HCV Core-antigen Assay as an Effective Alternative to HCV RNA Quantification in Patients With Hepatitis C

Abstract

Background/Aim: Hepatitis C virus (HCV) core antigen (Ag) test has been increasingly applied as an effective alternative to conventional molecular tests allowing rapid and affordable diagnosis, which is of paramount relevance to achieve global elimination of HCV infection.

Materials and Methods: ARCHITECT^® HCV Ag test was evaluated in comparison with HCV RNA quantification test (CAP/CTM) to calculate its sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and to determine their correlation level. Its performance, according to low and high viral load values and in different treatment stages [during treatment (T), at the end of the therapeutic protocol (EOT) and when sustained virological response (SVR) was evaluated].

Results: In total, 145 samples were included. Considering CAP/CTM, the sensitivity, specificity, PPV and NPV of the HCV-Ag test were 88.9%, 99.1%, 97.0% and 96.4%, respectively, and the correlation among tests was high (r=0.890), with only five discordant results. A decrease in sensitivity was found for low viral load values (<1,000 IU/ml), but the opposite was verified for high viral concentrations (≥1,000 IU/ml). A good agreement was verified for the T and EOT groups (k=0.789 and k=0.638) and an excellent agreement in the SVR group (k=1.000).

Conclusion: HCV-Ag seems to be an effective alternative that can be routinely combined with other faster and more accessible tests (e.g., HCV antibody tests) for the identification of new HCV infections in suspected patients, eventually reserving the molecular techniques for samples with discordant results.

Hepatitis C virus (HCV) is an RNA virus that belongs to the Flaviviridae family and is responsible for the development of the corresponding liver disease (1,2). The diagnosis of hepatitis C is challenging, as people often do not have symptoms in the acute phase of the disease, allowing the infection to progress undetected into a chronic phase (3). In fact, it is estimated that in 2015 71 million people were living with chronic HCV infection worldwide, with less than 20% were aware of their infected status (4). In addition, pathological processes such as cirrhosis, hepatocellular carcinoma, portal hypertension and liver decompensation can be observed in these patients, which may lead to the need for liver transplantation (5,6). To respond to this public health threat, one of the goals of the World Health Organization (WHO) is to globally eliminate hepatitis C by 2030 (7).

Currently, the diagnosis of HCV is based on the use of serological tests for the detection of anti-HCV antibodies and molecular tests for the detection and quantification of viral RNA (HCV RNA). Anti-HCV antibody tests, although cheaper and faster, do not distinguish between patients with active infection and those with resolved infection, still requiring HCV RNA tests to confirm the clinical status (8,9). HCV RNA tests are based on the amplification of nucleic acids using real-time polymerase chain reaction (RT-PCR), which makes the diagnostic process more expensive and complex, limiting its diagnostic application, especially in countries with less economic resources (10,11). To overcome the barriers imposed by the current diagnostic process, more efficient strategies have emerged, such as HCV core antigen quantification tests (12). In fact, several studies have been evaluating this technique, investigating its performance in diagnosing and monitoring these patients, as several systematic reviews and meta-analyses have recently highlighted (13,14).

The HCV core antigen test (HCV-Ag) has been considered as an excellent alternative to molecular tests as it is more stable, easier to use and associated with a lower cost (10-12). However, despite these advantages, it is still necessary to investigate its ability to identify active HCV infection and its applicability for monitoring patients during treatment (15).

Therefore, the aim of this study was to evaluate the performance of the ARCHITECT^® HCV Ag test (Abbott Diagnostics, Wiesbaden, Germany) in the detection of active infection and to investigate its potential use for monitoring patients with hepatitis C under treatment.

Materials and Methods

Study samples. For the purpose of this study, blood samples previously collected from patients at the Centro Hospitalar de Trás-os-Montes e Alto Douro, between January and October 2020, were used. The only eligibility criterion for the samples was that the HCV viral load had been quantified by an HCV RNA test.

After identifying the samples to be used for the viral quantification technique, the samples were frozen at -20°C for storage. Then, these same samples were used to perform the HCV core antigen test.

Samples from patients undergoing treatment were divided into three groups: the treatment (T) group that included samples collected during treatment, either at baseline or in subsequent reassessments; the end-of-treatment (EOT) group that included samples obtained at the end of treatment; and the sustained virological response (SVR) group that included samples from patients in the sustained virological response stage. Patients were considered to be on SVR if they had no detectable serum HCV RNA 24 weeks after the end of the therapeutic protocol (3).

This study was approved by the Ethics Committee of the Centro Hospitalar de Trás-os-Montes e Alto Douro, E.P.E.

Laboratory techniques. The first laboratory determination performed on the collected blood samples was the HCV RNA quantification, using COBAS^® AmpliPrep/COBAS^® TaqMan^® HCV v.2.0 test (CAP/CTM; Roche Molecular Systems, Pleasanton, CA, USA) in the COBAS^® AmpliPrep/COBAS^® TaqMan^® 48 analyzer (Roche Diagnostics, Indianapolis, IN, USA). The equipment’s lower and upper detection limits were 15 IU/ml and 10⁸ IU/ml, respectively.

In the second phase of the study, the detection of viral antigens was carried out in the same samples, that were previously stored frozen, using the ARCHITECT^® HCV Ag test (Abbott Diagnostics, Wiesbaden, Germany) conducted on the ARCHITECT^® i2000SR analyzer (Abbott Laboratories, Irving, TX, USA). This determination is based on the Chemiflex method through a chemiluminescence technology (16). Samples were considered non-reactive when concentrations were below 3 fmol/l and reactive above 10fmol/l. Whenever intermediate values (3 to 10 fmol/l) were obtained, a retest was performed, defining positive samples above 3 fmol/l and negative samples below 3fmol/l. The upper limit of the test was 20000 fmol/l. Finally, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated for this HCV Ag test, considering CAP/CTM as the standard test.

Statistical analysis. The SPSS^® (Statistical Package for the Social Sciences) software (Version 27.0; IBM Corp^®; Armonk, NY, USA) was used for statistical analysis. For the HCV-Ag test, a cut-off of 3 fmol/l was defined: equal or above (≥3 fmol/l) was considered as a positive result and below (<3 fmol/l) as negative. In turn, for the CAP/CTM test, a cut-off of 15 IU/ml was established and whenever a concentration above or below this cut-off was detected, a positive or a negative result was considered, respectively. Moreover, when it was not possible to quantify a result, it was excluded from the study. Spearman’s correlation coefficient was obtained to evaluate the association between the two tests, after the variable data were transformed into a logarithmic scale for the linear regression plot. In addition, the Cohen’s Kappa test was used to assess the degree of agreement between the two tests in the different treatment groups. Finally, associations were considered significant when p-values <0.05.

Results

Our population included 145 samples belonging to 113 patients: 80 (70.8%) were men and 33 (29.2%) were women. The mean age was 54.2 years, ranging from 29 to 91 years, and with a standard deviation of 13.7 years.

Clinical performance of the HCV-Ag test in comparison with the CAP/CTM test. The HCV-Ag test identified 112 samples as negative (<3 fmol/l) and 33 as positive (≥3 fmol/l), while the CAP/CTM test considered 109 samples as negative and 36 as positive. Regarding the latter technique, among the positive samples, 33 had a concentration equal to or greater than 15 IU/ml and 3 less than 15 IU/ml. In addition, 3 additional samples showed detectable but not quantifiable viral RNA.

Comparing the results of both tests, agreement was found in 140 samples (96.6%), of which 108 were negative and 32 were positive. Thus, four false negatives and one false positive were identified.

In order, to evaluate the performance of the HCV-Ag test, for low and high viral load values, a cut-off of 1,000 IU/ml was applied: <1,000 IU/ml versus ≥1,000 IU/ml. A decrease and an increase in sensitivity, respectively, were found for viral load values below 1,000 IU/ml and for viral load values equal or greater than 1,000 IU/ml.

Moreover, considering CAP/CTM as the standard test, the sensitivity, specificity, PPV and NPV of the HCV-Ag test were 88.9%, 99.1%, 97.0% and 96.4%, respectively (Table I).

Table I. HCV-Ag test performance analysis for low and high viral load values.

n: Number of patients; PPV: positive predictive value; NPV: negative predictive value.

For the correlation analysis between the tests, only samples with quantifiable results were considered, i.e., samples with viral load ≥15 IU/ml. In this case, Spearman’s correlation coefficient was high (r=0.890; p<0.001), which reflected a strong positive linear correlation between the results of both tests.

The relation between viral load and antigen concentration values can be expressed, according to linear regression, using the equation “Log (HCV Ag)=0.88×Log (Viral load) - 2.28”. Based on this equation we can consider that 1 fmol/l of HCV core antigen is equivalent to approximately 389.8 IU/ml of HCV RNA. Therefore, in this study, the detection limit of the HCV-Ag test was around 1169 IU/ml of HCV RNA (Figure 1).

Figure 1. Linear regression plot of the HCV-Ag versus CA/CT tests with a cut-off ≥15 UI/ml.

HCV-Ag test performance for different treatment groups. The performance of the HCV-Ag test was also assessed at different treatment stages. Of the 43 samples analyzed in the T group, both tests detected 12 as negative and 27 as positive. However, 4 discordant results were additionally found: 3 samples were found negative using the HCV-Ag test but positive using the CAP/CTM test, whereas the opposite occurred in one sample (Table II). Of the 17 EOT samples processed, 15 were found to be negative and 2 positive according to both laboratory tests. Only one sample gave a discordant result in this treatment stage; it was positive only in the CAP/CTM test. Finally, in the SVR group, all samples included were classified as negative using both tests, which revealed a 100% agreement at this stage.

Table II. Discordant samples analyzed by both tests (n=5).

*Viral RNA detected using CAP/CTM test, but not quantifiable. SVR: Sustained virological response; EOT: end of the therapeutic protocol.

According to the statistically significant results obtained using the Cohen’s Kappa test, a good agreement was observed for the T and EOT groups and an excellent agreement for the SVR group (Table III). In the SVR group, it was not possible to calculate the sensitivity, but the specificity of the HCV-Ag test was 100%.

Table III. HCV-Ag test performance analysis for different treatment groups.

n: Number of patients; p: statistical significance; PPV: positive predictive value; NPV: negative predictive value; SVR: sustained virological response; EOT: end of the therapeutic protocol.

Discussion

The viral antigen quantification test has been considered a good serological indicator of viral replication, although it is associated with a lower sensitivity compared to the RNA test (5,17,18). Despite that, our results showed that with a cut-off value of 3 fmol/l, the HCV-Ag test was able to detect active infection in 88.9% of the positive samples previously identified using the standard test, which is in line with the findings described by other authors (15,19-22). The specificity, PPV and NPV values obtained for this antigenic detection test were also in agreement with the results already reported by Park et al. (23). Beyond that, when the performance of the HCV-Ag test was investigated using samples with low viral load, the sensitivity of the test decreased to 50.0%, but when samples with high viral load were used its sensitivity increased from 88.9% in all samples to 93.8%, similar to the results previously described in the literature (21,24). A possible explanation for this decrease in sensitivity could have been the combination of false negatives and the presence of a few samples that were simultaneously positive for viral RNA and viral antigen.

A high correlation between HCV-Ag test and HCV PCR, such as the CAP/CTM test, has been described in the literature (12,21,23,25,26). Similar to Chang, et al. (12) and, more recently, to Iqbal, et al. (26), who found an excellent correlation (r=0.960 and r=0.950, respectively), our findings showed a strong positive association (r=0.890) between the two tests in all samples with quantifiable viral load, which suggests that the HCV-Ag test may be a good predictor for the identification of clinical cases of active hepatitis C infection.

Using linear regression, viral RNA load values have been extrapolated from the concentrations of viral antigens obtained using the HCV-Ag test. The concentration equivalence calculated in our study was within the range reported by other authors (5,15,18).

Our study also showed the usefulness of HCV-Ag in monitoring patients during treatment, as it produced quantitative results that can be compared between different treatment stages (18). In fact, although it was not possible to follow the patients from the beginning of the treatment until its end, biological samples at different time points were obtained, which allowed their categorization according to one of the three established stages (T, EOT or SVR). In the T group, the sensitivity was considered high (90.0%), but in the EOT group, it decreased to 50.0%, which may reflect the large number of negative results obtained in the two tests, as well as the false negatives found at this stage. In the SVR, the sensitivity was not achievable since all samples were classified as negative using both tests. Regarding specificity, it was 100% in the end-of-treatment and SVR stages, which is in accordance with previous findings (18,27).

Five discordant samples were found regarding samples of patients with low viral loads as reported by other authors (27,28). The low viral loads may be the result of a positive response to treatment, as these discordant values were detected in patients who were on treatment. This inequality of detection between the tests may be due to the fact that in patients undergoing treatment, free viral RNA is more abundant than complete virions (24). This hypothesis is supported by other authors such as Pérez-García, et al. (15) who obtained discordant results, particularly in samples obtained in the 4^th week of treatment. Moreover, the false positives detected using the HCV-Ag test can be explained by the presence of viral capsids without RNA (18), whereas the false negatives can result from mutations in the core region of the HCV genome, which may have limited the detection capacity of this test.

Our study has some limitations that should be highlighted, such as the unequal number of samples in each treatment group and the unavailability of complete follow-up of each patient during all these stages. However, these limitations portrayed the real scenario of hospital clinical research, and our study demonstrated the applicability of HCV-Ag in this context.

Thus, despite its lower sensitivity, the HCV-Ag has better stability in samples when stored at room temperature and allows for shorter response times and a lower cost compared to HCV RNA (21,27). In fact, when economic resources are limited and it is not possible to carry out molecular detection tests, the WHO recommend that the HCV-Ag test can be used to confirm viremia (29). Our results support this recommendation, as we believe that the increased routine application of this test will allow for a more accessible, systematic, and larger-scale diagnosis, contributing to a better clinical management of patients with hepatitis C.

In summary, our findings showed that HCV-Ag is an effective alternative to viral RNA quantification tests given its high sensitivity for detecting positive samples and its good performance in patients with high viral loads, as commonly found in an early stage of the therapeutic management.

Moreover, in light of our results and the current need to achieve more practical and more economical diagnostic techniques, we suggest that the HCV-Ag test can be combined with anti-HCV antibody tests for the identification of new hepatitis C infections in suspected patients, reserving molecular techniques only when there are discordant results between these two serological methods.

Conflicts of Interest

The Authors declare no conflicts of interest in relation to this study.

Authors’ Contributions

CC and FLQ conceived and designed the study; RV, CC, JAC and EC performed the experimental analysis and data collection; RV and FLQ analyzed and interpreted the results; RV and TRM wrote the manuscript; RV, CC, JAC, EC, TRM and FLQ reviewed and approved the submitted version.

Acknowledgements

This work was financed by National Funds (FCT/MCTES, Fundação para a Ciência e a Tecnologia and Ministério da Ciência, Tecnologia e Ensino Superior) under the project UIDB/00211/2020. Authors also want to thank the support received by projects UIDB/CVT/00772/2020 and LA/P/0059/2020, from FCT/MCTES. The authors would like to thank the Abbott Laboratories for providing the HCV Antigen Kit and to the Clinical Pathology Service of the Centro Hospitalar de Trás-os-Montes e Alto Douro for the availability of all the human and material resources that were essential for carrying out this study.