Abstract
An assay prototype designed to detect and quantify total hepatitis C virus [HCV] core antigen (HCV core Ag) protein in serum and plasma in the presence or absence of anti-HCV antibodies has been recently developed by Ortho-Clinical Diagnostics. The aim of the study was to evaluate the sensitivity, specificity, and reproducibility of the Total HCV core Ag assay in comparison with two quantitative assays for HCV RNA: Quantiplex HCV RNA 2.0 (bDNA v2.0) or Versant HCV RNA 3.0 (bDNA v3.0) assays and the Cobas Amplicor HCV Monitor version 2.0 (HCM v2.0) test. We have studied samples of a well-characterized panel and samples from patients with chronic hepatitis C treated with interferon alone or with ribavirin. We have also compared the kinetics of HCV core Ag and HCV RNA in the follow-up of treated patients. The HCV core Ag assay exhibited linear behavior across samples from different genotypes. The coefficients of variation for intra- and interassay performance were 5.11 and 9.95%, respectively. The specificity of the assay tested in blood donors was 99.5%. Samples from HCV-infected patients showed that the correlation between the HCV core Ag and the two HCV RNA quantitative assays (bDNA and HCM v2.0) was 0.8 and 0.7, respectively. This correlation was maintained across different genotypes of HCV (r2 = 0.64 to 0.94). Baseline HCV core Ag values were significantly lower in sustained responders to interferon (IFN) than in other groups of patients (5.31 log10 [104 pg/ml] versus 5.99 log10 [104 pg/ml]; P < 0.001). In patients treated with IFN or combination therapy, we found an association between a decrease of more than 2 log IU/ml in viral load, undetectable HCV core Ag, and sustained response. Among sustained responders to IFN alone or combination therapy and among relapsers after IFN alone, 84 out of 101 (83.2%) had undetectable HCV core Ag, and 76 out of 96 (79.2%) had a viral load decrease of ≥2 log IU/ml, after 1 month of treatment. In conclusion, the Total HCV core Ag assay is a new useful test for the detection of HCV viremia and the monitoring of patients treated with IFN alone or in combination with ribavirin.
Since its discovery in 1989 (8), it has been demonstrated that the hepatitis C virus (HCV) is a major cause of chronic hepatitis throughout the world. The routine virological diagnosis of this infection is based on the detection of HCV-specific antibodies (Ab). However, due to the absence of an efficient in vitro culture system for HCV or assays capable of detecting viral antigens, the direct detection of HCV has depended on methods of nucleic acid amplification or hybridization of HCV RNA. Indeed, Ab tests cannot distinguish between persons with anti-HCV Ab who have recovered and patients exhibiting an active infection, and they are not sufficient for the monitoring of therapy (17, 29, 30). The Amplicor HCV test (Roche Diagnostics Corporation, Branchburg, N.J.) and branched-chain DNA signal amplification assay (bDNA) (Bayer Corporation, Tarrytown, N.J.) have been used to detect HCV RNA in HCV Ab-positive individuals and to monitor the efficacy of therapy (6, 11, 13, 19, 21). Recently both assay systems have become semiautomated. However, several problems still persist when these methods are used for the screening of patients: the bDNA assay requires a long incubation period, while reverse transcription (RT)-PCR requires considerable skills and has limited reproducibility. In addition, both assays are costly. More recently, methods capable of detecting the core antigen (Ag) of HCV have been developed (2, 26). Preliminary reports suggest that these assays may be useful in monitoring interferon therapy (2, 5, 26).
A prototype enzyme immunoassay designed to detect and quantify total HCV core Ag protein in serum or plasma in the presence of anti-HCV Ab has been recently developed by Ortho-Clinical Diagnostics (Raritan, N.J.). The assay involves a 56°C sample pretreatment step using a denaturing solution to dissociate immune complexes and lyse viral particles. Free core Ag is captured on a 96-microwell plate coated with two murine monoclonal Ab (MAb) directed against different and conserved regions of the core protein. Detection of bound core protein is achieved by binding to two different murine MAb conjugated to horse-radish peroxidase. HCV core Ag concentrations are calculated against a standard curve.
Here, we present the results of a comparative study between the Total HCV core Ag assay and two quantitative assays for HCV RNA, bDNA v2.0 and v3.0 and the COBAS Amplicor HCV Monitor version 2.0 (HCM v2.0) test, in terms of specificity, sensitivity, and reproducibility in various clinical settings and in the follow-up of treated patients.
MATERIALS AND METHODS
World Health Organization HCV RNA 96/790 international standard.
The HCV RNA 96/790 standard prepared by the World Health Organization (WHO) was used to evaluate the sensitivity of the HCV core Ag assay. The undiluted standard should contain 50,000 IU/ml.
Fontevraud panel.
This quality control panel was prepared in our laboratory as previously described (20) and was used in the present study to evaluate the linearity and reproducibility of the Ag assay. Samples from patients chronically infected with different genotypes of HCV (Table 1) and exhibiting a range of viral load (range, 1,111,812 IU/ml to 12,199 IU/ml using bDNA v3.0 assay) were collected by plasmapheresis after obtaining written consent. Serial dilution of these samples was used to determine the linear behavior of the assay, while intra- and interassay reproducibility was assessed by testing the same samples in replicates of three.
TABLE 1.
Serial dilutions of the WHO HCV RNA 96/790 standard and plasmapheresis samples from Fontevraud panela
| Sample | Dilution | OD | HCV core concn (pg/ml) | HCV RNA level (IU/ml) | HCV core concn, corrected (pg/ml) |
|---|---|---|---|---|---|
| WHO Standard | 1/1 | 0.104 | 5.30 | 38,827 | 5.30 |
| 1/2 | 0.054 | 2.66 | (19,414) | 5.32 | |
| 1/4 | 0.031 | 1.48 | (9,707) | 5.94 | |
| 1, genotype 1 | 1/5 | 1.655 | 111.40 | 1,111,812 | 556.8 |
| 1/10 | 0.899 | 52.01 | (555,906) | 520.1 | |
| 1/20 | 0.531 | 27.00 | (277,953) | 539.3 | |
| 1/40 | 0.290 | 12.70 | (138,977) | 507.2 | |
| 1/80 | 0.166 | 6.30 | (69,488) | 505.7 | |
| 1/120 | 0.127 | 4.50 | (46,326) | 543.1 | |
| 2, genotype 3 | 1/10 | 0.859 | 42.99 | 298,376 | 429.9 |
| 1/20 | 0.472 | 20.56 | (149,188) | 411.2 | |
| 1/40 | 0.248 | 9.31 | (74,594) | 372.3 | |
| 1/80 | 0.127 | 4.08 | (37,297) | 326.6 | |
| 1/240 | 0.062 | 1.69 | (12,432) | 405.1 | |
| 3, genotype 4 | 1/1 | 1.752 | 103.41 | 731,398 | 103.4 |
| 1/10 | 0.300 | 11.77 | (73,140) | 117.7 | |
| 1/20 | 0.188 | 6.62 | (36,570) | 132.3 | |
| 1/40 | 0.109 | 3.38 | (18,285) | 135.3 | |
| 4, genotype 3 | 1/1 | 1.502 | 85.55 | 539,819 | 85.5 |
| 1/10 | 0.218 | 7.94 | (53,982) | 79.4 | |
| 1/20 | 0.132 | 4.28 | (26,991) | 85.6 | |
| 1/40 | 0.094 | 2.82 | (13,495) | 112.7 | |
| 5, genotype 2 | 1/1 | 0.166 | 6.33 | 39,237 | 6.3 |
| 1/2 | 0.138 | 4.98 | (19,618) | 9.9 | |
| 1/4 | 0.106 | 3.53 | (9,809) | 14.1 | |
| 6, genotype 2 | 1/1 | 0.108 | 3.62 | 12,199 | 3.6 |
| 1/2 | 0.090 | 2.85 | (6,099) | 5.7 |
Results obtained from different rounds. OD, optical density. HCV RNA levels were obtained from undiluted samples. HCV RNA levels in parentheses are expected values for each dilution. Mean of OD for negative control, 0.025 ± 0.010; mean of OD for standard, 1.5 pg/ml: 0.055 ± 0.011.
Negative controls.
We tested a total of 2,030 anti-HCV Ab and HCV RNA-negative blood donors from the Etablissement Français du Sang from Lille (France) (n = 1,008) and Strasbourg (France) (n = 1,022).
Anti-HCV-positive untreated patients.
One hundred forty anti-HCV Ab-positive patients were studied. Among these patients, 23 had serum alanine aminotransferase (ALT) levels above the upper normal limit. All the patients were HCV RNA-positive and had chronic hepatitis confirmed by liver biopsy.
Samples used for concentration studies.
In order to determine if a simple sample concentration procedure was capable of increasing the sensitivity of the HCV core Ag test, a total of six individual samples with different levels of core Ag (two HCV Ab negative and HCV RNA positive four HCV Ab and HCV RNA positive) were subjected to high-speed centrifugation. A total of six anti-HCV Ab-negative samples from healthy blood donors were used as negative controls and processed in a similar way.
Patients treated with IFN or with IFN/Rib combination therapy.
Prospectively collected samples from patients with chronic HCV infection enrolled in a therapeutic protocol, involving treatment with interferon (IFN) and ribavirin (Rib) combination therapy (IFN/Rib) were studied. Briefly, 144 patients received 3 million or 6 million IU of IFN three times a week for 3 months. The treatment was then adapted to the primary response after 2 months of IFN. Patients who still had detectable HCV RNA by the Amplicor HCV v1.2 assay (Roche Diagnostics) received Rib (1,000 to 1,200 mg per day) for an additional 9 months. The patients who had undetectable HCV RNA continued to receive IFN alone until month 12 (M12). HCV RNA and HCV Ag levels were measured before, during (M1, M2, M3, M4, and M6), and at the end (M12) of treatment and at M13 and M18 posttherapy.
Patients in whom both ALT levels remained normal and HCV RNA levels were undetectable 6 months after the end of therapy were defined as sustained responders (SR). Patients with undetectable HCV RNA at the end of therapy but detectable HCV RNA 6 months after the end of treatment were considered relapsers (R).
Nonresponders (NR) were defined as those patients who did achieve undetectable HCV RNA result during treatment.
Interestingly, R to IFN alone who were retreated with combination therapy during at least 6 months exhibited a sustained response and thus were included in the SR group in the HCV core Ag and HCV RNA kinetic study.
Detection of anti-HCV Ab.
Anti-HCV Ab were investigated by a third-generation HCV enzyme-linked immunosorbent assay (Axsym HCV version 3.0 [Abbott Diagnostics, Chicago, Ill.] and HCV 3.0 ELISA test system [Ortho-Clinical Diagnostics, Raritan, N.J.]) and by a third-generation recombinant immunoblot assay (RIBA HCV 3.0; Ortho-Clinical Diagnostics).
Quantification of total HCV core Ag.
A simple enzyme immunoassay for detection and quantification of total HCV core Ag has been recently developed (Ortho-Clinical Diagnostics). We used this test according to the manufacturer's recommendations. Briefly, 100 μl of one serum aliquot from each patient was pretreated with an immune complex dissociating buffer during 30 min at 56°C. Subsequently, samples and diluted standards were incubated during 1 h with shaking in a microwell coated with anti-core MAb at 25°C. After washing, an anti-core-specific conjugate was added and incubated for 30 min at 25°C. After a second wash, the Ab-core Ag conjugate was detected by the addition of o-phenylenediamine and hydrogen peroxide. Sample optical density was measured at 492 nm, and core Ag concentrations were calculated against a curve obtained from standards. Results were expressed in picograms per milliliter, with a limit of detection established by the manufacturer at 1.5 pg/ml.
Neutralization assay.
Positive samples for total HCV core Ag obtained in the specificity study were subjected to a neutralization procedure. Briefly, pretreated samples were incubated in quadruplicate for 1 h at 25°C with shaking. Before the addition of the conjugate, two wells were treated with 20 μl of neutralizing Ab reagent and the remaining two were treated with a similar volume of control Ab reagent. Following the addition of the conjugate, the assay was completed using the standard procedure. A sample was considered truly positive when a drop in concentration of ≥50% was observed in the neutralized versus the control treated sample.
Sample concentration.
Briefly, 800 μl of serum was centrifuged at 32,000 × g for 1 h at 4°C. After centrifugation, 700 μl of serum was carefully removed. The remaining 100 μl was used to perform a series of twofold dilutions (typically from 1:4 to 1:128) using HCV-negative serum. Similar dilutions were performed with the unprocessed sample. Both dilution series of concentrated and undiluted samples were processed according to the manufacturer's recommendations. The increase in concentration observed between the undiluted and concentrated sample for each dilution was expressed as percentage per 100 μl (concentration efficiency/100 μl). In order to identify a decrease in specificity due to this procedure, negative controls were processed in a similar fashion but no dilutions were performed.
Qualitative detection of HCV RNA.
HCV RNA was investigated using an in-house nested RT-PCR as previously described (9). This assay has a sensitivity ranging between 50 and 100 copies of HCV RNA per ml. The AMPLICOR HCV test, v1.2, and Cobas Amplicor HCV v2.0 assays were also used (29, 30). The HCM v2.0 assay has a lower limit of detection of 100 copies of HCV RNA per ml or 40 IU/ml.
Quantitative detection of HCV RNA.
Detection was performed using two commercial assays. (i) The HCM v2.0 assay (Roche Diagnostic Systems, Basel, Switzerland) is a quantitative assay with a lower detection limit of 600 IU per ml. We also used an ultrasensitive procedure on the HCM v2.0 assay, which allows detection of HCV RNA levels of 50 IU/ml. This method allowed 10-fold concentration of HCV RNA using a double volume of sera and fivefold less of sample diluent. (ii) The bDNA signal amplification assay (QUANTIPLEX HCV RNA 2.0 assay [Chiron Diagnostics, Emeryville, Calif.] and VERSANT HCV RNA 3.0 assay [Bayer Diagnostics, Emeryville, Calif.]), which is based on hybridization with specific probes located in the 5′ untranslated region. The lower detection limit of the bDNA v2.0 assay is 31,746 IU/ml, and the limit is 480 IU/ml for the bDNA v3.0 assay. The volume of serum required is 50 μl (28).
Genotyping.
The samples were genotyped using the second generation of Line Probe Assay (Innogenetics, Zwijnaarde, Belgium) (25). Genotypes 1a and 1b were grouped into genotype 1 due to the limited subtype discriminatory capacity of this assay (7).
Statistical analysis.
Statistical analysis for the measurement of agreement between the two RNA quantitative assays was performed as described by Bland and Altman (3), and the correlation between HCV core Ag and the two RNA quantitative assays was evaluated by linear regression. Univariate analysis was performed using a χ2 test and Student's t test. Results are presented as mean ± standard deviation log10. All statistical significance was assessed at the P ≤ 0.05 level. (For results under the cutoff value, we used the log10 of half the cutoff value).
RESULTS
The analysis of the WHO standard is summarized in Table 1. Positive results were obtained for HCV core Ag, in triplicate, up to the 1/4 dilution. The undiluted sample was quantified as having 5.30 pg of core Ag/ml and 38,827 IU/ml using the HCM v2.0 assay. The 1/4 dilution gave a core Ag concentration of 1.48 pg/ml, corresponding to 9,707 IU of HCV RNA/ml with the HCM v2.0 assay.
The reproducibility of the HCV core Ag assay was calculated using six samples of the Fontevraud panel tested in replicates of 10 in the same run and with the same six samples tested individually in three different runs. Intra- and interassay reproducibility results are shown in Tables 2 and 3, respectively. The intra-assay mean coefficient of variation (CV) was 5.11% for all six samples and varied from 1.40 to 5.51%, except for one plasmapheresis (CV = 11.65%; Table 2). The interassay CVs varied between 2.34 and 19.09% (mean = 9.95%; Table 3).
TABLE 2.
Intra-assay reproducibility of Total HCV core Ag assay resultsa
| Patient | Genotype | Result for sample:
|
Mean | CV (%)b | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| a | b | c | d | e | f | g | h | i | j | ||||
| 1 | 1 | 0.850 | 0.867 | 0.881 | 0.884 | 0.898 | 0.870 | 0.874 | 0.881 | 0.865 | 0.874 | 0.874 | 1.40 |
| 2 | 3 | 0.566 | 0.605 | 0.541 | 0.605 | 0.513 | 0.529 | 0.559 | 0.552 | 0.556 | 0.531 | 0.556 | 5.20 |
| 3 | 4 | 1.909 | 1.855 | 1.829 | 1.903 | 1.715 | 1.913 | 1.827 | 1.740 | 1.845 | 1.804 | 1.834 | 3.50 |
| 4 | 3 | 1.599 | 1.485 | 1.596 | 1.560 | 1.591 | 1.686 | 1.656 | 1.534 | 1.609 | 1.575 | 1.589 | 3.41 |
| 5 | 2 | 0.161 | 0.167 | 0.158 | 0.162 | 0.162 | 0.148 | 0.150 | 0.171 | 0.141 | 0.162 | 0.158 | 11.65 |
| 6 | 2 | 0.078 | 0.088 | 0.074 | 0.079 | 0.074 | 0.077 | 0.077 | 0.076 | 0.081 | 0.107 | 0.081 | 5.20 |
Total HCV core antigen is expressed in optical density. Sera from patients 1 and 2 were diluted at 1/10.
Mean CV, 5.11%.
TABLE 3.
Interassay reproducibility of Total HCV core Ag assay resultsa
| Patient | Genotype | Result for sample:
|
Mean | CV (%)b | ||
|---|---|---|---|---|---|---|
| A | B | C | ||||
| 1 | 1 | 1.007 | 1.365 | 0.876 | 1.083 | 19.09 |
| 2 | 3 | 0.593 | 0.613 | 0.579 | 0.595 | 2.34 |
| 3 | 4 | 1.524 | 1.540 | 1.874 | 1.646 | 9.80 |
| 4 | 3 | 1.377 | 1.276 | 1.560 | 1.404 | 8.37 |
| 5 | 2 | 0.164 | 0.130 | 0.162 | 0.152 | 10.25 |
| 6 | 2 | 0.079 | 0.064 | 0.080 | 0.074 | 9.85 |
Total HCV core Antigen is expressed in optical density. Sera from patients 1 and 2 were diluted at 1/10.
Mean CV, 9.95%.
Linearity was evaluated with the same samples used in the reproducibility study in three different runs and with the WHO HCV RNA 96/790 international standard. As shown in Table 1, the HCV core Ag assay exhibited linear behavior across a wide range of viremia levels. The coefficient of correlation was 0.99, and quantification of HCV core Ag was reliable up to the limit of detection of the test (1.5 pg/ml).
Among 2,030 anti-HCV Ab-negative samples tested from blood donors, 10 samples were repeatedly found positive with the HCV core Ag assay. When analyzed, these samples were found to be HCV RNA negative. Neutralization analysis (performed by Ortho-Clinical Diagnostics) showed that all 10 samples were in fact false positive. Thus, we observed a specificity of 99.5%.
The mean concentration efficiencies obtained in two independent experiments are summarized in Table 4. No significant increase in background was observed when six HCV Ab-negative samples were analyzed using the same procedure (data not shown).
TABLE 4.
Concentration of HCV core Ag by ultracentrifugation for the six HCV RNA-positive samplesa
| Sample | Concn of HCVcAg (pg/ml) | HCV Ab | % Mean concn efficiency/100 μl (range) for expt:
|
|
|---|---|---|---|---|
| 1 | 2 | |||
| a | 675.2 | Neg. | 18.9 (16.5-21.3) | 14 (NA) |
| b | 777.6 | Neg. | 17.9 (14.7-21.1) | Not done |
| c | 159.2 | Pos. | 24.3 (19-27.2) | 36.8 (34-39.6) |
| d | 54.8 | Pos. | 75.4 (NA) | 59 (58.5-59.5) |
| e | 99.6 | Pos. | Not done | 35.6 (33.8-37.5) |
| f | 4.2 | Pos. | Not done | 34 (NA) |
NA, not applicable; only one dilution was used to calculate the concentration efficiency. HCVcAg, HCV core Ag. Neg., negative; Pos., positive.
The comparison of the HCV core Ag assay with HCV RNA quantification tests with untreated and treated patients with chronic hepatitis C is shown in Tables 5 and 6. The levels of HCV core Ag and viremia were not significantly different between the three tests but were significantly different between the patients with normal ALT or elevated ALT (P < 0.01 for HCV core Ag assay; P < 0.02 for the two HCV RNA quantification assays).
TABLE 5.
Comparative results of the HCV core ag and HCV RNA quantitation assays for untreated patients with chronic hepatitis Ca
| Assay | Result (n) for patients with ALT levels:
|
P value | Result (n) for all patients (n = 140) | |
|---|---|---|---|---|
| ≤N (n = 23) | >N (n = 117) | |||
| HCV core Ag | 5.30 ± 0.94 (23) | 5.83 ± 0.82 (117) | <0.01 | 5.81 ± 0.86 (140) |
| bDNA HCV 2.0 or 3.0 | 5.28 ± 0.79 (22) | 5.76 ± 0.68 (115) | <0.02 | 5.73 ± 0.70 (137) |
| Monitor HCV 2.0 | 5.34 ± 1.46 (13) | 5.99 ± 0.66 (58) | <0.02 | 5.87 ± 0.90 (71) |
Results are expressed as mean ± standard deviation in log (104pg/ml) for HCV core antigen and in log (IU/ml) for HCV RNA levels. n, no. of samples. N, normal.
TABLE 6.
Comparative results of the HCV core Ag and HCV RNA quantitation assays for patients treated with IFN/Riba
| Assay | Result (n) for patient group
|
|||||
|---|---|---|---|---|---|---|
| Responders to IFN alone (n = 45) | Relapsers to IFN alone (n = 23) | Responders to IFN/Rib (n = 34) | Relapsers to IFN/Rib (n = 10) | Nonresponders (n = 32) | Total (n = 144) | |
| HCV core Ag | 5.28 ± 0.92 (45) | 6.07 ± 0.73 (23) | 5.92 ± 0.88 (34) | 5.79 ± 0.68 (10) | 6.08 ± 0.62 (32) | 5.77 ± 0.87 (144) |
| bDNA HCV 2.0 or 3.0 | 5.37 ± 0.70 (42) | 5.89 ± 0.69 (23) | 5.88 ± 0.73 (33) | 5.65 ± 0.70 (10) | 5.84 ± 0.60 (32) | 5.70 ± 0.71 (140) |
| Monitor HCV 2.0 | 5.42 ± 1.30 (19) | 6.21 ± 0.69 (15) | 6.18 ± 0.63 (15) | 5.90 ± 0.57 (10) | 5.84 ± 0.64 (13) | 5.88 ± 0.90 (72) |
Results are expressed as mean ± standard deviation in log (104 pg/ml) for HCV core antigen and in log (IU/ml) for HCV RNA levels. n, no. of samples. N, normal.
All the patients with normal ALT level (n = 23) tested HCV RNA positive using RT-PCR and Amplicor HCV v1.2. Among these 23 samples, 22 were detected by HCM v2.0 and bDNA v3.0 assays. The HCV core Ag assay failed to detect for four samples. The patient with a normal ALT level, whose infection was not detected with HCV core Ag assay, HCM v2.0, and bDNA v3.0 assays, was infected with a genotype 1 HCV isolate. The other three patients had low viral loads (patient 1, genotype 1, viral load of 1,368 IU/ml by bDNA v3.0 and 3,720 IU/ml by HCM v2.0; patient 2, genotype 1, viral load of 13,200 IU/ml by bDNA v3.0; patient 3, genotype 1, viral load of 33,692 IU/ml using bDNA v3.0).
All the patients with elevated ALT (n = 117) were HCV RNA positive. The HCM v2.0 assay and the bDNA v3.0 assay were able to detect infection in all the samples. The HCV core Ag assay failed to detect infection in two samples (patient 1: genotype 1, viral load = 35,692 IU/ml using bDNA v3.0; patient 2: genotype 3, viral load = 58,413 IU/ml using bDNA v3.0).
We did not find any correlation between HCV core Ag and ALT levels (r2 = 0.07), HCV core Ag and Knodell score (P = 0.786 [not significant]), or HCV core Ag and Metavir grade (activity, P = 0.129 [not significant]; fibrosis, P = 0.429 [not significant]).
HCV core Ag correlation with HCV RNA quantification assays.
We found a correlation coefficient of 0.80 and a slope of 0.998 with the bDNA v3.0 assay and a correlation coefficient of 0.70 and a slope of 0.923 with the HCM v2.0 assay. The measurement of agreement between the two HCV RNA quantitative assays was not good, with a mean of the difference of −0.36, after logarithmic transformation, and the limits of agreement (mean ± 2 times the standard deviation) were −1.36 and 0.65, respectively. The HCV core Ag assay exhibited similar efficiency for quantifying the five different HCV genotypes studied. We did not observed significant differences in the level of HCV core Ag and HCV RNA in patients infected with these HCV genotypes (genotype 1, 5.80 ± 0.89, 5.71 ± 0.76, and 5.74 ± 1.01 for HCV core Ag assay, bDNA assay, and HCM v2.0 assay, respectively; genotype 2, 5.43 ± 0.71, 5.36 ± 0.62, and 5.73 ± 0.54, respectively; genotype 3, 5.84 ± 0.81, 5.82 ± 0.66, and 6.21 ± 0.75, respectively; genotype 4, 5.67 ± 0.58, 5.70 ± 0.54, and 5.59 ± 0.41, respectively; genotype 5, 6.02 ± 0.90, 5.80 ± 0.74, and 6.07 ± 0.93, respectively). In addition, we found a good correlation between the HCV core Ag assay and the two HCV RNA quantification assays, with exception for genotype 2 (genotype 1, r2 = 0.847 and slope = 1.038; genotype 2, r2 = 0.645 and slope = 0.876; genotype 3, r2 = 0.704 and slope = 1.025; genotype 4, r2 = 0.928 and slope = 1.032; genotype 5, r2 = 0.949 and slope = 1.185 [Fig. 1]).
FIG. 1.
Correlation between HCV core antigen and HCV RNA load among different HCV genotypes. x axis, HCV RNA in log (IU/ml); y axis, HCV core antigen in log (104 pg/ml).
Decrease in HCV core Ag during treatment follow-up.
Before IFN treatment, all samples were found to be positive for HCV RNA using RT-PCR and the AMPLICOR HCV test v1.2. HCV core Ag was detectable in 138 of 144 patients (95.8%). HCV RNA was detectable, using the HCM v2.0 and bDNA v3.0 assays, in 143 of 144 patients (99.3%).
The HCV core Ag and HCV RNA kinetic profiles observed during treatment are summarized in Fig. 2A and B. We identified five different categories of patients based on treatment response: SR and R to IFN therapy and SR, R, and NR to IFN/Rib combination therapy.
FIG. 2.
Comparison of HCV core antigen (A) and HCV RNA (B) kinetics in five different response categories. x axis, months of treatment; y axis, mean of HCV core antigen and HCV RNA, in log (104 pg/ml) and log (IU/ml), respectively. ··········, limit of detection. SR1, SR to IFN alone (n = 45); R1, relapsers to IFN alone (n = 23); SR2, sustained responders to IFN/Rib (n = 34); R2, relapsers to IFN/Rib (n = 10); NR, nonresponders to IFN/Rib (n = 32).
Dynamics of Total HCV core Ag concentration. (i) IFN monotherapy and combination therapy in different categories of treatment response. (a) Baseline.
the proportion of patients with undetectable HCV core Ag before IFN therapy was higher in SR or R than in NR (11.1% versus 4.3 and 0%, respectively). The pretreatment HCV core Ag and HCV RNA levels were significantly lower in SR to IFN alone than in the other groups of patients (mean HCV core Ag = 5.31 ± 0.88 log10 (pg/ml × 10,000) versus 5.99 ± 0.75 log10 (pg/ml × 10 000) [P < 0.001], and mean HCV viral load = 5.35 ± 1.01 log IU/ml versus 5.75 ± 0.70 log IU/ml [P < 0.001]). No significant differences were observed in these two parameters between SR, R, and NR to combination therapy.
(b) One-month time point.
All SR and R to IFN monotherapy had undetectable HCV core Ag and a mean decrease in its concentration of 1.43 log and 2.18 log, respectively. SR and R to IFN/Rib also had a significant decrease in HCV core Ag concentration, but this was less impressive in the case of the R (1.69 and 1.42 log10, respectively). Patients who did not respond to combination therapy had a mean decrease of 0.57 log in HCV core Ag concentration.
(c) Three-month time point.
A significant decrease in HCV core Ag was observed only in SR with IFN/Rib treatment (1.92 log10) but not in the two others categories of patients treated with combination therapy (1.55 and 0.55 log10 for R and NR, respectively).
(d) Six-month time point (3 months of IFN/Rib treatment).
All SR and R to IFN/Rib treatment had undetectable HCV core Ag with a mean decrease of 2.04 and 1.91 log, respectively. NR also showed a substantial drop in HCV core Ag concentrations, but they increased at the end of treatment (1.42 and 0.54 log at M6 and M12, respectively).
(ii) IFN monotherapy and combination therapy in different categories of treatment response.
For monotherapy patients, SR and R to IFN showed a drastic drop in the mean HCV RNA load of 3.59 log and 3.45 log, respectively. For patients who did not respond to IFN alone initially and were subsequently placed on combination therapy after 3 months, there was a lesser decrease in the mean HCV RNA load after the first month of IFN monotherapy (1.81 log, 1.35 log, and 0.32 log, respectively, for SR, R, and NR to combination therapy). All patients treated with IFN alone had undetectable HCV RNA levels after 2 months of therapy, with a mean decrease of 3.86 log in SR and 4.49 log in R. SR to combination therapy had a mean decrease in HCV RNA of 3.05 log and 4.48 log at M3 and M6, respectively (all the patients who were tested with the HCM v2.0 test had viral loads under the assay's cutoff at M6). The combination therapy R group had a less impressive decrease in viral load after 3 months of IFN monotherapy (1.89 log) and after 3 months of combination therapy (4.02 log), but all of them had undetectable HCV RNA at the end of treatment. NR to combination therapy were characterized by an increase in mean HCV RNA load after 3 months of IFN therapy alone (+ 0.22 log). A decrease in HCV RNA until M6 (0.62 log) was observed under combination therapy, but the decrease in viral load did not achieved statistical significance at the end of treatment (0.23 log).
Predictive value comparison between HCV core Ag, HCV RNA levels, and qualitative HCV RNA detection.
Table 7 summarizes the predictive values obtained using the criteria of a ≥2 log decrease in the HCV RNA concentration, absence of detectable HCV core Ag levels, or undetectable HCV RNA at different time points after initiation of treatment.
TABLE 7.
Decrease of viral load (2 log) and undetectable HCV core Ag and HCV RNA after 1, 3, and 6 month(s) of treatment
| Month or response group | No. with result/no. tested (%)
|
||
|---|---|---|---|
| Decrease ≥2 log HCV RNA | Undetectable HCV core Ag | Undetectable HCV RNA | |
| M1 | |||
| Sustained response | |||
| Responders to IFN alone | 43/45 (96) | 44/44 (100) | 38/46 (83) |
| Relapsers to IFN alone | 21/22 (95) | 22/22 (100) | 11/23 (48) |
| Responders to IFN/Rib | 12/29 (41) | 17/34 (50) | 0/29 (03) |
| Total | 76/96 (79) | 83/100 (83) | 49/98 (50) |
| Relapsers to IFN/Rib | 3/10 (30) | 4/10 (40) | 0/10 (0) |
| Nonresponders to IFN/Rib | 1/32 (3) | 2/32 (6) | 0/35 (0) |
| M3 | |||
| Sustained response | |||
| Responders to IFN alone | 45/45 (100) | 45/45 (100) | 46/46 (100) |
| Relapsers to IFN alone | 23/23 (100) | 23/23 (100) | 23/23 (100) |
| Responders to IFN/Rib | 19/26 (73) | 30/34 (88) | 9/26 (34) |
| Total | 87/94 (93) | 98/102 (96) | 78/95 (82) |
| Relapsers to IFN/Rib | 4/10 (40) | 4/10 (40) | 0/10 (0) |
| Nonresponders to IFN/Rib | 0/14 (0) | 1/30 (3) | 0/35 (0) |
| M6 | |||
| Sustained response | |||
| Responders to IFN alone | 45/45 (100) | 45/45 (100) | 46/46 (100) |
| Relapsers to IFN alone | 23/23 (100) | 23/23 (100) | 23/23 (100) |
| Responders to IFN/Rib | 34/34 (100) | 34/34 (100) | 34/38 (89) |
| Total | 102/102 (100) | 102/102 (100) | 103/107 (96) |
| Relapsers to IFN/Rib | 10/10 (100) | 10/10 (100) | 8/10 (80) |
| Nonresponders to IFN/Rib | 0/14 (0) | 10/30 (33) | 0/35 (0) |
After 1 and 3 months of IFN therapy, HCV core Ag was overall a better predictor of SR (positive predictive value) than qualitative or quantitative HCV RNA testing (M1: 83, 50, and 79%, respectively; M3: 96, 82, and 93%, respectively).
After 3 months of combination therapy (M6), no differences in predictive value were observed after applying a ≥2 log decrease in the HCV RNA concentration or the disappearance of detectable HCV core Ag or HCV RNA criteria.
DISCUSSION
Since new molecular technologies to quantify HCV RNA became widely available, numerous studies have tried to establish a relationship between level of viremia and severity and progression of liver disease. In addition, investigators have tried to understand the likelihood of response to IFN/Rib therapy based on viremia levels. Some studies have found a correlation between disease severity (cirrhosis) and the level of viremia, while others have not (9, 12, 15). Several studies (6, 11, 13, 18, 21) have focused on the predictive value of the pretreatment HCV RNA level or infecting genotype in the response to IFN. Only 41% of the patients treated with IFN-Rib during 48 weeks show a long-term response (24). Patients with low HCV RNA level or infected by genotype 2 or 3 independently exhibit a sustained response more frequently than patients infected with genotype 1 or with a high level of HCV RNA. Toyoda et al. (27) first showed that patients with an HCV RNA level over 106 eq/ml as measured with bDNA v2.0 had only a 4% chance of responding to IFN. Several studies have evaluated different methods for the quantification of HCV RNA (10), such as quantitative RT-PCR (11, 17, 22) and branched-DNA (1, 12, 18), Amplicor Monitor (23), or NASBA (14, 16, 19, 20). In a recent study, Tanaka et al. showed that the new HCV core Ag assay was efficient for monitoring patients treated with IFN (26).
In this report, we have compared the performance of a new HCV core Ag assay with that of two commercially available quantitative assays for the assessment of HCV RNA levels. We used a panel containing dilutions of plasma samples (20), samples from anti-HCV positive patients, and samples from chronically HCV-infected patients undergoing different IFN treatment schedules. We found that the HCV core Ag assay exhibited linear behavior across a wide range of dilutions and genotypes. The mean intra- and interassay CVs obtained were 5.11 and 9.95%, respectively. The limit of detection of the HCV core Ag assay in International Units per milliliter of HCV RNA based on the WHO standard was established around 10,000. According to the linearity obtained, an increase in the WHO HCV RNA standard of 8,500 IU/ml corresponded to an increase of 1 pg/ml in core Ag. This is in agreement with the equivalency observed between numbers of picograms of core Ag and International Units of HCV RNA reported by Bouvier-Alias et al. (5). However, the same investigators observed that the RNA/core Ag ratio varies among some patients. The clinical significance of these fluctuations is currently not understood and is the subject of investigation. HCV core Ag detection could be enhanced by sample concentration in both anti-HCV-positive and -negative samples (Table 4). Interestingly, the efficiency of this procedure was sample dependent and greater for samples containing anti-HCV Ab. Comparing the reproducibility of the three assays, we found that our results were closely related to those expected with the Fontevraud panel (20). The reproducibility was very high regardless of genotype and viral load. Specificity assessed with HCV RNA-undetectable seropositive patients and with blood donors was also very high (99.5%).
In anti-HCV-positive untreated patients, we found that the HCV core Ag assay was capable of detecting most HCV RNA-positive samples (96.1%). When comparing patients with normal or elevated ALT, we found a significant difference in the levels of HCV core Ag as well as HCV RNA load. Conversely, there was no correlation between HCV core Ag and ALT levels and between patients with normal, mild, or severe liver lesions determined by liver biopsy. This observation reinforces the notion that HCV RNA levels in serum do not reflect the severity of the disease, and apparently serum HCV core Ag levels also do not.
When we compared the abilities of these tests to detect five different HCV genotypes, we found that the sensitivities of the three assays were comparable for all genotypes investigated. The HCV core Ag assay and the viral RNA assays correlated across all genotypes analyzed.
For treated patients, HCV core Ag decrease followed the same trend as, but became undetectable more frequently than, HCV RNA (HCV core Ag, 52.2%; bDNA v3.0, 72.8%; and HCM v2.0, 95.2%).
In the patients treated with IFN/Rib, we found an association between the pretreatment HCV core Ag and HCV RNA levels and the response to treatment. The association was stronger with viremia levels determined with the HCV core Ag and bDNA v3.0 assay (P < 0.001) than with the Amplicor Monitor v2.0 assay (P < 0.01). However, no differences between the HCV core Ag and HCV RNA levels were observed between sustained virological responders, R, and NR to IFN/Rib.
In conclusion, the performance characteristics exhibited by the HCV core Ag assay in this study support its use for the current indications requiring HCV viremia detection and quantification. An initially apparent drawback of this assay is its lower limit of detection (approximately 10,000 IU of HCV RNA/ml). Analysis of viral load distribution in chronic HCV infection indicates that most patients exhibit viral loads well above 10,000 IU/ml (4). This observation and our own results suggest that the Total core Ag assay meets the clinical sensitivity needs necessary to identify most (>95%) HCV viremic patients. In addition, the Total HCV core Ag test results could be obtained in routine laboratories without the need of special equipment or training in less than 3.5 h. Our results also showed that the HCV core assay has a high predictive value for treatment response at 1 and 3 months post-initiation of treatment for patients treated with IFN. Further prospective studies on large series of patients are needed in order to evaluate the predictive value of the HCV core Ag assay on new treatment strategies, such as use of pegylated IFN and Rib.
Acknowledgments
We thank the Fontevraud Study Group for their participation in the therapeutic protocol and comments and Christine Brault, Maryline Peigne, and Dominique Fray for technical help. We also thank Schering-Plough and SIDACTION/Ensemble contre le SIDA for supporting the therapeutic protocol and Ortho-Clinical Diagnostics for providing the Total HCV core antigen kits.
REFERENCES
- 1.Alter, H. J., R. Sanchez-Pescador, M. S. Urdea, J. C. Wilber, R. J. Lagier, A. M. Di Bisceglie, J. W. Shih, and P. D. Neuwald. 1995. Evaluation of branched-DNA signal amplification for the detection of hepatitis C virus RNA. J. Viral Hepatitis 2:121-132. [DOI] [PubMed] [Google Scholar]
- 2.Ayoagi, K., C. Ohue, K. Iida, T. Kimura, E. Tanaka, K. Kiyosawa, and S. Yagi. 1999. Development of a single and highly sensitive enzyme immunoassay for hepatitis C virus core antigen. J. Clin. Microbiol. 37:1802-1808. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Bland, J. M., and D. G. Altman. 1986. Statistical methods for assessing agreement between two methods of clinical measurements. Lancet i:307-310. [PubMed]
- 4.Blatt, L. M., M. G. Mutchnick, M. J. Tong, F. M. Klion, E. Lebovics, B. Freilich, N. Bach, C. Smith, J. Herrera, H. Tobias, A. Conrad, P. Schmid, and J. G. McHutchinson. 2000. Assessment of hepatitis C virus RNA and genotype from 6807 patients with chronic hepatitis C in the United States. J. Viral Hepatitis 7:196-202. [DOI] [PubMed] [Google Scholar]
- 5.Bouvier-Alias, M., K. Patel, H. Dahari, S. Beaucourt, P. Larderie, L. Blatt, C. Hezode, G. Picchio, D. Dhumeaux, A. U. Neumann, J. G. McHutchison, and J. M. Pawlotsky. 2002. Clinical utility of total HCV core antigen quantification: a new indirect marker of HCV replication. Hepatology 36:211-218. [DOI] [PubMed] [Google Scholar]
- 6.Chemello, L., L. Cavaletto, F. Noventa, P. Bonetti, C. Casarin, E. Bernardinello, P. Pontisso, C. Donada, P. Casarin, F. Belussi, M. Frezza, and A. Alberti. 1995. Predictors of sustained response, relapse and no response in patients with chronic hepatitis C treated with interferon alfa. J. Viral Hepatitis 2:91-96. [DOI] [PubMed] [Google Scholar]
- 7.Chen, Z., and K. E. Weck. 2002. Hepatitis C virus genotyping: interrogation of the 5′ untranslated region cannot accurately distinguish genotypes 1a and 1b. J. Clin. Microbiol. 40:3127-3134. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Choo, Q. L., G. Kuo, A. J. Weiner, L. R. Overby, D. W. Bradley, and M. Houghton. 1989. Isolation of a cDNA clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science 244:2451-2455. [DOI] [PubMed] [Google Scholar]
- 9.Gretch, D. R., L. Corey, J. Wilson, C. de la Rosa, R. Wilson, R. Carithers, Jr., M. Bush, J. Hart, M. Sayers, and J. Han. 1994. Assessment of hepatitis C virus RNA levels by quantitative competitive RNA polymerase chain reaction: high-titer viremia correlates with advanced stage of disease. J. Infect. Dis. 169:1219-1225. [DOI] [PubMed] [Google Scholar]
- 10.Gretch, D. R., C. de la Rosa, L. Corey, and R. L. Carithers. 1996. Assessment of hepatitis C viremia using molecular amplification technologies. Viral Hepatitis Rev. 2:85-96. [DOI] [PubMed] [Google Scholar]
- 11.Hagiwara, H., N. Hayashi, E. Mita, T. Takehara, A. Kasahara, H. Fusamoto, and T. Kamada. 1993. Quantitative analysis of hepatitis C virus RNA in serum during interferon alpha therapy. Gastroenterology 104:877-883. [DOI] [PubMed] [Google Scholar]
- 12.Hagiwara, H., N. Hayashi, E. Mita, T. Takehara, A. Kasahara, H. Fusamoto, and T. Kamada. 1993. Quantitation of hepatitis C virus RNA in serum of asymptomatic blood donors and patients with type C chronic liver disease. Hepatology 17:545-550. [DOI] [PubMed] [Google Scholar]
- 13.Hino, K., S. Sainokami, K. Shimoda, S. Iiono, Y. Wang, H. Okamoto, and Y. Miyakawa. 1994. Genotypes and titers of hepatitis C virus for predicting response to interferon in patients with chronic hepatitis C. J. Med. Virol. 42:299-305. [DOI] [PubMed] [Google Scholar]
- 14.Hollingsworth, R., P. Sillekens, P. Van Deursen, K. R. Neal, W. L. Irving, et al. 1996. Serum HCV RNA level assessed by quantitative NASBA: stability of viral load over time, and lack of correlation with liver disease. J. Hepatol. 25:301-306. [DOI] [PubMed] [Google Scholar]
- 15.Kato, N., O. Yokosuka, K. Hosoda, Y. Ito, M. Ohto, and M. Omata. 1993. Quantitation of hepatitis C virus by competitive reverse transcription-polymerase chain reaction: increase of the virus in advanced disease. Hepatology 18:16-20. [PubMed] [Google Scholar]
- 16.Kievits, T., B. Van Gemen, S. Van Strijp, R. Schukkink, M. Dircks, H. Adriaanse, L. Malek, R. Sooknanan, and P. Lens. 1991. NASBA™ isothermal enzymatic in vitro nucleic acid amplification optimized for the diagnosis of HIV-1 infection. J. Virol. Methods 35:273-286. [DOI] [PubMed] [Google Scholar]
- 17.Kobayashi, Y., S. Watanabe, M. Konishi, M. Yokoi, R. Kakahashi, M. Kaito, M. Kondo, Y. Hayashi, T. Jomori, and S. Suzuki. 1993. Quantitation and typing of serum hepatitis C virus RNA in patients with chronic hepatitis C treated with interferon beta. Hepatology 18:1319-1325. [PubMed] [Google Scholar]
- 18.Lau, J. Y., G. L. Davis, J. Kniffen, K. P. Quian, M. S. Urdea, C. Chan, M. Mizokami, P. D. Neuwald, and J. C. Wilber. 1993. Significance of serum hepatitis C virus RNA levels in chronic hepatitis. Lancet 341:1501-1504. [DOI] [PubMed] [Google Scholar]
- 19.Lunel, F., M. Mariotti, P. Cresta, I. De La Croix, J. M. Huraux, and J. J. Lefrère. 1995. Comparative study of conventional and novel strategies for the detection of hepatitis C virus RNA in serum: Amplicor, branched-DNA, NASBA and in-house PCR. J. Virol. Methods 54:159-171. [DOI] [PubMed] [Google Scholar]
- 20.Lunel, F., P. Cresta, D. Vitour, C. Payan, B. Dumont, L. Frangeul, D. Reboul, C. Brault, J. C. Piette, and J. M. Huraux. 1999. Comparative evaluation of hepatitis C virus RNA quantitation by branched DNA, NASBA and monitor assays. Hepatology 29:528-535. [DOI] [PubMed] [Google Scholar]
- 21.Martinot-Peignoux, M., P. Marcellin, M. Pouteau, C. Castelnau, N. Boyer, M. Poliquin, C. Degott, I. Descombes, V. Le Breton, V. Milotova, J. P. Benhamou, and S. Erlinger. 1995. Pretreatment serum hepatitis C virus RNA levels and hepatitis C virus genotype are main and independent prognostic factors of sustained response to interferon alfa therapy in chronic hepatitis C. Hepatology 22:1050-1056. [PubMed] [Google Scholar]
- 22.Naito, M., N. Hayashi, H. Hagiwara, K. Katayama, A. Kasahara, H. Fusamoto, M. Kato, M. Masuzawa, and T. Kamada. 1994. Serial quantitative analysis of serum hepatitis C virus RNA level in patients with acute and chronic hepatitis C. J. Hepatol. 20:755-759. [DOI] [PubMed] [Google Scholar]
- 23.Navas, S., and V. Carreno. 1995. Semiquantitation by Amplicor assay of hepatitis C virus genome during therapy. J. Hepatol. 22(Suppl.1):115-117. [PubMed] [Google Scholar]
- 24.Poynard, T., J. McHutchinson, Z. Goodman, M. H. Ling, H. J. Albrecht, et al.. 2000. Is an “A La Carte” combination interferon alfa-2b plus ribavirin regimen possible for the first line treatment in patients with chronic hepatitis C? Hepatology 31:211-218. [DOI] [PubMed] [Google Scholar]
- 25.Stuyver, L., R. Rossau, A. Wyseru, M. Duhamel, B. Vanderborght, H. Van Heuversmyn, and G. Maartens. 1993. Typing of hepatitis C virus isolates and characterization of new subtypes using a line probe assay. J. Gen. Virol. 74:1093-1102. [DOI] [PubMed] [Google Scholar]
- 26.Tanaka, E., C. Ohue, K. Ayoagi, K. Yamaguchi, S. Yagi, K. Kiyosawa, and H. J. Alter. 2000. Evaluation of a new immunoassay for hepatitis C virus (HCV) core antigen with clinical sensitivity approximating that of genomic amplification of HCV RNA. Hepatology 32:388-393. [DOI] [PubMed] [Google Scholar]
- 27.Toyoda, H., S. Nakano, T. Kumada, I. Takeda, K. Sugiyama, T. Osada, S. Kiriyama, E. Orito, and M. Mizokami. 1996. Comparison of serum hepatitis C virus RNA concentration by branched-DNA probe assay with competitive reverse transcription polymerase chain reaction as a predictor of response to interferon alpha therapy in chronic hepatitis C patients. J. Med. Virol. 48:354-359. [DOI] [PubMed] [Google Scholar]
- 28.Urdea, M. S. 1993. Synthesis and characterisation of branched DNA (bDNA) for the direct and quantitative detection of CMV, HBV, HCV and HIV. Clin. Chem. 39:725-726. [Google Scholar]
- 29.Wolfe, L., S. Tamatsukuri, C. Sayada, and J. C. Ryff. 1994. Detection of HCV RNA in serum using a single-tube enzyme PCR in combination with a colorimetric microwell assay, p. 83-94. In Groupe Français d'Etude Moléculaire des Hépatites (GEMHEP) (ed.), Hepatitis V virus: new diagnostic tools. John Libbey Eurotext, Paris, France.
- 30.Young, K. K. Y., R. M. Resnick, and T. W. Myers. 1993. Detection of hepatitis C virus RNA by a combined reverse transcription-polymerase chain reaction assay. J. Clin. Microbiol. 31:882-886. [DOI] [PMC free article] [PubMed] [Google Scholar]