Abbott RealTime HBV Assay Is More Sensitive in Detection of Low Viral Load and Little Impacted by Drug Resistant Mutation in Chronic Hepatitis B Patients under Nucleot(s)ide Analogues Therapy

Background Selection of drug-resistant strains may lead to failure of HBV antiviral therapy. There is little information whether there is detection difference in drug resistant mutations between different viral load assays of HBV. Objectives This study is aimed to investigate whether there is drug-resistant strains related detection difference between Abbott RealTime HBV (RealTime) and CobasAmpliPrep/CobasTaqMan HBV assays 2.0 (TaqMan). Study Design One hundred and thirty-four CHB patients who received HBV anti-viral therapy were enrolled. HBV virological markers were tested 3 months apart regularly. Serum HBV DNA levels were determined using the TaqMan and RealTime. YMDD (rt180M and rt204V) mutation was checked in patients who experienced virologic breakthrough (VBT). Results The correlation of HBV DNA observed between the RealTime and TaqMan was good for all 571 samples (R2 = 0.797; P<0.001). However, the correlation in the 434 samples with HBV DNA level <3 log10 IU/ml was not as good as in all samples (R2 = 0.457). Overall, 21.5% of samples had a detection difference of ≥1 log10 IU/ml with 91.9% of these having HBV DNA level <3 log10 IU/ml. Twenty-four patients experiencedVBT. Three of these patients had acquired the YMDD mutation and exhibited discordant viral load results between the two methods tested. In each case, persistent HBV DNA was detected by RealTime and undetectable with TaqMan. Of the patients who experienced a VBT and had acquired YMDD mutation, 4.7% had undetectable HBV DNA by TaqMan while all were detectable with RealTime. Conclusions RealTime assay is more sensitive and is little impacted by the development of drug resistant mutation.


Introduction
Hepatitis B virus (HBV) infection is a global health problem, with an estimated 350-400 million people being chronically infected [1,2]. HBV is a small partially double stranded DNA virus that replicates through an RNA intermediate. Because the HBVencoded polymerase is prone to error, the virus evolves quickly under the selection pressure like adaptive immune response and antiviral therapy. The estimated nucleotide substitution rate of HBV is between 1.5610 25 and 1.9610 24 substitutions per site per year [3,4,5]. A recent study demonstrated a higher substitution rate of 8.15610 24 substitutions/site/year in the HBV core gene compared to 4.57610 24 substitutions/site/year in the surface gene [6]. The number and position of these mutations may impact the viability of the virus. In a single study of patients with chronic HBV with active viral infection, 20 nucleotide changeshad accumulated between positions 1933-2048 of the core/pre-core region [7]. This indicates that considerable changes may occur in this region without stopping viral replication.
Assays for viral load quantificationwhich target core/precoremay be impacted by nucleotide changes if they resided in sites selected for primers and probes [8]. Differences of greater than 1 log 10 IU/ml for individual samples have been noted between commercial HBV viral load assays noted even when general correlation between the assays were considered good [9]. In our recent published study [10] evaluatingthe correlations between two commercially available HBV viral load assays (Abbott RealTime HBV assay, hereafter, ''RealTime'' and CobasAmpliPrep/CobasTaqMan HBV assay, hereafter, ''Taq-Man''), very good correlation was observed in untreated HBVinfected patients. However, 8.6% of patients in this study had discordant results of $1 log 10 IU/mlbetween the two assays, especially in patients with viral load of less than 3 log 10 IU/ml. Recently, a new developed ultra sensitive in-house real-time PCR assay for HBV DNA quantification was reported by Paraskevis et al. [11]. They found that the clinical performance of the ultra sensitive assay was tested similar to the ProcleixUltrio discriminatory HBV test in low-titer samples and the quantitative results by the ultra sensitive assay were also strongly correlated with COBAS TaqMan HBV assay.
In patients with HBV antiviral therapy, selection of drugresistant strains may lead to failure of HBV antiviral therapy. It occurred in approximately 16% to 43% of the patients treated with lamivudine for 1 to 2 years [12,13]. The 5-year cumulative incidence of resistance was only 0-1.2% in patients treated with entecavir and tenofovir, the current recommended first-line therapies with high potency and high genetic barrier [14]. Drug resistance is suspected with a rise in HBV viral load above the detection limit of the assay or $1 log 10 IU/ml from nadir. Currently, there is little information about HBV viral load assay concordance in samples harboring drug-resistant mutations.
This present study aimedto investigate the impact of drugresistant strainson detection difference between RealTime and TaqMan.

Patients and Methods
One hundred and thirty-four CHB patients, who had received treatment with an approved nucleot(s)ide analogue (NUC) for at least 3 months from August 2008 to October 2010 at Kaohsiung Medical University Hospital in Taiwan, were enrolled into this prospective study. All patients were seropositive for hepatitis B surface antigen (HBsAg) for more than 6 months. Anti-viral therapy for HBV and agents used for therapy were decided according to the criteria of the Bureau of National Health Insurance of Taiwan and physician's decision. Patients who were co-infected with hepatitis C virus (HCV), human immunedeficiency virus (HIV) or combined with other liver disease were excluded. HBV virological markers were tested at enrollment, and 3 months apart regularly during study period. Hepatitis B e antigen (HBeAg) was detected using commercially available enzyme-linked immunosorbent assay kits (Abbott Laboratories, North Chicago, IL, USA). Serum HBV DNA levels were determined using the TaqMan (CAP/CTM version 2.0, Roche Diagnostics, Indianapolis, IN, USA; dynamic range 20 IU/ml-1.7610 8 IU/ml) and RealTime (Abbott Laboratories, Abbott Park, IL, USA; dynamic range 10IU/ml-1610 9 IU/mL). The samples were simultaneously tested according to the manufacturer's instructions. Samples measured below lower limit of quantification with target detected were set as undetectable in current study.
YMDD (rt180M and rt204V) mutation would be checked in patients who experiencedvirologic breakthrough (VBT), defined as greater than 10 fold increase in serum HBV DNA from nadir or redetection of HBV DNA in serum after its initial disappearance in 2 consecutive measurements, at the time points of enrollment and VBT. YMDD mutation was determined by the Abbott HBV RUO Sequencing assay (Abbott Molecular Inc., Des Plaines, IL). The processing followed by real-time PCR amplification and detection was performed on the integrated Abbott m2000sp/ m2000rt instrument system (Abbott Molecular Inc.), using the m2000 HBV Sequencing Version 2.00 software application and HBV RUO Sequencing Application CD-ROM version 2.0 (Abbott Molecular Inc.). Specimen with a minimal volume of 0.8 ml was required for automated processing by the m 2000sp. Detection and quantification of the 1023-bp HBV amplification product was then performed with the m 2000rt. Four individual sequencing reactions (2 forward and 2 reverse) were performed for each sample to provide bidirectional sequences spanning the rt/ pol and S gene regions of interest. All cycle sequencing procedures were performed on a GeneAmp PCR System 9700 (Applied Biosystems, Foster City, CA, USA). The cycle sequencing products were generated with the Big Dye Terminator 3.1 Cycle Sequencing kit (Applied Biosystems) and then analyzed by capillary electrophoresis on an Applied Biosystems 3130xl Genetic Analyzer (AB 3130xl, Applied Biosystems) according to the assay manufacturer's instructions and analyzed with SeqScape Software, version 2.7 (Applied Biosystems).
The study was conducted according to the guidelines of the Declaration of Helsinki andthe principles of Good Clinical Practiceand was approved by the ethic committee of Kaohsiung Medical University Hospital (KMUH-IRB-970125). Written informed consent was obtained from all patients.

Statistical analysis
Continuous variables were expressed as median and25 th , 75 th percentile. Categorical variables were expressed as numbers and percentage. Correlation analysis was performed using Pearson correlation coefficients and linear regression. Bland-Altman plots were used to represent the degree of agreement. The mean log 10 paired difference and 95% limits of agreement are indicated in the figures by solid and dotted lines. All tests were two-sided, and p, 0.05 was considered significant. All analyses were performed using the SPSS ver17.0 statistical package (SPSS, Inc., Chicago, IL, USA).

Results
Demographics of the 134 CHB patients with antiviral therapy were shown on Table 1. Most of the patients were male with a median age of 45 years old. The median ALT level at enrollment was 34 U/L. The median time from treatment initiation to enrollment was 9 months and 67.2%, 13.4% and 19.4% of patients received lamivudine, telbivudine and entecavir treatment. HBeAg was positive in 45.1% of patients. The measured median HBV DNA level was 1.7 log 10 IU/ml by TaqMan and 1.9 log 10 IU/ml by RealTime. Overall, 571 serum samples were obtained during study period and the correlation of HBV DNA observed between the RealTime and TaqMan was good (R 2 = 0.797; P, 0.001)( Figure 1A). The mean difference between the two assays (RealTime-TaqMan) was 0.2860.82 log 10 IU/ml (limits of agreement, 21.36 to 1.91 log 10 IU/ml)( Figure 1B).
Among 571 serum samples, 434 (76.0%) samples had HBV DNA level ,3 log 10 IU/ml measured by both assays. The correlation between the two assays in the 434 samples was not as good as in all samples (R 2 = 0.457, P,0.001) (Figure 2A).The mean difference was 0.4060.77 log 10 IU/ml (limits of agreement, 21.15 to 1.95 log 10 IU/ml)( Figure 2B). Furthermore, 123 (21.5%) of the 571 samples had a detection difference of $1 log 10 IU/ml and most (91.9%) of them had HBV DNA level ,3 log 10 IU/ml measured by both of the assays. Figure 3 showed the mean HBV DNA levels of all (A) and VBT (B) patients according to the time points of NUC therapy. Generally, HBV DNA levels detected by RealTime were slightly higher than TaqMan although these differences were not statistically different. The HBV DNA level was found below lower limit of quantification in 250 and 150 samples, respectively, by TaqMan and RealTime assay. Of the 250 samples below lower limit of quantification by TaqMan, 174 (69.6%) samples showed ''target detected'' and the remained 76 (30.4%) samples showed ''target undetected''. Seventy-one (47.3%) of the 150 samples below lower limit of quantification by RealTime showed ''target detected'' and the remained 52.7% of samples showed ''target undetected''. Forty-one samples showed ''target detected'' by TaqMan, but ''target undetected'' by RealTime. By contrast, 24 samples showed ''target detected'' by RealTime, but ''target undetected'' by TaqMan. Notably, there were 105 samples having quantitative value by RealTime, but unquantitated by TaqMan. However, only 5 samples showed contrasty results.
During study period, 24 (17.9%) of the patients experienced VBT. We checked YMDD mutation in all 24 patients at the time point of enrollment and VBT. YMDD mutation was not detected in any of the patients at enrollment but was found in 22 (91.7%) of the patients following VBT. We further divided the patients into 4 groups according to the status of HBV DNA detection (persistent detectable, undetectable during study period) by the two assays. For the group of undetectable HBV DNA by both assays (n = 39), no VBT was observed. Persistent detectable DNA by both assays was observed in 69 (51.5%) of all patients in this study and 21 (30.4%) of thesepatients had VBT during study period. Most patients who experienced VBT (90.5%)had accumulated the

Discussion
In the present study, we demonstrated that there was a detection difference between RealTime and TaqMan in viral load less than 3 log 10 IU/ml and the detection difference may influence the identification of drug resistant mutation. Thus, RealTime HBV assay may have advantages when monitoring patients receiving NUC therapy for HBV infection.  HBV viral load is currently a standard marker to assess the efficacy of HBV antiviral therapy and also a marker for the progression of liver disease [15,16]. Sustained suppression of HBV DNA is one of the goals of HBV therapy in regional guidelines [17,18,19]. To achieve this goal, long term NUC therapy should be administered. The development of drug-resistance is a major concerned when placing a patient on long term NUC therapy. To facilitate the early detection of NUC drug-resistance, close monitoring of patient serum HBV DNA level is recommended.
Because low level HBV viremia is associated with the early stages of developing drug resistance, using a highly sensitive real time PCR assay to detect HBV DNA level is especially important. Generally, nucleotide sequencing of the PCR product was used to detect the development of resistant mutations. However, it could detect only when the minorities of viral quasispecies were greater than 25%. A recent publication by F. Ntziora et al. reported that using a sensitive amplification refractory mutation system real-time PCR combined with molecular beacon technology could effectively detect and quantify the YMDD mutation early before emerging to dominate the population [20].
RealTime assay and TaqMan assay are two highly sensitive HBV DNA quantitative assays and are widely used commercially. In a recent short communication by Morris CJ et al. [21], they compared the two assays in detection of HBV DNA in 215 CHB patients and found that the RealTime had a wider dynamic yielding absolute quantitative values. We also observeda good correlation between the two assays among serum samples from untreated chronic hepatitis B patients, irrespective of viral status [10]. However, a discordance of $1 log 10 IU/ml was also observedin 8.6% of samples and was found to be associated with HBV DNA level ,3 log 10 IU/ml. In the present study we found that 21.5% of the samples had a detection difference of $1 log 10 IU/ml between the two assays and most (91.9%) of them were HBV DNA level ,3 log 10 IU/ml measured by both of the assays. The result is compatible to our previous study and the difference especially found in samples with HBV DNA , 3 log may mainly related to the different lower limit of quantification by the two assays.
Introduction of a rescue therapy as early as possible in CHB patients with VBT has been demonstrated to provide benefits in long-term clinical outcomes [22]. Previous study by Thibault V. et al. [23] demonstrated that RealTime meets the requirements for the use in monitoring response to NUC therapy and its high sensitivity is adequate for the detection of VBT at an early stage. The detection difference in low viral load may influence the detection of VBT and drug resistant mutation. In the present study, HBV DNA levels waspersistently detected in 25 (18.7%) of the 134 patients by RealTime but were always undetected by TaqMan. The finding demonstrated that RealTime might be more sensitive in detection of very low viral load than TaqMan. The interesting finding was not reported before.
In addition, NUCs therapy may be stopped afterat least 12 months consolidation therapy when undetectable HBV DNA is achieved in CHB patients [24]. A sensitive PCR quantitative assay is important to identify the real status of DNA suppression. In the present study, none of the patients with undetectable HBV DNA by RealTime experienced a VBT compared to 4.7% of patients with undetectable HBV DNA by TaqMan who experienced VBT with the accumulation of YMDD mutation. The finding demonstrates that a proportion of patients who are considered to be achieving DNA suppression are not actually suppressed if a more sensitive HBV viral load assay is used and the risk of acquiring drug resistant mutation should be considered.
One limitation of our study is that we check YMDD mutation only in patients with VBT, not in all the patients with persistent detectable HBV DNA measured by either RealTime or TaqMan. Persistent detectable HBV DNA during NUC therapy is indicative of poor response to therapy with a high probability of developing drug-resistance. Drug-resistant mutations should also be checked in patients with persistent detectable HBV DNA. Another limitation is that the data of viral genotype is lacking. However, we believe that the majority of our patients are genotype B or C CHB. Thus, the results of the study may be drawn only on the 2 genotypes. The third limitation is that we tested the 24 patients experienced VBT for resistance mutations only in YMDD domain but not for substitutions in the region targeted by TaqMan assay. As previous mentioned that the substitution rate was higher in the HBV core gene than surface gene. Furthermore analysis of the samples that did not detect with TaqMan assay may be needed to clarify the issue.
In conclusion, these result demonstrated that Abbott RealTime HBV assay is more sensitive in detection of low viral load than CobasAmpliPrep/CobasTaqMan HBV assay. The acquisition of drug-resistance mutations is a constant consideration for patients in NUC therapy. Persistently detectable low levels of HBV DNAincreases the risk of acquiring drug resistant mutationsand the use of a highly sensitive, quantitative HBV DNA assay helps in the early detection of drug resistant mutations.