ABSTRACT
The detection and quantification of hepatitis B virus (HBV) DNA are essential for the diagnosis and treatment of chronic HBV infection. The use of real-time PCR assays for HBV DNA quantification is strongly recommended. The goal of this study was to evaluate the intrinsic characteristics and clinical performance of version 2.0 (v2.0) of the Cobas AmpliPrep/Cobas TaqMan (CAP/CTM) assay, a fully automated platform for HBV DNA quantification in serum or in plasma with a claimed lower limit of detection of 20 IU/ml and a claimed upper limit of quantification of 1.7 × 108 IU/ml. The specificity of the assay was 99% (95% confidence interval, 94.7 to 100%). Intra-assay and interassay coefficients of variation ranged from 0.21% to 2.67% and from 0.65% to 2.25%, respectively. The calibration of the assay was found to be satisfactory. Study of blood specimens from patients infected with HBV genotypes A to F showed good correspondence between HBV DNA levels measured by the CAP/CTM v2.0 assay, version 1.0 of the same assay, and the third-generation “branched DNA” assay. The CAP/CTM v2.0 assay quantified HBV DNA levels in serum or plasma from the same patients equally. In conclusion, the new version of the CAP/CTM assay is sensitive, specific, and reproducible. It accurately quantifies HBV DNA levels in patients chronically infected with HBV genotypes A to F. Improvements made to ensure equal quantification of HBV DNA in serum and plasma have been successful. Overall, the CAP/CTM assay, version 2.0, is well suited to monitoring clinical HBV DNA levels according to current clinical practice guidelines.
Chronic hepatitis B virus (HBV) infection is associated with a large spectrum of liver diseases, ranging from a low-viremia inactive carrier state to chronic active hepatitis, which may subsequently evolve toward cirrhosis and hepatocellular carcinoma (HCC). Morbidity and mortality are linked to the persistence of viral replication, and hepatic complications develop in 15% to 40% of patients chronically infected with HBV. Overall, HBV-related end-stage liver disease and HCC are responsible for more than 750,000 deaths worldwide per year (5).
The detection and quantification of HBV DNA are essential for diagnosing ongoing HBV infection and establishing the prognosis of related liver disease, influence the decision to treat, and are indispensable for monitoring the virological response to antiviral therapy and the emergence of resistance in order to tailor therapy (4). A number of HBV DNA detection and quantification assays are available. For many years, such methods were based on either hybrid capture, signal amplification by means of branched DNA (bDNA) technology, or classical PCR, all of which suffered from poor analytical sensitivity and a narrow range of HBV DNA quantification (3). More recently, assays based on real-time PCR quantification have been developed. Their use for the routine detection and quantification of HBV DNA is recommended, because of their excellent analytical sensitivity (lower limit of detection, 10 to 20 IU/ml), their specificity, their accuracy, and their broad dynamic range of linear quantification, which fully covers clinical needs (9). Among these assays, we recently evaluated the first-generation (V1.0) Cobas AmpliPrep/Cobas TaqMan (CAP/CTM; Roche Molecular Systems, Pleasanton, CA) assay. This assay was found to be sensitive, specific, and reproducible and to accurately quantify HBV DNA levels in patients chronically infected by HBV genotypes A to F (2). However, this assay could be used only for the quantification of HBV DNA in plasma.
A second version of the CAP/CTM assay (v2.0) has been released recently. Several changes have been made; in particular, the assay can be used on both serum and plasma, and it requires 650 μl of sample instead of 850 μl. Its claimed dynamic range of quantification is 20 IU/ml to 1.7 × 108 IU/ml (1.3 to 8.2 log10 IU/ml). The goal of this study was to evaluate the intrinsic characteristics and clinical performance of the CAP/CTM v2.0 assay.
MATERIALS AND METHODS
Materials. (i) Standards.A standard panel of HBV genotype A plasma samples (OptiQuant HBV DNA; AcroMetrix, Benicia, CA) was used to study the analytical performance of the assay. The panel was made of 7 samples (NAP-000 to NAP-HBV2E7) containing no HBV DNA and 2 × 102 IU/ml (2.3 log10 IU/ml), 2 × 103 IU/ml (3.3 log10 IU/ml), 2 × 104 IU/ml (4.3 log10 IU/ml), 2 × 105 IU/ml (5.3 log10 IU/ml), 2 × 106 IU/ml (6.3 log10 IU/ml), and 2 × 107 IU/ml (7.3 log10 IU/ml) of HBV DNA, respectively.
(ii) Clinical specimens.Plasma and serum samples were obtained from patients attending the Department of Hepatology and Gastroenterology of the Henri Mondor Hospital and from blood donors diagnosed with an HBV infection at the Institut National de la Transfusion Sanguine. Group A comprised 103 HBV-seronegative individuals (with no markers of past or ongoing HBV infection); group B comprised 97 patients with serological profiles of resolved HBV infection (the presence of both anti-HBc and anti-HBs antibodies). Group C comprised 51 patients with chronic HBV infections, all of whom had detectable HBsAg, anti-HBc antibodies, and HBV DNA. Based on the sequencing of a portion of the S gene followed by phylogenetic analysis, this group comprised 12 patients with HBV genotype A, 9 with genotype B, 8 with genotype C, 9 with genotype D, 10 with genotype E, and 3 with genotype F, as previously described (2). Group D included 16 patients receiving nucleoside/nucleotide analogue therapy who were serially sampled on treatment. They included 4 patients infected with genotype A, 5 with genotype C, 1 with genotype D, and 6 with genotype E. In total, 51 plasma and 51 serum specimens sampled at the same times were available from group D patients (3 time points per patient on average; range, 1 to 7).
Assessment of the performance of the CAP/CTM v2.0 assay. (i) Analytical sensitivity.To determine the analytical sensitivity of the assay, the NAP-HBV2E4 standard was serially diluted from 50 IU/ml (1.7 log10 IU/ml) to 6.25 IU/ml (0.8 log10 IU/ml). Twenty-one replicates of each HBV DNA concentration were tested in different experiments.
(ii) Specificity.The specificity of the CAP/CTM v2.0 assay was assessed by testing the 103 clinical specimens from group A and the 97 clinical specimens from group B.
(iii) Precision and reproducibility.To assess precision (intra-assay reproducibility), each sample in the OptiQuant HBV DNA standard panel was tested in triplicate. To assess interassay reproducibility, the low-positive control (LPC) and the high-positive control (HPC) provided with the kits were tested 18 times in corresponding runs on different days.
(iv) Linearity, accuracy, and influence of the HBV genotypes.The linearity of quantification by the CAP/CTM v2.0 assay was assessed by testing the 7 members of the OptiQuant HBV DNA standard panel. Each panel member was tested three times in the same experiment with the CAP/CTM v2.0 assay. The average measured values were then compared with the expected values. The 51 plasma specimens from group C were tested in parallel with the CAP/CTM v2.0, CAP/CTM v1.0, and bDNA assays. If needed, dilutions were made with the Nucleic Acid Test (NAT) dilution matrix (AcroMetrix), a defibrinated, delipidated normal human plasma sample.
(v) Assessment of equal quantification in serum and plasma.The 51 plasma specimens and 51 serum specimens from patients in group D sampled at the same time points were tested with version 2.0 of the CAP/CTM assay, and the results were compared. The plasma samples were tested in parallel by the CAP/CTM v1.0 assay; both the serum and the plasma samples were tested by the bDNA assay.
HBV DNA quantification. (i) CAP/CTM v1.0 and v2.0 assays.HBV DNA was extracted from 850 μl of plasma for version 1.0 and from 650 μl of plasma or serum for version 2.0 by use of the Cobas AmpliPrep automated extractor, according to the manufacturer's instructions. The Cobas TaqMan 96 analyzer was used for automated real-time PCR amplification and detection of PCR products according to the manufacturer's instructions. The data were analyzed with Amplilink software. HBV DNA levels were expressed in international units per milliliter. The dynamic ranges of quantification of the CAP/CTM v1.0 and CAP/CTM v2.0 assays are 54 to 110,000,000 IU/ml (1.7 to 8.0 log10 IU/ml) and 20 to 170,000,000 IU/ml (1.3 to 8.2 log10 IU/ml), respectively.
(ii) bDNA assay.In the Versant HBV DNA 3.0 assay (Siemens Medical Solutions Diagnostics, Tarrytown, NJ), HBV DNA was recovered from 50 μl of plasma or serum and was quantified by the semiautomated system 340 bDNA analyzer (Siemens Medical Solutions Diagnostics), according to the manufacturer's instructions. HBV DNA levels were expressed in international units per milliliter. The dynamic range of quantification of this assay is 357 to 17,857,000 IU/ml (2.5 to 7.3 log10 IU/ml).
Sequence analysis of the preC-C region.Sequence analysis of the preC-C gene, the target of PCR amplification and probe hybridization in the CAP/CTM v2.0 assay, was performed on one HBV DNA-positive sample from each patient from group D. A heminested PCR was performed to amplify a 772-bp fragment with sense primers HBPr86 and HBPr87 and antisense primer HBPr303, as previously described (10). Briefly, HBV DNA was extracted from 200 μl of serum or plasma using the QIAamp MinElute virus vacuum kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's instructions. Amplification included an initial denaturation step at 95°C for 10 min, followed by 35 cycles of denaturation at 95°C for 30 s, annealing at 50°C for 30 s, and elongation at 72°C for 1 min, with a final elongation step at 72°C for 7 min. The amplification products (5 μl) were run on 1.5% agarose gels. The gels were stained with SYBR Safe (Invitrogen, Carlsbad, CA). They were then purified with Montage PCR centrifugal filter devices (Millipore Corporation, Bedford, MA), according to the manufacturer's protocol. PCR products were directly sequenced with the BigDye Terminator cycle sequencing kit, v3.1, on an ABI 3100 sequencer (Applied Biosystems, Foster City, CA), according to the manufacturer's protocol.
Statistical analysis.Descriptive statistics are shown as the mean ± standard deviation (SD) or the median and interquartile range as appropriate. Comparisons between groups were made using the Kruskal-Wallis test or the Mann-Whitney test. The relationship between quantitative variables was studied by means of regression analysis. P values of <0.05 were considered significant.
RESULTS
Intrinsic performance of the CAP/CTM v2.0 assay. (i) Analytical sensitivity.The manufacturer claims that the lower limit of detection of the CAP/CTM v2.0 assay is 20 IU/ml. Twenty-one tested replicates containing 50 IU/ml (1.7 log10 IU/ml), 25 IU/ml (1.4 log10 IU/ml), or 12.5 IU/ml (1.1 log10 IU/ml) tested positive for HBV DNA by the CAP-CTM v2.0 assay. However, for 12 out of the 21 HBV DNA-positive replicates containing 25 IU/ml and for the 21 HBV DNA-positive replicates containing 12.5 IU/ml, the result was expressed as “HBV DNA positive, below 20 IU/ml.” Nineteen of the 21 replicates containing 6.25 IU/ml (0.8 log10 IU/ml) were HBV DNA positive by the CAP/CTM v2.0 assay. The result was expressed as “below 20 IU/ml” for all of them.
(ii) Specificity.Of the 103 HBV-seronegative specimens from group A, all tested HBV DNA negative by the CAP-CTM v2.0 assay except one, which tested HBV DNA positive (specificity, 99.0%; 95% confidence interval [95% CI], 94.7% to 100%). Ninety-six of the 97 specimens with a profile of resolved infection from group B (positive anti-HBs and anti-HBc antibodies) were found to be HBV DNA negative, while 1 sample tested HBV DNA positive (specificity, 99.0%; 95% CI, 94.4% to 100%). For both samples with detectable HBV DNA, the result was expressed as “target detected,” below the lower limit of detection of 20 IU/ml. Both samples were retested by another real-time PCR assay (m2000RT; Abbott Molecular, Des Plaines, IL) and were found to be HBV DNA negative by that assay.
(iii) Precision and reproducibility.As shown in Table 1, the intra-assay coefficients of variation (precision) and interassay coefficients of variation (reproducibility) ranged from 0.21% to 2.67% and from 0.65% to 2.25%, respectively.
Intra-assay (precision) and interassay reproducibility of the CAP/CTM v2.0 assay
Accuracy, linear quantification, and influence of the HBV genotype. (i) Linear quantification of standard panel dilutions.Quantification of the OptiQuant HBV DNA panel in triplicate showed a significant relationship between the average HBV DNA levels measured by the CAP/CTM v2.0 assay and the expected HBV DNA levels (r = 0.9987; P < 0.0001). The differences between the average measured and expected HBV DNA levels ranged from 0.01 to 0.36 log10 IU/ml, with a modest underestimation by the CAP/CTM v2.0 assay when HBV DNA levels exceeded 5 log10 IU/ml (Fig. 1).
Quantification by the CAP/CTM v2.0 assay of HBV DNA levels in a commercial standard panel containing 2 × 102 (2.3 log10) to 2 × 107 (7.3 log10) IU of HBV DNA/ml (OptiQuant HBV DNA; AcroMetrix, Benicia, CA). The average measured values are shown as a function of the expected values (the actual HBV DNA contents of the panel members). The dashed line is the equality line.
(ii) Quantification of HBV DNA in clinical samples containing HBV genotypes A to F.HBV DNA levels in the 51 samples from patients from group C, infected with HBV genotypes A to F, were measured. Thirty-eight of these samples (75%) fell within the dynamic range of quantification of the CAP/CTM v2.0 assay; the remaining 13 samples (25%) had to be retested after dilution. As shown in Fig. 2, there was a significant relationship between the HBV DNA levels obtained by the CAP/CTM v2.0 and bDNA assays (Fig. 2A) and between those obtained by the CAP/CTM v1.0 and v2.0 assays (Fig. 2B), regardless of the HBV genotype.
Correlation between the HBV DNA levels determined by the CAP/CTM v2.0 and bDNA assays (A) or by the CAP/CTM v2.0 and v1.0 assays (B) in 51 clinical samples (group C) containing HBV genotypes A (n = 12), B (n = 9), C (n = 8), D (n = 9), E (n = 10), and F (n = 3).
Figure 3 A shows a Bland-Altman plot of HBV DNA levels measured by the second-generation CAP/CTM assay and the bDNA method. The figure plots the difference between the two measured values (the CAP/CTM v2.0 assay value minus the bDNA assay value) as a function of the mean of the two measurements. A moderate underestimation of HBV DNA levels by the CAP/CTM v2.0 assay compared to the bDNA method was observed in 34 (66.7%) of the 51 samples containing HBV genotypes A to F (median difference [CAP/CTM minus bDNA values], −0.16 log10 IU/ml). HBV DNA levels were underestimated by the CAP/CTM v2.0 assay compared to the bDNA assay in almost all samples with more than 5 log10 IU/ml (mean difference, −0.29 ± 0.18 log10 IU/ml). Below 5 log10 IU/ml, HBV DNA levels were often moderately overestimated by the CAP/CTM v2.0 assay compared to the bDNA method (mean difference, +0.11 ± 0.21 log10 IU/ml) (Fig. 3A). Six samples had a difference more than 1.96 times the mean difference, including three (two with genotype A and one with genotype B) that were underquantified and three (one with genotype D and two with genotype E) that were overquantified by the CAP/CTM v2.0 assay relative to the bDNA assay. However, the individual differences between the CAP/CTM v2.0 and bDNA values were always below 1.0 log10 IU/ml in these samples.
(A) Bland-Altman plot of HBV DNA levels measured by the CAP/CTM v2.0 and bDNA assays in the 51 group B samples. The difference between the HBV DNA levels obtained by the CAP/CTM v2.0 assay and those obtained by the bDNA assay is plotted as a function of the mean of the two values. Different genotypes are represented by different colors. The shaded area corresponds to the mean difference ± 1.96 standard deviation. (B) Distribution of the differences between the HBV DNA levels obtained by the CAP/CTM v2.0 assay and those obtained by the bDNA assay for the same samples, according to the HBV genotype (genotypes A to E). The difference was not significant.
Box plots of individual differences between the two methods are shown for each genotype in Fig. 3B. They confirm the global, moderate underestimation of HBV DNA levels by the CAP/CTM v2.0 assay compared to the bDNA method, independent of the HBV genotype. The median differences were −0.17 log10 IU/ml for genotype A, −0.37 log10 IU/ml for genotype B, −0.32 log10 IU/ml for genotype C, +0.05 log10 IU/ml for genotype D, and −0.03 log10 IU/ml for genotype E (not significant). Results for HBV genotype F are not shown in Fig. 3B, because only three samples were tested.
(iii) HBV DNA monitoring of patients receiving antiviral therapy.Sixteen patients chronically infected with HBV (4 with genotype A, 5 with genotype C, 1 with genotype D, and 6 with genotype E), included in group D, were serially sampled while on treatment with nucleoside/nucleotide analogues. For these patients, 51 plasma and 51 serum specimens sampled at the same times were available. The 51 plasma samples were tested in parallel by the CAP/CTM v1.0, CAP/CTM v2.0, and bDNA assays, while the 51 serum samples were tested by the CAP/CTM v2.0 and bDNA assays. As shown in Table 2, the CAP/CTM v2.0 assay quantified HBV DNA equally in plasma and serum (mean difference range, 0.04 to 0.11 log10 IU/ml, according to the HBV genotype). As shown in Table 3, the two versions of the CAP/CTM assay quantified HBV DNA levels equally in plasma (mean difference range, 0.04 to 0.20 log10 IU/ml, according to the HBV genotype).
Mean differences between HBV DNA levels measured with the CAP/CTM v2.0 assay in 51 serum and 51 plasma specimens from the 16 patients with chronic HBV infection from group Da
Mean differences between HBV DNA levels measured with the CAP/CTM v1.0 and v2.0 assays in 51 plasma specimens from the 16 patients with chronic HBV infection from group Da
Figure 4 shows individual examples of the kinetics of HBV DNA levels measured by the two versions of the CAP/CTM assay and the bDNA assay for patients infected with HBV genotypes A, C, D, and E. The differences between the CAP/CTM v2.0 and bDNA assays were always less than 0.5 log10 IU/ml, except for two patients. Indeed, for one patient infected with genotype C (patient [Pt] 3) and for one patient infected with genotype E (Pt 7), the CAP/CTM v2.0 assay underestimated HBV DNA levels by −1.38 and −1.41 log10 IU/ml on average, respectively, relative to the levels found by the third-generation bDNA assay (Fig. 4).
Kinetics of HBV DNA levels measured in individual patients' plasma specimens with the CAP/CTM v1.0 assay (solid lines and open circles), the CAP/CTM v2.0 assay (solid lines and filled circles), and the bDNA assay (dashed lines). The top and bottom shaded areas correspond to the lower limits of detection of the bDNA (2.55 log10 IU/ml) and CAP/CTM v2.0 (1.30 log10 IU/ml) assays, respectively. HBV DNA levels (log10 IU/ml) are given on the y axis.
(iv) Target sequence analysis in clinical specimens from patients included in the longitudinal analysis.The nearly full-length preC-C region of the HBV genome, which is the target of CAP/CTM primers and probes, was sequenced for the 16 patients from group D in order to determine the possible role of nucleotide sequence polymorphisms in the underestimation of HBV DNA levels for patients 3 and 7. Sequences from these two patients were compared with those from the patients whose HBV DNA levels were quantified equally by the different assays. In spite of the presence of polymorphisms in the pre-C region, no nucleotide sequence signature was found to explain the underestimation of HBV DNA levels for the two patients (data not shown).
DISCUSSION
Real-time PCR assays have become the reference for the quantification of viral genomes in clinical practice. The most recent international guidelines on the management of HBV therapy recommend the use of real-time PCR assays to detect and quantify HBV DNA in order to diagnose HBV infection, establish the indication for therapy, and monitor antiviral treatment responses and resistance (4, 8). The first-generation CAP/CTM assay showed excellent analytical performance (2) but could be used only on plasma specimens. A second-generation assay was recently developed. In this second-generation assay, HBV DNA levels can be measured in both plasma and serum from only 650 μl of material, and the assay has a claimed lower limit of detection of 20 IU/ml.
In the present study, we showed that the CAP/CTM v2.0 assay has excellent analytical sensitivity, with a lower limit of detection close to that claimed by the manufacturer, i.e., 20 IU/ml. This assay was also precise and reproducible, as previously reported (6). In addition, we observed a strong, significant relationship between HBV DNA levels obtained by the CAP/CTM v2.0 assay and those obtained by the third-generation bDNA-based assay. The use of the third-generation bDNA assay as a comparator was justified by the facts that this assay is accurate, precise, reproducible, and well calibrated to the World Health Organization HBV DNA standard and that it quantifies HBV DNA levels independently of the HBV genotype, due to the presence of a large number of capture and extender probes located at various positions along the HBV genome (7, 11). There was also an excellent correspondence between versions 1.0 and 2.0 of the CAP/CTM assay in plasma in our experiments, regardless of the HBV genotype.
As with the first-generation CAP/CTM assay, we observed a modest underestimation of HBV DNA levels in the 6 members of the standard panel, mainly in those with HBV DNA levels above 5.0 log10 IU/ml. This finding was confirmed in clinical samples when values obtained by the CAP/CTM v2.0 assay were compared to those obtained with the bDNA assay. This modest underestimation was independent of the HBV genotype and has no implications for clinical practice. Indeed, in contrast with HCV therapy, where accurate quantification is needed in order to make treatment decisions (3), this technical issue has little clinical importance in HBV therapy with nucleoside/nucleotide analogues, since the goal of this treatment is to maintain HBV DNA levels below the lower limit of quantification of the assay in the long term.
In two cases, however, we observed a substantial underestimation of HBV DNA levels by the CAP/CTM assay relative to the bDNA assay. These samples contained HBV genotypes C and E, respectively. The underestimations were consistent in several serial samples from the two patients, suggesting that they were due to the nature of the infecting HBV strains in these patients. This may be seen when mismatches occur between the primers and/or the TaqMan probe and the target viral sequence as a result of natural nucleotide polymorphisms (1). We therefore sequenced the target region (preC-C) in order to determine whether nucleotide polymorphisms could distinguish the patients whose HBV DNA levels were quantified equally by the CAP/CTM and bDNA assays from these two individuals. No sequence signature has been identified that could explain the underestimation of HBV DNA levels for these two patients.
The main improvement in the second generation of the CAP/CTM assay is the possibility of quantifying HBV DNA in both serum and plasma, whereas the first-generation assay was for plasma only. We thus assessed whether the values obtained in plasma with both versions of the assay and in serum and plasma with the CAP/CTM v2.0 assay were concordant. As shown here, there was an excellent correspondence between the values obtained in different matrices. In addition, serial follow-up of patients from group D showed almost perfect superimposition in all cases, confirming that the CAP/CTM v2.0 assay can be used equally on serum or plasma samples in clinical practice.
In conclusion, this study shows that the new version of the CAP/CTM assay is sensitive, specific, and reproducible and that it accurately quantifies HBV DNA in both plasma and serum samples from patients with chronic HBV infection. Quantification is linear over the full dynamic range of quantification, which covers values observed in both treated and untreated patients with chronic hepatitis B. However, the upper limit of quantification (8.2 log10 IU/ml) is still too low to cover the full range of possible values. Any sample falling above the upper limit of quantification of the assay needs to be retested after dilution, a step that does not affect quantification. In our hands, the new version of the CAP/CTM assay appeared to be suitable for large-scale routine analysis of samples containing HBV genotypes A to F. Broad use of fully automated real-time PCR assays, as recommended in recent international guidelines for the management of HBV therapy (4, 8), will improve the management of patients with chronic HBV infection, as well as the monitoring of antiviral responses and drug resistance.
ACKNOWLEDGMENTS
The CAP/CTM kits used in this study were kindly provided by Roche Diagnostics (Meylan, France). We are grateful to Françoise Darthuy, Alexandre Soulier, and Françoise Bouchardeau for technical assistance.
FOOTNOTES
- Received 28 June 2010.
- Returned for modification 28 July 2010.
- Accepted 5 August 2010.
- Copyright © 2010 American Society for Microbiology
