Potential conflict of interest: Nothing to report.
Although adefovir dipivoxil (ADV) has a unique profile of delayed and infrequent resistance in treatment-naïve chronic hepatitis B patients, the association of ADV resistance with previous lamivudine (LAM) resistance is not well understood. We compared the emergence of the ADV-resistant mutations rtA181V/T and rtN236T between LAM-resistant patients and treatment-naïve patients at 48 weeks of ADV monotherapy. Fifty-seven LAM-resistant patients and 38 treatment-naïve patients were treated with 10 mg/d ADV for more than 48 weeks. Both baseline and 48-week blood samples were analyzed for ADV-resistant mutations via restriction fragment mass polymorphism analysis. Antiviral responses were evaluated according to changes in serum HBV DNA (measured via real-time polymerase chain reaction) and alanine aminotransferase (ALT) levels and loss of hepatitis B e antigen (HBeAg). After 48 weeks, 10 (18%) of the 57 LAM-resistant patients were found to have developed ADV-resistant mutations, whereas none of the 38 treatment-naïve patients developed such mutations (P < .01). Among LAM-resistant patients, the reduction in serum HBV DNA levels was significantly lower in patients with ADV-resistant mutations than in those without such mutations (−1.04 vs. −2.63 log10 copies/mL) (P = .01). However, the rates of serum ALT normalization (60% vs. 55%) and HBeAg loss (14% vs. 21%) were not significantly different between the 2 groups (P > .05). In conclusion, the emergence of the rtA181V/T and rtN236T mutations was more common in LAM-resistant patients than in treatment-naïve patients after 48 weeks of ADV therapy and was associated with reduced antiviral efficacy to drug treatment. (HEPATOLOGY 2006;43:1385–1391.)
The goals of therapy in patients with chronic hepatitis B (CHB) are sustained hepatitis B virus (HBV) DNA suppression, normalization of serum alanine aminotransferase (ALT) levels, and improvement in liver necroinflammation. Long-term objectives include prevention of cirrhosis, end-stage liver disease, and hepatocellular carcinoma and subsequent prolongation of survival.1, 2 To ensure a favorable risk–benefit profile, any treatment regimen must provide durable efficacy and limited toxicity with either minimal or no emergence of viral resistance.2 Lamivudine (LAM) has been shown to achieve virological, biochemical, and histological responses in patients with CHB and is well tolerated.3–6 However, long-term LAM therapy induces drug-resistant mutations in the YMDD motif of HBV DNA polymerase.6, 7 The emergence of LAM-resistant mutations increases with extended treatment duration (16%-32% at 1 year to 70% at 4 years of therapy).7, 8 Mutations associated with LAM resistance result in both a reduction of disease susceptibility to LAM and progression of liver disease in a significant number of patients.9
Adefovir dipivoxil (ADV), an oral pro-drug of adefovir, has an antiviral activity against not only wild-type HBV but also LAM-resistant HBV mutants in vitro and in vivo.10–13 Therefore, a switch to ADV is usually recommended for patients with LAM-resistant CHB.14 In contrast to LAM, ADV therapy in treatment-naïve patients is associated with delayed and infrequent emergence of drug-resistant mutations. Two mutations (alanine to valine or threonine [rtA181V/T] substitutions and asparagine to threonine [rtN236T] substitution) have been reported to confer resistance to ADV.15, 16 The cumulative percentages of patients who showed ADV resistance have been reported to be 0% in year 1, 3% in year 2, 11% in year 3, 18% in year 4, and 28% in year 5.17, 18 Although ADV has a unique profile of delayed and infrequent resistance in treatment-naïve patients, ADV resistance in patients with LAM resistance is not well understood. In this study, we compared the emergence of ADV-resistant mutations between LAM-resistant patients and treatment-naïve patients. We further investigated the effects of the mutation emergence on both virological and biochemical responses to ADV.
CHB, chronic hepatitis B; HBV, hepatitis B virus; ALT, alanine aminotransferase; LAM, lamivudine; ADV, adefovir dipivoxil; HBeAg, hepatitis B e antigen; RFMP, restriction fragment mass polymorphism.
Patients and Methods
Ninety-five patients with CHB were enrolled in the study, 57 of whom were LAM-resistant and 38 of whom were treatment-naïve. Before enrollment, the 57 LAM-resistant patients had documented breakthrough defined as reappearance of serum HBV DNA using a hybrid capture assay on 2 or more occasions after its initial disappearance during prolonged LAM therapy. Serum hepatitis B surface antigen and HBV DNA had been positive for more than 6 months before antiviral therapy in all patients. A diagnosis of CHB was made either histologically or clinically. The clinical criteria for CHB included elevated serum ALT levels over a 6-month period and the absence of any clinical evidence of portal hypertension, such as esophageal varices, ascites, hepatic encephalopathy, and imaging features that are suggestive of cirrhosis on sonographic examination. All patients had well-preserved liver function.
Laboratory and Virological Studies.
Treatment-naïve patients were treated with ADV alone after diagnosis of CHB; LAM-resistant patients were switched to ADV therapy from LAM without an overlapping period. ADV was administered orally at a dosage of 10 mg/d for more than 48 weeks; blood samples were selected for analysis during this same period. For some patients developing ADV-resistant mutations, samples were analyzed for 84 weeks of follow-up. Laboratory tests were performed at baseline and every 12 weeks during ADV therapy. The laboratory tests included serum ALT level, hepatitis B e antigen (HBeAg), and antibodies to HBeAg as determined via radioimmunoassay (Abbott Laboratories, Abbott Park, IL).
HBV DNA Quantification.
Serum HBV DNA levels were quantified at baseline and then every 12 weeks during ADV therapy via real-time polymerase chain reaction. HBV DNA was extracted from 200 μL serum using the QIAamp DNA Mini Kit (Qiagen, Valencia, CA) according to the manufacturer's instructions. HBV DNA samples were amplified using an ABI 7300 instrument (Applied Biosystems, Foster City, CA), and polymerase chain reaction was performed under the following conditions: 50°C for 2 minutes, 95°C for 10 minutes, and 45 cycles of 95°C for 15 seconds and 62°C for 1 minute. The sequences of the forward primer, reverse primer, and TaqMan probe (PE Biosystems, Foster City, CA), respectively, were as follows: 5′-CTCCCCGTCTGTKCCTTCTCATC-3′ (HBVRT1F, nucleotides 1542-1564; K = G or T), 5′-GGCGTTCACGGTGGTCTCCATGC-3′ (HBVRT1R, nucleotides 1625-1603), and 5′ FAM-CCGTGTGCACTTCGCTTCACCTCTGC-TAMRA 3′ (HBV1TAQ, nucleotides 1575-1600). Nucleotide sequence positions were numbered according to Ono et al.19 Standard plasmid DNA was purified by CsCl gradient ultracentrifugation and calibrated using the World Health Organization International Standard for HBV DNA (NIBSC, Potters Bar, UK). Tenfold, serially diluted HBV DNAs ranging from 2.4 × 101 to 2.4 × 1010 IU/mL were used to generate a standard curve for each reaction. One IU/mL was converted into 5 copies/mL according to Saldanha et al.20 The concentrations of HBV DNA in clinical specimens were calculated using the standard curve equation: Ct = slope × log copy + intercept. The dynamic range of detection was found to be 366≈3.66 × 1011 copies/mL (r2 = 0.998).
Detection of Mutations via Restriction Fragment Mass Polymorphism Analysis.
Mutations at rt181 and rt236 were identified both at baseline and after 48 weeks of ADV therapy using restriction fragment mass polymorphism (RFMP) analysis as previously described.21 For RFMP genotyping, viral DNA was amplified via polymerase chain reaction. The sequences of the forward and reverse primers used in polymerase chain reaction are described in Table 1. Sequences with lowercase letters were engineered to insert the MmeI and HaeIII and the MmeI and SspI restriction recognition sites in the amplicons for rtA181V/T and rtN236T mutations, respectively. After complete enzyme digestion, oligonucleotide fragments were analyzed using mass spectrometry. The methodology was validated by a clonal sequencing assay for successful detection of rtM204I/V mutations in LAM-resistant patients.21
Table 1. Primers Used for Amplification in RFMP Assays of the rtA181V/T and rtN236T Mutations
NOTE. Nucleotide sequence positions are numbered according to Ono et al.19
Data are expressed as the mean ± SD or median (range). Group comparisons were performed using the Student t test or the Mann-Whitney U test and χ2 tests. A P value of less than .05 was considered statistically significant.
The baseline characteristics of the 57 LAM-resistant patients and 38 treatment-naïve patients are shown in Table 2. Treatment-naïve patients had higher mean HBV DNA levels and HBeAg-positive rate than LAM-resistant patients. Among the 57 LAM-resistant patients, there were no significant differences in baseline characteristics (age, sex, serum ALT level, and rate of positive HBeAg) between patients with (group A) and without (group B) ADV-resistant mutations after 48 weeks of ADV therapy; however, the mean HBV DNA level was higher in group A than in group B (Table 2). The rt204 mutations were identified in 50 (88%) of the 57 LAM-resistant patients at baseline (8 in group A and 42 in group B). However, none of the treatment-naïve patients had mutations at rt204 at baseline (Table 2).
NOTE. P < .05 for a vs. b, c vs. e, f vs. g; P < .01 for d vs. e, h vs. i; P > .05 for other comparisons.
Abbreviations: YIDD, methionine to isoleucine substitution at rt204; YVDD, methionine to valine substitution at rt204.
Patients developing ADV-resistant mutations.
Patients not developing ADV-resistant mutations after 48 weeks of ADV therapy.
Age in years, median (range)
Serum ALT in IU/L, median (range)
Serum HBV DNA level (log10 copies/mL), mean ± SD
7.50 ± 0.66c
6.83 ± 0.60d
7.79 ± 1.56e
HBeAg positivity, n (%)
YMDD mutation, n (%)
Genotypes at Baseline and at Week 48.
At baseline, the rtA181T mutation was identified in 3 (5%) of the 57 LAM-resistant patients, but the rtA181V and rtN236T mutations were observed in none. Of the 38 treatment-naïve patients, none showed mutations at rt181 and rt236 at baseline (Table 3). After 48 weeks of ADV therapy, 10 (18%) of the 57 LAM-resistant patients developed the rtA181V/T and/or rtN236T mutation; however, none (0%) of 38 treatment-naïve patients developed these mutations (P < .01). As shown in Table 3, the rtN236T single mutation was observed in 3 patients, the rtA181T single mutation was observed in 3 patients, and the rtA181V single mutation was observed in 1 patient. Dual rtA181V/T and rtN236T mutations were detected in 3 patients. Of these 3 patients, 1 had an rtA181V + rtN236T mutation and 2 developed rtA181T + rtN236T mutations. One of the 3 patients in whom the rtA181T mutation was detected at baseline maintained the mutation at 48 weeks, whereas the other 2 lost it. Two patients with the rtA181T mutation and the rtN236T mutation had only minor populations of resistant viruses (approximately 10%).
Table 3. Patterns of ADV-Resistant Mutations at Baseline and Week 48 After ADV Therapy in LAM-Resistant Patients and Treatment-Naïve Patients
LAM-Resistant Patients (n = 57)
Treatment-Naïve Patients (n = 38)
NOTE. rtN236T, asparagines to threonine substitution at rt236; rtA181V, alanine to valine substitution at rt181; rtA181T, alanine to threonine substitution at rt181. P < .01 for a vs. b.
One patient maintained the rtA181T mutation after 48 weeks of ADV therapy, but the other 2 lost it.
After 48 weeks of ADV therapy, LAM-resistant patients showed a smaller reduction in serum HBV DNA levels than treatment-naïve patients (2.39 and 4.04 log10 copies/mL, respectively, P < .01) (Table 4). Similarly, LAM-resistant patients tended to have a lower rate of serum ALT normalization than treatment-naïve patients, but the differences were not statistically significant (60% and 79%, respectively; P = .07). There were no differences in the rates of undetectable HBV DNA (19% vs. 29%) and HBeAg loss (20% vs. 14%) at week 48 between the 2 groups (Table 4). In the LAM-resistant group, the patients who developed ADV-resistant mutations had much smaller reductions in serum HBV DNA levels compared with those who did not develop such mutations (−1.04 and −2.63 log10 copies/mL, respectively; P = .01) (Table 5). However, there were no significant differences in terms of undetectable HBV DNA (0% vs. 23%), HBeAg loss (14% vs. 21%), or serum ALT normalization (60% vs. 55%) between patients developing and not developing ADV-resistant mutations (P > .05) (Table 5). In comparison with changes of the mean HBV DNA levels during 48 weeks of ADV therapy, the LAM-resistant patients developing ADV-resistant mutations showed the poorest virological responses throughout 48 weeks of ADV therapy compared with the LAM-resistant patients not developing ADV-resistant mutations and the treatment-naïve patients. Rather, the mean HBV DNA level was increased at 48 weeks during ADV therapy in the patients developing ADV-resistant mutations (6.27 log10 copies/mL at 36 weeks vs. 6.52 log10 copies/mL at week 48), while it was steadily decreased in the other 2 groups during 48-week ADV therapy (Fig. 1).
Table 4. Comparisons of Antiviral Responses at 48 Weeks of ADV Therapy Between LAM-Resistant Patients and Treatment-Naïve Patients
LAM-Resistant Patients (n = 57)
Treatment-Naïve Patients (n = 38)
Detection limit via real-time polymerase chain reaction with a dynamic range of 366∼3.66 × 1011 copies/mL (r2 = 0.998).
Nine of 45 patients who tested positive for HBeAg at baseline.
Five of 37 patients who tested positive for HBeAg at baseline.
Decrease of serum HBV DNA levels (log10 copies/mL), mean ± SD
Antiviral Responses in Patients Developing ADV-Resistant Mutations.
All 10 patients developing ADV-resistant mutations at 48 weeks showed variable patterns of virological and biochemical responses during the median follow-up of 84 weeks. Of these, 4 patients—one each with the rtA181T+rtN236T, rtA181V+rtN236T, rtA181V, and rtN236T mutation—subsequently rebounded in HBV DNA with increases of more than 1 log10 copies/mL on 1 or 2 consecutive tests in serum HBV DNA levels at 24, 48, 48, and 60 weeks of ADV therapy, respectively. All had elevated ALT level as well (Fig. 2B). Two patients with the rtA181T mutation showed poor virological responses with reduction of less than 1 log10 copies/mL throughout the follow-up periods.
In this study, ADV-resistant mutations were observed in 18% of LAM-resistant patients but were absent in treatment-naïve patients after 48 weeks of ADV therapy. Our data are in agreement with previous studies that show ADV-resistant mutations are infrequent and delayed in treatment-naïve patients.10, 11, 22 However, in contrast to previous reports in which no ADV-resistant mutations were detected in LAM-resistant patients after 1 year of ADV therapy,23, 24 our study indicated that the emergence of those mutations in LAM-resistant patients was frequently detected after 1 year of ADV therapy.
Our results strongly suggest that patients with LAM resistance may be vulnerable to the emergence of ADV resistance. Fung et al.25 recently reported higher probability of ADV resistance (0% at 1 year, 22% at 2 years) in patients with LAM resistance compared with previous results in treatment-naïve patients. Similarly, a 2-year entecavir study has shown that entecavir resistance was found in 10% of LAM-resistant patients after 2 years of entecavir therapy, but not in treatment-naïve patients.26 In addition, it was demonstrated that resistance to ADV arose only in patients who received ADV monotherapy and not in patients who received both ADV and LAM together.18, 25 One in vitro study suggested that the subsequent use of nucleotide/nucleoside analogs could lead to the selection of multiple drug-resistant HBV strains.27 Other studies indicated that the monotherapy of HBV, including sequential monotherapy, encourages the selection of viral resistance.28 Therefore, ADV/LAM combination therapy may be a better option to prevent or delay the emergence of ADV-resistant mutations in patients with LAM-resistant CHB.
Development of ADV-resistant mutations was associated with considerably reduced antiviral suppression in this study. Treatment-naïve patients showed a reduction of more than 4 log10 copies/mL in serum HBV DNA levels at 48 weeks of ADV therapy, much more than LAM-resistant patients (< 3 log10 copies/mL). In the LAM-resistant group, patients who developed ADV-resistant mutations showed a much smaller decrease in serum HBV DNA levels during 48 weeks of ADV therapy compared with patients who did not develop such mutations. All patients who developed ADV-resistant mutations showed poor virological responses throughout follow-up periods of a median of 84 weeks. Among them, 4 patients showed HBV DNA rebound as well as ALT elevation during 24 to 60 weeks after ADV therapy. Because genotypic analyses were not performed during the entire follow-up period, it was difficult to predict the exact time when the mutations occurred. However, as Fig. 1 shows, the early rebound of HBV DNA suggests that ADV-resistant mutations may occur much earlier than 48 weeks of ADV therapy.
None of the patients in our study developed clinical deterioration to a decompensated state throughout the follow-up periods. Other studies have shown that an increase of serum ALT after the emergence of ADV resistance is mild to moderate and is rarely associated with hepatic decompensation.2, 29 Whereas in vitro data have suggested that the rtA181V/T and rtN236T mutations lead to only minor reductions in sensitivity to ADV (≈3- to 13-fold),30, 31 clinical studies have indicated that viral rebound and hepatic decompensation may occur in some patients with advanced disease.23 Therefore, additional incidence of ADV-resistant mutations, particularly in patients who do demonstrate LAM resistance, must be carefully assessed. Given the detection of a mixed population of ADV-resistant mutant and wild-type viruses in association with poor virological responses (Fig. 2), regular testing for emerging mutations in patients with LAM resistance may be of value.
MALDI-TOF MS-based RFMP assay was shown to identify viral variants present at concentrations as low as 1% of total virus populations with defined mixtures and showed good concordance with clonal sequencing results. It has been demonstrated that there is a good correlation between estimated peak heights and real proportions in mixed genotype pools, indicating the RFMP assay enables better quantitative detection of mixed populations without the need for population-based cloning and subsequent sequencing.21, 32, 33 Thus, we performed genotype analyses using RFMP assay and confirmed genotype results by comparing them with those from direct sequence analysis on the baseline and 48-week sera of the 12 patients with ADV-resistant mutations. There was a good concordance in most samples between the 2 methods. In only 2 samples with minor populations of resistant viruses (approximately 10%), RFMP identified mutant viruses that the direct sequencing method did not. This underlines the sensitivity of the RFMP assay and demonstrated that valuable additional information can be obtained via RFMP analysis over conventional direct sequencing. However, it cannot be excluded that ADV-resistant HBV exists at baseline and results in increased frequency of ADV resistance in patients with LAM resistance, because current genotyping analyses are limited by a sensitivity, and minute amounts of mutants (i.e., <0.1%) would not be detected.
In conclusion, our study suggests that ADV monotherapy in patients with LAM resistance may induce frequent and early emergence of ADV-resistant mutations compared with ADV monotherapy in treatment-naïve patients, and development of such mutations may be associated with reduced susceptibility to ADV. Further studies should aim to assess the long-term profiles of ADV resistance and its clinical course, especially in patients with LAM resistance.
We gratefully acknowledge the valuable assistance of Dr. William Folk in reviewing the manuscript.