-first authorship.
    David Smookler
    Footnotes
    a Authors share co-first authorship.
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Author Footnotes
    a Authors share co-first authorship.
    Aaron Vanderhoff
    Footnotes
    a Authors share co-first authorship.
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Author Footnotes
    a Authors share co-first authorship.
    Mia J. Biondi
    Footnotes
    a Authors share co-first authorship.
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Jorge Valencia
    Affiliations
    Harm Reduction Unit (SMASD), Subdireccion General de Adicciones, Madrid, Spain
    Search for articles by this author
  • Pablo Ryan
    Affiliations
    Internal Medicine/HIV Unit, Infanta Leonor Hospital, Madrid, Spain
    Search for articles by this author
  • Joel Karkada
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Rachel Hong
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Izza Sattar
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Erin Mandel
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Martina Gjevori
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Julia Casey
    Affiliations
    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • David Fletcher
    Affiliations
    Maple Leaf Clinic, Toronto, Ontario, Canada
    Search for articles by this author
  • Hemant Shah
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Bettina E. Hansen
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada

    Institute of Health, Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Camelia Capraru
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Harry L.A. Janssen
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Jeffrey V. Lazarus
    Affiliations
    Barcelona Institute for Global Health (ISGlobal), Barcelona, Spain
    Search for articles by this author
  • Jordan J. Feld
    Correspondence
    Reprint requests Address requests for reprints to: Jordan J. Feld, MD, MPH, Toronto Centre for Liver Disease, Toronto General Hospital, University Health Network, 200 Elizabeth Street 9EB-240, Toronto, Ontario, Canada M5G 2C4. fax: (416) 340-5133.
    Affiliations
    Viral Hepatitis Care Network (VIRCAN), Toronto, Ontario, Canada

    Toronto Centre for Liver Disease, University Health Network, University of Toronto, Toronto, Ontario, Canada
    Search for articles by this author
  • Author Footnotes
    a Authors share co-first authorship.
Open AccessPublished:August 03, 2020DOI:https://doi.org/10.1016/j.cgh.2020.07.058

      Background & Aims

      Global elimination of hepatitis C virus (HCV) will require increases in diagnosis. Point of care (POC) tests that detect antibodies against HCV can be useful for testing large and difficult to reach populations. The most accurate POC test requires a 20 min read time to identify antibody-positive samples. We investigated whether viremic patients could be identified using a shorter read time, to increase efficiency and reduce the need for reflex tests (a follow-up test for HCV RNA on the same specimen to confirm viremia).

      Methods

      Patients with past or current HCV infections provided samples at 2 clinics in Canada for evaluation by the OraQuick HCV Rapid antibody POC test. A community HCV-screening program in Madrid, Spain (real-world cohort) invited people to be tested for HCV with the same OraQuick test. Patients provided samples of whole blood, via finger prick. Fingerprick samples were tested immediately after collection. In the clinic cohort, photographs of the developing test were taken at 15 second intervals, and blinded readers recorded the time to positivity. In the real-world cohort, readers recorded the OraQuick result at 5 minutes, and each minute after, up to 10 minutes, and then again at 20 minutes; viremia was then evaluated using a POC HCV RNA test (GeneXpert HCV Viral Load Assay). Sera from viremic and non-viremic clinic patients were used to quantify antibody titers to investigate the relationship between the time of band appearance and antibody concentration. Fisher’s exact test and exact logistic regression were used to determine factors associated with a positive result at 5 minutes.

      Results

      Blood from all viremic patients produced a positive result in the antibody POC test by 5 min. Median time to a positive result for 171 viremic patients was 2.6 min (range, 1.8–4.6 min), vs 4.1 min (range, 2.3–14.4 min) for 108 patients with resolved infection (P < .001). The 5-min threshold identified all viremic cases among 176 HCV antibody–positive patients in the real-world cohort, confirmed by testing for HCV RNA. In the pooled cohorts, antibody positivity at 5 min identified viremic patients with 100% sensitivity (95% CI, 98.4%–100%); the negative predictive value was 100% (95% CI, 94.9%–100%). The positive predictive value at 5 min was 62.0% (95% CI, 56.7%–67.0%) and therefore insufficient alone to detect viremia; an HCV RNA test would still be necessary to confirm active infection.

      Conclusions

      The wait time for the OraQuick HCV Rapid antibody POC blood test can be reduced from 20 min to 5 min and continue to reliably identify patients with HCV infection. Shortening the test time could increase high-throughput screening, reduce loss to follow up, and reduce the need for reflex HCV RNA testing.

      Key words

      Abbreviations used in this paper:

      CI (confidence interval), DAA (direct-acting antiviral), HCV (hepatitis C virus), HIV (human immunodeficiency virus), L-SVR (long-term sustained virologic response), NPV (negative predictive value), OQ test (OraQuick hepatitis C virus rapid antibody test), POC (point of care), PPV (positive predictive value), R-SVR (recently achieved sustained virologic response), SVR (sustained virologic response), TCLD (Toronto Centre for Liver Disease), tpos (time to positivity), VIRCAN (Viral Hepatitis Care Network)

       Background

      Worldwide elimination of hepatitis C virus (HCV) will require increases in diagnosis. Point-of-care tests for antibodies against HCV can be useful for screening large and difficult-to-reach populations. However, the most accurate point-of-care test requires a 20-minute read time to identify antibody-positive samples; it is not clear whether these results can be read in a shorter time period for viremic patients.

       Findings

      Reducing read time of the OraQuick HCV rapid antibody test from 20 minutes to 5 minutes causes no loss of antibody detection in viremic patients. Antibody positivity at 5 minutes identified viremic patients with 100% sensitivity and a negative predictive value of 100%.

       Implications for patient care

      Shortening of the OraQuick HCV rapid antibody point-of-care test from 20 minutes to 5 minutes could improve high-throughput screening, reduce loss to follow-up, and reduce the need for follow-up HCV RNA tests.
      Following the success of direct-acting antiviral (DAA) treatment for hepatitis C virus (HCV), the World Health Organization proposed the goal of eliminating viral hepatitis as a public health threat by 2030.
      • Cooke G.S.
      • Andrieux-Meyer I.
      • Applegate T.L.
      • et al.
      Accelerating the elimination of viral hepatitis: a Lancet Gastroenterology and Hepatology Commission.
      This will require dramatic increases in screening efforts world-wide. Simplification of the HCV diagnostic pathway would enhance overall linkage to care, particularly among priority populations with a high burden of HCV infection, such as people who inject or people who use drugs.
      • Lazarus J.V.
      • Pericàs J.M.
      • Picchio C.
      • et al.
      We know DAA s work, so now what? Simplifying models of care to enhance the hepatitis C cascade.
      ,
      • Bajis S.
      • Dore G.J.
      • Hajarizadeh B.
      • Cunningham E.B.
      • Maher L.
      • Grebely J.
      Interventions to enhance testing, linkage to care and treatment uptake for hepatitis C virus infection among people who inject drugs: a systematic review.
      Point-of-care (POC) HCV antibody tests are designed to be rapid, easy to use, and minimally invasive. These tests avoid the need for phlebotomy, typically requiring only fingerprick blood; as such, the tests are readily administered by individuals with limited medical training, allowing for task shifting to community workers and peers in outreach settings, greatly expanding the reach of HCV testing.
      • Bajis S.
      • Dore G.J.
      • Hajarizadeh B.
      • Cunningham E.B.
      • Maher L.
      • Grebely J.
      Interventions to enhance testing, linkage to care and treatment uptake for hepatitis C virus infection among people who inject drugs: a systematic review.
      • Keats J.
      • Micallef M.
      • Grebely J.
      • et al.
      Assessment and delivery of treatment for hepatitis C virus infection in an opioid substitution treatment clinic with integrated peer-based support in Newcastle, Australia.
      • Morgan K.
      • Lee J.
      • Sebar B.
      Community health workers: A bridge to healthcare for people who inject drugs.
      • Treloar C.
      • Rance J.
      • Bath N.
      • et al.
      Evaluation of two community-controlled peer support services for assessment and treatment of hepatitis C virus infection in opioid substitution treatment clinics: the ETHOS study, Australia.
      • Henderson C.
      • Madden A.
      • Kelsall J.
      ‘Beyond the willing &the waiting’ — the role of peer-based approaches in hepatitis C diagnosis and treatment.
      A systematic review of 14 high-performance POC tests for HCV antibody rated the OraQuick HCV rapid antibody test (OQ test) (OraQuick, Bethlehem, PA) highest for sensitivity (99.5%) and specificity (99.8%).
      • Khuroo M.S.
      • Khuroo N.S.
      • Khuroo M.S.
      Diagnostic accuracy of point of care tests for hepatitis C virus infection: a systematic review and meta-analysis.
      However, the OQ test is read between 20 and 40 minutes,
      OraSure Technologies - OraQuick HCV Rapid Antibody Test.
      making it one of the slowest “rapid” tests available. Among populations that are difficult to reach, a 20-minute delay may discourage testing or contribute to attrition.
      We hypothesized that such a long delay may be unnecessary for viremic patients. Because antibody levels typically decrease in the absence of viral antigen,
      • Amanna I.J.
      • Carlson N.E.
      • Slifka M.K.
      Duration of humoral immunity to common viral and vaccine antigens.
      individuals with active infection generally have higher antibody titers than those who cleared the virus years before, either spontaneously or with antiviral treatment. When using the OQ dipstick, a test band with any discernible contrast is considered a positive result.
      OraQuick HCV link to package insert PDF.
      We hypothesized higher titers in viremic individuals would produce a positive test faster than those who had cleared HCV. The aim of our study was to determine whether the OQ test could be read before 20 minutes, and whether there was an early read time that would successfully identify all viremic individuals.

      Materials and Methods

       Clinic Cohorts

      Patients at the Toronto Centre for Liver Disease (TCLD) or the Maple Leaf Medical Clinic were tested with the OQ test if they were known to have past or current HCV infection. Viremic patients had a positive HCV RNA within 6 months of the OQ test. Those with resolved infection had (1) a sustained virological response (SVR) following recent treatment in the last 3 years (recently achieved SVR [R-SVR]), 2) achieved SVR 3 years or longer prior to the collection date (long-term SVR [L-SVR]); or 3) spontaneously cleared HCV infection. Clinical (age, sex, treatment status, cirrhosis status) and virological (HCV RNA level, HCV genotype) characteristics were recorded. Patients provided samples for fingerprick whole blood, phlebotomy whole blood, and serum. Oral (crevicular) fluid was also assessed (see supplemental section). Fingerprick and oral fluid samples were immediately tested; whole blood from phlebotomy was collected in an EDTA (ethylenediaminetetraacetic acid) tube and tested within 8 hours; serum was frozen at –80°C and tested at a later date. Additionally, frozen sera from TCLD’s biorepository of patients with known HCV or HCV/human immunodeficiency virus (HIV) coinfection were evaluated. Results from frozen serum and fresh whole blood were compared (Supplemental Figure 5).

       Real-World Cohort

      To validate the early read-time developed in the clinic cohort, a community HCV screening program, operated from a mobile clinic in Madrid, Spain was recruited. To validate the 5-minute read time, testers offered routine screening using the OQ test, which, if positive, was followed by a POC HCV RNA test (GeneXpert HCV Viral Load Assay; Cepheid, Sunnyvale, CA). Rapid screening for HIV (First Response HIV 1-2.O Rapid Whole Blood Test from Premier Medical Corporation Private Limited, Mumbai, India) was also offered. Readers recorded OQ result each minute from 5 to 10 minutes, with a final read at 5 minutes and each minute after, to 20 minutes.

       Test Performance and Results Recording

      An individual OQ kit (OraSure, Bethlehem, PA) was used for each test and carried out according to the manufacturer’s instructions.
      OraQuick HCV link to package insert PDF.
      Any detectable test band was counted as positive. Serial photographs using a smartphone (16-megapixel, f/2.0, 27-mm Nokia 6.1 camera running Android One operating system, version 9; Nokia, Espoo, Finland) were taken at 15-second intervals from the time of test initiation up to 20 minutes. Shutter frequency was controlled by the Open Camera app (version 1.44.1 Google Play). Time to positivity (tpos) was recorded as the first photograph in which a test band was visible. Results were recorded by 2 observers blinded to the viremic status of the sample donors. For disputed results, a third observer was consulted, or the later observation was chosen.

       Quantification of HCV Antibody Titers

      Sera from viremic and nonviremic patients were selected to investigate the relationship between the time of band appearance and antibody titer. Titers were semi-quantified using an HCV antibody enzyme-linked immunosorbent assay kit (Abnova, Walnut, CA) according to the manufacturer’s instructions. Samples were plated at 10-fold serial dilutions until not detected; the penultimate concentrations were then plotted against tpos.

       Statistical Analysis

      Fisher’s exact test and exact logistic regression were used to determine factors associated with the 5-minute threshold reported by odds ratio with 95% confidence interval (CI). See Supplemental Materials for further details.

       Ethics

      All patients provided written informed consent and the protocol was approved by the Research Ethics Board of the University Health Network. The protocol for the Madrid validation cohort was reviewed by the Comité de Ética del Hospital Gregorio Marañón, and all individuals provided written informed consent.

       Role of the Funding Source

      The Viral Hepatitis Care Network (VIRCAN) is a nonprofit organization. VIRCAN team members were involved in data collection, analysis, interpretation of the data, writing the report and the decision to publish. OraSure provided testing kits in kind but had no access to study data.
      All authors had access to the study data and reviewed and approved the final manuscript.

      Results

       Clinic Cohort

      The TCLD cohort included 171 viremic and 108 nonviremic individuals. Nonviremic individuals included 27 with R-SVR, 63 with L-SVR, and 18 with spontaneous clearance. Males predominated slightly among viremic (60%) and SVR (59%) patients, compared with spontaneous clearers (50%). The median age was higher in the SVR group than in viremic individuals or spontaneous clearers (63.5 years of age vs 56.0 years of age vs 44.5 years of age). Genotype distribution was similar between viremic and SVR groups, with genotypes 1 and 3 being most common (Table 1). In the viremic population, the median HCV RNA was 1.48× 106 log IU/mL. Among SVR patients, 100% of R-SVR patients and 32% of L-SVR patients were treated with DAAs, the remaining having received interferon-based therapy. A high proportion of SVR patients had cirrhosis (78% vs 38% in viremic patients), as only SVR patients with cirrhosis routinely return to the clinic after treatment. Four viremic and 9 SVR patients had HIV coinfection. Spontaneous clearers were confirmed as previously exposed to HCV using 2 or more tests (Supplemental Methods, Supplemental Table 1, Supplemental Figures 2 and 3).
      Table 1Demographics of Clinic Cohorts From Patients of the TCLD and Maple Leaf Clinic
      Viremic (n = 171)R-SVR (<3 y) (n = 27)L-SVR (≥3 y) (n = 63)Spontaneous Clearance (n = 18)Total (N = 279)
      Sex
      Male103 (60.2)17 (63.0)36 (57.1)9 (50.0)165 (59.1)
      Female68 (39.8)10 (37.0)27 (42.9)9 (50.0)114 (40.9)
      Age
      18–29 y7 (4.1)0 (0.0)0 (0.0)3 (16.7)10 (3.6)
      30–49 y37 (21.6)3 (11.1)4 (6.3)8 (44.4)52 (18.6)
      50–64 y100 (58.5)15 (55.6)28 (44.4)4 (22.2)147 (52.7)
      65+ y27 (15.8)9 (33.3)31 (49.2)3 (16.7)70 (25.1)
      Genotype
      195 (55.6)17 (63.0)48 (76.2)2 (11.1)162 (58.1)
      212 (7.0)5 (18.5)6 (9.5)0 (0.0)23 (8.2)
      339 (22.8)4 (14.8)7 (11.1)0 (0.0)50 (17.9)
      47 (4.1)0 (0.0)0 (0.0)0 (0.0)7 (2.5)
      52 (1.2)0 (0.0)0 (0.0)0 (0.0)2 (0.7)
      66 (3.5)0 (0.0)0 (0.0)0 (0.0)6 (2.2)
      Multiple8 (4.7)1 (3.7)1 (1.6)0 (0.0)10 (3.6)
      Unknown2 (1.2)0 (0.0)1 (1.6)16 (88.9)19 (6.8)
      Mean years since SVRN/A1.1 (0–3.0)8.9 (3.2–19.5)21.6 (0.3–59.3)N/A
      Median log10 HCV RNA1.48 × 106 (2.72 × 105 to 3.38 × 106)N/AN/AN/AN/A
      Treatment
      IFNN/A5 (18.5)43 (68.3)N/A48 (17.2)
      DAAN/A22 (81.5)20 (31.7)N/A42 (15.1)
      Cirrhosis
      Cirrhotic65 (38.0)14 (51.9)56 (88.9)1 (5.6)136 (48.7)
      Noncirrhotic106 (62.0)13 (48.1)7 (11.1)17 (94.4)143 (51.3)
      NOTE. Values are n (%), mean (range), or median (interquartile range).
      DAA, direct-acting antiviral; HCV, hepatitis C virus; IFN, interferon; L-SVR, long-term sustained virologic response; N/A, not available; R-SVR, recently achieved sustained virologic response; SVR, sustained virologic response.

       Time to Positivity (tpos)

      All 171 viremic samples showed a positive band within 5 minutes (Figures 1 and 2) leading to a sensitivity for viremia of 100% at 5 minutes (95% CI, 97.9%–100%) and a negative predictive value (NPV) for viremia of 100% at 5 minutes (95% CI, 92.3%–100%). Median tpos was 2.6 (range, 1.8–4.6) minutes in the viremic group versus 4.1 (range, 2.3–14.4) minutes in the nonviremic group (P < .001). Within the nonviremic group, median tpos was 3.5 (range, 2.3–5.4 minutes) minutes for the R-SVR group (P < .001) and 4.8 (range, 2.3–14.4) minutes for L-SVR patients (P < .001). Six L-SVR samples never developed a positive test band (positive antibody results confirmed from same-day samples using Siemens anti-HCV assay [Siemens Healthcare, Erlangen, Germany]), and as such were excluded from statistical analysis of tpos. Of the spontaneous clearers (n = 18), the majority (n = 13) never developed a positive test band on the OQ test; the median tpos for the remaining samples from spontaneous clearers (n = 5) was 3.4 (range, 2.2–12.5) minutes. In both viremic and nonviremic samples, once a band became visible, it remained visible at 20 minutes. Table 2 summarizes the specificity, sensitivity, and predictive values found in the clinic, real-world, and pooled cohorts.
      Figure thumbnail gr1
      Figure 1Percentage of samples with a positive test band over time. All viremic samples (orange) developed a positive test band before 5 minutes, and all recent SVR samples (blue) before 5.4 minutes. Several (n = 6) L-SVR samples (green) failed to show a positive band within 20 minutes. Of spontaneous clearers (yellow), less than one-third (n = 5) showed a positive result within 20 minutes. All samples were from fingerprick or serum.
      Figure thumbnail gr2
      Figure 2Median and mean tpos for viremic (V), R-SVR, and L-SVR samples. Viremic samples (n = 171) had a median tpos (green diamond) of 2.6 (range, 1.8–4.6) minutes. Median tpos for L-SVRs with positive results (n = 57 of 63) was 5.5 (range, 3.0–19.0) minutes. Mean tpos for R-SVR (n = 29) was 3.5 (range, 2.3–5.4) minutes. Red circles represent outliers beyond 1.5× the interquartile range. Tests that did not become positive within 20 minutes: For L-SVR samples, 6 of 63 tests did not become positive. For spontaneous clearers (SC), 13 of 18 tests did not become positive.
      Table 2Sensitivity, Specificity, Positive, and Negative Predictive Values of the OraQuick HCV Antibody Test at 5 Minutes
      Clinic cohortReal-World cohortPooled cohorts
      Positive at 5 minNegative at 5 minPositive at 5 minNegative at 5 minPositive at 5 minNegative at 5 min
      HCV RNA (+)17105602270
      HCV RNA (–)6246772513971
      PPV (95% CI)73.4 (67.2–78.9)42.1 (33.6–51.0)62.0 (56.8–67.0)
      NPV (95% CI)100.0 (92.3–100.0)100.0 (86.3–100.0)100.0 (94.9–100.0)
      Sensitivity (95% CI)100.0 (97.9–100.0)100.0 (93.6–100.0)100.0 (98.4–100.0)
      Specificity (95% CI)42.6 (33.1–52.4)25.3 (16.5–34.0)33.8 (27.4–40.6)
      CI, confidence interval; HCV, hepatitis C virus; NPV, negative predictive value; PPV, positive predictive value
      For nonviremic samples, 62 of 108 (57.4%) showed a positive OQ result by 5 minutes, including 26 of 27 (96.2%) R-SVRs, 33 of 63 (52.4%) L-SVRs, and 3 of 18 (16.7%) spontaneous clearers. Collectively, the positive predictive value (PPV) of the 5-minute cutoff for viremia in the TCLD cohort was 73.4% (95% CI, 67.2%–79.0%). Secondary analysis excluding R-SVRs increased the PPV for viremia to 82.6% (95% CI, 77.1%–87.9%), as nearly all R-SVR tests appear before 5 minutes. Time to positivity did not strongly correlate with time since SVR (R2 = 0.1), however the range of values becomes considerably greater after 3 or more years following SVR (Supplemental Figure 3). Time-to-positivity using oral fluid did not distinguish viremic from nonviremic individuals (Supplemental section).

       Factors Affecting tpos

      Female sex was associated with a lower likelihood of testing HCV antibody–positive by 5 minutes (odds ratio, 0.51; P = .04), however after controlling for the presence of viremia, this association was no longer significant. Age, HCV genotype, type of HCV therapy (DAA vs interferon), and the presence of cirrhosis were not associated with positivity at 5 minutes (data not shown).

       Correlation of Time to OQ Test Positivity With HCV Antibody Titers

      To investigate the relationship between tpos and HCV antibody titer, we measured the concentration of HCV core antibody by enzyme-linked immunosorbent assay. Serum samples from 11 viremic and 13 nonviremic patients were selected to reflect a wide range in tpos (105–215 seconds for viremic and 156–430 seconds for nonviremic). Overall, viremic samples yielded positive results at higher dilutions (1:1000–1:100,000) than nonviremic samples (1:10–1:1000), confirming the presence of higher antibody titers in viremic samples (Figure 3).
      Figure thumbnail gr3
      Figure 3Anti-HCV antibody concentration and tpos. Sera from viremic and L-SVRs samples were analyzed by enzyme-linked immunosorbent assay for anti-HCV antibody concentration against core HCV antigen. Samples were plated at 10-fold serial dilutions until not detected; the minimum detectable dilution (y-axis) was then plotted against tpos. ∗L-SVRs with no tpos (ie, not within 20 minutes) but which appeared positive on enzyme-linked immunosorbent assay at 1000× and 100× dilutions.

       Real-World Cohort in Madrid, Spain

      To investigate whether 5 minutes was sufficient to identify all viremic samples in settings outside the hospital, a real-world validation study was carried out in Madrid with 176 antibody-positive individuals (135 men and 41 women; median 49 years of age). Of these, 56 (31.8%) were found to be HCV viremic, 102 (58.0%) were nonviremic, 3 (1.7%) had indeterminate results, and 15 (8.5%) had no follow-up HCV RNA testing. All 56 viremic individuals were HCV antibody–positive within 5 minutes, confirming 100% sensitivity (95% CI, 93.6%–100%) for viremia at the 5-minute read time, and 100% NPV (95% CI, 86.3%–100%) for absence of viremia with a negative OQ test at 5 minutes. Among nonviremic individuals, 77 (75.5%) patients tested positive by 5 minutes, and 100 (98.0%) tested positive by 10 minutes. Collectively, in the Madrid cohort, the PPV of a positive OQ test at 5 minutes for viremia was 42.1% (95% CI, 33.6%–51.0%). Similar to the clinic cohort, all OQ results that became positive prior to 20 minutes, remained positive at the 20-minute read time. When combined with the samples from the clinical cohort, all 227 viremic individuals tested positive within 5 minutes, confirming the 100% (95% CI, 98.4%–100%) sensitivity for viremia and the NPV of 100% (95% CI, 94.9%–100%) for viremia of a negative test at 5 minutes.

       Results of HIV-Positive Samples From HCV Antibody-Positive Donors, in the Combined Cohorts

      The 13 HIV-positive patients from the clinic cohort consisted of 4 viremic, 6 R-SVR, and 3 L-SVR (mean tpos of 3.3, 3.5, and 4.5 minutes respectively). Most (n = 10, 79.6%) were on antiretroviral therapy at the time of sampling, with mean CD4 count of 532 (range, 306–860) cells/μL. All HCV viremic and R-SVR samples from this group became positive in <5 minutes. An additional 46 HIV-positive/HCV antibody–positive participants were tested in the Madrid cohort: 33 nonviremic, 10 viremic, and 3 unknown. All 10 viremic samples had a positive OQ test before 5 minutes, as did the 3 with unknown HCV RNA status. Of the remaining 33 HCV nonviremic samples, 26 were positive before 5 minutes, with 7 positive after 5 minutes.

      Discussion

      Rapid POC testing has been used widely for many years to overcome obstacles in infectious disease screening.
      • Johannessen A.
      Where we are with point of care testing.
      Additionally, many studies have demonstrated that using peer/community members
      • Latham N.H.
      • Pedrana A.
      • Doyle J.S.
      • et al.
      Community-based, point-of-care hepatitis C testing: perspectives and preferences of people who inject drugs.
      for HCV POC testing, or testing in which people receive other services
      • Williams B.
      • Howell J.
      • Doyle J.
      • et al.
      Point of care hepatitis C testing from needle and syringe programs: an Australian feasibility study.
      • Dong B.J.
      • Lopez M.
      • Cocohoba J.
      Pharmacists performing hepatitis C antibody point of care screening in a community pharmacy: a pilot project.
      • Morano J.P.
      • Zelenev A.
      • Lombard A.
      • Marcus R.
      • Gibson B.A.
      • Altice F.L.
      Strategies for hepatitis C testing and linkage to care for vulnerable populations: point of care and standard HCV testing in a mobile medical clinic.
      improves the diagnostic cascade of care for HCV. However, further improvements are possible: shortening the 20-minute read time of the OQ test could significantly increase its utility in both high- and low-income settings. We have shown that reading the OQ test at 5 minutes identifies all HCV viremic individuals, with the potential to increase throughput and reduce the need for HCV RNA testing.
      Shortening read time potentially decreases loss to follow-up. The VIRCAN program has tested 1525 individuals in various screening settings using the OQ test; 282 (18.5%) did not return at 20 minutes to receive their results (unpublished data). With a 5-minute read time, tested individuals would not leave the testing site. In low-income, high-prevalence settings, the main advantage of the early read time is the reduction in the number of expensive reflex HCV RNA tests required. The 100% sensitivity allows those negative at 5 minutes to be reassured that they do not have active HCV infection; tests that become positive after 5 minutes do not require HCV RNA testing. The early read time also reduces overall testing time, allowing for higher screening throughput.
      We are confident in the reliability of the 5-minute test to exclude viremia because of the breadth of our approach. We evaluated the test in a large clinical cohort with well-pedigreed samples from fingerprick whole blood, and frozen serum; we subsequently validated the strategy in a real-world cohort, testing individuals at high risk for HCV, including people with treated and untreated HIV coinfection. In both groups, all viremic patients had a positive test by 5 minutes, resulting in a sensitivity of 100% with a very narrow confidence interval (95% CI, 0.98–1.00); combining the laboratory and real world cohorts, the NPV for viremia of a negative result at 5 minutes remained 100%. We also showed that the antibody titers were higher in viremic patients correlating with the short tpos, strengthening the scientific rationale for the approach.
      Checking earlier than 5 minutes, we found no threshold that had adequate PPV to forego HCV RNA testing. For many individuals who clear the virus, antibody titers drop over time; however, for many, even years after viral clearance, antibody titers remain high. Furthermore, the PPV would be even lower in a population with lower viremic prevalence, as was seen in our real-world cohort. As such, any positive OQ result by 5 minutes still requires an HCV RNA test to confirm active infection.
      However, reducing testing to a 5-minute cutoff eliminates the requirement for HCV RNA testing of a large proportion of samples (42.6% TCLD cohort and 24.5% Spanish real-world cohort) that would otherwise be counted positive. This could markedly reduce the overall cost of testing in large screening programs. The 5-minute read time increases the number of tests that can be completed in a given time frame, and also reduce loss to follow-up. With a positive antibody test at 5 minutes, a tester may immediately progress to collection for either dried blood spot HCV RNA testing,
      • Soulier A.
      • Poiteau L.
      • Rosa I.
      • et al.
      Dried blood spots: a tool to ensure broad access to hepatitis C screening, diagnosis, and treatment monitoring.
      or for on-site rapid HCV RNA testing such as the POC GeneXpert.
      • Latham N.H.
      • Pedrana A.
      • Doyle J.S.
      • et al.
      Community-based, point-of-care hepatitis C testing: perspectives and preferences of people who inject drugs.
      ,
      • Grebely J.
      • Lamoury F.M.J.
      • Hajarizadeh B.
      • et al.
      Evaluation of the Xpert HCV Viral Load point of care assay from venepuncture-collected and finger-stick capillary whole-blood samples: a cohort study.
      For the oral fluid test, our findings are in line with previous studies reporting that the test is more sensitive for viremic individuals than for nonviremic individuals but less reliable than blood testing (see Supplemental Materials).
      • Tang W.
      • Chen W.
      • Amini A.
      • et al.
      Diagnostic accuracy of tests to detect Hepatitis C antibody: a meta-analysis and review of the literature.
      ,
      • Pallarés C.
      • Carvalho-Gomes Â
      • Hontangas V.
      • et al.
      Performance of the OraQuick Hepatitis C virus antibody test in oral fluid and fingerstick blood before and after treatment-induced viral clearance.
      After the 20-minute read time, even for viremic individuals, sensitivity was still not 100%, suggesting that the oral test should be reserved for use in which it is not possible to obtain a blood sample, at the cost of slightly reduced sensitivity, and no time saving.
      The study has some limitations. The real-world cohort included only 56 viremic individuals, owing to the relatively low RNA positivity rate among those who were HCV antibody–positive. However, it is reassuring that all 227 viremic patients in the clinical and real-world cohorts tested positive within 5 minutes, providing a very narrow CI (98.4%–100%) around the 100% sensitivity estimate. While additional real-world validation would clearly be helpful, the benefits in terms of reducing loss to follow-up (>18% in our experience) justify consideration of early adoption of the proposed early read time.
      The time-to-positivity analysis was based on the read time of the photographs, rather than on the direct read at the time of collection. Photos were used to ensure readers were blinded to clinical status of the participant and to allow for 2 readers. When compared, the time to positivity was similar between direct read and the time detected by reading the photographs. Differences in read time were small, typically <15 seconds, with the direct read reporting slightly earlier results, likely owing to the 15-second interval used between photographs. The real-world cohort results were based on direct read times. As such, the 5-minute rule performs well with direct read, as it will be used in clinical practice.
      In the clinical cohort, 38% (n = 65 of 171) of the viremic cohort was cirrhotic; however, no significant difference was noted in the tpos in this group, compared with viremic patients without cirrhosis, giving us confidence that the 5-minute threshold will work in patients with compensated cirrhosis. No patients with decompensated cirrhosis were included in the study; however, individuals with very advanced liver disease are unlikely to be first diagnosed in a screening context.
      There were relatively few individuals with HIV coinfection, and most in the clinical cohort were on antiretroviral therapy with preserved CD4 counts. However, in the real-world cohort, people were tested for HIV as well as HCV, and all those who were viremic for HCV had a positive OQ test within 5 minutes, including the 10 HIV-positive individuals. Further validation with more people living with HIV would be valuable.
      A limitation of the 5-minute rule unrelated to the study design is the possibility of missing recently infected individuals who have yet to develop detectable levels of antibodies. While antibody titers are still rising, it is possible that some individuals with viremia may not test positive by 5 minutes. However, in early infection, people may test antibody negative regardless of the assay; if acute infection is of clinical concern, HCV RNA testing should be performed, or patients should return for retesting.
      In conclusion, we have shown that the OQ test can be reduced from 20–40 minutes to 5 minutes. Patients testing negative at 5 minutes can be reliably told that they do not have active HCV infection. Reducing test time improves throughput, substantially decreases the need for reflex RNA testing, and would avoid the loss to follow-up rates observed even with a 20-minute wait time, providing benefit in both low-income and high-income settings; these simple improvements to an existing test would contribute to the high-sensitivity, high-throughput testing required to achieve the World Health Organization goals.

      Supplementary Material Methods

       Methods

       Quantification of Band Intensity

      Using densitometry, we quantified band intensities over time, comparing serial photos at each minute to the final 20-minute photo for each test. Photos were converted to grayscale and auto-reshaped to a standardized rectangle (Perspective Crop, Photoshop CS6 Extended; Adobe, San Jose, CA) before analysis (Bio-Rad ImageLab version 6.0.1; Bio-Rad, Hercules, CA). After subtraction of background, band intensity was expressed as a fraction of the final 20-minute intensity.

       Hepatitis C Virus Enzyme-Linked ImmunoSpot of Peripheral Blood Mononuclear Cells From Spontaneous Clearers

      All spontaneous clearers were confirmed to have been previously infected with hepatitis C virus (HCV) using 2 antibody tests: the tests were done on 2 separate commercial assays performed at the Public Health Ontario Laboratory, followed by 2 negative HCV RNA tests. In addition, 9 of the 18 spontaneous clearers provided samples that could be tested for the presence of HCV-specific lymphocytes cells, using an enzyme-linked immunospot assay.
      Patients with spontaneous clearance provided blood samples collected in acid-citrate-dextrose, from which peripheral blood mononuclear cells were isolated using Ficoll Paque PLUS density gradients (Cytiva, Marlborough, MA. Cells were washed twice then cryopreserved in medium containing 90% fetal calf serum (Hyclone Laboratories, Logan, UT) and 10% dimethyl sulfoxide (Sigma-Aldrich, St. Louis, MO). HCV-specific immune responses were evaluated by human interferon gamma single-color enzyme-linked immunospot assay. Peripheral blood mononuclear cells were seeded in 96-well plates at 200,000 cells per well and stimulated with 10 pools containing overlapping peptides spanning the entire HCV genome (2 μg/peptide/mL). All HCV peptides were obtained from BEI Resources (Manassas, VA) (peptide arrays, hepatitis C virus, H77) (Supplementary Table 1, Supplementary Figures 2 and 3).

       Statistical Analysis

      Continuous measurements were compared using the Student’s t or Wilcoxon rank sum as appropriate. Factors associated with time to positivity were evaluated by linear regression. Percentage of positive tests at the 5-minute threshold were assessed and the binomial exact 95% confidence intervals (CIs) calculated. Fisher’s exact test and exact logistic regression were used to determine factors associated with the 5-minute threshold reported by odds ratios with 95% CIs determined using the Clopper-Pearson (exact) method. Analysis was carried out with SAS version 9.4 (SAS Institute, Cary, NC); 2-sided P values below .05 were considered statistically significant.
      Sample size was calculated to minimize the 95% CI around the sensitivity for the 5-minute read time for active viremia, a threshold, based on a pilot study of 17 viremic samples, as one that appeared to be well outside the results found for the initial tests, all 17 samples had an OraQuick HCV rapid antibody test (OQ test) (OraQuick, Bethlehem, PA) time to positivity (tpos) well below this chosen threshold. With a sample size of 150 viremic patients, 100% sensitivity of a positive test at 5 minutes would have an acceptably narrow 95% CI of 97.6%–100%. We therefore aimed to include minimum 150 viremic patients and at least 100 patients in the nonviremic group for adequate comparison of tpos.

      Results

       Assessment of Band Visibility Across Test Duration

      To characterize band visibility, we sequentially photographed viremic samples (n = 20) at 1-minute intervals and analyzed them by densitometry. Mean test band intensities at 5 minutes were 52% (range, 29%–77%) of their final 20-minute intensity. A representative 20-minute photo strip of a viremic sample appears in Supplementary Figure 1.

       Comparison of Serum and Fingerprick Whole Blood

      We investigated whether frozen serum provided similar results to fingerprick whole blood, to determine whether it was acceptable to include sera from our frozen serum library in the clinical cohort. To make the comparison, we drew venipuncture whole blood from patients (n = 72) on the same day that they provided fingerprick samples. The blood was serum-separated, frozen and subsequently thawed, tested, and compared with the fingerprick results (Supplementary Figure 5). For the viremic matched pairs (n = 40), the median tpos for fingerprick was 2.7 minutes vs 2.6 minutes for serum, a difference of 9 seconds (P = .14). For the nonviremic matched pairs (n = 32), median tpos for fingerprick was 4.4 minutes vs 3.7 minutes for serum, a difference of 40 seconds (P < .01).

       Sensitivity Using Oral (Crevicular) Fluid

      We evaluated the sensitivity and tpos using the OQ test in oral fluid (n = 72), comparing viremic and nonviremic individuals. Among viremic patients (n = 33), the OQ test was positive in 32 (97.0%; 95% CI, 84.2%–99.9%) patients by 20 minutes, with a median tpos of 5.53 (interquartile range, 3.0–19.0) minutes. For long-term sustained virologic resistance (n = 39), 28 (71.7%; 95% CI, 55.1%–85.0%) appeared positive by 20 minutes, while the remaining 11 (28.2%) failed to produce a band by 20 minutes.
      Figure thumbnail fx1
      Supplementary Figure 1Average band intensity over time. Band intensities are expressed as a percentage of final intensity at 20 minutes. The average intensity of band appearance for viremic samples is shown for each minute (blue line). Error bars = SD. Photos: serial 1-minute photographs of a representative viremic sample.
      Figure thumbnail fx2
      Supplementary Figure 2Magnitude of enzyme-linked immunospot (ELISPOT) response to each peptide pool. The number of interferon gamma (IFN-γ) spot-forming units (SFUs), minus unstimulated controls, for each HCV antigen peptide pool. Each bar is divided into individual patients denoted by color (n = 9 patients tested).
      Figure thumbnail fx3
      Supplementary Figure 3Number of patients with positive response to each peptide pool. Positive response defined as 2 times background with at least 10 spot-forming units.
      Figure thumbnail fx4
      Supplementary Figure 4Distribution of tpos versus years since sustained virologic response (SVR). tpos weakly correlated with time since SVR (R2 = 0.1). L-SVR, long-term sustained virologic response; R-SVR, recently achieved sustained virologic response.
      Figure thumbnail fx5
      Supplementary Figure 5Comparison using matched samples for viremic and nonviremic samples from HCV antibody–positive patients. (Left) Fingerprick from viremic patients (n = 40) had a nonsignificant median difference in tpos of +9 seconds, compared with serum (n = 40) (P = .14). (Right) Nonviremic samples of fingerprick (n = 32) vs matched samples (n = 32); median tpos of serum samples is 40 seconds faster than fingerprick samples (P < .01).
      Supplementary Table 1Hepatitis C Virus Peptides
      Pool NumberHCV ProteinGenotype 1a (H77)Genotype 1b (J4)Genotype 3a (K3a/650)
      1Core1-281-281-29
      2NS2 + p7NS2(1–32) + p7(1–8)NS2(1–32) + p7(1–8)NS2(1–33) + p7(1–8)
      3NS31-491-491-49
      4NS350-9850-9850-97
      5NS4A + NS4BNS4A(1–7) NS4B(1–40)NS4A(1–7) NS4B(1–40)NS4A(1–7) NS4B(1-39)
      6E11-291-291-28
      7E21-561-551-57
      8NS5A1-711-711-72
      9NS5B1-491-491-49
      10NS5B50-9050-9050-90

      References

        • Cooke G.S.
        • Andrieux-Meyer I.
        • Applegate T.L.
        • et al.
        Accelerating the elimination of viral hepatitis: a Lancet Gastroenterology and Hepatology Commission.
        Lancet Gastroenterol Hepatol. 2019; 4: 135-184
        • Lazarus J.V.
        • Pericàs J.M.
        • Picchio C.
        • et al.
        We know DAA s work, so now what? Simplifying models of care to enhance the hepatitis C cascade.
        J Intern Med. 2019; 286: 503-525
        • Bajis S.
        • Dore G.J.
        • Hajarizadeh B.
        • Cunningham E.B.
        • Maher L.
        • Grebely J.
        Interventions to enhance testing, linkage to care and treatment uptake for hepatitis C virus infection among people who inject drugs: a systematic review.
        Int J Drug Policy. 2017; 47: 34-46
        • Keats J.
        • Micallef M.
        • Grebely J.
        • et al.
        Assessment and delivery of treatment for hepatitis C virus infection in an opioid substitution treatment clinic with integrated peer-based support in Newcastle, Australia.
        Int J Drug Policy. 2015; 26: 999-1006
        • Morgan K.
        • Lee J.
        • Sebar B.
        Community health workers: A bridge to healthcare for people who inject drugs.
        Int J Drug Policy. 2015; 26: 380-387
        • Treloar C.
        • Rance J.
        • Bath N.
        • et al.
        Evaluation of two community-controlled peer support services for assessment and treatment of hepatitis C virus infection in opioid substitution treatment clinics: the ETHOS study, Australia.
        Int J Drug Policy. 2015; 26: 992-998
        • Henderson C.
        • Madden A.
        • Kelsall J.
        ‘Beyond the willing &the waiting’ — the role of peer-based approaches in hepatitis C diagnosis and treatment.
        Int J Drug Policy. 2017; 50: 111-115
        • Khuroo M.S.
        • Khuroo N.S.
        • Khuroo M.S.
        Diagnostic accuracy of point of care tests for hepatitis C virus infection: a systematic review and meta-analysis.
        PLoS One. 2015; 10e0121450
      1. OraSure Technologies - OraQuick HCV Rapid Antibody Test.
        (Available at:)
        • Amanna I.J.
        • Carlson N.E.
        • Slifka M.K.
        Duration of humoral immunity to common viral and vaccine antigens.
        N Engl J Med. 2007; 357: 1903-1915
      2. OraQuick HCV link to package insert PDF.
        (Available at:)
        • Johannessen A.
        Where we are with point of care testing.
        J Viral Hepat. 2015; 22: 362-365
        • Latham N.H.
        • Pedrana A.
        • Doyle J.S.
        • et al.
        Community-based, point-of-care hepatitis C testing: perspectives and preferences of people who inject drugs.
        J Viral Hepat. 2019; 26: 919-922
        • Williams B.
        • Howell J.
        • Doyle J.
        • et al.
        Point of care hepatitis C testing from needle and syringe programs: an Australian feasibility study.
        Int J Drug Policy. 2019; 72: 91-98
        • Dong B.J.
        • Lopez M.
        • Cocohoba J.
        Pharmacists performing hepatitis C antibody point of care screening in a community pharmacy: a pilot project.
        J Am Pharm Assoc. 2017; 57: 510-515.e2
        • Morano J.P.
        • Zelenev A.
        • Lombard A.
        • Marcus R.
        • Gibson B.A.
        • Altice F.L.
        Strategies for hepatitis C testing and linkage to care for vulnerable populations: point of care and standard HCV testing in a mobile medical clinic.
        J Community Health. 2014; 39: 922-934
        • Soulier A.
        • Poiteau L.
        • Rosa I.
        • et al.
        Dried blood spots: a tool to ensure broad access to hepatitis C screening, diagnosis, and treatment monitoring.
        J Infect Dis. 2016; 213: 1087-1095
        • Grebely J.
        • Lamoury F.M.J.
        • Hajarizadeh B.
        • et al.
        Evaluation of the Xpert HCV Viral Load point of care assay from venepuncture-collected and finger-stick capillary whole-blood samples: a cohort study.
        Lancet Gastroenterol Hepatol. 2017; 2: 514-520
        • Tang W.
        • Chen W.
        • Amini A.
        • et al.
        Diagnostic accuracy of tests to detect Hepatitis C antibody: a meta-analysis and review of the literature.
        BMC Infect Dis. 2017; 17: 695
        • Pallarés C.
        • Carvalho-Gomes Â
        • Hontangas V.
        • et al.
        Performance of the OraQuick Hepatitis C virus antibody test in oral fluid and fingerstick blood before and after treatment-induced viral clearance.
        J Clin Virol. 2018; 102: 77-83