|Home | About | Journals | Submit | Contact Us | Français|
Improved sensitivity of HBV-DNA tests is of critical importance for the management of HBV infection. Our aim was to develop and assess a new ultra sensitive in-house real-time PCR assay for HBV-DNA quantification (ultra sensitive RTQ-PCR).
Previously used HBV-DNA standards were calibrated against the WHO 1st International Standard for HBV-DNA (OptiQuant® HBV-DNA Quantification Panel, Accrometrix Europe B.V.). The 95% and 50% HBV-DNA detection end-point of the assay were 22.2 and 8.4 IU/mL. According to the calibration results, 1 IU/mL equals 2.8 copies/mL. Importantly the clinical performance of the ultra sensitive real-time PCR was tested similar (67%) to the Procleix Ultrio discriminatory HBV test (dHBV) (70%) in low-titer samples from patients with occult Hepatitis B. Finally, in the comparison of ultra sensitive RTQ-PCR with the commercially available COBAS TaqMan HBV Test, the in-house assay identified 94.7% of the 94 specimens as positive versus 90.4% identified by TaqMan, while the quantitative results that were positive by both assay were strongly correlated (r = 0.979).
We report a new ultra sensitive real time PCR molecular beacon based assay with remarkable analytical and clinical sensitivity, calibrated against the WHO 1st International standard.
Chronic hepatitis B virus (HBV) infection can be assessed by evaluating clinical features and biochemical, virologic, and histologic parameters. Knowledge of HBV-DNA viral load is useful for predicting disease prognosis, determining infectivity, evaluating indications for treatment, assessing response to treatment identifying emergence of resistance, and diagnosing occult HBV infection [1-5].
There are several commercially available HBV-DNA tests. The oldest versions suffered from poor sensitivity (approximately 5·105 copies/mL) and a narrow dynamic range of HBV-DNA quantification (~3-4 log10), while the current limit of detection of the newer assays is lower than 50 IU/mL with a dynamic range of approximately 8 log10 [6-13]. Additionally several in-house quantitative assays for HBV-DNA have been developed, based mostly on real-time PCR methodology, showing a remarkable sensitivity and a wide linear range for quantification [6,7,11-14].
Importantly, improved sensitivity of HBV-DNA tests is of critical importance for the management of HBV infection as well as for the diagnosis of occult HBV infection [HBsAg(-), HBV-DNA(+)] .
The objective of this study was to develop a new ultra sensitive real-time PCR assay based on the knowhow of the existing real-time PCR assay for the quantification of the HBV-DNA  and, also, to assess its performance characteristics.
The HBV-DNA was extracted from 0.5 mL of serum/plasma using the QIAamp UltraSens Virus Kit (QIAGEN Inc., Valencia CA) and the DNA was eluted in 60 μl of elution buffer. For each ultra sensitive RTQ-PCR HBV-DNA quantification measurement, a positive HBV-DNA sample with known HBV-DNA titer (~104 IU/mL) and one negative control (negative human plasma) (Procleix® Negative Calibrator, Chiron/GenProbe, CA) were extracted additional to the unknown samples.
The assay was carried out into a LightCycler® 2.0 apparatus. The reaction mixture contained 1× reaction buffer (LightCycler DNA Master HybProbe; Roche Applied Science. Germany), 5 mM MgCl2, 0.5 μM primer hbv305 (sense) 5'-GCCAAAATTCGCAGTCCC-3', 0.5 μM primer hbv460 (antisense) 5'-GATAGTCCAGAAGAACCAACAAGAAG-3', 0.4 μM molecular beacon 5'-CGCGCGATGAGGCATAGCAGCAGGATGAAGAACGCGCG-3' labelled with FAM and Dabcyl at the 5' and 3' ends, respectively, Taq DNA polymerase, dNTP mix (with dUTP instead of dTTP), and 1 U of uracil-DNA glycosylase (Roche Applied Science. Germany). The amplification profile was as follows: One cycle of initial denaturation (95°C for 10 min) followed by 40 cycles of amplification at 95°C for 0 sec, 55°C for 10 sec, and at 72°C for 8 sec. The final reaction volume was 60 μl containing 30 μl of extracted viral DNA.
The sensitivity together with the 95% and 50% HBV-DNA detection limits of the assay were estimated by using the WHO 1st International Standard for HBV-DNA (OptiQuant® HBV-DNA Quantification Panel, Accrometrix Europe B.V.), tested in 8 replicas. We performed serial dilutions with human negative plasma (Procleix® Negative Calibrator, Chiron/GenProbe, CA) of the WHO standard at final concentrations of 100, 50, 25, 10 and 5 IU/mL. The 95% and 50% HBV-DNA detection limits were calculated by probit analysis (8 replicas).
The clinical performance of the assay was tested on 27 samples drawn from 22 individuals with occult Hepatitis B (HBs Ag negative, consistently reactive by multiplex Procleix Ultrio HIV-1/HCV/HBV and Procleix Ultrio discriminatory HIV and HCV negative) screened with the discriminatory HBV assay (dHBV) (Chiron/GenProbe, Emeryville/San Diego, CA).
The calibration of the HBV-DNA values was performed against the WHO 1st International Standard for HBV-DNA (OptiQuant® HBV-DNA Quantification Panel, Accrometrix Europe B.V.). Specifically, serial dilutions of the WHO standard ranging from 2·102 to 2·106 IU/mL) in 5 individual sample preparations were tested by the ultra sensitive RTQ-PCR method. Based on a linear regression, a conversion formula was calculated for the in-house measurements (copies/mL) to the international standard units (IU/mL).
To compare the ultra sensitive RTQ-PCR assay with the previously reported test, HBV-DNA derived from 25 randomly selected samples was quantified using both methods. The correlation of the HBV-DNA quantifications by using the current and the previous in-house assay was also estimated.
To compare the ultra sensitive RTQ-PCR assay with the commercially available COBAS TaqMan HBV Test, plasma samples collected from 94 different HBV-infected patients were tested. The correlation of the HBV-DNA quantifications by using the in-house and the commercial test was also estimated.
Pearson's correlation coefficient was used to assess the strength of the linear association between the log10-transformed values of the estimated (copies/mL) versus the expected values (IU/mL) and also between HBV-DNA quantifications using the two in house real-time PCR assays and COBAS TaqMan as well. To compare the quantitative results obtained for the samples found positive by the two different methods (ultra sensitive RTQ-PCR vs COBAS TaqMan HBV Test) the fitted regression line was compared to the line of equality by testing the two-tailed hypothesis of slope = 1 and the intercept = 0.
The ultra sensitive RTQ-PCR was performed in a LightCycler® 2.0 apparatus using a plasmid HBV-DNA standard as described previously. We should note that the specificity of the assay was 100% as shown previously . The differences between the newly developed method (ultra sensitive RTQ-PCR) and the previous one  were: 1) the utilization of a molecular beacon instead of 2 hybridization probes, 2) extraction of HBV-DNA from a larger volume of serum/plasma (0.5 versus 0.2 mL), 3) use of 30 μl instead of 10 μl extracted DNA, 4) modified RTQ-PCR conditions, and 5) all experiments were performed in a LightCycler® 2.0 apparatus because the latter shows improved performance characteristics compared to LightCycler® 1.0. Moreover reaction volume in LightCycler® 2.0 can be as large as 100 μL compared to 20 μL in the LighCycler® 1.0.
To determine the analytical sensitivity of the ultra sensitive RTQ-PCR, 5 dilutions of the WHO 1st International Standard for HBV-DNA were tested (OptiQuant® HBV-DNA Quantification Panel, Accrometrix Europe B.V.). DNA concentrations above 25 IU/mL were detected on all 8 occasions; concentrations of 10 IU/mL were tested on 4 of 8 occasions (50%), while 5 IU/mL were detected on 2 out of 8 occasions (25%). The 95% and 50% detection end-point of the assay were 22.2 and 8.4 IU/mL respectively.
The performance of the assay was tested on 27 samples collected from 22 individuals with occult Hepatitis B (HBsAg negative, HBV-DNA consistently reactive by Ultrio, Ultrio HIV and HCV negative). This analysis was part of another study concerning the molecular typing of occult Hepatitis B in blood donors across Greece . The characteristics of these patients are shown in Table Table1.1. These samples were selected because of the low HBV-DNA titer. Among the 27 samples tested with the dHBV test, 19 (70%) were found to be above the threshold of detection, while a similar sensitivity was observed for the ultra sensitive RTQ-PCR method (18 samples tested positive out of 27, 67%). The mean HBV-DNA was 59.1 IU/mL (range, 10.2 to 346.7 IU/mL). We should note, however, that the dHBV assay was applied multiple times to 18 samples (a sample was considered positive, if it was tested above the threshold of detection at least once), while ultra sensitive RTQ-PCR was performed only once. As shown in table table2,2, 14 samples were detectable by both methods, while 4 and 5 samples were tested positive only by ultra sensitive RTQ-PCR and the dHBV test, respectively. These qualitative findings suggest that the in-house assay performed equally well as the commercial test, although the samples were tested multiple times by the latter method.
The calibration of the in-house assay was done against the WHO 1st International Standard for HBV-DNA (OptiQuant® HBV-DNA Quantification Panel, Accrometrix Europe B.V.) DNA was extracted and HBV-DNA was tested in 5 replicates by using the plasmid DNA as a standard. The correlation plot of the expected (log10 HBV-DNA IU/mL) versus the estimated values by ultra sensitive RTQ-PCR log10 HBV-DNA copies/mL) is shown in Fig. Fig.1.1. The fitted regression lines between IU/mL and copies/mL was given by the following equation: log10 (IU/mL) = -0.4475 + 1.0096·log10(ultra sensitive RTQ-PCR copies/mL), suggesting that in our setting, 1 IU/mL = 2.8 copies/mL. The 95% CI for the estimated intercept was -0.69, -0.21. The 95% CI for the estimated slope was: 0.96, 1.06. The correlation coefficient between the expected and the estimated values was very good r = 0.9937 (p < .001) (Fig. (Fig.11).
To assess the correlation between the previous and the new assay, we re-tested 25 previously quantified samples. The average log10 HBV-DNA was 5.38 copies/mL (range, 2.31 to 10.51 log10 copies/mL). Importantly, the linearity was very good between the HBV-DNA values quantified by the two assays (r = 0.985, p < .001) over the whole range of quantification. The fitted regression line between the ultra sensitive RTQ-PCR and the previous test was: log10 HBV-DNA (copies/mL) with the ultra sensitive RTQ-PCR = -0.531 + 1.015x log10 (copies/mL) with the previous real-time PCR test. The 95% CI for the estimated intercept was -0.92, -0.14. The 95% CI for the estimated slope was: 0.95, 1.08 (data not shown).
Among the 94 samples tested with the TaqMan test, 85 (90.4%) were found to be above the threshold of detection, while a higher sensitivity was observed for the ultra sensitive RTQ-PCR method (89 samples tested positive out of 94, 94.7%). As shown in Table Table3,3, 83 (88.3%) samples were detectable by both methods, while 6 (6.4%) and 2 (2.1%) samples were tested positive only by the ultra sensitive RTQ-PCR and the TaqMan test, respectively. These qualitative findings suggest that the new version of the in-house assay performed equally well as the commercially available TaqMan Test (McNemar's p = 0.157). The in-house assay found positive six samples (ranging from 1.37 to 1.83 log10 IU/mL) with undetectable viral load levels with COBAS TaqMan HBV Test, while it did not quantify two samples (with viral load levels 1.08 and 1.77 log10 IU/mL, respectively).
Fig. Fig.22 shows the scatter plot of log10 HBV-DNA IU/mL determined by the ultra sensitive RTQ-PCR and the TaqMan assays, using specimens with detectable HBV-DNA by both assays (N = 83). The results of the two assays were linearly correlated (r = 0.979, p < 0.001). The fitted regression line differs significantly from the line of equality and was described by the equation: log10 (TaqMan, IU/mL) = -0.457 + 0.987 log10 (ultra sensitive RTQ-PCR, IU/mL) (joint test of intercept = 0 and slope = 1: p < 0.001).
Sensitive detection and quantification of HBV-DNA is essential for monitoring response to therapy and, also, for assessment of occult HBV infection. Therefore, ultra-sensitive assays for HBV-DNA quantification are needed in clinical practice.
In the current study, we describe the characteristics of a new ultra sensitive in-house RTQ-PCR performed in a LightCycler® 2.0 instrument (Roche, Molecular Biochemicals, Mannheim, Germany). The new assay showed improved sensitivity of 22 and 8 IU/mL as 95% and 50% detection end-points, respectively, versus the 94 IU/mL (250 copies/mL) of the previously reported quantitative test . The sensitivity of the assay was improved mainly because of a larger volume of HBV-DNA used in ultra sensitive RTQ-PCR, extracted also from a larger volume of serum/plasma. Additionally, the new assay was carried out in the LightCycler® 2.0 that shows improved performance than the LightCycler® 1.0. We should note that the sensitivity of the assay can be potentially improved by using a larger than 0.5 mL volume of plasma, after a concentration step performed before the standard procedure. Moreover, the plasmid standard calibrated against the WHO 1st International standard, thus suggesting that the HBV-DNA values are reported in the international format (IU/mL).
Importantly, the ultra sensitive RTQ-PCR performed equally well as the qualitative dHBV assay tested on low-titer HBV-DNA samples collected from individuals with occult Hepatitis B. These findings are in accordance with the previously reported 95% detection limit of the dHBV, 19 IU/mL , which is almost identical to the 95% detection limit (22 IU/mL) of the ultra sensitive RTQ-PCR assay. These findings suggest that the ultra sensitive RTQ-PCR assay is equally reliable as the dHBV for detecting HBV-DNA in low-titer samples as those collected from patients with occult Hepatitis B. We should note however that the ultra sensitive RTQ-PCR is a quantitative test and therefore it shouldn't be used for diagnostic purposes.
The HBV-DNA quantified values between the ultra sensitive RTQ-PCR and the previously reported HBV-DNA test correlated very well, suggesting that the new version of the test can be utilized in occasions in which the original test was used. Furthermore, the ultra sensitive RTQ-PCR was found to be highly correlated with commercial available COBAS TaqMan HBV Test (r = 0.979, p < 0.001).
In the era of highly potent antivirals (e.g. tenofovir, entecavir) approved for the treatment of chronic Hepatitis B, it is of crucial importance to utilize highly sensitive assays for detecting and quantifying HBV-DNA. We report the new ultra sensitive assay, using molecular beacon as a detection system, with remarkable analytical and clinical sensitivity, calibrated against the WHO 1st International standard.
The authors declare that they have no competing interests.
DP participated to the study design and coordination, he prepared the manuscript and the reply to the reviewer's comments; AB and CH carried out the experiments; AK, ZM, HH, AV were responsible for the part of the study concerning the occult Hepatitis B infection; VS did the statistical analysis; AH was the study coordinator and participated to the writing and editing of the manuscript. All authors read and approved the final manuscript.
This study was supported by Gilead. DP, AK, HH and VS were supported by the Hellenic Center for Infectious Diseases Control (HCIDC) of the Ministry of Health & Welfare.